From: Keith Busch keith.busch@intel.com

mainline inclusion from mainline-v4.20-rc1 commit 27c1ce8bbed7e7f0e4a87cf4a93f09be26d62ada category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

The kernel provides a generic FIFO implementation, so no need to reinvent that capability in a driver. Replace the AER-specific implementation with the kernel-provided kfifo. Since the interrupt handler producer and work queue consumer run single threaded, there is no need for additional locking, so remove that lock, too.

Signed-off-by: Keith Busch keith.busch@intel.com Signed-off-by: Bjorn Helgaas bhelgaas@google.com Signed-off-by: liyouhong liyouhong@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xiongfeng Wang wangxiongfeng2@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/pci/pcie/aer.c | 88 ++++++------------------------------------ 1 file changed, 11 insertions(+), 77 deletions(-)

diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c index 52aa1555620a8..14acf51cd9c31 100644 --- a/drivers/pci/pcie/aer.c +++ b/drivers/pci/pcie/aer.c @@ -30,7 +30,7 @@ #include "../pci.h" #include "portdrv.h"

-#define AER_ERROR_SOURCES_MAX 100 +#define AER_ERROR_SOURCES_MAX 128

#define AER_MAX_TYPEOF_COR_ERRS 16 /* as per PCI_ERR_COR_STATUS */ #define AER_MAX_TYPEOF_UNCOR_ERRS 26 /* as per PCI_ERR_UNCOR_STATUS*/ @@ -43,14 +43,8 @@ struct aer_err_source { struct aer_rpc { struct pci_dev *rpd; /* Root Port device */ struct work_struct dpc_handler; - struct aer_err_source e_sources[AER_ERROR_SOURCES_MAX]; - unsigned short prod_idx; /* Error Producer Index */ - unsigned short cons_idx; /* Error Consumer Index */ + DECLARE_KFIFO(aer_fifo, struct aer_err_source, AER_ERROR_SOURCES_MAX); int isr; - spinlock_t e_lock; /* - * Lock access to Error Status/ID Regs - * and error producer/consumer index - */ struct mutex rpc_mutex; /* * only one thread could do * recovery on the same @@ -1213,35 +1207,6 @@ static void aer_isr_one_error(struct aer_rpc *rpc, } }

-/** - * get_e_source - retrieve an error source - * @rpc: pointer to the root port which holds an error - * @e_src: pointer to store retrieved error source - * - * Return 1 if an error source is retrieved, otherwise 0. - * - * Invoked by DPC handler to consume an error. - */ -static int get_e_source(struct aer_rpc *rpc, struct aer_err_source *e_src) -{ - unsigned long flags; - - /* Lock access to Root error producer/consumer index */ - spin_lock_irqsave(&rpc->e_lock, flags); - if (rpc->prod_idx == rpc->cons_idx) { - spin_unlock_irqrestore(&rpc->e_lock, flags); - return 0; - } - - *e_src = rpc->e_sources[rpc->cons_idx]; - rpc->cons_idx++; - if (rpc->cons_idx == AER_ERROR_SOURCES_MAX) - rpc->cons_idx = 0; - spin_unlock_irqrestore(&rpc->e_lock, flags); - - return 1; -} - /** * aer_isr - consume errors detected by root port * @work: definition of this work item @@ -1254,7 +1219,7 @@ static void aer_isr(struct work_struct *work) struct aer_err_source uninitialized_var(e_src);

mutex_lock(&rpc->rpc_mutex); - while (get_e_source(rpc, &e_src)) + while (kfifo_get(&rpc->aer_fifo, &e_src)) aer_isr_one_error(rpc, &e_src); mutex_unlock(&rpc->rpc_mutex); } @@ -1268,51 +1233,23 @@ static void aer_isr(struct work_struct *work) */ irqreturn_t aer_irq(int irq, void *context) { - unsigned int status, id; struct pcie_device *pdev = (struct pcie_device *)context; struct aer_rpc *rpc = get_service_data(pdev); - int next_prod_idx; - unsigned long flags; - int pos; - - pos = pdev->port->aer_cap; - /* - * Must lock access to Root Error Status Reg, Root Error ID Reg, - * and Root error producer/consumer index - */ - spin_lock_irqsave(&rpc->e_lock, flags); + struct pci_dev *rp = rpc->rpd; + struct aer_err_source e_src = {}; + int pos = rp->aer_cap;

- /* Read error status */ - pci_read_config_dword(pdev->port, pos + PCI_ERR_ROOT_STATUS, &status); - if (!(status & (PCI_ERR_ROOT_UNCOR_RCV|PCI_ERR_ROOT_COR_RCV))) { - spin_unlock_irqrestore(&rpc->e_lock, flags); + pci_read_config_dword(rp, pos + PCI_ERR_ROOT_STATUS, &e_src.status); + if (!(e_src.status & (PCI_ERR_ROOT_UNCOR_RCV|PCI_ERR_ROOT_COR_RCV))) return IRQ_NONE; - }

- /* Read error source and clear error status */ - pci_read_config_dword(pdev->port, pos + PCI_ERR_ROOT_ERR_SRC, &id); - pci_write_config_dword(pdev->port, pos + PCI_ERR_ROOT_STATUS, status); + pci_read_config_dword(rp, pos + PCI_ERR_ROOT_ERR_SRC, &e_src.id); + pci_write_config_dword(rp, pos + PCI_ERR_ROOT_STATUS, e_src.status);

- /* Store error source for later DPC handler */ - next_prod_idx = rpc->prod_idx + 1; - if (next_prod_idx == AER_ERROR_SOURCES_MAX) - next_prod_idx = 0; - if (next_prod_idx == rpc->cons_idx) { - /* - * Error Storm Condition - possibly the same error occurred. - * Drop the error. - */ - spin_unlock_irqrestore(&rpc->e_lock, flags); + if (!kfifo_put(&rpc->aer_fifo, e_src)) return IRQ_HANDLED; - } - rpc->e_sources[rpc->prod_idx].status = status; - rpc->e_sources[rpc->prod_idx].id = id; - rpc->prod_idx = next_prod_idx; - spin_unlock_irqrestore(&rpc->e_lock, flags);

- /* Invoke DPC handler */ schedule_work(&rpc->dpc_handler); - return IRQ_HANDLED; } EXPORT_SYMBOL_GPL(aer_irq); @@ -1437,9 +1374,6 @@ static struct aer_rpc *aer_alloc_rpc(struct pcie_device *dev) if (!rpc) return NULL;

- /* Initialize Root lock access, e_lock, to Root Error Status Reg */ - spin_lock_init(&rpc->e_lock); - rpc->rpd = dev->port; INIT_WORK(&rpc->dpc_handler, aer_isr); mutex_init(&rpc->rpc_mutex);

From: Yicong Yang yangyicong@hisilicon.com

mainline inclusion from mainline-v5.6-rc1 commit 01daacfb9035e5b86d43a01f11a0614648f306c1 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

PCI error recovery will fail if any device under the Root Port doesn't have an error_detected callback. Currently only the failure result is printed, which is not enough to identify the driver that lacks the callback.

Log a message to identify the device with no error_detected callback.

[bhelgaas: tweak log message] Link: https://lore.kernel.org/r/1576237474-32021-1-git-send-email-yangyicong@hisil... Signed-off-by: Yicong Yang yangyicong@hisilicon.com Signed-off-by: Bjorn Helgaas bhelgaas@google.com Signed-off-by: liyouhong liyouhong@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xiongfeng Wang wangxiongfeng2@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/pci/pcie/err.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/pci/pcie/err.c b/drivers/pci/pcie/err.c index 2c3b5bd59b18f..571ac3d9c3265 100644 --- a/drivers/pci/pcie/err.c +++ b/drivers/pci/pcie/err.c @@ -69,10 +69,12 @@ static int report_error_detected(struct pci_dev *dev, void *data) * error callbacks of "any" device in the subtree, and will * exit in the disconnected error state. */ - if (dev->hdr_type != PCI_HEADER_TYPE_BRIDGE) + if (dev->hdr_type != PCI_HEADER_TYPE_BRIDGE) { vote = PCI_ERS_RESULT_NO_AER_DRIVER; - else + pci_info(dev, "AER: Can't recover (no error_detected callback)\n"); + } else { vote = PCI_ERS_RESULT_NONE; + } } else { err_handler = dev->driver->err_handler; vote = err_handler->error_detected(dev, result_data->state);

From: Yicong Yang yangyicong@hisilicon.com

mainline inclusion from mainline-v5.7-rc1 commit 9cb3985af63555810bb07de50acdf4170771451d category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

Link speed 32.0 GT/s is supported in PCIe r5.0. Add this speed to PCIE_SPEED2STR() and PCIE_SPEED2MBS_ENC() to correctly decode it.

This is complementary to de76cda215d5 ("PCI: Decode PCIe 32 GT/s link speed").

Link: https://lore.kernel.org/r/1581937984-40353-2-git-send-email-yangyicong@hisil... Signed-off-by: Yicong Yang yangyicong@hisilicon.com Signed-off-by: Bjorn Helgaas bhelgaas@google.com Signed-off-by: liyouhong liyouhong@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xiongfeng Wang wangxiongfeng2@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/pci/pci.h | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h index 091a36413863f..918673f78865d 100644 --- a/drivers/pci/pci.h +++ b/drivers/pci/pci.h @@ -255,7 +255,8 @@ void pci_disable_bridge_window(struct pci_dev *dev);

/* PCIe link information */ #define PCIE_SPEED2STR(speed) \ - ((speed) == PCIE_SPEED_16_0GT ? "16 GT/s" : \ + ((speed) == PCIE_SPEED_32_0GT ? "32 GT/s" : \ + (speed) == PCIE_SPEED_16_0GT ? "16 GT/s" : \ (speed) == PCIE_SPEED_8_0GT ? "8 GT/s" : \ (speed) == PCIE_SPEED_5_0GT ? "5 GT/s" : \ (speed) == PCIE_SPEED_2_5GT ? "2.5 GT/s" : \ @@ -263,7 +264,8 @@ void pci_disable_bridge_window(struct pci_dev *dev);

/* PCIe speed to Mb/s reduced by encoding overhead */ #define PCIE_SPEED2MBS_ENC(speed) \ - ((speed) == PCIE_SPEED_16_0GT ? 16000*128/130 : \ + ((speed) == PCIE_SPEED_32_0GT ? 32000*128/130 : \ + (speed) == PCIE_SPEED_16_0GT ? 16000*128/130 : \ (speed) == PCIE_SPEED_8_0GT ? 8000*128/130 : \ (speed) == PCIE_SPEED_5_0GT ? 5000*8/10 : \ (speed) == PCIE_SPEED_2_5GT ? 2500*8/10 : \

From: Bjorn Helgaas bhelgaas@google.com

mainline inclusion from mainline-v5.7-rc1 commit 6348a34dcb98d8e285685a205f2a601817fa2d38 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

Previously some PCI speed strings came from pci_speed_string(), some came from the PCIe-specific PCIE_SPEED2STR(), and some came from a PCIe-specific switch statement. These methods were inconsistent:

pci_speed_string() PCIE_SPEED2STR() switch ------------------ ---------------- ------ 33 MHz PCI ... 2.5 GT/s PCIe 2.5 GT/s 2.5 GT/s 5.0 GT/s PCIe 5 GT/s 5 GT/s 8.0 GT/s PCIe 8 GT/s 8 GT/s 16.0 GT/s PCIe 16 GT/s 16 GT/s 32.0 GT/s PCIe 32 GT/s 32 GT/s

Standardize on pci_speed_string() as the single source of these strings.

Note that this adds ".0" and "PCIe" to some messages, including sysfs "max_link_speed" files, a brcmstb "link up" message, and the link status dmesg logging, e.g.,

nvme 0000:01:00.0: 16.000 Gb/s available PCIe bandwidth, limited by 5.0 GT/s PCIe x4 link at 0000:00:01.1 (capable of 31.504 Gb/s with 8.0 GT/s PCIe x4 link)

I think it's better to standardize on a single version of the speed text. Previously we had strings like this:

/sys/bus/pci/slots/0/cur_bus_speed: 8.0 GT/s PCIe /sys/bus/pci/slots/0/max_bus_speed: 8.0 GT/s PCIe /sys/devices/pci0000:00/0000:00:1c.0/current_link_speed: 8 GT/s /sys/devices/pci0000:00/0000:00:1c.0/max_link_speed: 8 GT/s

This changes the latter two to match the slots files:

/sys/devices/pci0000:00/0000:00:1c.0/current_link_speed: 8.0 GT/s PCIe /sys/devices/pci0000:00/0000:00:1c.0/max_link_speed: 8.0 GT/s PCIe

Based-on-patch by: Yicong Yang yangyicong@hisilicon.com Signed-off-by: Bjorn Helgaas bhelgaas@google.com Signed-off-by: liyouhong liyouhong@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xiongfeng Wang wangxiongfeng2@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/pci/pci-sysfs.c | 27 +++++---------------------- drivers/pci/pci.c | 6 +++--- drivers/pci/pci.h | 9 --------- 3 files changed, 8 insertions(+), 34 deletions(-)

diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index e10fb16f8b135..a62349a44c8c0 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -156,7 +156,8 @@ static ssize_t max_link_speed_show(struct device *dev, { struct pci_dev *pdev = to_pci_dev(dev);

- return sprintf(buf, "%s\n", PCIE_SPEED2STR(pcie_get_speed_cap(pdev))); + return sprintf(buf, "%s\n", + pci_speed_string(pcie_get_speed_cap(pdev))); } static DEVICE_ATTR_RO(max_link_speed);

@@ -175,33 +176,15 @@ static ssize_t current_link_speed_show(struct device *dev, struct pci_dev *pci_dev = to_pci_dev(dev); u16 linkstat; int err; - const char *speed; + enum pci_bus_speed speed;

err = pcie_capability_read_word(pci_dev, PCI_EXP_LNKSTA, &linkstat); if (err) return -EINVAL;

- switch (linkstat & PCI_EXP_LNKSTA_CLS) { - case PCI_EXP_LNKSTA_CLS_32_0GB: - speed = "32 GT/s"; - break; - case PCI_EXP_LNKSTA_CLS_16_0GB: - speed = "16 GT/s"; - break; - case PCI_EXP_LNKSTA_CLS_8_0GB: - speed = "8 GT/s"; - break; - case PCI_EXP_LNKSTA_CLS_5_0GB: - speed = "5 GT/s"; - break; - case PCI_EXP_LNKSTA_CLS_2_5GB: - speed = "2.5 GT/s"; - break; - default: - speed = "Unknown speed"; - } + speed = pcie_link_speed[linkstat & PCI_EXP_LNKSTA_CLS];

- return sprintf(buf, "%s\n", speed); + return sprintf(buf, "%s\n", pci_speed_string(speed)); } static DEVICE_ATTR_RO(current_link_speed);

diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index cf2fa9f656177..949a485aa68b1 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -5667,14 +5667,14 @@ void __pcie_print_link_status(struct pci_dev *dev, bool verbose) if (bw_avail >= bw_cap && verbose) pci_info(dev, "%u.%03u Gb/s available PCIe bandwidth (%s x%d link)\n", bw_cap / 1000, bw_cap % 1000, - PCIE_SPEED2STR(speed_cap), width_cap); + pci_speed_string(speed_cap), width_cap); else if (bw_avail < bw_cap) pci_info(dev, "%u.%03u Gb/s available PCIe bandwidth, limited by %s x%d link at %s (capable of %u.%03u Gb/s with %s x%d link)\n", bw_avail / 1000, bw_avail % 1000, - PCIE_SPEED2STR(speed), width, + pci_speed_string(speed), width, limiting_dev ? pci_name(limiting_dev) : "<unknown>", bw_cap / 1000, bw_cap % 1000, - PCIE_SPEED2STR(speed_cap), width_cap); + pci_speed_string(speed_cap), width_cap); }

/** diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h index cbf2ebab757ae..8ecc8a5bd0ff3 100644 --- a/drivers/pci/pci.h +++ b/drivers/pci/pci.h @@ -253,15 +253,6 @@ bool pci_bus_clip_resource(struct pci_dev *dev, int idx); void pci_reassigndev_resource_alignment(struct pci_dev *dev); void pci_disable_bridge_window(struct pci_dev *dev);

-/* PCIe link information */ -#define PCIE_SPEED2STR(speed) \ - ((speed) == PCIE_SPEED_32_0GT ? "32 GT/s" : \ - (speed) == PCIE_SPEED_16_0GT ? "16 GT/s" : \ - (speed) == PCIE_SPEED_8_0GT ? "8 GT/s" : \ - (speed) == PCIE_SPEED_5_0GT ? "5 GT/s" : \ - (speed) == PCIE_SPEED_2_5GT ? "2.5 GT/s" : \ - "Unknown speed") - /* PCIe speed to Mb/s reduced by encoding overhead */ #define PCIE_SPEED2MBS_ENC(speed) \ ((speed) == PCIE_SPEED_32_0GT ? 32000*128/130 : \

From: Jay Fang f.fangjian@huawei.com

mainline inclusion from mainline-v5.8-rc1 commit 5dda3ba6fc9c5d784b48687a3f3003023a0d7c74 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

Fix kernel-doc of the "srv" parameter to pcie_pme_resume() and pcie_pme_remove(). Building with W=1 produced these warnings:

drivers/pci/pcie/pme.c:414: warning: Function parameter or member 'srv' not described in 'pcie_pme_resume' drivers/pci/pcie/pme.c:437: warning: Function parameter or member 'srv' not described in 'pcie_pme_remove'

Link: https://lore.kernel.org/r/1589612414-61682-1-git-send-email-f.fangjian@huawe... Signed-off-by: Jay Fang f.fangjian@huawei.com Signed-off-by: Bjorn Helgaas bhelgaas@google.com Signed-off-by: liyouhong liyouhong@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xiongfeng Wang wangxiongfeng2@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/pci/pcie/pme.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/pci/pcie/pme.c b/drivers/pci/pcie/pme.c index 54d593d10396f..94479ec041c22 100644 --- a/drivers/pci/pcie/pme.c +++ b/drivers/pci/pcie/pme.c @@ -406,7 +406,7 @@ static int pcie_pme_suspend(struct pcie_device *srv)

/** * pcie_pme_resume - Resume PCIe PME service device. - * @srv - PCIe service device to resume. + * @srv: PCIe service device to resume. */ static int pcie_pme_resume(struct pcie_device *srv) { @@ -429,7 +429,7 @@ static int pcie_pme_resume(struct pcie_device *srv)

/** * pcie_pme_remove - Prepare PCIe PME service device for removal. - * @srv - PCIe service device to remove. + * @srv: PCIe service device to remove. */ static void pcie_pme_remove(struct pcie_device *srv) {

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.2-rc1 commit 6017608936c1825ff5d7325270484042f597edff category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

Add Icelake core PMU perf code, including constraint tables and the main enable code.

Icelake expanded the generic counters to always 8 even with HT on, but a range of events cannot be scheduled on the extra 4 counters. Add new constraint ranges to describe this to the scheduler. The number of constraints that need to be checked is larger now than with earlier CPUs. At some point we may need a new data structure to look them up more efficiently than with linear search. So far it still seems to be acceptable however.

Icelake added a new fixed counter SLOTS. Full support for it is added later in the patch series.

The cache events table is identical to Skylake.

Compare to PEBS instruction event on generic counter, fixed counter 0 has less skid. Force instruction:ppp always in fixed counter 0.

Originally-by: Andi Kleen ak@linux.intel.com Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Arnaldo Carvalho de Melo acme@redhat.com Cc: Jiri Olsa jolsa@redhat.com Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Stephane Eranian eranian@google.com Cc: Thomas Gleixner tglx@linutronix.de Cc: Vince Weaver vincent.weaver@maine.edu Cc: acme@kernel.org Cc: jolsa@kernel.org Link: https://lkml.kernel.org/r/20190402194509.2832-9-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/intel/core.c | 110 ++++++++++++++++++++++++++++++ arch/x86/events/intel/ds.c | 25 ++++++- arch/x86/events/perf_event.h | 2 + arch/x86/include/asm/intel_ds.h | 2 +- arch/x86/include/asm/perf_event.h | 2 +- 5 files changed, 137 insertions(+), 4 deletions(-)

diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 710c24846df24..63a140bc3adda 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -238,6 +238,35 @@ static struct extra_reg intel_skl_extra_regs[] __read_mostly = { EVENT_EXTRA_END };

+static struct event_constraint intel_icl_event_constraints[] = { + FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */ + INTEL_UEVENT_CONSTRAINT(0x1c0, 0), /* INST_RETIRED.PREC_DIST */ + FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */ + FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */ + FIXED_EVENT_CONSTRAINT(0x0400, 3), /* SLOTS */ + INTEL_EVENT_CONSTRAINT_RANGE(0x03, 0x0a, 0xf), + INTEL_EVENT_CONSTRAINT_RANGE(0x1f, 0x28, 0xf), + INTEL_EVENT_CONSTRAINT(0x32, 0xf), /* SW_PREFETCH_ACCESS.* */ + INTEL_EVENT_CONSTRAINT_RANGE(0x48, 0x54, 0xf), + INTEL_EVENT_CONSTRAINT_RANGE(0x60, 0x8b, 0xf), + INTEL_UEVENT_CONSTRAINT(0x04a3, 0xff), /* CYCLE_ACTIVITY.STALLS_TOTAL */ + INTEL_UEVENT_CONSTRAINT(0x10a3, 0xff), /* CYCLE_ACTIVITY.STALLS_MEM_ANY */ + INTEL_EVENT_CONSTRAINT(0xa3, 0xf), /* CYCLE_ACTIVITY.* */ + INTEL_EVENT_CONSTRAINT_RANGE(0xa8, 0xb0, 0xf), + INTEL_EVENT_CONSTRAINT_RANGE(0xb7, 0xbd, 0xf), + INTEL_EVENT_CONSTRAINT_RANGE(0xd0, 0xe6, 0xf), + INTEL_EVENT_CONSTRAINT_RANGE(0xf0, 0xf4, 0xf), + EVENT_CONSTRAINT_END +}; + +static struct extra_reg intel_icl_extra_regs[] __read_mostly = { + INTEL_UEVENT_EXTRA_REG(0x01b7, MSR_OFFCORE_RSP_0, 0x3fffff9fffull, RSP_0), + INTEL_UEVENT_EXTRA_REG(0x01bb, MSR_OFFCORE_RSP_1, 0x3fffff9fffull, RSP_1), + INTEL_UEVENT_PEBS_LDLAT_EXTRA_REG(0x01cd), + INTEL_UEVENT_EXTRA_REG(0x01c6, MSR_PEBS_FRONTEND, 0x7fff17, FE), + EVENT_EXTRA_END +}; + EVENT_ATTR_STR(mem-loads, mem_ld_nhm, "event=0x0b,umask=0x10,ldlat=3"); EVENT_ATTR_STR(mem-loads, mem_ld_snb, "event=0xcd,umask=0x1,ldlat=3"); EVENT_ATTR_STR(mem-stores, mem_st_snb, "event=0xcd,umask=0x2"); @@ -3240,6 +3269,9 @@ static struct event_constraint counter0_constraint = static struct event_constraint counter2_constraint = EVENT_CONSTRAINT(0, 0x4, 0);

+static struct event_constraint fixed0_constraint = + FIXED_EVENT_CONSTRAINT(0x00c0, 0); + static struct event_constraint * hsw_get_event_constraints(struct cpu_hw_events *cpuc, int idx, struct perf_event *event) @@ -3258,6 +3290,21 @@ hsw_get_event_constraints(struct cpu_hw_events *cpuc, int idx, return c; }

+static struct event_constraint * +icl_get_event_constraints(struct cpu_hw_events *cpuc, int idx, + struct perf_event *event) +{ + /* + * Fixed counter 0 has less skid. + * Force instruction:ppp in Fixed counter 0 + */ + if ((event->attr.precise_ip == 3) && + constraint_match(&fixed0_constraint, event->hw.config)) + return &fixed0_constraint; + + return hsw_get_event_constraints(cpuc, idx, event); +} + static struct event_constraint * glp_get_event_constraints(struct cpu_hw_events *cpuc, int idx, struct perf_event *event) @@ -3924,6 +3971,42 @@ static struct attribute *hsw_tsx_events_attrs[] = { NULL };

+EVENT_ATTR_STR(tx-capacity-read, tx_capacity_read, "event=0x54,umask=0x80"); +EVENT_ATTR_STR(tx-capacity-write, tx_capacity_write, "event=0x54,umask=0x2"); +EVENT_ATTR_STR(el-capacity-read, el_capacity_read, "event=0x54,umask=0x80"); +EVENT_ATTR_STR(el-capacity-write, el_capacity_write, "event=0x54,umask=0x2"); + +static struct attribute *icl_events_attrs[] = { + EVENT_PTR(mem_ld_hsw), + EVENT_PTR(mem_st_hsw), + NULL, +}; + +static struct attribute *icl_tsx_events_attrs[] = { + EVENT_PTR(tx_start), + EVENT_PTR(tx_abort), + EVENT_PTR(tx_commit), + EVENT_PTR(tx_capacity_read), + EVENT_PTR(tx_capacity_write), + EVENT_PTR(tx_conflict), + EVENT_PTR(el_start), + EVENT_PTR(el_abort), + EVENT_PTR(el_commit), + EVENT_PTR(el_capacity_read), + EVENT_PTR(el_capacity_write), + EVENT_PTR(el_conflict), + EVENT_PTR(cycles_t), + EVENT_PTR(cycles_ct), + NULL, +}; + +static __init struct attribute **get_icl_events_attrs(void) +{ + return boot_cpu_has(X86_FEATURE_RTM) ? + merge_attr(icl_events_attrs, icl_tsx_events_attrs) : + icl_events_attrs; +} + static __init struct attribute **get_hsw_events_attrs(void) { return boot_cpu_has(X86_FEATURE_RTM) ? @@ -4478,6 +4561,33 @@ __init int intel_pmu_init(void) name = "skylake"; break;

+ case INTEL_FAM6_ICELAKE_MOBILE: + x86_pmu.late_ack = true; + memcpy(hw_cache_event_ids, skl_hw_cache_event_ids, sizeof(hw_cache_event_ids)); + memcpy(hw_cache_extra_regs, skl_hw_cache_extra_regs, sizeof(hw_cache_extra_regs)); + hw_cache_event_ids[C(ITLB)][C(OP_READ)][C(RESULT_ACCESS)] = -1; + intel_pmu_lbr_init_skl(); + + x86_pmu.event_constraints = intel_icl_event_constraints; + x86_pmu.pebs_constraints = intel_icl_pebs_event_constraints; + x86_pmu.extra_regs = intel_icl_extra_regs; + x86_pmu.pebs_aliases = NULL; + x86_pmu.pebs_prec_dist = true; + x86_pmu.flags |= PMU_FL_HAS_RSP_1; + x86_pmu.flags |= PMU_FL_NO_HT_SHARING; + + x86_pmu.hw_config = hsw_hw_config; + x86_pmu.get_event_constraints = icl_get_event_constraints; + extra_attr = boot_cpu_has(X86_FEATURE_RTM) ? + hsw_format_attr : nhm_format_attr; + extra_attr = merge_attr(extra_attr, skl_format_attr); + x86_pmu.cpu_events = get_icl_events_attrs(); + x86_pmu.lbr_pt_coexist = true; + intel_pmu_pebs_data_source_skl(false); + pr_cont("Icelake events, "); + name = "icelake"; + break; + default: switch (x86_pmu.version) { case 1: diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index 0405ed81a791d..2de55e1bef8ec 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -849,6 +849,26 @@ struct event_constraint intel_skl_pebs_event_constraints[] = { EVENT_CONSTRAINT_END };

+struct event_constraint intel_icl_pebs_event_constraints[] = { + INTEL_FLAGS_UEVENT_CONSTRAINT(0x1c0, 0x100000000ULL), /* INST_RETIRED.PREC_DIST */ + INTEL_FLAGS_UEVENT_CONSTRAINT(0x0400, 0x400000000ULL), /* SLOTS */ + + INTEL_PLD_CONSTRAINT(0x1cd, 0xff), /* MEM_TRANS_RETIRED.LOAD_LATENCY */ + INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x1d0, 0xf), /* MEM_INST_RETIRED.LOAD */ + INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x2d0, 0xf), /* MEM_INST_RETIRED.STORE */ + + INTEL_FLAGS_EVENT_CONSTRAINT_DATALA_LD_RANGE(0xd1, 0xd4, 0xf), /* MEM_LOAD_*_RETIRED.* */ + + INTEL_FLAGS_EVENT_CONSTRAINT(0xd0, 0xf), /* MEM_INST_RETIRED.* */ + + /* + * Everything else is handled by PMU_FL_PEBS_ALL, because we + * need the full constraints from the main table. + */ + + EVENT_CONSTRAINT_END +}; + struct event_constraint *intel_pebs_constraints(struct perf_event *event) { struct event_constraint *c; @@ -960,7 +980,7 @@ void intel_pmu_pebs_enable(struct perf_event *event)

cpuc->pebs_enabled |= 1ULL << hwc->idx;

- if (event->hw.flags & PERF_X86_EVENT_PEBS_LDLAT) + if ((event->hw.flags & PERF_X86_EVENT_PEBS_LDLAT) && (x86_pmu.version < 5)) cpuc->pebs_enabled |= 1ULL << (hwc->idx + 32); else if (event->hw.flags & PERF_X86_EVENT_PEBS_ST) cpuc->pebs_enabled |= 1ULL << 63; @@ -1004,7 +1024,8 @@ void intel_pmu_pebs_disable(struct perf_event *event)

cpuc->pebs_enabled &= ~(1ULL << hwc->idx);

- if (event->hw.flags & PERF_X86_EVENT_PEBS_LDLAT) + if ((event->hw.flags & PERF_X86_EVENT_PEBS_LDLAT) && + (x86_pmu.version < 5)) cpuc->pebs_enabled &= ~(1ULL << (hwc->idx + 32)); else if (event->hw.flags & PERF_X86_EVENT_PEBS_ST) cpuc->pebs_enabled &= ~(1ULL << 63); diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index aec00e9f263e6..ad7290d51a3b1 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -974,6 +974,8 @@ extern struct event_constraint intel_bdw_pebs_event_constraints[];

extern struct event_constraint intel_skl_pebs_event_constraints[];

+extern struct event_constraint intel_icl_pebs_event_constraints[]; + struct event_constraint *intel_pebs_constraints(struct perf_event *event);

void intel_pmu_pebs_add(struct perf_event *event); diff --git a/arch/x86/include/asm/intel_ds.h b/arch/x86/include/asm/intel_ds.h index ae26df1c27896..8380c3ddd4b2e 100644 --- a/arch/x86/include/asm/intel_ds.h +++ b/arch/x86/include/asm/intel_ds.h @@ -8,7 +8,7 @@

/* The maximal number of PEBS events: */ #define MAX_PEBS_EVENTS 8 -#define MAX_FIXED_PEBS_EVENTS 3 +#define MAX_FIXED_PEBS_EVENTS 4

/* * A debug store configuration. diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h index 620c3e3564bef..7a5ca63a3cca6 100644 --- a/arch/x86/include/asm/perf_event.h +++ b/arch/x86/include/asm/perf_event.h @@ -7,7 +7,7 @@ */

#define INTEL_PMC_MAX_GENERIC 32 -#define INTEL_PMC_MAX_FIXED 3 +#define INTEL_PMC_MAX_FIXED 4 #define INTEL_PMC_IDX_FIXED 32

#define X86_PMC_IDX_MAX 64

From: Dave Jiang dave.jiang@intel.com

mainline inclusion from mainline-v5.1-rc1 commit 4d75873f814055359bb6722c4e35a185d02157a8 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 4d75873f814055359bb6722c4e35a185d02157a8 upsream.

Add Snowridge Xeon-D ioatdma PCI device id. Also applies for Icelake SP Xeon. This introduces ioatdma v3.4 platform. Also bumping driver version to 5.0 since we are adding additional code for 3.4 support.

Signed-off-by: Dave Jiang dave.jiang@intel.com Signed-off-by: Vinod Koul vkoul@kernel.org Signed-off-by: Lin Wang lin.x.wang@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/dma/ioat/dma.h | 2 +- drivers/dma/ioat/hw.h | 2 ++ drivers/dma/ioat/init.c | 3 +++ 3 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h index b061db2013324..a575e4ba1efef 100644 --- a/drivers/dma/ioat/dma.h +++ b/drivers/dma/ioat/dma.h @@ -27,7 +27,7 @@ #include "registers.h" #include "hw.h"

-#define IOAT_DMA_VERSION "4.00" +#define IOAT_DMA_VERSION "5.00"

#define IOAT_DMA_DCA_ANY_CPU ~0

diff --git a/drivers/dma/ioat/hw.h b/drivers/dma/ioat/hw.h index abcc51b343cec..3e38877cae6b9 100644 --- a/drivers/dma/ioat/hw.h +++ b/drivers/dma/ioat/hw.h @@ -66,6 +66,8 @@

#define PCI_DEVICE_ID_INTEL_IOAT_SKX 0x2021

+#define PCI_DEVICE_ID_INTEL_IOAT_ICX 0x0b00 + #define IOAT_VER_1_2 0x12 /* Version 1.2 */ #define IOAT_VER_2_0 0x20 /* Version 2.0 */ #define IOAT_VER_3_0 0x30 /* Version 3.0 */ diff --git a/drivers/dma/ioat/init.c b/drivers/dma/ioat/init.c index 0fec3c554fe35..ec5d0ac295d92 100644 --- a/drivers/dma/ioat/init.c +++ b/drivers/dma/ioat/init.c @@ -119,6 +119,9 @@ static const struct pci_device_id ioat_pci_tbl[] = { { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_BDXDE2) }, { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_BDXDE3) },

+ /* I/OAT v3.4 platforms */ + { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_ICX) }, + { 0, } }; MODULE_DEVICE_TABLE(pci, ioat_pci_tbl);

From: Dave Jiang dave.jiang@intel.com

mainline inclusion from mainline-v5.1-rc1 commit e0100d40906d5dbe6d09d31083c1a5aaccc947fa category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit e0100d40906d5dbe6d09d31083c1a5aaccc947fa upstream.

Adding support for new feature on ioatdma 3.4 hardware that provides descriptor pre-fetching in order to reduce small DMA latencies.

Signed-off-by: Dave Jiang dave.jiang@intel.com Signed-off-by: Vinod Koul vkoul@kernel.org Signed-off-by: Lin Wang lin.x.wang@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/dma/ioat/dma.c | 12 ++++++++++++ drivers/dma/ioat/init.c | 8 ++++++-- drivers/dma/ioat/registers.h | 10 ++++++++++ 3 files changed, 28 insertions(+), 2 deletions(-)

diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c index 890cadf3ec5d8..48d62d8e2c532 100644 --- a/drivers/dma/ioat/dma.c +++ b/drivers/dma/ioat/dma.c @@ -384,6 +384,7 @@ struct ioat_ring_ent ** ioat_alloc_ring(struct dma_chan *c, int order, gfp_t flags) { struct ioatdma_chan *ioat_chan = to_ioat_chan(c); + struct ioatdma_device *ioat_dma = ioat_chan->ioat_dma; struct ioat_ring_ent **ring; int total_descs = 1 << order; int i, chunks; @@ -450,6 +451,17 @@ ioat_alloc_ring(struct dma_chan *c, int order, gfp_t flags) } ring[i]->hw->next = ring[0]->txd.phys;

+ /* setup descriptor pre-fetching for v3.4 */ + if (ioat_dma->cap & IOAT_CAP_DPS) { + u16 drsctl = IOAT_CHAN_DRSZ_2MB | IOAT_CHAN_DRS_EN; + + if (chunks == 1) + drsctl |= IOAT_CHAN_DRS_AUTOWRAP; + + writew(drsctl, ioat_chan->reg_base + IOAT_CHAN_DRSCTL_OFFSET); + + } + return ring; }

diff --git a/drivers/dma/ioat/init.c b/drivers/dma/ioat/init.c index 9f9e9488fd185..aaa904c088024 100644 --- a/drivers/dma/ioat/init.c +++ b/drivers/dma/ioat/init.c @@ -138,10 +138,10 @@ static int ioat3_dma_self_test(struct ioatdma_device *ioat_dma); static int ioat_dca_enabled = 1; module_param(ioat_dca_enabled, int, 0644); MODULE_PARM_DESC(ioat_dca_enabled, "control support of dca service (default: 1)"); -int ioat_pending_level = 4; +int ioat_pending_level = 7; module_param(ioat_pending_level, int, 0644); MODULE_PARM_DESC(ioat_pending_level, - "high-water mark for pushing ioat descriptors (default: 4)"); + "high-water mark for pushing ioat descriptors (default: 7)"); static char ioat_interrupt_style[32] = "msix"; module_param_string(ioat_interrupt_style, ioat_interrupt_style, sizeof(ioat_interrupt_style), 0644); @@ -1188,6 +1188,10 @@ static int ioat3_dma_probe(struct ioatdma_device *ioat_dma, int dca) if (err) return err;

+ if (ioat_dma->cap & IOAT_CAP_DPS) + writeb(ioat_pending_level + 1, + ioat_dma->reg_base + IOAT_PREFETCH_LIMIT_OFFSET); + return 0; }

diff --git a/drivers/dma/ioat/registers.h b/drivers/dma/ioat/registers.h index 2f3bbc88ff2a0..2b517d6db5fd3 100644 --- a/drivers/dma/ioat/registers.h +++ b/drivers/dma/ioat/registers.h @@ -84,6 +84,9 @@ #define IOAT_CAP_PQ 0x00000200 #define IOAT_CAP_DWBES 0x00002000 #define IOAT_CAP_RAID16SS 0x00020000 +#define IOAT_CAP_DPS 0x00800000 + +#define IOAT_PREFETCH_LIMIT_OFFSET 0x4C /* CHWPREFLMT */

#define IOAT_CHANNEL_MMIO_SIZE 0x80 /* Each Channel MMIO space is this size */

@@ -243,4 +246,11 @@

#define IOAT_CHANERR_MASK_OFFSET 0x2C /* 32-bit Channel Error Register */

+#define IOAT_CHAN_DRSCTL_OFFSET 0xB6 +#define IOAT_CHAN_DRSZ_4KB 0x0000 +#define IOAT_CHAN_DRSZ_8KB 0x0001 +#define IOAT_CHAN_DRSZ_2MB 0x0009 +#define IOAT_CHAN_DRS_EN 0x0100 +#define IOAT_CHAN_DRS_AUTOWRAP 0x0200 + #endif /* _IOAT_REGISTERS_H_ */

From: Dan Williams dan.j.williams@intel.com

mainline inclusion from mainline-v5.1-rc1 commit 21b9e979501fdb5f6797193d70428a2b00bd5247 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 21b9e979501fdb5f6797193d70428a2b00bd5247 upstream.

Commit bbb3be170ac2 "device-dax: fix sysfs duplicate warnings" arranged for passing a dax instance-id to devm_create_dax_dev(), rather than generating one internally. Remove the dax_region ida and related code.

Signed-off-by: Dan Williams dan.j.williams@intel.com Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/dax/dax-private.h | 3 --- drivers/dax/device.c | 24 +++--------------------- 2 files changed, 3 insertions(+), 24 deletions(-)

diff --git a/drivers/dax/dax-private.h b/drivers/dax/dax-private.h index b6fc4f04636de..d1b36a42132f5 100644 --- a/drivers/dax/dax-private.h +++ b/drivers/dax/dax-private.h @@ -28,7 +28,6 @@ */ struct dax_region { int id; - struct ida ida; void *base; struct kref kref; struct device *dev; @@ -42,7 +41,6 @@ struct dax_region { * @region - parent region * @dax_dev - core dax functionality * @dev - device core - * @id - child id in the region * @num_resources - number of physical address extents in this device * @res - array of physical address ranges */ @@ -50,7 +48,6 @@ struct dev_dax { struct dax_region *region; struct dax_device *dax_dev; struct device dev; - int id; int num_resources; struct resource res[0]; }; diff --git a/drivers/dax/device.c b/drivers/dax/device.c index a89ebd94c6702..f8f6ee0b2c1b7 100644 --- a/drivers/dax/device.c +++ b/drivers/dax/device.c @@ -128,7 +128,6 @@ struct dax_region *alloc_dax_region(struct device *parent, int region_id, dax_region->pfn_flags = pfn_flags; kref_init(&dax_region->kref); dax_region->id = region_id; - ida_init(&dax_region->ida); dax_region->align = align; dax_region->dev = parent; dax_region->base = addr; @@ -580,8 +579,6 @@ static void dev_dax_release(struct device *dev) struct dax_region *dax_region = dev_dax->region; struct dax_device *dax_dev = dev_dax->dax_dev;

- if (dev_dax->id >= 0) - ida_simple_remove(&dax_region->ida, dev_dax->id); dax_region_put(dax_region); put_dax(dax_dev); kfree(dev_dax); @@ -640,19 +637,7 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, }

if (i < count) - goto err_id; - - if (id < 0) { - id = ida_simple_get(&dax_region->ida, 0, 0, GFP_KERNEL); - dev_dax->id = id; - if (id < 0) { - rc = id; - goto err_id; - } - } else { - /* region provider owns @id lifetime */ - dev_dax->id = -1; - } + goto err;

/* * No 'host' or dax_operations since there is no access to this @@ -661,7 +646,7 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, dax_dev = alloc_dax(dev_dax, NULL, NULL); if (!dax_dev) { rc = -ENOMEM; - goto err_dax; + goto err; }

/* from here on we're committed to teardown via dax_dev_release() */ @@ -698,10 +683,7 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region,

return dev_dax;

- err_dax: - if (dev_dax->id >= 0) - ida_simple_remove(&dax_region->ida, dev_dax->id); - err_id: + err: kfree(dev_dax);

return ERR_PTR(rc);

From: Dan Williams dan.j.williams@intel.com

mainline inclusion from mainline-v5.1-rc1 commit 753a0850e707e9a8c5861356222f9b9e4eba7945 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 753a0850e707e9a8c5861356222f9b9e4eba7945 upstream.

The multi-resource implementation anticipated discontiguous sub-division support. That has not yet materialized, delete the infrastructure and related code.

Signed-off-by: Dan Williams dan.j.williams@intel.com Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/dax/dax-private.h | 4 --- drivers/dax/device-dax.h | 3 +-- drivers/dax/device.c | 49 ++++++---------------------------- drivers/dax/pmem.c | 3 +-- tools/testing/nvdimm/dax-dev.c | 16 +++-------- 5 files changed, 13 insertions(+), 62 deletions(-)

diff --git a/drivers/dax/dax-private.h b/drivers/dax/dax-private.h index 9b393c218fe4d..dbd077653b5c9 100644 --- a/drivers/dax/dax-private.h +++ b/drivers/dax/dax-private.h @@ -39,14 +39,10 @@ struct dax_region { * @region - parent region * @dax_dev - core dax functionality * @dev - device core - * @num_resources - number of physical address extents in this device - * @res - array of physical address ranges */ struct dev_dax { struct dax_region *region; struct dax_device *dax_dev; struct device dev; - int num_resources; - struct resource res[0]; }; #endif diff --git a/drivers/dax/device-dax.h b/drivers/dax/device-dax.h index 4f1c69e1b3a24..e9be99584b927 100644 --- a/drivers/dax/device-dax.h +++ b/drivers/dax/device-dax.h @@ -19,6 +19,5 @@ struct dax_region; void dax_region_put(struct dax_region *dax_region); struct dax_region *alloc_dax_region(struct device *parent, int region_id, struct resource *res, unsigned int align, unsigned long flags); -struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, - int id, struct resource *res, int count); +struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, int id); #endif /* __DEVICE_DAX_H__ */ diff --git a/drivers/dax/device.c b/drivers/dax/device.c index 59518cefd0b2d..d3706a8f0ea5f 100644 --- a/drivers/dax/device.c +++ b/drivers/dax/device.c @@ -151,11 +151,7 @@ static ssize_t size_show(struct device *dev, struct device_attribute *attr, char *buf) { struct dev_dax *dev_dax = to_dev_dax(dev); - unsigned long long size = 0; - int i; - - for (i = 0; i < dev_dax->num_resources; i++) - size += resource_size(&dev_dax->res[i]); + unsigned long long size = resource_size(&dev_dax->region->res);

return sprintf(buf, "%llu\n", size); } @@ -224,21 +220,11 @@ static int check_vma(struct dev_dax *dev_dax, struct vm_area_struct *vma, __weak phys_addr_t dax_pgoff_to_phys(struct dev_dax *dev_dax, pgoff_t pgoff, unsigned long size) { - struct resource *res; - /* gcc-4.6.3-nolibc for i386 complains that this is uninitialized */ - phys_addr_t uninitialized_var(phys); - int i; - - for (i = 0; i < dev_dax->num_resources; i++) { - res = &dev_dax->res[i]; - phys = pgoff * PAGE_SIZE + res->start; - if (phys >= res->start && phys <= res->end) - break; - pgoff -= PHYS_PFN(resource_size(res)); - } + struct resource *res = &dev_dax->region->res; + phys_addr_t phys;

- if (i < dev_dax->num_resources) { - res = &dev_dax->res[i]; + phys = pgoff * PAGE_SIZE + res->start; + if (phys >= res->start && phys <= res->end) { if (phys + size - 1 <= res->end) return phys; } @@ -606,8 +592,7 @@ static void unregister_dev_dax(void *dev) put_device(dev); }

-struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, - int id, struct resource *res, int count) +struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, int id) { struct device *parent = dax_region->dev; struct dax_device *dax_dev; @@ -615,29 +600,12 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, struct inode *inode; struct device *dev; struct cdev *cdev; - int rc, i; + int rc;

- if (!count) - return ERR_PTR(-EINVAL); - - dev_dax = kzalloc(struct_size(dev_dax, res, count), GFP_KERNEL); + dev_dax = kzalloc(sizeof(*dev_dax), GFP_KERNEL); if (!dev_dax) return ERR_PTR(-ENOMEM);

- for (i = 0; i < count; i++) { - if (!IS_ALIGNED(res[i].start, dax_region->align) - || !IS_ALIGNED(resource_size(&res[i]), - dax_region->align)) { - rc = -EINVAL; - break; - } - dev_dax->res[i].start = res[i].start; - dev_dax->res[i].end = res[i].end; - } - - if (i < count) - goto err; - /* * No 'host' or dax_operations since there is no access to this * device outside of mmap of the resulting character device. @@ -657,7 +625,6 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, cdev_init(cdev, &dax_fops); cdev->owner = parent->driver->owner;

- dev_dax->num_resources = count; dev_dax->dax_dev = dax_dev; dev_dax->region = dax_region; kref_get(&dax_region->kref); diff --git a/drivers/dax/pmem.c b/drivers/dax/pmem.c index 72a76105eb024..6c61a019f9971 100644 --- a/drivers/dax/pmem.c +++ b/drivers/dax/pmem.c @@ -129,8 +129,7 @@ static int dax_pmem_probe(struct device *dev) if (!dax_region) return -ENOMEM;

- /* TODO: support for subdividing a dax region... */ - dev_dax = devm_create_dev_dax(dax_region, id, &res, 1); + dev_dax = devm_create_dev_dax(dax_region, id);

/* child dev_dax instances now own the lifetime of the dax_region */ dax_region_put(dax_region); diff --git a/tools/testing/nvdimm/dax-dev.c b/tools/testing/nvdimm/dax-dev.c index 36ee3d8797c3b..f36e708265b85 100644 --- a/tools/testing/nvdimm/dax-dev.c +++ b/tools/testing/nvdimm/dax-dev.c @@ -17,20 +17,11 @@ phys_addr_t dax_pgoff_to_phys(struct dev_dax *dev_dax, pgoff_t pgoff, unsigned long size) { - struct resource *res; + struct resource *res = &dev_dax->region->res; phys_addr_t addr; - int i;

- for (i = 0; i < dev_dax->num_resources; i++) { - res = &dev_dax->res[i]; - addr = pgoff * PAGE_SIZE + res->start; - if (addr >= res->start && addr <= res->end) - break; - pgoff -= PHYS_PFN(resource_size(res)); - } - - if (i < dev_dax->num_resources) { - res = &dev_dax->res[i]; + addr = pgoff * PAGE_SIZE + res->start; + if (addr >= res->start && addr <= res->end) { if (addr + size - 1 <= res->end) { if (get_nfit_res(addr)) { struct page *page; @@ -44,6 +35,5 @@ phys_addr_t dax_pgoff_to_phys(struct dev_dax *dev_dax, pgoff_t pgoff, return addr; } } - return -1; }

From: Dan Williams dan.j.williams@intel.com

mainline inclusion from mainline-v5.1-rc1 commit 9567da0b408a2553d32ca83cba4f1fc5a8aad459 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 9567da0b408a2553d32ca83cba4f1fc5a8aad459 upstream.

In support of multiple device-dax instances per device-dax-region and allowing the 'kmem' driver to attach to dax-instances instead of the current device-node access, convert the dax sub-system from a class to a bus. Recall that the kmem driver takes reserved / special purpose memories and assigns them to be managed by the core-mm.

Aside from the fact the device-dax instances are registered and probed on a bus, two other lifetime-management changes are made:

1/ Delay attaching a cdev until driver probe time

2/ A new run_dax() helper is introduced to allow restoring dax-operation after a kill_dax() event. So, at driver ->probe() time we run_dax() and at ->remove() time we kill_dax() and invalidate all mappings.

Signed-off-by: Dan Williams dan.j.williams@intel.com Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/dax/bus.c | 133 +++++++++++++++++++++++++++++++++++--- drivers/dax/bus.h | 16 +++++ drivers/dax/dax-private.h | 6 +- drivers/dax/device.c | 95 +++++++++------------------ drivers/dax/super.c | 40 ++++++++---- 5 files changed, 203 insertions(+), 87 deletions(-)

diff --git a/drivers/dax/bus.c b/drivers/dax/bus.c index 8a398e8e1956f..0cff32102c4c8 100644 --- a/drivers/dax/bus.c +++ b/drivers/dax/bus.c @@ -6,6 +6,33 @@ #include "dax-private.h" #include "bus.h"

+static int dax_bus_uevent(struct device *dev, struct kobj_uevent_env *env) +{ + /* + * We only ever expect to handle device-dax instances, i.e. the + * @type argument to MODULE_ALIAS_DAX_DEVICE() is always zero + */ + return add_uevent_var(env, "MODALIAS=" DAX_DEVICE_MODALIAS_FMT, 0); +} + +static int dax_bus_match(struct device *dev, struct device_driver *drv); + +static struct bus_type dax_bus_type = { + .name = "dax", + .uevent = dax_bus_uevent, + .match = dax_bus_match, +}; + +static int dax_bus_match(struct device *dev, struct device_driver *drv) +{ + /* + * The drivers that can register on the 'dax' bus are private to + * drivers/dax/ so any device and driver on the bus always + * match. + */ + return 1; +} + /* * Rely on the fact that drvdata is set before the attributes are * registered, and that the attributes are unregistered before drvdata @@ -142,11 +169,10 @@ static const struct attribute_group dev_dax_attribute_group = { .attrs = dev_dax_attributes, };

-const struct attribute_group *dax_attribute_groups[] = { +static const struct attribute_group *dax_attribute_groups[] = { &dev_dax_attribute_group, NULL, }; -EXPORT_SYMBOL_GPL(dax_attribute_groups);

void kill_dev_dax(struct dev_dax *dev_dax) { @@ -158,17 +184,108 @@ void kill_dev_dax(struct dev_dax *dev_dax) } EXPORT_SYMBOL_GPL(kill_dev_dax);

-void unregister_dev_dax(void *dev) +static void dev_dax_release(struct device *dev) { struct dev_dax *dev_dax = to_dev_dax(dev); + struct dax_region *dax_region = dev_dax->region; struct dax_device *dax_dev = dev_dax->dax_dev; - struct inode *inode = dax_inode(dax_dev); - struct cdev *cdev = inode->i_cdev;

- dev_dbg(dev, "trace\n"); + dax_region_put(dax_region); + put_dax(dax_dev); + kfree(dev_dax); +} + +static void unregister_dev_dax(void *dev) +{ + struct dev_dax *dev_dax = to_dev_dax(dev); + + dev_dbg(dev, "%s\n", __func__);

kill_dev_dax(dev_dax); - cdev_device_del(cdev, dev); + device_del(dev); put_device(dev); } -EXPORT_SYMBOL_GPL(unregister_dev_dax); + +struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, int id) +{ + struct device *parent = dax_region->dev; + struct dax_device *dax_dev; + struct dev_dax *dev_dax; + struct inode *inode; + struct device *dev; + int rc = -ENOMEM; + + if (id < 0) + return ERR_PTR(-EINVAL); + + dev_dax = kzalloc(sizeof(*dev_dax), GFP_KERNEL); + if (!dev_dax) + return ERR_PTR(-ENOMEM); + + /* + * No 'host' or dax_operations since there is no access to this + * device outside of mmap of the resulting character device. + */ + dax_dev = alloc_dax(dev_dax, NULL, NULL); + if (!dax_dev) + goto err; + + /* a device_dax instance is dead while the driver is not attached */ + kill_dax(dax_dev); + + /* from here on we're committed to teardown via dax_dev_release() */ + dev = &dev_dax->dev; + device_initialize(dev); + + dev_dax->dax_dev = dax_dev; + dev_dax->region = dax_region; + kref_get(&dax_region->kref); + + inode = dax_inode(dax_dev); + dev->devt = inode->i_rdev; + dev->bus = &dax_bus_type; + dev->parent = parent; + dev->groups = dax_attribute_groups; + dev->release = dev_dax_release; + dev_set_name(dev, "dax%d.%d", dax_region->id, id); + + rc = device_add(dev); + if (rc) { + kill_dev_dax(dev_dax); + put_device(dev); + return ERR_PTR(rc); + } + + rc = devm_add_action_or_reset(dax_region->dev, unregister_dev_dax, dev); + if (rc) + return ERR_PTR(rc); + + return dev_dax; + + err: + kfree(dev_dax); + + return ERR_PTR(rc); +} +EXPORT_SYMBOL_GPL(devm_create_dev_dax); + +int __dax_driver_register(struct device_driver *drv, + struct module *module, const char *mod_name) +{ + drv->owner = module; + drv->name = mod_name; + drv->mod_name = mod_name; + drv->bus = &dax_bus_type; + return driver_register(drv); +} +EXPORT_SYMBOL_GPL(__dax_driver_register); + +int __init dax_bus_init(void) +{ + return bus_register(&dax_bus_type); +} + +void __exit dax_bus_exit(void) +{ + bus_unregister(&dax_bus_type); +} diff --git a/drivers/dax/bus.h b/drivers/dax/bus.h index 840865aa69e81..ea509504df3aa 100644 --- a/drivers/dax/bus.h +++ b/drivers/dax/bus.h @@ -11,5 +11,21 @@ void dax_region_put(struct dax_region *dax_region); struct dax_region *alloc_dax_region(struct device *parent, int region_id, struct resource *res, unsigned int align, unsigned long flags); struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, int id); +int __dax_driver_register(struct device_driver *drv, + struct module *module, const char *mod_name); +#define dax_driver_register(driver) \ + __dax_driver_register(driver, THIS_MODULE, KBUILD_MODNAME) void kill_dev_dax(struct dev_dax *dev_dax); + +/* + * While run_dax() is potentially a generic operation that could be + * defined in include/linux/dax.h we don't want to grow any users + * outside of drivers/dax/ + */ +void run_dax(struct dax_device *dax_dev); + +#define MODULE_ALIAS_DAX_DEVICE(type) \ + MODULE_ALIAS("dax:t" __stringify(type) "*") +#define DAX_DEVICE_MODALIAS_FMT "dax:t%d" + #endif /* __DAX_BUS_H__ */ diff --git a/drivers/dax/dax-private.h b/drivers/dax/dax-private.h index 620c3f4eefe7d..c3a1217008373 100644 --- a/drivers/dax/dax-private.h +++ b/drivers/dax/dax-private.h @@ -20,10 +20,8 @@ struct dax_device; struct dax_device *inode_dax(struct inode *inode); struct inode *dax_inode(struct dax_device *dax_dev); - -/* temporary until devm_create_dax_dev moves to bus.c */ -extern const struct attribute_group *dax_attribute_groups[]; -void unregister_dev_dax(void *dev); +int dax_bus_init(void); +void dax_bus_exit(void);

/** * struct dax_region - mapping infrastructure for dax devices diff --git a/drivers/dax/device.c b/drivers/dax/device.c index f2ce8af2e6cc2..20759c3f0b918 100644 --- a/drivers/dax/device.c +++ b/drivers/dax/device.c @@ -13,8 +13,6 @@ #include "dax-private.h" #include "bus.h"

-static struct class *dax_class; - static int check_vma(struct dev_dax *dev_dax, struct vm_area_struct *vma, const char *func) { @@ -402,93 +400,64 @@ static const struct file_operations dax_fops = { .mmap_supported_flags = MAP_SYNC, };

-static void dev_dax_release(struct device *dev) +static void dev_dax_cdev_del(void *cdev) { - struct dev_dax *dev_dax = to_dev_dax(dev); - struct dax_region *dax_region = dev_dax->region; - struct dax_device *dax_dev = dev_dax->dax_dev; + cdev_del(cdev); +}

- dax_region_put(dax_region); - put_dax(dax_dev); - kfree(dev_dax); +static void dev_dax_kill(void *dev_dax) +{ + kill_dev_dax(dev_dax); }

-struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, int id) +static int dev_dax_probe(struct device *dev) { - struct device *parent = dax_region->dev; - struct dax_device *dax_dev; - struct dev_dax *dev_dax; + struct dev_dax *dev_dax = to_dev_dax(dev); + struct dax_device *dax_dev = dev_dax->dax_dev; struct inode *inode; - struct device *dev; struct cdev *cdev; int rc;

- dev_dax = kzalloc(sizeof(*dev_dax), GFP_KERNEL); - if (!dev_dax) - return ERR_PTR(-ENOMEM); - - /* - * No 'host' or dax_operations since there is no access to this - * device outside of mmap of the resulting character device. - */ - dax_dev = alloc_dax(dev_dax, NULL, NULL); - if (!dax_dev) { - rc = -ENOMEM; - goto err; - } - - /* from here on we're committed to teardown via dax_dev_release() */ - dev = &dev_dax->dev; - device_initialize(dev); - inode = dax_inode(dax_dev); cdev = inode->i_cdev; cdev_init(cdev, &dax_fops); - cdev->owner = parent->driver->owner; - - dev_dax->dax_dev = dax_dev; - dev_dax->region = dax_region; - kref_get(&dax_region->kref); - - dev->devt = inode->i_rdev; - dev->class = dax_class; - dev->parent = parent; - dev->groups = dax_attribute_groups; - dev->release = dev_dax_release; - dev_set_name(dev, "dax%d.%d", dax_region->id, id); - - rc = cdev_device_add(cdev, dev); - if (rc) { - kill_dev_dax(dev_dax); - put_device(dev); - return ERR_PTR(rc); - } - - rc = devm_add_action_or_reset(dax_region->dev, unregister_dev_dax, dev); + cdev->owner = dev->driver->owner; + cdev_set_parent(cdev, &dev->kobj); + rc = cdev_add(cdev, dev->devt, 1); if (rc) - return ERR_PTR(rc); + return rc;

- return dev_dax; + rc = devm_add_action_or_reset(dev, dev_dax_cdev_del, cdev); + if (rc) + return rc;

- err: - kfree(dev_dax); + run_dax(dax_dev); + return devm_add_action_or_reset(dev, dev_dax_kill, dev_dax); +}

- return ERR_PTR(rc); +static int dev_dax_remove(struct device *dev) +{ + /* all probe actions are unwound by devm */ + return 0; } -EXPORT_SYMBOL_GPL(devm_create_dev_dax); + +static struct device_driver device_dax_driver = { + .probe = dev_dax_probe, + .remove = dev_dax_remove, +};

static int __init dax_init(void) { - dax_class = class_create(THIS_MODULE, "dax"); - return PTR_ERR_OR_ZERO(dax_class); + return dax_driver_register(&device_dax_driver); }

static void __exit dax_exit(void) { - class_destroy(dax_class); + driver_unregister(&device_dax_driver); }

MODULE_AUTHOR("Intel Corporation"); MODULE_LICENSE("GPL v2"); -subsys_initcall(dax_init); +module_init(dax_init); module_exit(dax_exit); +MODULE_ALIAS_DAX_DEVICE(0); diff --git a/drivers/dax/super.c b/drivers/dax/super.c index 0ecc1a2cf1ccd..ccb22d8db3a26 100644 --- a/drivers/dax/super.c +++ b/drivers/dax/super.c @@ -366,11 +366,15 @@ void kill_dax(struct dax_device *dax_dev) spin_lock(&dax_host_lock); hlist_del_init(&dax_dev->list); spin_unlock(&dax_host_lock); - - dax_dev->private = NULL; } EXPORT_SYMBOL_GPL(kill_dax);

+void run_dax(struct dax_device *dax_dev) +{ + set_bit(DAXDEV_ALIVE, &dax_dev->flags); +} +EXPORT_SYMBOL_GPL(run_dax); + static struct inode *dax_alloc_inode(struct super_block *sb) { struct dax_device *dax_dev; @@ -585,6 +589,8 @@ EXPORT_SYMBOL_GPL(dax_inode);

void *dax_get_private(struct dax_device *dax_dev) { + if (!test_bit(DAXDEV_ALIVE, &dax_dev->flags)) + return NULL; return dax_dev->private; } EXPORT_SYMBOL_GPL(dax_get_private); @@ -598,7 +604,7 @@ static void init_once(void *_dax_dev) inode_init_once(inode); }

-static int __dax_fs_init(void) +static int dax_fs_init(void) { int rc;

@@ -630,35 +636,45 @@ static int __dax_fs_init(void) return rc; }

-static void __dax_fs_exit(void) +static void dax_fs_exit(void) { kern_unmount(dax_mnt); unregister_filesystem(&dax_fs_type); kmem_cache_destroy(dax_cache); }

-static int __init dax_fs_init(void) +static int __init dax_core_init(void) { int rc;

- rc = __dax_fs_init(); + rc = dax_fs_init(); if (rc) return rc;

rc = alloc_chrdev_region(&dax_devt, 0, MINORMASK+1, "dax"); if (rc) - __dax_fs_exit(); - return rc; + goto err_chrdev; + + rc = dax_bus_init(); + if (rc) + goto err_bus; + return 0; + +err_bus: + unregister_chrdev_region(dax_devt, MINORMASK+1); +err_chrdev: + dax_fs_exit(); + return 0; }

-static void __exit dax_fs_exit(void) +static void __exit dax_core_exit(void) { unregister_chrdev_region(dax_devt, MINORMASK+1); ida_destroy(&dax_minor_ida); - __dax_fs_exit(); + dax_fs_exit(); }

MODULE_AUTHOR("Intel Corporation"); MODULE_LICENSE("GPL v2"); -subsys_initcall(dax_fs_init); -module_exit(dax_fs_exit); +subsys_initcall(dax_core_init); +module_exit(dax_core_exit);

From: Dan Williams dan.j.williams@intel.com

mainline inclusion from mainline-v5.1-rc1 commit d200781ef237a354d918ceff5cee350d88a93d42 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit d200781ef237a354d918ceff5cee350d88a93d42 upstream.

Introduce the 'new_id' concept for enabling a custom device-driver attach policy for dax-bus drivers. The intended use is to have a mechanism for hot-plugging device-dax ranges into the page allocator on-demand. With this in place the default policy of using device-dax for performance differentiated memory can be overridden by user-space policy that can arrange for the memory range to be managed as 'System RAM' with user-defined NUMA and other performance attributes.

Signed-off-by: Dan Williams dan.j.williams@intel.com Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/dax/bus.c | 145 +++++++++++++++++++++++++++++++++++++++++-- drivers/dax/bus.h | 10 ++- drivers/dax/device.c | 11 ++-- 3 files changed, 156 insertions(+), 10 deletions(-)

diff --git a/drivers/dax/bus.c b/drivers/dax/bus.c index 69aae2cbd45f1..17af6fbc3be5b 100644 --- a/drivers/dax/bus.c +++ b/drivers/dax/bus.c @@ -2,11 +2,21 @@ /* Copyright(c) 2017-2018 Intel Corporation. All rights reserved. */ #include <linux/memremap.h> #include <linux/device.h> +#include <linux/mutex.h> +#include <linux/list.h> #include <linux/slab.h> #include <linux/dax.h> #include "dax-private.h" #include "bus.h"

+static DEFINE_MUTEX(dax_bus_lock); + +#define DAX_NAME_LEN 30 +struct dax_id { + struct list_head list; + char dev_name[DAX_NAME_LEN]; +}; + static int dax_bus_uevent(struct device *dev, struct kobj_uevent_env *env) { /* @@ -16,22 +26,115 @@ static int dax_bus_uevent(struct device *dev, struct kobj_uevent_env *env) return add_uevent_var(env, "MODALIAS=" DAX_DEVICE_MODALIAS_FMT, 0); }

+static struct dax_device_driver *to_dax_drv(struct device_driver *drv) +{ + return container_of(drv, struct dax_device_driver, drv); +} + +static struct dax_id *__dax_match_id(struct dax_device_driver *dax_drv, + const char *dev_name) +{ + struct dax_id *dax_id; + + lockdep_assert_held(&dax_bus_lock); + + list_for_each_entry(dax_id, &dax_drv->ids, list) + if (sysfs_streq(dax_id->dev_name, dev_name)) + return dax_id; + return NULL; +} + +static int dax_match_id(struct dax_device_driver *dax_drv, struct device *dev) +{ + int match; + + mutex_lock(&dax_bus_lock); + match = !!__dax_match_id(dax_drv, dev_name(dev)); + mutex_unlock(&dax_bus_lock); + + return match; +} + +static ssize_t do_id_store(struct device_driver *drv, const char *buf, + size_t count, bool add) +{ + struct dax_device_driver *dax_drv = to_dax_drv(drv); + unsigned int region_id, id; + char devname[DAX_NAME_LEN]; + struct dax_id *dax_id; + ssize_t rc = count; + int fields; + + fields = sscanf(buf, "dax%d.%d", &region_id, &id); + if (fields != 2) + return -EINVAL; + sprintf(devname, "dax%d.%d", region_id, id); + if (!sysfs_streq(buf, devname)) + return -EINVAL; + + mutex_lock(&dax_bus_lock); + dax_id = __dax_match_id(dax_drv, buf); + if (!dax_id) { + if (add) { + dax_id = kzalloc(sizeof(*dax_id), GFP_KERNEL); + if (dax_id) { + strncpy(dax_id->dev_name, buf, DAX_NAME_LEN); + list_add(&dax_id->list, &dax_drv->ids); + } else + rc = -ENOMEM; + } else + /* nothing to remove */; + } else if (!add) { + list_del(&dax_id->list); + kfree(dax_id); + } else + /* dax_id already added */; + mutex_unlock(&dax_bus_lock); + return rc; +} + +static ssize_t new_id_store(struct device_driver *drv, const char *buf, + size_t count) +{ + return do_id_store(drv, buf, count, true); +} +static DRIVER_ATTR_WO(new_id); + +static ssize_t remove_id_store(struct device_driver *drv, const char *buf, + size_t count) +{ + return do_id_store(drv, buf, count, false); +} +static DRIVER_ATTR_WO(remove_id); + +static struct attribute *dax_drv_attrs[] = { + &driver_attr_new_id.attr, + &driver_attr_remove_id.attr, + NULL, +}; +ATTRIBUTE_GROUPS(dax_drv); + static int dax_bus_match(struct device *dev, struct device_driver *drv);

static struct bus_type dax_bus_type = { .name = "dax", .uevent = dax_bus_uevent, .match = dax_bus_match, + .drv_groups = dax_drv_groups, };

static int dax_bus_match(struct device *dev, struct device_driver *drv) { + struct dax_device_driver *dax_drv = to_dax_drv(drv); + /* - * The drivers that can register on the 'dax' bus are private to - * drivers/dax/ so any device and driver on the bus always - * match. + * All but the 'device-dax' driver, which has 'match_always' + * set, requires an exact id match. */ - return 1; + if (dax_drv->match_always) + return 1; + + return dax_match_id(dax_drv, dev); }

/* @@ -273,17 +376,49 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, int id, } EXPORT_SYMBOL_GPL(devm_create_dev_dax);

-int __dax_driver_register(struct device_driver *drv, +static int match_always_count; + +int __dax_driver_register(struct dax_device_driver *dax_drv, struct module *module, const char *mod_name) { + struct device_driver *drv = &dax_drv->drv; + int rc = 0; + + INIT_LIST_HEAD(&dax_drv->ids); drv->owner = module; drv->name = mod_name; drv->mod_name = mod_name; drv->bus = &dax_bus_type; + + /* there can only be one default driver */ + mutex_lock(&dax_bus_lock); + match_always_count += dax_drv->match_always; + if (match_always_count > 1) { + match_always_count--; + WARN_ON(1); + rc = -EINVAL; + } + mutex_unlock(&dax_bus_lock); + if (rc) + return rc; return driver_register(drv); } EXPORT_SYMBOL_GPL(__dax_driver_register);

+void dax_driver_unregister(struct dax_device_driver *dax_drv) +{ + struct dax_id *dax_id, *_id; + + mutex_lock(&dax_bus_lock); + match_always_count -= dax_drv->match_always; + list_for_each_entry_safe(dax_id, _id, &dax_drv->ids, list) { + list_del(&dax_id->list); + kfree(dax_id); + } + mutex_unlock(&dax_bus_lock); +} +EXPORT_SYMBOL_GPL(dax_driver_unregister); + int __init dax_bus_init(void) { return bus_register(&dax_bus_type); diff --git a/drivers/dax/bus.h b/drivers/dax/bus.h index e08e0c394983d..395ab812367c8 100644 --- a/drivers/dax/bus.h +++ b/drivers/dax/bus.h @@ -12,10 +12,18 @@ struct dax_region *alloc_dax_region(struct device *parent, int region_id, struct resource *res, unsigned int align, unsigned long flags); struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region, int id, struct dev_pagemap *pgmap); -int __dax_driver_register(struct device_driver *drv, + +struct dax_device_driver { + struct device_driver drv; + struct list_head ids; + int match_always; +}; + +int __dax_driver_register(struct dax_device_driver *dax_drv, struct module *module, const char *mod_name); #define dax_driver_register(driver) \ __dax_driver_register(driver, THIS_MODULE, KBUILD_MODNAME) +void dax_driver_unregister(struct dax_device_driver *dax_drv); void kill_dev_dax(struct dev_dax *dev_dax);

/* diff --git a/drivers/dax/device.c b/drivers/dax/device.c index d164d39fb401f..7b64fd8280084 100644 --- a/drivers/dax/device.c +++ b/drivers/dax/device.c @@ -502,9 +502,12 @@ static int dev_dax_remove(struct device *dev) return 0; }

-static struct device_driver device_dax_driver = { - .probe = dev_dax_probe, - .remove = dev_dax_remove, +static struct dax_device_driver device_dax_driver = { + .drv = { + .probe = dev_dax_probe, + .remove = dev_dax_remove, + }, + .match_always = 1, };

static int __init dax_init(void) @@ -514,7 +517,7 @@ static int __init dax_init(void)

static void __exit dax_exit(void) { - driver_unregister(&device_dax_driver); + dax_driver_unregister(&device_dax_driver); }

MODULE_AUTHOR("Intel Corporation");

From: Dan Williams dan.j.williams@intel.com

mainline inclusion from mainline-v5.1-rc1 commit 8fc5c73554db0ac18c0c6ac5b2099ab917f83bdf category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 8fc5c73554db0ac18c0c6ac5b2099ab917f83bdf upstream.

Persistent memory, as described by the ACPI NFIT (NVDIMM Firmware Interface Table), is the first known instance of a memory range described by a unique "target" proximity domain. Where "initiator" and "target" proximity domains is an approach that the ACPI HMAT (Heterogeneous Memory Attributes Table) uses to described the unique performance properties of a memory range relative to a given initiator (e.g. CPU or DMA device).

Currently the numa-node for a /dev/pmemX block-device or /dev/daxX.Y char-device follows the traditional notion of 'numa-node' where the attribute conveys the closest online numa-node. That numa-node attribute is useful for cpu-binding and memory-binding processes *near* the device. However, when the memory range backing a 'pmem', or 'dax' device is onlined (memory hot-add) the memory-only-numa-node representing that address needs to be differentiated from the set of online nodes. In other words, the numa-node association of the device depends on whether you can bind processes *near* the cpu-numa-node in the offline device-case, or bind process *on* the memory-range directly after the backing address range is onlined.

Allow for the case that platform firmware describes persistent memory with a unique proximity domain, i.e. when it is distinct from the proximity of DRAM and CPUs that are on the same socket. Plumb the Linux numa-node translation of that proximity through the libnvdimm region device to namespaces that are in device-dax mode. With this in place the proposed kmem driver [1] can optionally discover a unique numa-node number for the address range as it transitions the memory from an offline state managed by a device-driver to an online memory range managed by the core-mm.

[1]: https://lore.kernel.org/lkml/20181022201317.8558C1D8@viggo.jf.intel.com

Reported-by: Fan Du fan.du@intel.com Cc: Michael Ellerman mpe@ellerman.id.au Cc: "Oliver O'Halloran" oohall@gmail.com Cc: Dave Hansen dave.hansen@linux.intel.com Cc: Jérôme Glisse jglisse@redhat.com Reviewed-by: Yang Shi yang.shi@linux.alibaba.com Signed-off-by: Dan Williams dan.j.williams@intel.com Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/acpi/nfit/core.c | 8 ++++++-- drivers/acpi/numa.c | 1 + drivers/dax/bus.c | 4 +++- drivers/dax/bus.h | 3 ++- drivers/dax/dax-private.h | 4 ++++ drivers/dax/pmem/core.c | 4 +++- drivers/nvdimm/e820.c | 1 + drivers/nvdimm/nd.h | 2 +- drivers/nvdimm/of_pmem.c | 1 + drivers/nvdimm/region_devs.c | 1 + include/linux/acpi.h | 5 +++++ include/linux/libnvdimm.h | 1 + 12 files changed, 29 insertions(+), 6 deletions(-)

diff --git a/drivers/acpi/nfit/core.c b/drivers/acpi/nfit/core.c index cb88f3b43a940..9fa7d19fe2bb6 100644 --- a/drivers/acpi/nfit/core.c +++ b/drivers/acpi/nfit/core.c @@ -2814,11 +2814,15 @@ static int acpi_nfit_register_region(struct acpi_nfit_desc *acpi_desc, ndr_desc->res = &res; ndr_desc->provider_data = nfit_spa; ndr_desc->attr_groups = acpi_nfit_region_attribute_groups; - if (spa->flags & ACPI_NFIT_PROXIMITY_VALID) + if (spa->flags & ACPI_NFIT_PROXIMITY_VALID) { ndr_desc->numa_node = acpi_map_pxm_to_online_node( spa->proximity_domain); - else + ndr_desc->target_node = acpi_map_pxm_to_node( + spa->proximity_domain); + } else { ndr_desc->numa_node = NUMA_NO_NODE; + ndr_desc->target_node = NUMA_NO_NODE; + }

/* * Persistence domain bits are hierarchical, if diff --git a/drivers/acpi/numa.c b/drivers/acpi/numa.c index a28ff3cfbc296..63d4f6b15e4db 100644 --- a/drivers/acpi/numa.c +++ b/drivers/acpi/numa.c @@ -85,6 +85,7 @@ int acpi_map_pxm_to_node(int pxm)

return node; } +EXPORT_SYMBOL(acpi_map_pxm_to_node);

/** * acpi_map_pxm_to_online_node - Map proximity ID to online node diff --git a/drivers/dax/bus.c b/drivers/dax/bus.c index 5681685002175..c620ad52d7e5c 100644 --- a/drivers/dax/bus.c +++ b/drivers/dax/bus.c @@ -214,7 +214,7 @@ static void dax_region_unregister(void *region) }

struct dax_region *alloc_dax_region(struct device *parent, int region_id, - struct resource *res, unsigned int align, + struct resource *res, int target_node, unsigned int align, unsigned long pfn_flags) { struct dax_region *dax_region; @@ -244,6 +244,7 @@ struct dax_region *alloc_dax_region(struct device *parent, int region_id, dax_region->id = region_id; dax_region->align = align; dax_region->dev = parent; + dax_region->target_node = target_node; if (sysfs_create_groups(&parent->kobj, dax_region_attribute_groups)) { kfree(dax_region); return NULL; @@ -348,6 +349,7 @@ struct dev_dax *__devm_create_dev_dax(struct dax_region *dax_region, int id,

dev_dax->dax_dev = dax_dev; dev_dax->region = dax_region; + dev_dax->target_node = dax_region->target_node; kref_get(&dax_region->kref);

inode = dax_inode(dax_dev); diff --git a/drivers/dax/bus.h b/drivers/dax/bus.h index ce977552ffb5c..8619e32999436 100644 --- a/drivers/dax/bus.h +++ b/drivers/dax/bus.h @@ -10,7 +10,8 @@ struct dax_device; struct dax_region; void dax_region_put(struct dax_region *dax_region); struct dax_region *alloc_dax_region(struct device *parent, int region_id, - struct resource *res, unsigned int align, unsigned long flags); + struct resource *res, int target_node, unsigned int align, + unsigned long flags);

enum dev_dax_subsys { DEV_DAX_BUS, diff --git a/drivers/dax/dax-private.h b/drivers/dax/dax-private.h index a82ce48f58844..a45612148ca02 100644 --- a/drivers/dax/dax-private.h +++ b/drivers/dax/dax-private.h @@ -26,6 +26,7 @@ void dax_bus_exit(void); /** * struct dax_region - mapping infrastructure for dax devices * @id: kernel-wide unique region for a memory range + * @target_node: effective numa node if this memory range is onlined * @kref: to pin while other agents have a need to do lookups * @dev: parent device backing this region * @align: allocation and mapping alignment for child dax devices @@ -34,6 +35,7 @@ void dax_bus_exit(void); */ struct dax_region { int id; + int target_node; struct kref kref; struct device *dev; unsigned int align; @@ -46,6 +48,7 @@ struct dax_region { * data while the device is activated in the driver. * @region - parent region * @dax_dev - core dax functionality + * @target_node: effective numa node if dev_dax memory range is onlined * @dev - device core * @pgmap - pgmap for memmap setup / lifetime (driver owned) * @ref: pgmap reference count (driver owned) @@ -54,6 +57,7 @@ struct dax_region { struct dev_dax { struct dax_region *region; struct dax_device *dax_dev; + int target_node; struct device dev; struct dev_pagemap pgmap; struct percpu_ref ref; diff --git a/drivers/dax/pmem/core.c b/drivers/dax/pmem/core.c index bdcff1b14e95c..f71019ce06470 100644 --- a/drivers/dax/pmem/core.c +++ b/drivers/dax/pmem/core.c @@ -20,6 +20,7 @@ struct dev_dax *__dax_pmem_probe(struct device *dev, enum dev_dax_subsys subsys) struct nd_namespace_common *ndns; struct nd_dax *nd_dax = to_nd_dax(dev); struct nd_pfn *nd_pfn = &nd_dax->nd_pfn; + struct nd_region *nd_region = to_nd_region(dev->parent);

ndns = nvdimm_namespace_common_probe(dev); if (IS_ERR(ndns)) @@ -52,7 +53,8 @@ struct dev_dax *__dax_pmem_probe(struct device *dev, enum dev_dax_subsys subsys) memcpy(&res, &pgmap.res, sizeof(res)); res.start += offset; dax_region = alloc_dax_region(dev, region_id, &res, - le32_to_cpu(pfn_sb->align), PFN_DEV|PFN_MAP); + nd_region->target_node, le32_to_cpu(pfn_sb->align), + PFN_DEV|PFN_MAP); if (!dax_region) return ERR_PTR(-ENOMEM);

diff --git a/drivers/nvdimm/e820.c b/drivers/nvdimm/e820.c index 521eaf53a52aa..36be9b6191876 100644 --- a/drivers/nvdimm/e820.c +++ b/drivers/nvdimm/e820.c @@ -47,6 +47,7 @@ static int e820_register_one(struct resource *res, void *data) ndr_desc.res = res; ndr_desc.attr_groups = e820_pmem_region_attribute_groups; ndr_desc.numa_node = e820_range_to_nid(res->start); + ndr_desc.target_node = ndr_desc.numa_node; set_bit(ND_REGION_PAGEMAP, &ndr_desc.flags); if (!nvdimm_pmem_region_create(nvdimm_bus, &ndr_desc)) return -ENXIO; diff --git a/drivers/nvdimm/nd.h b/drivers/nvdimm/nd.h index 01e194a5824e4..5259b8953cc64 100644 --- a/drivers/nvdimm/nd.h +++ b/drivers/nvdimm/nd.h @@ -157,7 +157,7 @@ struct nd_region { u16 ndr_mappings; u64 ndr_size; u64 ndr_start; - int id, num_lanes, ro, numa_node; + int id, num_lanes, ro, numa_node, target_node; void *provider_data; struct kernfs_node *bb_state; struct badblocks bb; diff --git a/drivers/nvdimm/of_pmem.c b/drivers/nvdimm/of_pmem.c index 0a701837dfc0b..ecaaa27438e25 100644 --- a/drivers/nvdimm/of_pmem.c +++ b/drivers/nvdimm/of_pmem.c @@ -68,6 +68,7 @@ static int of_pmem_region_probe(struct platform_device *pdev) memset(&ndr_desc, 0, sizeof(ndr_desc)); ndr_desc.attr_groups = region_attr_groups; ndr_desc.numa_node = dev_to_node(&pdev->dev); + ndr_desc.target_node = ndr_desc.numa_node; ndr_desc.res = &pdev->resource[i]; ndr_desc.of_node = np; set_bit(ND_REGION_PAGEMAP, &ndr_desc.flags); diff --git a/drivers/nvdimm/region_devs.c b/drivers/nvdimm/region_devs.c index 609fc450522a1..a5c80767d81b9 100644 --- a/drivers/nvdimm/region_devs.c +++ b/drivers/nvdimm/region_devs.c @@ -1064,6 +1064,7 @@ static struct nd_region *nd_region_create(struct nvdimm_bus *nvdimm_bus, nd_region->flags = ndr_desc->flags; nd_region->ro = ro; nd_region->numa_node = ndr_desc->numa_node; + nd_region->target_node = ndr_desc->target_node; ida_init(&nd_region->ns_ida); ida_init(&nd_region->btt_ida); ida_init(&nd_region->pfn_ida); diff --git a/include/linux/acpi.h b/include/linux/acpi.h index 4b9c0ec79b645..99b361e48d1b9 100644 --- a/include/linux/acpi.h +++ b/include/linux/acpi.h @@ -400,12 +400,17 @@ extern bool acpi_osi_is_win8(void);

#ifdef CONFIG_ACPI_NUMA int acpi_map_pxm_to_online_node(int pxm); +int acpi_map_pxm_to_node(int pxm); int acpi_get_node(acpi_handle handle); #else static inline int acpi_map_pxm_to_online_node(int pxm) { return 0; } +static inline int acpi_map_pxm_to_node(int pxm) +{ + return 0; +} static inline int acpi_get_node(acpi_handle handle) { return 0; diff --git a/include/linux/libnvdimm.h b/include/linux/libnvdimm.h index 097072c5a852f..941102c0c81fc 100644 --- a/include/linux/libnvdimm.h +++ b/include/linux/libnvdimm.h @@ -124,6 +124,7 @@ struct nd_region_desc { void *provider_data; int num_lanes; int numa_node; + int target_node; unsigned long flags; struct device_node *of_node; };

From: Dan Williams dan.j.williams@intel.com

mainline inclusion from mainline-v5.1-rc1 commit 21c75763a3ae18679e5c4e2260aa9379b073566b category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 21c75763a3ae18679e5c4e2260aa9379b073566b upstream.

The target-node attribute is the Linux numa-node that a device-dax instance may create when it is online. Prior to being online the device's 'numa_node' property reflects the closest online cpu node which is the typical expectation of a device 'numa_node'. Once it is online it becomes its own distinct numa node, i.e. 'target_node'.

Export the 'target_node' property to give userspace tooling the ability to predict the effective numa-node from a device-dax instance configured to provide 'System RAM' capacity.

Cc: Vishal Verma vishal.l.verma@intel.com Reported-by: Dave Hansen dave.hansen@linux.intel.com Signed-off-by: Dan Williams dan.j.williams@intel.com Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/dax/bus.c | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+)

diff --git a/drivers/dax/bus.c b/drivers/dax/bus.c index a410154d75fb9..28c3324271acd 100644 --- a/drivers/dax/bus.c +++ b/drivers/dax/bus.c @@ -279,13 +279,41 @@ static ssize_t size_show(struct device *dev, } static DEVICE_ATTR_RO(size);

+static int dev_dax_target_node(struct dev_dax *dev_dax) +{ + struct dax_region *dax_region = dev_dax->region; + + return dax_region->target_node; +} + +static ssize_t target_node_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct dev_dax *dev_dax = to_dev_dax(dev); + + return sprintf(buf, "%d\n", dev_dax_target_node(dev_dax)); +} +static DEVICE_ATTR_RO(target_node); + +static umode_t dev_dax_visible(struct kobject *kobj, struct attribute *a, int n) +{ + struct device *dev = container_of(kobj, struct device, kobj); + struct dev_dax *dev_dax = to_dev_dax(dev); + + if (a == &dev_attr_target_node.attr && dev_dax_target_node(dev_dax) < 0) + return 0; + return a->mode; +} + static struct attribute *dev_dax_attributes[] = { &dev_attr_size.attr, + &dev_attr_target_node.attr, NULL, };

static const struct attribute_group dev_dax_attribute_group = { .attrs = dev_dax_attributes, + .is_visible = dev_dax_visible, };

static const struct attribute_group *dax_attribute_groups[] = {

From: Dave Hansen dave.hansen@linux.intel.com

mainline inclusion from mainline-v5.1-rc1 commit b926b7f3baecb2a855db629e6822e1a85212e91c category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit b926b7f3baecb2a855db629e6822e1a85212e91c upstream.

HMM consumes physical address space for its own use, even though nothing is mapped or accessible there. It uses a special resource description (IORES_DESC_DEVICE_PRIVATE_MEMORY) to uniquely identify these areas.

When HMM consumes address space, it makes a best guess about what to consume. However, it is possible that a future memory or device hotplug can collide with the reserved area. In the case of these conflicts, there is an error message in register_memory_resource().

Later patches in this series move register_memory_resource() from using request_resource_conflict() to __request_region(). Unfortunately, __request_region() does not return the conflict like the previous function did, which makes it impossible to check for IORES_DESC_DEVICE_PRIVATE_MEMORY in a conflicting resource.

Instead of warning in register_memory_resource(), move the check into the core resource code itself (__request_region()) where the conflicting resource _is_ available. This has the added bonus of producing a warning in case of HMM conflicts with devices *or* RAM address space, as opposed to the RAM- only warnings that were there previously.

Signed-off-by: Dave Hansen dave.hansen@linux.intel.com Reviewed-by: Jerome Glisse jglisse@redhat.com Cc: Dan Williams dan.j.williams@intel.com Cc: Dave Jiang dave.jiang@intel.com Cc: Ross Zwisler zwisler@kernel.org Cc: Vishal Verma vishal.l.verma@intel.com Cc: Tom Lendacky thomas.lendacky@amd.com Cc: Andrew Morton akpm@linux-foundation.org Cc: Michal Hocko mhocko@suse.com Cc: linux-nvdimm@lists.01.org Cc: linux-kernel@vger.kernel.org Cc: linux-mm@kvack.org Cc: Huang Ying ying.huang@intel.com Cc: Fengguang Wu fengguang.wu@intel.com Cc: Keith Busch keith.busch@intel.com Signed-off-by: Dan Williams dan.j.williams@intel.com Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- kernel/resource.c | 9 +++++++++ mm/memory_hotplug.c | 5 ----- 2 files changed, 9 insertions(+), 5 deletions(-)

diff --git a/kernel/resource.c b/kernel/resource.c index 7dd11c78c7385..2cf0edd8223f2 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -1126,6 +1126,15 @@ struct resource * __request_region(struct resource *parent, conflict = __request_resource(parent, res); if (!conflict) break; + /* + * mm/hmm.c reserves physical addresses which then + * become unavailable to other users. Conflicts are + * not expected. Warn to aid debugging if encountered. + */ + if (conflict->desc == IORES_DESC_DEVICE_PRIVATE_MEMORY) { + pr_warn("Unaddressable device %s %pR conflicts with %pR", + conflict->name, conflict, res); + } if (conflict != parent) { if (!(conflict->flags & IORESOURCE_BUSY)) { parent = conflict; diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 0031c5063aa11..b8dac58215fac 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -111,11 +111,6 @@ static struct resource *register_memory_resource(u64 start, u64 size) res->flags = IORESOURCE_SYSTEM_RAM | IORESOURCE_BUSY; conflict = request_resource_conflict(&iomem_resource, res); if (conflict) { - if (conflict->desc == IORES_DESC_DEVICE_PRIVATE_MEMORY) { - pr_debug("Device unaddressable memory block " - "memory hotplug at %#010llx !\n", - (unsigned long long)start); - } pr_debug("System RAM resource %pR cannot be added\n", res); kfree(res); return ERR_PTR(-EEXIST);

From: Dave Hansen dave.hansen@linux.intel.com

mainline inclusion from mainline-v5.1-rc1 commit 2b539aefe9e48e3908cff02699aa63a8b9bd268e category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

In the process of onlining memory, we use walk_system_ram_range() to find the actual RAM areas inside of the area being onlined.

However, it currently only finds memory resources which are "top-level" iomem_resources. Children are not currently searched which causes it to skip System RAM in areas like this (in the format of /proc/iomem):

a0000000-bfffffff : Persistent Memory (legacy) a0000000-afffffff : System RAM

Changing the true->false here allows children to be searched as well. We need this because we add a new "System RAM" resource underneath the "persistent memory" resource when we use persistent memory in a volatile mode.

Signed-off-by: Dave Hansen dave.hansen@linux.intel.com Cc: Keith Busch keith.busch@intel.com Cc: Dan Williams dan.j.williams@intel.com Cc: Dave Jiang dave.jiang@intel.com Cc: Ross Zwisler zwisler@kernel.org Cc: Vishal Verma vishal.l.verma@intel.com Cc: Tom Lendacky thomas.lendacky@amd.com Cc: Andrew Morton akpm@linux-foundation.org Cc: Michal Hocko mhocko@suse.com Cc: linux-nvdimm@lists.01.org Cc: linux-kernel@vger.kernel.org Cc: linux-mm@kvack.org Cc: Huang Ying ying.huang@intel.com Cc: Fengguang Wu fengguang.wu@intel.com Cc: Borislav Petkov bp@suse.de Cc: Bjorn Helgaas bhelgaas@google.com Cc: Yaowei Bai baiyaowei@cmss.chinamobile.com Cc: Takashi Iwai tiwai@suse.de Cc: Jerome Glisse jglisse@redhat.com Signed-off-by: Dan Williams dan.j.williams@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- kernel/resource.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/kernel/resource.c b/kernel/resource.c index 2cf0edd8223f2..55a3c60106827 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -448,6 +448,9 @@ int walk_mem_res(u64 start, u64 end, void *arg, * This function calls the @func callback against all memory ranges of type * System RAM which are marked as IORESOURCE_SYSTEM_RAM and IORESOUCE_BUSY. * It is to be used only for System RAM. + * + * This will find System RAM ranges that are children of top-level resources + * in addition to top-level System RAM resources. */ int walk_system_ram_range(unsigned long start_pfn, unsigned long nr_pages, void *arg, int (*func)(unsigned long, unsigned long, void *)) @@ -463,7 +466,7 @@ int walk_system_ram_range(unsigned long start_pfn, unsigned long nr_pages, flags = IORESOURCE_SYSTEM_RAM | IORESOURCE_BUSY; while (start < end && !find_next_iomem_res(start, end, flags, IORES_DESC_NONE, - true, &res)) { + false, &res)) { pfn = (res.start + PAGE_SIZE - 1) >> PAGE_SHIFT; end_pfn = (res.end + 1) >> PAGE_SHIFT; if (end_pfn > pfn)

From: Erik Schmauss erik.schmauss@intel.com

mainline inclusion from mainline-v5.1-rc1 commit 9a8d961f1ef835b0d338fbe13da03cb424e87ae5 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 9a8d961f1ef835b0d338fbe13da03cb424e87ae5 upstream.

ACPICA commit a216e8ca9f7c79f90788b193e2e61151b2c973c0

This change reserves several field and renames subtable 0 to "memory proximity domain attributes"

Link: https://github.com/acpica/acpica/commit/a216e8ca Signed-off-by: Erik Schmauss erik.schmauss@intel.com Signed-off-by: Bob Moore robert.moore@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- include/acpi/actbl1.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/include/acpi/actbl1.h b/include/acpi/actbl1.h index ab424509cae9b..7df78a9bddb19 100644 --- a/include/acpi/actbl1.h +++ b/include/acpi/actbl1.h @@ -1406,17 +1406,17 @@ struct acpi_hmat_structure { * HMAT Structures, correspond to Type in struct acpi_hmat_structure */

-/* 0: Memory subystem address range */ +/* 0: Memory proximity domain attributes */

-struct acpi_hmat_address_range { +struct acpi_hmat_proximity_domain { struct acpi_hmat_structure header; u16 flags; u16 reserved1; u32 processor_PD; /* Processor proximity domain */ u32 memory_PD; /* Memory proximity domain */ u32 reserved2; - u64 physical_address_base; /* Physical address range base */ - u64 physical_address_length; /* Physical address range length */ + u64 reserved3; + u64 reserved4; };

/* Masks for Flags field above */

From: Zheng Zengkai zhengzengkai@huawei.com

hulk inclusion category: Feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

-------------------------------------------------

60574d1e05b0 ("acpi: Create subtable parsing infrastructure") modifies the argument 1 type of acpi_tbl_entry_handler from struct acpi_subtable_header * to union acpi_subtable_headers *, leading to compilation errors in original ft2500 irqchip driver.

Adapt the phytium ft2500 irqchip driver using the new argument type.

Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/irqchip/irq-gic-phytium-2500-its.c | 6 +++--- drivers/irqchip/irq-gic-phytium-2500.c | 10 +++++----- 2 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/drivers/irqchip/irq-gic-phytium-2500-its.c b/drivers/irqchip/irq-gic-phytium-2500-its.c index cdb58bf99c7c3..fff1c8546d230 100644 --- a/drivers/irqchip/irq-gic-phytium-2500-its.c +++ b/drivers/irqchip/irq-gic-phytium-2500-its.c @@ -3944,13 +3944,13 @@ static int __init acpi_get_its_numa_node(u32 its_id) return NUMA_NO_NODE; }

-static int __init gic_acpi_match_srat_its(struct acpi_subtable_header *header, +static int __init gic_acpi_match_srat_its(union acpi_subtable_headers *header, const unsigned long end) { return 0; }

-static int __init gic_acpi_parse_srat_its(struct acpi_subtable_header *header, +static int __init gic_acpi_parse_srat_its(union acpi_subtable_headers *header, const unsigned long end) { int node; @@ -4017,7 +4017,7 @@ static int __init acpi_get_its_numa_node(u32 its_id) { return NUMA_NO_NODE; } static void __init acpi_its_srat_maps_free(void) { } #endif

-static int __init gic_acpi_parse_madt_its(struct acpi_subtable_header *header, +static int __init gic_acpi_parse_madt_its(union acpi_subtable_headers *header, const unsigned long end) { struct acpi_madt_generic_translator *its_entry; diff --git a/drivers/irqchip/irq-gic-phytium-2500.c b/drivers/irqchip/irq-gic-phytium-2500.c index fc69057e53f68..8674463a08c60 100644 --- a/drivers/irqchip/irq-gic-phytium-2500.c +++ b/drivers/irqchip/irq-gic-phytium-2500.c @@ -1823,7 +1823,7 @@ gic_acpi_register_redist(phys_addr_t phys_base, void __iomem *redist_base) }

static int __init -gic_acpi_parse_madt_redist(struct acpi_subtable_header *header, +gic_acpi_parse_madt_redist(union acpi_subtable_headers *header, const unsigned long end) { struct acpi_madt_generic_redistributor *redist = @@ -1841,7 +1841,7 @@ gic_acpi_parse_madt_redist(struct acpi_subtable_header *header, }

static int __init -gic_acpi_parse_madt_gicc(struct acpi_subtable_header *header, +gic_acpi_parse_madt_gicc(union acpi_subtable_headers *header, const unsigned long end) { struct acpi_madt_generic_interrupt *gicc = @@ -1883,14 +1883,14 @@ static int __init gic_acpi_collect_gicr_base(void) return -ENODEV; }

-static int __init gic_acpi_match_gicr(struct acpi_subtable_header *header, +static int __init gic_acpi_match_gicr(union acpi_subtable_headers *header, const unsigned long end) { /* Subtable presence means that redist exists, that's it */ return 0; }

-static int __init gic_acpi_match_gicc(struct acpi_subtable_header *header, +static int __init gic_acpi_match_gicc(union acpi_subtable_headers *header, const unsigned long end) { struct acpi_madt_generic_interrupt *gicc = @@ -1960,7 +1960,7 @@ static bool __init acpi_validate_gic_table(struct acpi_subtable_header *header, return true; }

-static int __init gic_acpi_parse_virt_madt_gicc(struct acpi_subtable_header *header, +static int __init gic_acpi_parse_virt_madt_gicc(union acpi_subtable_headers *header, const unsigned long end) { struct acpi_madt_generic_interrupt *gicc =

From: Keith Busch keith.busch@intel.com

mainline inclusion from mainline-v5.2-rc1 commit 3accf7ae37a96c3bf4b51999f3c395ac5ffcd6d4 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 3accf7ae37a96c3bf4b51999f3c395ac5ffcd6d4 upstream.

Systems may provide different memory types and export this information in the ACPI Heterogeneous Memory Attribute Table (HMAT). Parse these tables provided by the platform and report the memory access and caching attributes to the kernel messages.

Reviewed-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Acked-by: Jonathan Cameron Jonathan.Cameron@huawei.com Tested-by: Jonathan Cameron Jonathan.Cameron@huawei.com Signed-off-by: Keith Busch keith.busch@intel.com Tested-by: Brice Goglin Brice.Goglin@inria.fr Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/acpi/Kconfig | 1 + drivers/acpi/Makefile | 1 + drivers/acpi/hmat/Kconfig | 7 ++ drivers/acpi/hmat/Makefile | 1 + drivers/acpi/hmat/hmat.c | 236 +++++++++++++++++++++++++++++++++++++ 5 files changed, 246 insertions(+) create mode 100644 drivers/acpi/hmat/Kconfig create mode 100644 drivers/acpi/hmat/Makefile create mode 100644 drivers/acpi/hmat/hmat.c

diff --git a/drivers/acpi/Kconfig b/drivers/acpi/Kconfig index e955a30f1a52f..89d5fd8020ecd 100644 --- a/drivers/acpi/Kconfig +++ b/drivers/acpi/Kconfig @@ -470,6 +470,7 @@ config ACPI_REDUCED_HARDWARE_ONLY If you are unsure what to do, do not enable this option.

source "drivers/acpi/nfit/Kconfig" +source "drivers/acpi/hmat/Kconfig"

source "drivers/acpi/apei/Kconfig" source "drivers/acpi/dptf/Kconfig" diff --git a/drivers/acpi/Makefile b/drivers/acpi/Makefile index 347d1eac72ae3..1864d17573387 100644 --- a/drivers/acpi/Makefile +++ b/drivers/acpi/Makefile @@ -84,6 +84,7 @@ obj-$(CONFIG_ACPI_PROCESSOR) += processor.o obj-$(CONFIG_ACPI) += container.o obj-$(CONFIG_ACPI_THERMAL) += thermal.o obj-$(CONFIG_ACPI_NFIT) += nfit/ +obj-$(CONFIG_ACPI_HMAT) += hmat/ obj-$(CONFIG_ACPI) += acpi_memhotplug.o obj-$(CONFIG_ACPI_HOTPLUG_IOAPIC) += ioapic.o obj-$(CONFIG_ACPI_BATTERY) += battery.o diff --git a/drivers/acpi/hmat/Kconfig b/drivers/acpi/hmat/Kconfig new file mode 100644 index 0000000000000..2f7111b7af623 --- /dev/null +++ b/drivers/acpi/hmat/Kconfig @@ -0,0 +1,7 @@ +# SPDX-License-Identifier: GPL-2.0 +config ACPI_HMAT + bool "ACPI Heterogeneous Memory Attribute Table Support" + depends on ACPI_NUMA + help + If set, this option has the kernel parse and report the + platform's ACPI HMAT (Heterogeneous Memory Attributes Table). diff --git a/drivers/acpi/hmat/Makefile b/drivers/acpi/hmat/Makefile new file mode 100644 index 0000000000000..e909051d3d00f --- /dev/null +++ b/drivers/acpi/hmat/Makefile @@ -0,0 +1 @@ +obj-$(CONFIG_ACPI_HMAT) := hmat.o diff --git a/drivers/acpi/hmat/hmat.c b/drivers/acpi/hmat/hmat.c new file mode 100644 index 0000000000000..4758beb3b2c10 --- /dev/null +++ b/drivers/acpi/hmat/hmat.c @@ -0,0 +1,236 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (c) 2019, Intel Corporation. + * + * Heterogeneous Memory Attributes Table (HMAT) representation + * + * This program parses and reports the platform's HMAT tables, and registers + * the applicable attributes with the node's interfaces. + */ + +#include <linux/acpi.h> +#include <linux/bitops.h> +#include <linux/device.h> +#include <linux/init.h> +#include <linux/list.h> +#include <linux/node.h> +#include <linux/sysfs.h> + +static __initdata u8 hmat_revision; + +static __init const char *hmat_data_type(u8 type) +{ + switch (type) { + case ACPI_HMAT_ACCESS_LATENCY: + return "Access Latency"; + case ACPI_HMAT_READ_LATENCY: + return "Read Latency"; + case ACPI_HMAT_WRITE_LATENCY: + return "Write Latency"; + case ACPI_HMAT_ACCESS_BANDWIDTH: + return "Access Bandwidth"; + case ACPI_HMAT_READ_BANDWIDTH: + return "Read Bandwidth"; + case ACPI_HMAT_WRITE_BANDWIDTH: + return "Write Bandwidth"; + default: + return "Reserved"; + } +} + +static __init const char *hmat_data_type_suffix(u8 type) +{ + switch (type) { + case ACPI_HMAT_ACCESS_LATENCY: + case ACPI_HMAT_READ_LATENCY: + case ACPI_HMAT_WRITE_LATENCY: + return " nsec"; + case ACPI_HMAT_ACCESS_BANDWIDTH: + case ACPI_HMAT_READ_BANDWIDTH: + case ACPI_HMAT_WRITE_BANDWIDTH: + return " MB/s"; + default: + return ""; + } +} + +static __init u32 hmat_normalize(u16 entry, u64 base, u8 type) +{ + u32 value; + + /* + * Check for invalid and overflow values + */ + if (entry == 0xffff || !entry) + return 0; + else if (base > (UINT_MAX / (entry))) + return 0; + + /* + * Divide by the base unit for version 1, convert latency from + * picosenonds to nanoseconds if revision 2. + */ + value = entry * base; + if (hmat_revision == 1) { + if (value < 10) + return 0; + value = DIV_ROUND_UP(value, 10); + } else if (hmat_revision == 2) { + switch (type) { + case ACPI_HMAT_ACCESS_LATENCY: + case ACPI_HMAT_READ_LATENCY: + case ACPI_HMAT_WRITE_LATENCY: + value = DIV_ROUND_UP(value, 1000); + break; + default: + break; + } + } + return value; +} + +static __init int hmat_parse_locality(union acpi_subtable_headers *header, + const unsigned long end) +{ + struct acpi_hmat_locality *hmat_loc = (void *)header; + unsigned int init, targ, total_size, ipds, tpds; + u32 *inits, *targs, value; + u16 *entries; + u8 type; + + if (hmat_loc->header.length < sizeof(*hmat_loc)) { + pr_notice("HMAT: Unexpected locality header length: %d\n", + hmat_loc->header.length); + return -EINVAL; + } + + type = hmat_loc->data_type; + ipds = hmat_loc->number_of_initiator_Pds; + tpds = hmat_loc->number_of_target_Pds; + total_size = sizeof(*hmat_loc) + sizeof(*entries) * ipds * tpds + + sizeof(*inits) * ipds + sizeof(*targs) * tpds; + if (hmat_loc->header.length < total_size) { + pr_notice("HMAT: Unexpected locality header length:%d, minimum required:%d\n", + hmat_loc->header.length, total_size); + return -EINVAL; + } + + pr_info("HMAT: Locality: Flags:%02x Type:%s Initiator Domains:%d Target Domains:%d Base:%lld\n", + hmat_loc->flags, hmat_data_type(type), ipds, tpds, + hmat_loc->entry_base_unit); + + inits = (u32 *)(hmat_loc + 1); + targs = inits + ipds; + entries = (u16 *)(targs + tpds); + for (init = 0; init < ipds; init++) { + for (targ = 0; targ < tpds; targ++) { + value = hmat_normalize(entries[init * tpds + targ], + hmat_loc->entry_base_unit, + type); + pr_info(" Initiator-Target[%d-%d]:%d%s\n", + inits[init], targs[targ], value, + hmat_data_type_suffix(type)); + } + } + + return 0; +} + +static __init int hmat_parse_cache(union acpi_subtable_headers *header, + const unsigned long end) +{ + struct acpi_hmat_cache *cache = (void *)header; + u32 attrs; + + if (cache->header.length < sizeof(*cache)) { + pr_notice("HMAT: Unexpected cache header length: %d\n", + cache->header.length); + return -EINVAL; + } + + attrs = cache->cache_attributes; + pr_info("HMAT: Cache: Domain:%d Size:%llu Attrs:%08x SMBIOS Handles:%d\n", + cache->memory_PD, cache->cache_size, attrs, + cache->number_of_SMBIOShandles); + + return 0; +} + +static int __init hmat_parse_proximity_domain(union acpi_subtable_headers *header, + const unsigned long end) +{ + struct acpi_hmat_proximity_domain *p = (void *)header; + + if (p->header.length != sizeof(*p)) { + pr_notice("HMAT: Unexpected address range header length: %d\n", + p->header.length); + return -EINVAL; + } + + if (hmat_revision == 1) + pr_info("HMAT: Memory (%#llx length %#llx) Flags:%04x Processor Domain:%d Memory Domain:%d\n", + p->reserved3, p->reserved4, p->flags, p->processor_PD, + p->memory_PD); + else + pr_info("HMAT: Memory Flags:%04x Processor Domain:%d Memory Domain:%d\n", + p->flags, p->processor_PD, p->memory_PD); + + return 0; +} + +static int __init hmat_parse_subtable(union acpi_subtable_headers *header, + const unsigned long end) +{ + struct acpi_hmat_structure *hdr = (void *)header; + + if (!hdr) + return -EINVAL; + + switch (hdr->type) { + case ACPI_HMAT_TYPE_ADDRESS_RANGE: + return hmat_parse_proximity_domain(header, end); + case ACPI_HMAT_TYPE_LOCALITY: + return hmat_parse_locality(header, end); + case ACPI_HMAT_TYPE_CACHE: + return hmat_parse_cache(header, end); + default: + return -EINVAL; + } +} + +static __init int hmat_init(void) +{ + struct acpi_table_header *tbl; + enum acpi_hmat_type i; + acpi_status status; + + if (srat_disabled()) + return 0; + + status = acpi_get_table(ACPI_SIG_HMAT, 0, &tbl); + if (ACPI_FAILURE(status)) + return 0; + + hmat_revision = tbl->revision; + switch (hmat_revision) { + case 1: + case 2: + break; + default: + pr_notice("Ignoring HMAT: Unknown revision:%d\n", hmat_revision); + goto out_put; + } + + for (i = ACPI_HMAT_TYPE_ADDRESS_RANGE; i < ACPI_HMAT_TYPE_RESERVED; i++) { + if (acpi_table_parse_entries(ACPI_SIG_HMAT, + sizeof(struct acpi_table_hmat), i, + hmat_parse_subtable, 0) < 0) { + pr_notice("Ignoring HMAT: Invalid table"); + goto out_put; + } + } +out_put: + acpi_put_table(tbl); + return 0; +} +subsys_initcall(hmat_init);

From: Keith Busch keith.busch@intel.com

mainline inclusion from mainline-v5.2-rc1 commit e1cf33aafb8462c7d0a0e6349925870316f040ee category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit e1cf33aafb8462c7d0a0e6349925870316f040ee upstream.

Heterogeneous memory systems provide memory nodes with different latency and bandwidth performance attributes. Provide a new kernel interface for subsystems to register the attributes under the memory target node's initiator access class. If the system provides this information, applications may query these attributes when deciding which node to request memory.

The following example shows the new sysfs hierarchy for a node exporting performance attributes:

# tree -P "read*|write*"/sys/devices/system/node/nodeY/accessZ/initiators/ /sys/devices/system/node/nodeY/accessZ/initiators/ |-- read_bandwidth |-- read_latency |-- write_bandwidth `-- write_latency

The bandwidth is exported as MB/s and latency is reported in nanoseconds. The values are taken from the platform as reported by the manufacturer.

Memory accesses from an initiator node that is not one of the memory's access "Z" initiator nodes linked in the same directory may observe different performance than reported here. When a subsystem makes use of this interface, initiators of a different access number may not have the same performance relative to initiators in other access numbers, or omitted from the any access class' initiators.

Descriptions for memory access initiator performance access attributes are added to sysfs stable documentation.

Acked-by: Jonathan Cameron Jonathan.Cameron@huawei.com Tested-by: Jonathan Cameron Jonathan.Cameron@huawei.com Signed-off-by: Keith Busch keith.busch@intel.com Reviewed-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Tested-by: Brice Goglin Brice.Goglin@inria.fr Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- Documentation/ABI/stable/sysfs-devices-node | 28 ++++++++++ drivers/base/Kconfig | 8 +++ drivers/base/node.c | 59 +++++++++++++++++++++ include/linux/node.h | 26 +++++++++ 4 files changed, 121 insertions(+)

diff --git a/Documentation/ABI/stable/sysfs-devices-node b/Documentation/ABI/stable/sysfs-devices-node index 433bcc04e542f..735a40a3f9b23 100644 --- a/Documentation/ABI/stable/sysfs-devices-node +++ b/Documentation/ABI/stable/sysfs-devices-node @@ -114,3 +114,31 @@ Contact: Keith Busch keith.busch@intel.com Description: The directory containing symlinks to memory targets that this initiator node has class "Y" access. + +What: /sys/devices/system/node/nodeX/accessY/initiators/read_bandwidth +Date: December 2018 +Contact: Keith Busch keith.busch@intel.com +Description: + This node's read bandwidth in MB/s when accessed from + nodes found in this access class's linked initiators. + +What: /sys/devices/system/node/nodeX/accessY/initiators/read_latency +Date: December 2018 +Contact: Keith Busch keith.busch@intel.com +Description: + This node's read latency in nanoseconds when accessed + from nodes found in this access class's linked initiators. + +What: /sys/devices/system/node/nodeX/accessY/initiators/write_bandwidth +Date: December 2018 +Contact: Keith Busch keith.busch@intel.com +Description: + This node's write bandwidth in MB/s when accessed from + found in this access class's linked initiators. + +What: /sys/devices/system/node/nodeX/accessY/initiators/write_latency +Date: December 2018 +Contact: Keith Busch keith.busch@intel.com +Description: + This node's write latency in nanoseconds when access + from nodes found in this class's linked initiators. diff --git a/drivers/base/Kconfig b/drivers/base/Kconfig index 3e63a900b330e..32dc81bd7056e 100644 --- a/drivers/base/Kconfig +++ b/drivers/base/Kconfig @@ -149,6 +149,14 @@ config DEBUG_TEST_DRIVER_REMOVE unusable. You should say N here unless you are explicitly looking to test this functionality.

+config HMEM_REPORTING + bool + default n + depends on NUMA + help + Enable reporting for heterogenous memory access attributes under + their non-uniform memory nodes. + source "drivers/base/test/Kconfig"

config SYS_HYPERVISOR diff --git a/drivers/base/node.c b/drivers/base/node.c index 7f7dbf20c1019..fc199e6b432c2 100644 --- a/drivers/base/node.c +++ b/drivers/base/node.c @@ -71,6 +71,9 @@ struct node_access_nodes { struct device dev; struct list_head list_node; unsigned access; +#ifdef CONFIG_HMEM_REPORTING + struct node_hmem_attrs hmem_attrs; +#endif }; #define to_access_nodes(dev) container_of(dev, struct node_access_nodes, dev)

@@ -148,6 +151,62 @@ static struct node_access_nodes *node_init_node_access(struct node *node, return NULL; }

+#ifdef CONFIG_HMEM_REPORTING +#define ACCESS_ATTR(name) \ +static ssize_t name##_show(struct device *dev, \ + struct device_attribute *attr, \ + char *buf) \ +{ \ + return sprintf(buf, "%u\n", to_access_nodes(dev)->hmem_attrs.name); \ +} \ +static DEVICE_ATTR_RO(name); + +ACCESS_ATTR(read_bandwidth) +ACCESS_ATTR(read_latency) +ACCESS_ATTR(write_bandwidth) +ACCESS_ATTR(write_latency) + +static struct attribute *access_attrs[] = { + &dev_attr_read_bandwidth.attr, + &dev_attr_read_latency.attr, + &dev_attr_write_bandwidth.attr, + &dev_attr_write_latency.attr, + NULL, +}; + +/** + * node_set_perf_attrs - Set the performance values for given access class + * @nid: Node identifier to be set + * @hmem_attrs: Heterogeneous memory performance attributes + * @access: The access class the for the given attributes + */ +void node_set_perf_attrs(unsigned int nid, struct node_hmem_attrs *hmem_attrs, + unsigned access) +{ + struct node_access_nodes *c; + struct node *node; + int i; + + if (WARN_ON_ONCE(!node_online(nid))) + return; + + node = node_devices[nid]; + c = node_init_node_access(node, access); + if (!c) + return; + + c->hmem_attrs = *hmem_attrs; + for (i = 0; access_attrs[i] != NULL; i++) { + if (sysfs_add_file_to_group(&c->dev.kobj, access_attrs[i], + "initiators")) { + pr_info("failed to add performance attribute to node %d\n", + nid); + break; + } + } +} +#endif + #define K(x) ((x) << (PAGE_SHIFT - 10)) static ssize_t node_read_meminfo(struct device *dev, struct device_attribute *attr, char *buf) diff --git a/include/linux/node.h b/include/linux/node.h index 1f25e9655d11e..526f1523c0f15 100644 --- a/include/linux/node.h +++ b/include/linux/node.h @@ -20,6 +20,32 @@ #include <linux/list.h> #include <linux/workqueue.h>

+/** + * struct node_hmem_attrs - heterogeneous memory performance attributes + * + * @read_bandwidth: Read bandwidth in MB/s + * @write_bandwidth: Write bandwidth in MB/s + * @read_latency: Read latency in nanoseconds + * @write_latency: Write latency in nanoseconds + */ +struct node_hmem_attrs { + unsigned int read_bandwidth; + unsigned int write_bandwidth; + unsigned int read_latency; + unsigned int write_latency; +}; + +#ifdef CONFIG_HMEM_REPORTING +void node_set_perf_attrs(unsigned int nid, struct node_hmem_attrs *hmem_attrs, + unsigned access); +#else +static inline void node_set_perf_attrs(unsigned int nid, + struct node_hmem_attrs *hmem_attrs, + unsigned access) +{ +} +#endif + struct node { struct device dev; struct list_head access_list;

From: Keith Busch keith.busch@intel.com

mainline inclusion from mainline-v5.2-rc1 commit 665ac7e92757fb74d249b2817dd91b3d9439d0a0 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 665ac7e92757fb74d249b2817dd91b3d9439d0a0 upstream.

If the HMAT Subsystem Address Range provides a valid processor proximity domain for a memory domain, or a processor domain matches the performance access of the valid processor proximity domain, register the memory target with that initiator so this relationship will be visible under the node's sysfs directory.

Since HMAT requires valid address ranges have an equivalent SRAT entry, verify each memory target satisfies this requirement.

Reviewed-by: Jonathan Cameron Jonathan.Cameron@huawei.com Signed-off-by: Keith Busch keith.busch@intel.com Acked-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Reviewed-by: Brice Goglin Brice.Goglin@inria.fr Tested-by: Brice Goglin Brice.Goglin@inria.fr Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/acpi/hmat/Kconfig | 3 +- drivers/acpi/hmat/hmat.c | 392 +++++++++++++++++++++++++++++++++++++- 2 files changed, 393 insertions(+), 2 deletions(-)

diff --git a/drivers/acpi/hmat/Kconfig b/drivers/acpi/hmat/Kconfig index 2f7111b7af623..13cddd612a529 100644 --- a/drivers/acpi/hmat/Kconfig +++ b/drivers/acpi/hmat/Kconfig @@ -4,4 +4,5 @@ config ACPI_HMAT depends on ACPI_NUMA help If set, this option has the kernel parse and report the - platform's ACPI HMAT (Heterogeneous Memory Attributes Table). + platform's ACPI HMAT (Heterogeneous Memory Attributes Table), + and register memory initiators with their targets. diff --git a/drivers/acpi/hmat/hmat.c b/drivers/acpi/hmat/hmat.c index 4758beb3b2c10..01a6eddac6f7a 100644 --- a/drivers/acpi/hmat/hmat.c +++ b/drivers/acpi/hmat/hmat.c @@ -13,11 +13,105 @@ #include <linux/device.h> #include <linux/init.h> #include <linux/list.h> +#include <linux/list_sort.h> #include <linux/node.h> #include <linux/sysfs.h>

static __initdata u8 hmat_revision;

+static __initdata LIST_HEAD(targets); +static __initdata LIST_HEAD(initiators); +static __initdata LIST_HEAD(localities); + +/* + * The defined enum order is used to prioritize attributes to break ties when + * selecting the best performing node. + */ +enum locality_types { + WRITE_LATENCY, + READ_LATENCY, + WRITE_BANDWIDTH, + READ_BANDWIDTH, +}; + +static struct memory_locality *localities_types[4]; + +struct memory_target { + struct list_head node; + unsigned int memory_pxm; + unsigned int processor_pxm; + struct node_hmem_attrs hmem_attrs; +}; + +struct memory_initiator { + struct list_head node; + unsigned int processor_pxm; +}; + +struct memory_locality { + struct list_head node; + struct acpi_hmat_locality *hmat_loc; +}; + +static __init struct memory_initiator *find_mem_initiator(unsigned int cpu_pxm) +{ + struct memory_initiator *initiator; + + list_for_each_entry(initiator, &initiators, node) + if (initiator->processor_pxm == cpu_pxm) + return initiator; + return NULL; +} + +static __init struct memory_target *find_mem_target(unsigned int mem_pxm) +{ + struct memory_target *target; + + list_for_each_entry(target, &targets, node) + if (target->memory_pxm == mem_pxm) + return target; + return NULL; +} + +static __init void alloc_memory_initiator(unsigned int cpu_pxm) +{ + struct memory_initiator *initiator; + + if (pxm_to_node(cpu_pxm) == NUMA_NO_NODE) + return; + + initiator = find_mem_initiator(cpu_pxm); + if (initiator) + return; + + initiator = kzalloc(sizeof(*initiator), GFP_KERNEL); + if (!initiator) + return; + + initiator->processor_pxm = cpu_pxm; + list_add_tail(&initiator->node, &initiators); +} + +static __init void alloc_memory_target(unsigned int mem_pxm) +{ + struct memory_target *target; + + if (pxm_to_node(mem_pxm) == NUMA_NO_NODE) + return; + + target = find_mem_target(mem_pxm); + if (target) + return; + + target = kzalloc(sizeof(*target), GFP_KERNEL); + if (!target) + return; + + target->memory_pxm = mem_pxm; + target->processor_pxm = PXM_INVAL; + list_add_tail(&target->node, &targets); +} + static __init const char *hmat_data_type(u8 type) { switch (type) { @@ -89,14 +183,83 @@ static __init u32 hmat_normalize(u16 entry, u64 base, u8 type) return value; }

+static __init void hmat_update_target_access(struct memory_target *target, + u8 type, u32 value) +{ + switch (type) { + case ACPI_HMAT_ACCESS_LATENCY: + target->hmem_attrs.read_latency = value; + target->hmem_attrs.write_latency = value; + break; + case ACPI_HMAT_READ_LATENCY: + target->hmem_attrs.read_latency = value; + break; + case ACPI_HMAT_WRITE_LATENCY: + target->hmem_attrs.write_latency = value; + break; + case ACPI_HMAT_ACCESS_BANDWIDTH: + target->hmem_attrs.read_bandwidth = value; + target->hmem_attrs.write_bandwidth = value; + break; + case ACPI_HMAT_READ_BANDWIDTH: + target->hmem_attrs.read_bandwidth = value; + break; + case ACPI_HMAT_WRITE_BANDWIDTH: + target->hmem_attrs.write_bandwidth = value; + break; + default: + break; + } +} + +static __init void hmat_add_locality(struct acpi_hmat_locality *hmat_loc) +{ + struct memory_locality *loc; + + loc = kzalloc(sizeof(*loc), GFP_KERNEL); + if (!loc) { + pr_notice_once("Failed to allocate HMAT locality\n"); + return; + } + + loc->hmat_loc = hmat_loc; + list_add_tail(&loc->node, &localities); + + switch (hmat_loc->data_type) { + case ACPI_HMAT_ACCESS_LATENCY: + localities_types[READ_LATENCY] = loc; + localities_types[WRITE_LATENCY] = loc; + break; + case ACPI_HMAT_READ_LATENCY: + localities_types[READ_LATENCY] = loc; + break; + case ACPI_HMAT_WRITE_LATENCY: + localities_types[WRITE_LATENCY] = loc; + break; + case ACPI_HMAT_ACCESS_BANDWIDTH: + localities_types[READ_BANDWIDTH] = loc; + localities_types[WRITE_BANDWIDTH] = loc; + break; + case ACPI_HMAT_READ_BANDWIDTH: + localities_types[READ_BANDWIDTH] = loc; + break; + case ACPI_HMAT_WRITE_BANDWIDTH: + localities_types[WRITE_BANDWIDTH] = loc; + break; + default: + break; + } +} + static __init int hmat_parse_locality(union acpi_subtable_headers *header, const unsigned long end) { struct acpi_hmat_locality *hmat_loc = (void *)header; + struct memory_target *target; unsigned int init, targ, total_size, ipds, tpds; u32 *inits, *targs, value; u16 *entries; - u8 type; + u8 type, mem_hier;

if (hmat_loc->header.length < sizeof(*hmat_loc)) { pr_notice("HMAT: Unexpected locality header length: %d\n", @@ -105,6 +268,7 @@ static __init int hmat_parse_locality(union acpi_subtable_headers *header, }

type = hmat_loc->data_type; + mem_hier = hmat_loc->flags & ACPI_HMAT_MEMORY_HIERARCHY; ipds = hmat_loc->number_of_initiator_Pds; tpds = hmat_loc->number_of_target_Pds; total_size = sizeof(*hmat_loc) + sizeof(*entries) * ipds * tpds + @@ -123,6 +287,7 @@ static __init int hmat_parse_locality(union acpi_subtable_headers *header, targs = inits + ipds; entries = (u16 *)(targs + tpds); for (init = 0; init < ipds; init++) { + alloc_memory_initiator(inits[init]); for (targ = 0; targ < tpds; targ++) { value = hmat_normalize(entries[init * tpds + targ], hmat_loc->entry_base_unit, @@ -130,9 +295,18 @@ static __init int hmat_parse_locality(union acpi_subtable_headers *header, pr_info(" Initiator-Target[%d-%d]:%d%s\n", inits[init], targs[targ], value, hmat_data_type_suffix(type)); + + if (mem_hier == ACPI_HMAT_MEMORY) { + target = find_mem_target(targs[targ]); + if (target && target->processor_pxm == inits[init]) + hmat_update_target_access(target, type, value); + } } }

+ if (mem_hier == ACPI_HMAT_MEMORY) + hmat_add_locality(hmat_loc); + return 0; }

@@ -160,6 +334,7 @@ static int __init hmat_parse_proximity_domain(union acpi_subtable_headers *heade const unsigned long end) { struct acpi_hmat_proximity_domain *p = (void *)header; + struct memory_target *target;

if (p->header.length != sizeof(*p)) { pr_notice("HMAT: Unexpected address range header length: %d\n", @@ -175,6 +350,23 @@ static int __init hmat_parse_proximity_domain(union acpi_subtable_headers *heade pr_info("HMAT: Memory Flags:%04x Processor Domain:%d Memory Domain:%d\n", p->flags, p->processor_PD, p->memory_PD);

+ if (p->flags & ACPI_HMAT_MEMORY_PD_VALID) { + target = find_mem_target(p->memory_PD); + if (!target) { + pr_debug("HMAT: Memory Domain missing from SRAT\n"); + return -EINVAL; + } + } + if (target && p->flags & ACPI_HMAT_PROCESSOR_PD_VALID) { + int p_node = pxm_to_node(p->processor_PD); + + if (p_node == NUMA_NO_NODE) { + pr_debug("HMAT: Invalid Processor Domain\n"); + return -EINVAL; + } + target->processor_pxm = p_node; + } + return 0; }

@@ -198,6 +390,191 @@ static int __init hmat_parse_subtable(union acpi_subtable_headers *header, } }

+static __init int srat_parse_mem_affinity(union acpi_subtable_headers *header, + const unsigned long end) +{ + struct acpi_srat_mem_affinity *ma = (void *)header; + + if (!ma) + return -EINVAL; + if (!(ma->flags & ACPI_SRAT_MEM_ENABLED)) + return 0; + alloc_memory_target(ma->proximity_domain); + return 0; +} + +static __init u32 hmat_initiator_perf(struct memory_target *target, + struct memory_initiator *initiator, + struct acpi_hmat_locality *hmat_loc) +{ + unsigned int ipds, tpds, i, idx = 0, tdx = 0; + u32 *inits, *targs; + u16 *entries; + + ipds = hmat_loc->number_of_initiator_Pds; + tpds = hmat_loc->number_of_target_Pds; + inits = (u32 *)(hmat_loc + 1); + targs = inits + ipds; + entries = (u16 *)(targs + tpds); + + for (i = 0; i < ipds; i++) { + if (inits[i] == initiator->processor_pxm) { + idx = i; + break; + } + } + + if (i == ipds) + return 0; + + for (i = 0; i < tpds; i++) { + if (targs[i] == target->memory_pxm) { + tdx = i; + break; + } + } + if (i == tpds) + return 0; + + return hmat_normalize(entries[idx * tpds + tdx], + hmat_loc->entry_base_unit, + hmat_loc->data_type); +} + +static __init bool hmat_update_best(u8 type, u32 value, u32 *best) +{ + bool updated = false; + + if (!value) + return false; + + switch (type) { + case ACPI_HMAT_ACCESS_LATENCY: + case ACPI_HMAT_READ_LATENCY: + case ACPI_HMAT_WRITE_LATENCY: + if (!*best || *best > value) { + *best = value; + updated = true; + } + break; + case ACPI_HMAT_ACCESS_BANDWIDTH: + case ACPI_HMAT_READ_BANDWIDTH: + case ACPI_HMAT_WRITE_BANDWIDTH: + if (!*best || *best < value) { + *best = value; + updated = true; + } + break; + } + + return updated; +} + +static int initiator_cmp(void *priv, struct list_head *a, struct list_head *b) +{ + struct memory_initiator *ia; + struct memory_initiator *ib; + unsigned long *p_nodes = priv; + + ia = list_entry(a, struct memory_initiator, node); + ib = list_entry(b, struct memory_initiator, node); + + set_bit(ia->processor_pxm, p_nodes); + set_bit(ib->processor_pxm, p_nodes); + + return ia->processor_pxm - ib->processor_pxm; +} + +static __init void hmat_register_target_initiators(struct memory_target *target) +{ + static DECLARE_BITMAP(p_nodes, MAX_NUMNODES); + struct memory_initiator *initiator; + unsigned int mem_nid, cpu_nid; + struct memory_locality *loc = NULL; + u32 best = 0; + int i; + + mem_nid = pxm_to_node(target->memory_pxm); + /* + * If the Address Range Structure provides a local processor pxm, link + * only that one. Otherwise, find the best performance attributes and + * register all initiators that match. + */ + if (target->processor_pxm != PXM_INVAL) { + cpu_nid = pxm_to_node(target->processor_pxm); + register_memory_node_under_compute_node(mem_nid, cpu_nid, 0); + return; + } + + if (list_empty(&localities)) + return; + + /* + * We need the initiator list sorted so we can use bitmap_clear for + * previously set initiators when we find a better memory accessor. + * We'll also use the sorting to prime the candidate nodes with known + * initiators. + */ + bitmap_zero(p_nodes, MAX_NUMNODES); + list_sort(p_nodes, &initiators, initiator_cmp); + for (i = WRITE_LATENCY; i <= READ_BANDWIDTH; i++) { + loc = localities_types[i]; + if (!loc) + continue; + + best = 0; + list_for_each_entry(initiator, &initiators, node) { + u32 value; + + if (!test_bit(initiator->processor_pxm, p_nodes)) + continue; + + value = hmat_initiator_perf(target, initiator, loc->hmat_loc); + if (hmat_update_best(loc->hmat_loc->data_type, value, &best)) + bitmap_clear(p_nodes, 0, initiator->processor_pxm); + if (value != best) + clear_bit(initiator->processor_pxm, p_nodes); + } + if (best) + hmat_update_target_access(target, loc->hmat_loc->data_type, best); + } + + for_each_set_bit(i, p_nodes, MAX_NUMNODES) { + cpu_nid = pxm_to_node(i); + register_memory_node_under_compute_node(mem_nid, cpu_nid, 0); + } +} + +static __init void hmat_register_targets(void) +{ + struct memory_target *target; + + list_for_each_entry(target, &targets, node) + hmat_register_target_initiators(target); +} + +static __init void hmat_free_structures(void) +{ + struct memory_target *target, *tnext; + struct memory_locality *loc, *lnext; + struct memory_initiator *initiator, *inext; + + list_for_each_entry_safe(target, tnext, &targets, node) { + list_del(&target->node); + kfree(target); + } + + list_for_each_entry_safe(initiator, inext, &initiators, node) { + list_del(&initiator->node); + kfree(initiator); + } + + list_for_each_entry_safe(loc, lnext, &localities, node) { + list_del(&loc->node); + kfree(loc); + } +} + static __init int hmat_init(void) { struct acpi_table_header *tbl; @@ -207,6 +584,17 @@ static __init int hmat_init(void) if (srat_disabled()) return 0;

+ status = acpi_get_table(ACPI_SIG_SRAT, 0, &tbl); + if (ACPI_FAILURE(status)) + return 0; + + if (acpi_table_parse_entries(ACPI_SIG_SRAT, + sizeof(struct acpi_table_srat), + ACPI_SRAT_TYPE_MEMORY_AFFINITY, + srat_parse_mem_affinity, 0) < 0) + goto out_put; + acpi_put_table(tbl); + status = acpi_get_table(ACPI_SIG_HMAT, 0, &tbl); if (ACPI_FAILURE(status)) return 0; @@ -229,7 +617,9 @@ static __init int hmat_init(void) goto out_put; } } + hmat_register_targets(); out_put: + hmat_free_structures(); acpi_put_table(tbl); return 0; }

From: Keith Busch keith.busch@intel.com

mainline inclusion from mainline-v5.2-rc1 commit d9e8844c7d8165d97848950ae6bf66b2be86ef06 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit d9e8844c7d8165d97848950ae6bf66b2be86ef06 upstream.

Register memory side cache attributes with the memory's node if HMAT provides the side cache iniformation table.

Acked-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Keith Busch keith.busch@intel.com Tested-by: Brice Goglin Brice.Goglin@inria.fr Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org Signed-off-by: Fan Du fan.du@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/acpi/hmat/hmat.c | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+)

diff --git a/drivers/acpi/hmat/hmat.c b/drivers/acpi/hmat/hmat.c index 7a3a2b50cadde..b7824a0309f78 100644 --- a/drivers/acpi/hmat/hmat.c +++ b/drivers/acpi/hmat/hmat.c @@ -314,6 +314,7 @@ static __init int hmat_parse_cache(union acpi_subtable_headers *header, const unsigned long end) { struct acpi_hmat_cache *cache = (void *)header; + struct node_cache_attrs cache_attrs; u32 attrs;

if (cache->header.length < sizeof(*cache)) { @@ -327,6 +328,37 @@ static __init int hmat_parse_cache(union acpi_subtable_headers *header, cache->memory_PD, cache->cache_size, attrs, cache->number_of_SMBIOShandles);

+ cache_attrs.size = cache->cache_size; + cache_attrs.level = (attrs & ACPI_HMAT_CACHE_LEVEL) >> 4; + cache_attrs.line_size = (attrs & ACPI_HMAT_CACHE_LINE_SIZE) >> 16; + + switch ((attrs & ACPI_HMAT_CACHE_ASSOCIATIVITY) >> 8) { + case ACPI_HMAT_CA_DIRECT_MAPPED: + cache_attrs.indexing = NODE_CACHE_DIRECT_MAP; + break; + case ACPI_HMAT_CA_COMPLEX_CACHE_INDEXING: + cache_attrs.indexing = NODE_CACHE_INDEXED; + break; + case ACPI_HMAT_CA_NONE: + default: + cache_attrs.indexing = NODE_CACHE_OTHER; + break; + } + + switch ((attrs & ACPI_HMAT_WRITE_POLICY) >> 12) { + case ACPI_HMAT_CP_WB: + cache_attrs.write_policy = NODE_CACHE_WRITE_BACK; + break; + case ACPI_HMAT_CP_WT: + cache_attrs.write_policy = NODE_CACHE_WRITE_THROUGH; + break; + case ACPI_HMAT_CP_NONE: + default: + cache_attrs.write_policy = NODE_CACHE_WRITE_OTHER; + break; + } + + node_add_cache(pxm_to_node(cache->memory_PD), &cache_attrs); return 0; }

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.2-rc1 commit 878068ea270ea82767ff1d26c91583263c81fba0 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 878068ea270ea82767ff1d26c91583263c81fba0 upstream.

Starting from Icelake, XMM registers can be collected in PEBS record. But current code only output the pt_regs.

Add a new struct x86_perf_regs for both pt_regs and xmm_regs. The xmm_regs will be used later to keep a pointer to PEBS record which has XMM information.

XMM registers are 128 bit. To simplify the code, they are handled like two different registers, which means setting two bits in the register bitmap. This also allows only sampling the lower 64bit bits in XMM.

The index of XMM registers starts from 32. There are 16 XMM registers. So all reserved space for regs are used. Remove REG_RESERVED.

Add PERF_REG_X86_XMM_MAX, which stands for the max number of all x86 regs including both GPRs and XMM.

Add REG_NOSUPPORT for 32bit to exclude unsupported registers.

Previous platforms can not collect XMM information in PEBS record. Adding pebs_no_xmm_regs to indicate the unsupported platforms.

The common code still validates the supported registers. However, it cannot check model specific registers, e.g. XMM. Add extra check in x86_pmu_hw_config() to reject invalid config of regs_user and regs_intr. The regs_user never supports XMM collection. The regs_intr only supports XMM collection when sampling PEBS event on icelake and later platforms.

Originally-by: Andi Kleen ak@linux.intel.com Suggested-by: Peter Zijlstra (Intel) peterz@infradead.org Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Arnaldo Carvalho de Melo acme@redhat.com Cc: Jiri Olsa jolsa@redhat.com Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Stephane Eranian eranian@google.com Cc: Thomas Gleixner tglx@linutronix.de Cc: Vince Weaver vincent.weaver@maine.edu Cc: acme@kernel.org Cc: jolsa@kernel.org Link: https://lkml.kernel.org/r/20190402194509.2832-3-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/core.c | 15 ++++++++++++ arch/x86/events/intel/ds.c | 2 ++ arch/x86/events/perf_event.h | 33 +++++++++++++++++++++------ arch/x86/include/asm/perf_event.h | 5 ++++ arch/x86/include/uapi/asm/perf_regs.h | 23 ++++++++++++++++++- arch/x86/kernel/perf_regs.c | 27 ++++++++++++++++------ 6 files changed, 90 insertions(+), 15 deletions(-)

diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c index 640f85da2b342..206ce31116b23 100644 --- a/arch/x86/events/core.c +++ b/arch/x86/events/core.c @@ -565,6 +565,21 @@ int x86_pmu_hw_config(struct perf_event *event) return -EINVAL; }

+ /* sample_regs_user never support XMM registers */ + if (unlikely(event->attr.sample_regs_user & PEBS_XMM_REGS)) + return -EINVAL; + /* + * Besides the general purpose registers, XMM registers may + * be collected in PEBS on some platforms, e.g. Icelake + */ + if (unlikely(event->attr.sample_regs_intr & PEBS_XMM_REGS)) { + if (x86_pmu.pebs_no_xmm_regs) + return -EINVAL; + + if (!event->attr.precise_ip) + return -EINVAL; + } + return x86_setup_perfctr(event); }

diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index 2de55e1bef8ec..f62fd8f54ac14 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -1651,6 +1651,8 @@ void __init intel_ds_init(void) x86_pmu.bts = boot_cpu_has(X86_FEATURE_BTS); x86_pmu.pebs = boot_cpu_has(X86_FEATURE_PEBS); x86_pmu.pebs_buffer_size = PEBS_BUFFER_SIZE; + if (x86_pmu.version <= 4) + x86_pmu.pebs_no_xmm_regs = 1; if (x86_pmu.pebs) { char pebs_type = x86_pmu.intel_cap.pebs_trap ? '+' : '-'; int format = x86_pmu.intel_cap.pebs_format; diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index ad7290d51a3b1..97745c45334c2 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -123,6 +123,24 @@ struct amd_nb { (1ULL << PERF_REG_X86_R14) | \ (1ULL << PERF_REG_X86_R15))

+#define PEBS_XMM_REGS \ + ((1ULL << PERF_REG_X86_XMM0) | \ + (1ULL << PERF_REG_X86_XMM1) | \ + (1ULL << PERF_REG_X86_XMM2) | \ + (1ULL << PERF_REG_X86_XMM3) | \ + (1ULL << PERF_REG_X86_XMM4) | \ + (1ULL << PERF_REG_X86_XMM5) | \ + (1ULL << PERF_REG_X86_XMM6) | \ + (1ULL << PERF_REG_X86_XMM7) | \ + (1ULL << PERF_REG_X86_XMM8) | \ + (1ULL << PERF_REG_X86_XMM9) | \ + (1ULL << PERF_REG_X86_XMM10) | \ + (1ULL << PERF_REG_X86_XMM11) | \ + (1ULL << PERF_REG_X86_XMM12) | \ + (1ULL << PERF_REG_X86_XMM13) | \ + (1ULL << PERF_REG_X86_XMM14) | \ + (1ULL << PERF_REG_X86_XMM15)) + /* * Per register state. */ @@ -639,13 +657,14 @@ struct x86_pmu { /* * Intel DebugStore bits */ - unsigned int bts :1, - bts_active :1, - pebs :1, - pebs_active :1, - pebs_broken :1, - pebs_prec_dist :1, - pebs_no_tlb :1; + unsigned int bts :1, + bts_active :1, + pebs :1, + pebs_active :1, + pebs_broken :1, + pebs_prec_dist :1, + pebs_no_tlb :1, + pebs_no_xmm_regs :1; int pebs_record_size; int pebs_buffer_size; void (*drain_pebs)(struct pt_regs *regs); diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h index 7a5ca63a3cca6..af66f97544a2a 100644 --- a/arch/x86/include/asm/perf_event.h +++ b/arch/x86/include/asm/perf_event.h @@ -252,6 +252,11 @@ extern void perf_events_lapic_init(void); #define PERF_EFLAGS_VM (1UL << 5)

struct pt_regs; +struct x86_perf_regs { + struct pt_regs regs; + u64 *xmm_regs; +}; + extern unsigned long perf_instruction_pointer(struct pt_regs *regs); extern unsigned long perf_misc_flags(struct pt_regs *regs); #define perf_misc_flags(regs) perf_misc_flags(regs) diff --git a/arch/x86/include/uapi/asm/perf_regs.h b/arch/x86/include/uapi/asm/perf_regs.h index f3329cabce5c6..ac67bbea10cae 100644 --- a/arch/x86/include/uapi/asm/perf_regs.h +++ b/arch/x86/include/uapi/asm/perf_regs.h @@ -27,8 +27,29 @@ enum perf_event_x86_regs { PERF_REG_X86_R13, PERF_REG_X86_R14, PERF_REG_X86_R15, - + /* These are the limits for the GPRs. */ PERF_REG_X86_32_MAX = PERF_REG_X86_GS + 1, PERF_REG_X86_64_MAX = PERF_REG_X86_R15 + 1, + + /* These all need two bits set because they are 128bit */ + PERF_REG_X86_XMM0 = 32, + PERF_REG_X86_XMM1 = 34, + PERF_REG_X86_XMM2 = 36, + PERF_REG_X86_XMM3 = 38, + PERF_REG_X86_XMM4 = 40, + PERF_REG_X86_XMM5 = 42, + PERF_REG_X86_XMM6 = 44, + PERF_REG_X86_XMM7 = 46, + PERF_REG_X86_XMM8 = 48, + PERF_REG_X86_XMM9 = 50, + PERF_REG_X86_XMM10 = 52, + PERF_REG_X86_XMM11 = 54, + PERF_REG_X86_XMM12 = 56, + PERF_REG_X86_XMM13 = 58, + PERF_REG_X86_XMM14 = 60, + PERF_REG_X86_XMM15 = 62, + + /* These include both GPRs and XMMX registers */ + PERF_REG_X86_XMM_MAX = PERF_REG_X86_XMM15 + 2, }; #endif /* _ASM_X86_PERF_REGS_H */ diff --git a/arch/x86/kernel/perf_regs.c b/arch/x86/kernel/perf_regs.c index c06c4c16c6b69..07c30ee174254 100644 --- a/arch/x86/kernel/perf_regs.c +++ b/arch/x86/kernel/perf_regs.c @@ -59,18 +59,34 @@ static unsigned int pt_regs_offset[PERF_REG_X86_MAX] = {

u64 perf_reg_value(struct pt_regs *regs, int idx) { + struct x86_perf_regs *perf_regs; + + if (idx >= PERF_REG_X86_XMM0 && idx < PERF_REG_X86_XMM_MAX) { + perf_regs = container_of(regs, struct x86_perf_regs, regs); + if (!perf_regs->xmm_regs) + return 0; + return perf_regs->xmm_regs[idx - PERF_REG_X86_XMM0]; + } + if (WARN_ON_ONCE(idx >= ARRAY_SIZE(pt_regs_offset))) return 0;

return regs_get_register(regs, pt_regs_offset[idx]); }

-#define REG_RESERVED (~((1ULL << PERF_REG_X86_MAX) - 1ULL)) - #ifdef CONFIG_X86_32 +#define REG_NOSUPPORT ((1ULL << PERF_REG_X86_R8) | \ + (1ULL << PERF_REG_X86_R9) | \ + (1ULL << PERF_REG_X86_R10) | \ + (1ULL << PERF_REG_X86_R11) | \ + (1ULL << PERF_REG_X86_R12) | \ + (1ULL << PERF_REG_X86_R13) | \ + (1ULL << PERF_REG_X86_R14) | \ + (1ULL << PERF_REG_X86_R15)) + int perf_reg_validate(u64 mask) { - if (!mask || mask & REG_RESERVED) + if (!mask || (mask & REG_NOSUPPORT)) return -EINVAL;

return 0; @@ -96,10 +112,7 @@ void perf_get_regs_user(struct perf_regs *regs_user,

int perf_reg_validate(u64 mask) { - if (!mask || mask & REG_RESERVED) - return -EINVAL; - - if (mask & REG_NOSUPPORT) + if (!mask || (mask & REG_NOSUPPORT)) return -EINVAL;

return 0;

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.2-rc1 commit 477f00f9617009a9a3a9271885231573b728ca4f category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 477f00f9617009a9a3a9271885231573b728ca4f upstream.

The drain_pebs() could be called twice in a short period for auto-reload event in pmu::read(). The intel_pmu_save_and_restart_reload() should be called to update the event->count.

This case should also be handled on Icelake. Extract the code for later reuse.

Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Arnaldo Carvalho de Melo acme@redhat.com Cc: Jiri Olsa jolsa@redhat.com Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Stephane Eranian eranian@google.com Cc: Thomas Gleixner tglx@linutronix.de Cc: Vince Weaver vincent.weaver@maine.edu Cc: acme@kernel.org Cc: jolsa@kernel.org Link: https://lkml.kernel.org/r/20190402194509.2832-5-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/intel/ds.c | 33 ++++++++++++++++++++------------- 1 file changed, 20 insertions(+), 13 deletions(-)

diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index e5ffc12ced017..7c046c2c05885 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -1514,6 +1514,25 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) __intel_pmu_pebs_event(event, iregs, at, top, 0, n); }

+static void intel_pmu_pebs_event_update_no_drain(struct cpu_hw_events *cpuc, int size) +{ + struct perf_event *event; + int bit; + + /* + * The drain_pebs() could be called twice in a short period + * for auto-reload event in pmu::read(). There are no + * overflows have happened in between. + * It needs to call intel_pmu_save_and_restart_reload() to + * update the event->count for this case. + */ + for_each_set_bit(bit, (unsigned long *)&cpuc->pebs_enabled, size) { + event = cpuc->events[bit]; + if (event->hw.flags & PERF_X86_EVENT_AUTO_RELOAD) + intel_pmu_save_and_restart_reload(event, 0); + } +} + static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) { struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); @@ -1541,19 +1560,7 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) }

if (unlikely(base >= top)) { - /* - * The drain_pebs() could be called twice in a short period - * for auto-reload event in pmu::read(). There are no - * overflows have happened in between. - * It needs to call intel_pmu_save_and_restart_reload() to - * update the event->count for this case. - */ - for_each_set_bit(bit, (unsigned long *)&cpuc->pebs_enabled, - size) { - event = cpuc->events[bit]; - if (event->hw.flags & PERF_X86_EVENT_AUTO_RELOAD) - intel_pmu_save_and_restart_reload(event, 0); - } + intel_pmu_pebs_event_update_no_drain(cpuc, size); return; }

From: Andi Kleen ak@linux.intel.com

mainline inclusion from mainline-v5.2-rc1 commit d3617b98b04583df222f34992e65712862a77bf1 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit d3617b98b04583df222f34992e65712862a77bf1 upstream.

With adaptive PEBS the CPU can directly supply the LBR information, so we don't need to read it again. But the LBRs still need to be enabled. Add a special count to the cpuc that distinguishes these two cases, and avoid reading the LBRs unnecessarily when PEBS is active.

Signed-off-by: Andi Kleen ak@linux.intel.com Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Arnaldo Carvalho de Melo acme@redhat.com Cc: Jiri Olsa jolsa@redhat.com Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Stephane Eranian eranian@google.com Cc: Thomas Gleixner tglx@linutronix.de Cc: Vince Weaver vincent.weaver@maine.edu Cc: acme@kernel.org Cc: jolsa@kernel.org Link: https://lkml.kernel.org/r/20190402194509.2832-7-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/intel/lbr.c | 13 ++++++++++++- arch/x86/events/perf_event.h | 1 + 2 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/arch/x86/events/intel/lbr.c b/arch/x86/events/intel/lbr.c index 58889f9529593..b8cac31f089c3 100644 --- a/arch/x86/events/intel/lbr.c +++ b/arch/x86/events/intel/lbr.c @@ -488,6 +488,8 @@ void intel_pmu_lbr_add(struct perf_event *event) * be 'new'. Conversely, a new event can get installed through the * context switch path for the first time. */ + if (x86_pmu.intel_cap.pebs_baseline && event->attr.precise_ip > 0) + cpuc->lbr_pebs_users++; perf_sched_cb_inc(event->ctx->pmu); if (!cpuc->lbr_users++ && !event->total_time_running) intel_pmu_lbr_reset(); @@ -507,8 +509,11 @@ void intel_pmu_lbr_del(struct perf_event *event) task_ctx->lbr_callstack_users--; }

+ if (x86_pmu.intel_cap.pebs_baseline && event->attr.precise_ip > 0) + cpuc->lbr_pebs_users--; cpuc->lbr_users--; WARN_ON_ONCE(cpuc->lbr_users < 0); + WARN_ON_ONCE(cpuc->lbr_pebs_users < 0); perf_sched_cb_dec(event->ctx->pmu); }

@@ -658,7 +663,13 @@ void intel_pmu_lbr_read(void) { struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);

- if (!cpuc->lbr_users) + /* + * Don't read when all LBRs users are using adaptive PEBS. + * + * This could be smarter and actually check the event, + * but this simple approach seems to work for now. + */ + if (!cpuc->lbr_users || cpuc->lbr_users == cpuc->lbr_pebs_users) return;

if (x86_pmu.intel_cap.lbr_format == LBR_FORMAT_32) diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index d825eb137d458..12db03091acfd 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -241,6 +241,7 @@ struct cpu_hw_events { * Intel LBR bits */ int lbr_users; + int lbr_pebs_users; struct perf_branch_stack lbr_stack; struct perf_branch_entry lbr_entries[MAX_LBR_ENTRIES]; struct er_account *lbr_sel;

From: Andi Kleen ak@linux.intel.com

mainline inclusion from mainline-v5.2-rc1 commit ca138a7aabc68bf727918bb40ce08157cd5ec0a5 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit ca138a7aabc68bf727918bb40ce08157cd5ec0a5 upstream.

Icelake and later platforms support collecting XMM registers with PEBS event.

Add support for 'perf script' to dump them, and support for the register parser in 'perf record -I=' ... to configure them.

For now they are just printed in hex, we could potentially later add other formats too.

Committer testing:

Before:

# perf record -IXMM0 Warning: unknown register XMM0, check man page or run 'perf record -I?'

Usage: perf record [<options>] [<command>] or: perf record [<options>] -- <command> [<options>]

# # perf record -I? available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10 R11 R12 R13 R14 R15

Usage: perf record [<options>] [<command>] or: perf record [<options>] -- <command> [<options>] #

After:

# perf record -IXMM0 Error: The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (cycles). /bin/dmesg | grep -i perf may provide additional information.

# # perf record -I? available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10 R11 R12 R13 R14 R15 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 XMM9 XMM10 XMM11 XMM12 XMM13 XMM14 XMM15

Usage: perf record [<options>] [<command>] or: perf record [<options>] -- <command> [<options>]

-I, --intr-regs[=<any register>] sample selected machine registers on interrupt, use -I ? to list register names #

More work is needed to, when faced with such error, warn the user that that register is not available on the running platform.

Signed-off-by: Andi Kleen ak@linux.intel.com Tested-by: Arnaldo Carvalho de Melo acme@redhat.com Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Jiri Olsa jolsa@kernel.org Cc: Peter Zijlstra peterz@infradead.org Link: http://lkml.kernel.org/r/20190506141926.13659-1-kan.liang@linux.intel.com Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo acme@redhat.com Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- tools/perf/arch/x86/include/perf_regs.h | 25 +++++++++++++++++++++++-- tools/perf/arch/x86/util/perf_regs.c | 16 ++++++++++++++++ tools/perf/util/perf_regs.h | 1 + 3 files changed, 40 insertions(+), 2 deletions(-)

diff --git a/tools/perf/arch/x86/include/perf_regs.h b/tools/perf/arch/x86/include/perf_regs.h index 7f6d538f8a895..b7321337d1003 100644 --- a/tools/perf/arch/x86/include/perf_regs.h +++ b/tools/perf/arch/x86/include/perf_regs.h @@ -8,9 +8,9 @@

void perf_regs_load(u64 *regs);

+#define PERF_REGS_MAX PERF_REG_X86_XMM_MAX #ifndef HAVE_ARCH_X86_64_SUPPORT #define PERF_REGS_MASK ((1ULL << PERF_REG_X86_32_MAX) - 1) -#define PERF_REGS_MAX PERF_REG_X86_32_MAX #define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_32 #else #define REG_NOSUPPORT ((1ULL << PERF_REG_X86_DS) | \ @@ -18,7 +18,6 @@ void perf_regs_load(u64 *regs); (1ULL << PERF_REG_X86_FS) | \ (1ULL << PERF_REG_X86_GS)) #define PERF_REGS_MASK (((1ULL << PERF_REG_X86_64_MAX) - 1) & ~REG_NOSUPPORT) -#define PERF_REGS_MAX PERF_REG_X86_64_MAX #define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_64 #endif #define PERF_REG_IP PERF_REG_X86_IP @@ -77,6 +76,28 @@ static inline const char *perf_reg_name(int id) case PERF_REG_X86_R15: return "R15"; #endif /* HAVE_ARCH_X86_64_SUPPORT */ + +#define XMM(x) \ + case PERF_REG_X86_XMM ## x: \ + case PERF_REG_X86_XMM ## x + 1: \ + return "XMM" #x; + XMM(0) + XMM(1) + XMM(2) + XMM(3) + XMM(4) + XMM(5) + XMM(6) + XMM(7) + XMM(8) + XMM(9) + XMM(10) + XMM(11) + XMM(12) + XMM(13) + XMM(14) + XMM(15) +#undef XMM default: return NULL; } diff --git a/tools/perf/arch/x86/util/perf_regs.c b/tools/perf/arch/x86/util/perf_regs.c index fead6b3b4206e..71d7604dbf0bc 100644 --- a/tools/perf/arch/x86/util/perf_regs.c +++ b/tools/perf/arch/x86/util/perf_regs.c @@ -31,6 +31,22 @@ const struct sample_reg sample_reg_masks[] = { SMPL_REG(R14, PERF_REG_X86_R14), SMPL_REG(R15, PERF_REG_X86_R15), #endif + SMPL_REG2(XMM0, PERF_REG_X86_XMM0), + SMPL_REG2(XMM1, PERF_REG_X86_XMM1), + SMPL_REG2(XMM2, PERF_REG_X86_XMM2), + SMPL_REG2(XMM3, PERF_REG_X86_XMM3), + SMPL_REG2(XMM4, PERF_REG_X86_XMM4), + SMPL_REG2(XMM5, PERF_REG_X86_XMM5), + SMPL_REG2(XMM6, PERF_REG_X86_XMM6), + SMPL_REG2(XMM7, PERF_REG_X86_XMM7), + SMPL_REG2(XMM8, PERF_REG_X86_XMM8), + SMPL_REG2(XMM9, PERF_REG_X86_XMM9), + SMPL_REG2(XMM10, PERF_REG_X86_XMM10), + SMPL_REG2(XMM11, PERF_REG_X86_XMM11), + SMPL_REG2(XMM12, PERF_REG_X86_XMM12), + SMPL_REG2(XMM13, PERF_REG_X86_XMM13), + SMPL_REG2(XMM14, PERF_REG_X86_XMM14), + SMPL_REG2(XMM15, PERF_REG_X86_XMM15), SMPL_REG_END };

diff --git a/tools/perf/util/perf_regs.h b/tools/perf/util/perf_regs.h index f732e3af2bd47..9f9b25bf71064 100644 --- a/tools/perf/util/perf_regs.h +++ b/tools/perf/util/perf_regs.h @@ -12,6 +12,7 @@ struct sample_reg { uint64_t mask; }; #define SMPL_REG(n, b) { .name = #n, .mask = 1ULL << (b) } +#define SMPL_REG2(n, b) { .name = #n, .mask = 3ULL << (b) } #define SMPL_REG_END { .name = NULL }

extern const struct sample_reg sample_reg_masks[];

From: Arnaldo Carvalho de Melo acme@redhat.com

mainline inclusion from mainline-v5.2-rc1 commit 4c1cf20334ae6e390e1ea787ef76c40a161370d1 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 4c1cf20334ae6e390e1ea787ef76c40a161370d1 upstream.

Add quotes around the register name and suggest using 'perf record -I?' to get the list of available registers.

Before:

# perf record -Idi,xmm20,xmm1 Warning: unknown register xmm20, check man page

Usage: perf record [<options>] [<command>] or: perf record [<options>] -- <command> [<options>]

-I, --intr-regs[=<any register>] sample selected machine registers on interrupt, use -I ? to list register names # # perf record -Idi,xmm20,xmm1 Warning: unknown register "xmm20", check man page or run "perf record -I?"

Usage: perf record [<options>] [<command>] or: perf record [<options>] -- <command> [<options>]

-I, --intr-regs[=<any register>] sample selected machine registers on interrupt, use -I ? to list register names #

Cc: Adrian Hunter adrian.hunter@intel.com Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Andi Kleen ak@linux.intel.com Cc: Jiri Olsa jolsa@kernel.org Cc: Kan Liang kan.liang@linux.intel.com Cc: Namhyung Kim namhyung@kernel.org Cc: Peter Zijlstra peterz@infradead.org Cc: Stephane Eranian eranian@google.com Link: https://lkml.kernel.org/n/tip-9a9hyuum8c0oggg86xd3sxc5@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo acme@redhat.com Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- tools/perf/util/parse-regs-options.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/tools/perf/util/parse-regs-options.c b/tools/perf/util/parse-regs-options.c index e5ad120e7f69a..409745663718f 100644 --- a/tools/perf/util/parse-regs-options.c +++ b/tools/perf/util/parse-regs-options.c @@ -48,8 +48,7 @@ parse_regs(const struct option *opt, const char *str, int unset) break; } if (!r->name) { - ui__warning("unknown register %s," - " check man page\n", s); + ui__warning("Unknown register "%s", check man page or run "perf record -I?"\n", s); goto error; }

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.2-rc1 commit af785e75bf616704cab031e66403b6adcf5b700a category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit af785e75bf616704cab031e66403b6adcf5b700a upstream.

There may be different register mask for use with intr or user on some platforms, e.g. Icelake.

Add weak functions arch__intr_reg_mask() and arch__user_reg_mask() to return intr and user register mask respectively.

Check mask before printing or comparing the register name.

Generic code always return PERF_REGS_MASK. No functional change.

Suggested-by: Arnaldo Carvalho de Melo acme@redhat.com Signed-off-by: Kan Liang kan.liang@linux.intel.com Tested-by: Ravi Bangoria ravi.bangoria@linux.ibm.com Cc: Andi Kleen ak@linux.intel.com Cc: Jiri Olsa jolsa@kernel.org Link: http://lkml.kernel.org/r/1557865174-56264-2-git-send-email-kan.liang@linux.i... Signed-off-by: Arnaldo Carvalho de Melo acme@redhat.com Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- tools/perf/util/parse-regs-options.c | 13 ++++++++++--- tools/perf/util/perf_regs.c | 10 ++++++++++ tools/perf/util/perf_regs.h | 2 ++ 3 files changed, 22 insertions(+), 3 deletions(-)

diff --git a/tools/perf/util/parse-regs-options.c b/tools/perf/util/parse-regs-options.c index f3b8f4f760b92..b82e2a097df85 100644 --- a/tools/perf/util/parse-regs-options.c +++ b/tools/perf/util/parse-regs-options.c @@ -12,6 +12,7 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr) const struct sample_reg *r; char *s, *os = NULL, *p; int ret = -1; + uint64_t mask;

if (unset) return 0; @@ -22,6 +23,11 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr) if (*mode) return -1;

+ if (intr) + mask = arch__intr_reg_mask(); + else + mask = arch__user_reg_mask(); + /* str may be NULL in case no arg is passed to -I */ if (str) { /* because str is read-only */ @@ -37,14 +43,15 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr) if (!strcmp(s, "?")) { fprintf(stderr, "available registers: "); for (r = sample_reg_masks; r->name; r++) { - fprintf(stderr, "%s ", r->name); + if (r->mask & mask) + fprintf(stderr, "%s ", r->name); } fputc('\n', stderr); /* just printing available regs */ goto error; } for (r = sample_reg_masks; r->name; r++) { - if (!strcasecmp(s, r->name)) + if ((r->mask & mask) && !strcasecmp(s, r->name)) break; } if (!r->name) { @@ -65,7 +72,7 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)

/* default to all possible regs */ if (*mode == 0) - *mode = PERF_REGS_MASK; + *mode = mask; error: free(os); return ret; diff --git a/tools/perf/util/perf_regs.c b/tools/perf/util/perf_regs.c index 2acfcc527caca..2774cec1f15fa 100644 --- a/tools/perf/util/perf_regs.c +++ b/tools/perf/util/perf_regs.c @@ -13,6 +13,16 @@ int __weak arch_sdt_arg_parse_op(char *old_op __maybe_unused, return SDT_ARG_SKIP; }

+uint64_t __weak arch__intr_reg_mask(void) +{ + return PERF_REGS_MASK; +} + +uint64_t __weak arch__user_reg_mask(void) +{ + return PERF_REGS_MASK; +} + #ifdef HAVE_PERF_REGS_SUPPORT int perf_reg_value(u64 *valp, struct regs_dump *regs, int id) { diff --git a/tools/perf/util/perf_regs.h b/tools/perf/util/perf_regs.h index 9f9b25bf71064..2f41b10698ad5 100644 --- a/tools/perf/util/perf_regs.h +++ b/tools/perf/util/perf_regs.h @@ -23,6 +23,8 @@ enum { };

int arch_sdt_arg_parse_op(char *old_op, char **new_op); +uint64_t arch__intr_reg_mask(void); +uint64_t arch__user_reg_mask(void);

#ifdef HAVE_PERF_REGS_SUPPORT #include <perf_regs.h>

From: Alexander Shishkin alexander.shishkin@linux.intel.com

mainline inclusion from mainline-v5.1 commit 26ae4f4406f88d82d79c85c11ac5fae18213cd38 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 26ae4f4406f88d82d79c85c11ac5fae18213cd38 upstream.

This recent commit:

5768402fd9c6e87 ("perf/ring_buffer: Use high order allocations for AUX buffers optimistically")

overlooked the fact that the previous one page granularity of the AUX buffer provided an implicit double buffering capability to the PMU driver, which went away when the entire buffer became one high-order page.

Always make the full-trace mode AUX allocation at least two-part to preserve the previous behavior and allow the implicit double buffering to continue.

Reported-by: Ammy Yi ammy.yi@intel.com Signed-off-by: Alexander Shishkin alexander.shishkin@linux.intel.com Acked-by: Peter Zijlstra a.p.zijlstra@chello.nl Cc: Arnaldo Carvalho de Melo acme@redhat.com Cc: Jiri Olsa jolsa@redhat.com Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Stephane Eranian eranian@google.com Cc: Thomas Gleixner tglx@linutronix.de Cc: Vince Weaver vincent.weaver@maine.edu Cc: adrian.hunter@intel.com Fixes: 5768402fd9c6e87 ("perf/ring_buffer: Use high order allocations for AUX buffers optimistically") Link: http://lkml.kernel.org/r/20190503085536.24119-2-alexander.shishkin@linux.int... Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- kernel/events/ring_buffer.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c index 12f351b253bbb..34e1667e42549 100644 --- a/kernel/events/ring_buffer.c +++ b/kernel/events/ring_buffer.c @@ -630,8 +630,7 @@ int rb_alloc_aux(struct ring_buffer *rb, struct perf_event *event, * PMU requests more than one contiguous chunks of memory * for SW double buffering */ - if ((event->pmu->capabilities & PERF_PMU_CAP_AUX_SW_DOUBLEBUF) && - !overwrite) { + if (!overwrite) { if (!max_order) return -EINVAL;

From: Andrew Murray andrew.murray@arm.com

mainline inclusion from mainline-v5.1-rc1 commit 486efe9f8e30bac1e236f867df164f4966f3e207 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 486efe9f8e30bac1e236f867df164f4966f3e207 upstream.

Add a function that tests if any of the perf event exclusion flags are set on a given event.

Signed-off-by: Andrew Murray andrew.murray@arm.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Arnaldo Carvalho de Melo acme@kernel.org Cc: Benjamin Herrenschmidt benh@kernel.crashing.org Cc: Borislav Petkov bp@alien8.de Cc: Ivan Kokshaysky ink@jurassic.park.msu.ru Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Mark Rutland mark.rutland@arm.com Cc: Matt Turner mattst88@gmail.com Cc: Michael Ellerman mpe@ellerman.id.au Cc: Paul Mackerras paulus@samba.org Cc: Peter Zijlstra peterz@infradead.org Cc: Richard Henderson rth@twiddle.net Cc: Russell King linux@armlinux.org.uk Cc: Sascha Hauer s.hauer@pengutronix.de Cc: Shawn Guo shawnguo@kernel.org Cc: Thomas Gleixner tglx@linutronix.de Cc: Will Deacon will.deacon@arm.com Cc: linux-arm-kernel@lists.infradead.org Cc: linuxppc-dev@lists.ozlabs.org Cc: robin.murphy@arm.com Cc: suzuki.poulose@arm.com Link: https://lkml.kernel.org/r/1547128414-50693-3-git-send-email-andrew.murray@ar... Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- include/linux/perf_event.h | 9 +++++++++ 1 file changed, 9 insertions(+)

diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index 64bfe0ac02673..8b6f2aeed9752 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -1018,6 +1018,15 @@ perf_event__output_id_sample(struct perf_event *event, extern void perf_log_lost_samples(struct perf_event *event, u64 lost);

+static inline bool event_has_any_exclude_flag(struct perf_event *event) +{ + struct perf_event_attr *attr = &event->attr; + + return attr->exclude_idle || attr->exclude_user || + attr->exclude_kernel || attr->exclude_hv || + attr->exclude_guest || attr->exclude_host; +} + static inline bool is_sampling_event(struct perf_event *event) { return event->attr.sample_period != 0;

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.2-rc7 commit e321d02db87af7840da29ef833a2a71fc0eab198 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit e321d02db87af7840da29ef833a2a71fc0eab198 upstream.

The perf fuzzer caused Skylake machine to crash:

[ 9680.085831] Call Trace: [ 9680.088301] <IRQ> [ 9680.090363] perf_output_sample_regs+0x43/0xa0 [ 9680.094928] perf_output_sample+0x3aa/0x7a0 [ 9680.099181] perf_event_output_forward+0x53/0x80 [ 9680.103917] __perf_event_overflow+0x52/0xf0 [ 9680.108266] ? perf_trace_run_bpf_submit+0xc0/0xc0 [ 9680.113108] perf_swevent_hrtimer+0xe2/0x150 [ 9680.117475] ? check_preempt_wakeup+0x181/0x230 [ 9680.122091] ? check_preempt_curr+0x62/0x90 [ 9680.126361] ? ttwu_do_wakeup+0x19/0x140 [ 9680.130355] ? try_to_wake_up+0x54/0x460 [ 9680.134366] ? reweight_entity+0x15b/0x1a0 [ 9680.138559] ? __queue_work+0x103/0x3f0 [ 9680.142472] ? update_dl_rq_load_avg+0x1cd/0x270 [ 9680.147194] ? timerqueue_del+0x1e/0x40 [ 9680.151092] ? __remove_hrtimer+0x35/0x70 [ 9680.155191] __hrtimer_run_queues+0x100/0x280 [ 9680.159658] hrtimer_interrupt+0x100/0x220 [ 9680.163835] smp_apic_timer_interrupt+0x6a/0x140 [ 9680.168555] apic_timer_interrupt+0xf/0x20 [ 9680.172756] </IRQ>

The XMM registers can only be collected by PEBS hardware events on the platforms with PEBS baseline support, e.g. Icelake, not software/probe events.

Add capabilities flag PERF_PMU_CAP_EXTENDED_REGS to indicate the PMU which support extended registers. For X86, the extended registers are XMM registers.

Add has_extended_regs() to check if extended registers are applied.

The generic code define the mask of extended registers as 0 if arch headers haven't overridden it.

Originally-by: Peter Zijlstra (Intel) peterz@infradead.org Reported-by: Vince Weaver vincent.weaver@maine.edu Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Arnaldo Carvalho de Melo acme@redhat.com Cc: Jiri Olsa jolsa@redhat.com Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Stephane Eranian eranian@google.com Cc: Thomas Gleixner tglx@linutronix.de Fixes: 878068ea270e ("perf/x86: Support outputting XMM registers") Link: https://lkml.kernel.org/r/1559081314-9714-1-git-send-email-kan.liang@linux.i... Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/core.c | 5 +++++ arch/x86/events/intel/ds.c | 1 + arch/x86/events/perf_event.h | 1 + arch/x86/include/uapi/asm/perf_regs.h | 3 +++ include/linux/perf_event.h | 1 + include/linux/perf_regs.h | 8 ++++++++ kernel/events/core.c | 18 ++++++++++++++---- 7 files changed, 33 insertions(+), 4 deletions(-)

diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c index 206ce31116b23..082fe2b988a1d 100644 --- a/arch/x86/events/core.c +++ b/arch/x86/events/core.c @@ -681,6 +681,11 @@ static inline int is_x86_event(struct perf_event *event) return event->pmu == &pmu; }

+struct pmu *x86_get_pmu(void) +{ + return &pmu; +} + /* * Event scheduler state: * diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index 8323b64385466..8ee52e1a062e8 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -2020,6 +2020,7 @@ void __init intel_ds_init(void) PERF_SAMPLE_TIME; x86_pmu.flags |= PMU_FL_PEBS_ALL; pebs_qual = "-baseline"; + x86_get_pmu()->capabilities |= PERF_PMU_CAP_EXTENDED_REGS; } else { /* Only basic record supported */ x86_pmu.pebs_no_xmm_regs = 1; diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index 12db03091acfd..ca41b57350d30 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -773,6 +773,7 @@ static struct perf_pmu_events_ht_attr event_attr_##v = { \ .event_str_ht = ht, \ }

+struct pmu *x86_get_pmu(void); extern struct x86_pmu x86_pmu __read_mostly;

static inline bool x86_pmu_has_lbr_callstack(void) diff --git a/arch/x86/include/uapi/asm/perf_regs.h b/arch/x86/include/uapi/asm/perf_regs.h index ac67bbea10cae..7c9d2bb3833bd 100644 --- a/arch/x86/include/uapi/asm/perf_regs.h +++ b/arch/x86/include/uapi/asm/perf_regs.h @@ -52,4 +52,7 @@ enum perf_event_x86_regs { /* These include both GPRs and XMMX registers */ PERF_REG_X86_XMM_MAX = PERF_REG_X86_XMM15 + 2, }; + +#define PERF_REG_EXTENDED_MASK (~((1ULL << PERF_REG_X86_XMM0) - 1)) + #endif /* _ASM_X86_PERF_REGS_H */ diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index ae938925442db..e27d517bd0fe4 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -240,6 +240,7 @@ struct perf_event; #define PERF_PMU_CAP_NO_INTERRUPT 0x01 #define PERF_PMU_CAP_NO_NMI 0x02 #define PERF_PMU_CAP_AUX_NO_SG 0x04 +#define PERF_PMU_CAP_EXTENDED_REGS 0x08 #define PERF_PMU_CAP_EXCLUSIVE 0x10 #define PERF_PMU_CAP_ITRACE 0x20 #define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x40 diff --git a/include/linux/perf_regs.h b/include/linux/perf_regs.h index 476747456bca5..2d12e97d5e7bc 100644 --- a/include/linux/perf_regs.h +++ b/include/linux/perf_regs.h @@ -11,6 +11,11 @@ struct perf_regs {

#ifdef CONFIG_HAVE_PERF_REGS #include <asm/perf_regs.h> + +#ifndef PERF_REG_EXTENDED_MASK +#define PERF_REG_EXTENDED_MASK 0 +#endif + u64 perf_reg_value(struct pt_regs *regs, int idx); int perf_reg_validate(u64 mask); u64 perf_reg_abi(struct task_struct *task); @@ -18,6 +23,9 @@ void perf_get_regs_user(struct perf_regs *regs_user, struct pt_regs *regs, struct pt_regs *regs_user_copy); #else + +#define PERF_REG_EXTENDED_MASK 0 + static inline u64 perf_reg_value(struct pt_regs *regs, int idx) { return 0; diff --git a/kernel/events/core.c b/kernel/events/core.c index df8df121e1f0a..667e0afe7ddd1 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -9878,6 +9878,12 @@ void perf_pmu_unregister(struct pmu *pmu) } EXPORT_SYMBOL_GPL(perf_pmu_unregister);

+static inline bool has_extended_regs(struct perf_event *event) +{ + return (event->attr.sample_regs_user & PERF_REG_EXTENDED_MASK) || + (event->attr.sample_regs_intr & PERF_REG_EXTENDED_MASK); +} + static int perf_try_init_event(struct pmu *pmu, struct perf_event *event) { struct perf_event_context *ctx = NULL; @@ -9909,12 +9915,16 @@ static int perf_try_init_event(struct pmu *pmu, struct perf_event *event) perf_event_ctx_unlock(event->group_leader, ctx);

if (!ret) { + if (!(pmu->capabilities & PERF_PMU_CAP_EXTENDED_REGS) && + has_extended_regs(event)) + ret = -EOPNOTSUPP; + if (pmu->capabilities & PERF_PMU_CAP_NO_EXCLUDE && - event_has_any_exclude_flag(event)) { - if (event->destroy) - event->destroy(event); + event_has_any_exclude_flag(event)) ret = -EINVAL; - } + + if (ret && event->destroy) + event->destroy(event); }

if (ret)

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.2-rc7 commit dce86ac75d772047e9bc606154704aa73bfd4c83 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit dce86ac75d772047e9bc606154704aa73bfd4c83 upstream.

Use generic macro PERF_REG_EXTENDED_MASK to replace PEBS_XMM_REGS to avoid duplication.

Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Arnaldo Carvalho de Melo acme@redhat.com Cc: Jiri Olsa jolsa@redhat.com Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Stephane Eranian eranian@google.com Cc: Thomas Gleixner tglx@linutronix.de Cc: Vince Weaver vincent.weaver@maine.edu Link: https://lkml.kernel.org/r/1559081314-9714-3-git-send-email-kan.liang@linux.i... Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/core.c | 4 ++-- arch/x86/events/intel/ds.c | 2 +- arch/x86/events/perf_event.h | 18 ------------------ 3 files changed, 3 insertions(+), 21 deletions(-)

diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c index 082fe2b988a1d..d940e7bc2d25e 100644 --- a/arch/x86/events/core.c +++ b/arch/x86/events/core.c @@ -566,13 +566,13 @@ int x86_pmu_hw_config(struct perf_event *event) }

/* sample_regs_user never support XMM registers */ - if (unlikely(event->attr.sample_regs_user & PEBS_XMM_REGS)) + if (unlikely(event->attr.sample_regs_user & PERF_REG_EXTENDED_MASK)) return -EINVAL; /* * Besides the general purpose registers, XMM registers may * be collected in PEBS on some platforms, e.g. Icelake */ - if (unlikely(event->attr.sample_regs_intr & PEBS_XMM_REGS)) { + if (unlikely(event->attr.sample_regs_intr & PERF_REG_EXTENDED_MASK)) { if (x86_pmu.pebs_no_xmm_regs) return -EINVAL;

diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index 8ee52e1a062e8..606728f3d188b 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -987,7 +987,7 @@ static u64 pebs_update_adaptive_cfg(struct perf_event *event) pebs_data_cfg |= PEBS_DATACFG_GP;

if ((sample_type & PERF_SAMPLE_REGS_INTR) && - (attr->sample_regs_intr & PEBS_XMM_REGS)) + (attr->sample_regs_intr & PERF_REG_EXTENDED_MASK)) pebs_data_cfg |= PEBS_DATACFG_XMMS;

if (sample_type & PERF_SAMPLE_BRANCH_STACK) { diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index ca41b57350d30..22d8568ecd4b8 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -123,24 +123,6 @@ struct amd_nb { (1ULL << PERF_REG_X86_R14) | \ (1ULL << PERF_REG_X86_R15))

-#define PEBS_XMM_REGS \ - ((1ULL << PERF_REG_X86_XMM0) | \ - (1ULL << PERF_REG_X86_XMM1) | \ - (1ULL << PERF_REG_X86_XMM2) | \ - (1ULL << PERF_REG_X86_XMM3) | \ - (1ULL << PERF_REG_X86_XMM4) | \ - (1ULL << PERF_REG_X86_XMM5) | \ - (1ULL << PERF_REG_X86_XMM6) | \ - (1ULL << PERF_REG_X86_XMM7) | \ - (1ULL << PERF_REG_X86_XMM8) | \ - (1ULL << PERF_REG_X86_XMM9) | \ - (1ULL << PERF_REG_X86_XMM10) | \ - (1ULL << PERF_REG_X86_XMM11) | \ - (1ULL << PERF_REG_X86_XMM12) | \ - (1ULL << PERF_REG_X86_XMM13) | \ - (1ULL << PERF_REG_X86_XMM14) | \ - (1ULL << PERF_REG_X86_XMM15)) - /* * Per register state. */

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.2-rc7 commit 8b12b812f5367c2469fb937da7e28dd321ad8d7b category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 8b12b812f5367c2469fb937da7e28dd321ad8d7b upstream.

Use the macro defined in kernel ABI header to replace the local name.

No functional change.

Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Alexander Shishkin alexander.shishkin@linux.intel.com Cc: Arnaldo Carvalho de Melo acme@redhat.com Cc: Jiri Olsa jolsa@redhat.com Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Stephane Eranian eranian@google.com Cc: Thomas Gleixner tglx@linutronix.de Cc: Vince Weaver vincent.weaver@maine.edu Link: https://lkml.kernel.org/r/1559081314-9714-5-git-send-email-kan.liang@linux.i... Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Shen, Xiaochen xiaochen.shen@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- tools/arch/x86/include/uapi/asm/perf_regs.h | 3 +++ tools/perf/arch/x86/include/perf_regs.h | 1 - tools/perf/arch/x86/util/perf_regs.c | 4 ++-- 3 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/tools/arch/x86/include/uapi/asm/perf_regs.h b/tools/arch/x86/include/uapi/asm/perf_regs.h index f3329cabce5c6..62a6965637fd7 100644 --- a/tools/arch/x86/include/uapi/asm/perf_regs.h +++ b/tools/arch/x86/include/uapi/asm/perf_regs.h @@ -31,4 +31,7 @@ enum perf_event_x86_regs { PERF_REG_X86_32_MAX = PERF_REG_X86_GS + 1, PERF_REG_X86_64_MAX = PERF_REG_X86_R15 + 1, }; + +#define PERF_REG_EXTENDED_MASK (~((1ULL << PERF_REG_X86_XMM0) - 1)) + #endif /* _ASM_X86_PERF_REGS_H */ diff --git a/tools/perf/arch/x86/include/perf_regs.h b/tools/perf/arch/x86/include/perf_regs.h index b7cd91a9014f9..b7321337d1003 100644 --- a/tools/perf/arch/x86/include/perf_regs.h +++ b/tools/perf/arch/x86/include/perf_regs.h @@ -9,7 +9,6 @@ void perf_regs_load(u64 *regs);

#define PERF_REGS_MAX PERF_REG_X86_XMM_MAX -#define PERF_XMM_REGS_MASK (~((1ULL << PERF_REG_X86_XMM0) - 1)) #ifndef HAVE_ARCH_X86_64_SUPPORT #define PERF_REGS_MASK ((1ULL << PERF_REG_X86_32_MAX) - 1) #define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_32 diff --git a/tools/perf/arch/x86/util/perf_regs.c b/tools/perf/arch/x86/util/perf_regs.c index 7886ca5263e31..3666c0076df94 100644 --- a/tools/perf/arch/x86/util/perf_regs.c +++ b/tools/perf/arch/x86/util/perf_regs.c @@ -277,7 +277,7 @@ uint64_t arch__intr_reg_mask(void) .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES, .sample_type = PERF_SAMPLE_REGS_INTR, - .sample_regs_intr = PERF_XMM_REGS_MASK, + .sample_regs_intr = PERF_REG_EXTENDED_MASK, .precise_ip = 1, .disabled = 1, .exclude_kernel = 1, @@ -293,7 +293,7 @@ uint64_t arch__intr_reg_mask(void) fd = sys_perf_event_open(&attr, 0, -1, -1, 0); if (fd != -1) { close(fd); - return (PERF_XMM_REGS_MASK | PERF_REGS_MASK); + return (PERF_REG_EXTENDED_MASK | PERF_REGS_MASK); }

return PERF_REGS_MASK;

From: Len Brown len.brown@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 7745f03eb39587dd15a1fb26e6223678b8e906d2 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 7745f03eb39587dd15a1fb26e6223678b8e906d2 upstream.

Some new systems have multiple software-visible die within each package.

Update Linux parsing of the Intel CPUID "Extended Topology Leaf" to handle either CPUID.B, or the new CPUID.1F.

Add cpuinfo_x86.die_id and cpuinfo_x86.max_dies to store the result.

die_id will be non-zero only for multi-die/package systems.

Signed-off-by: Len Brown len.brown@intel.com Signed-off-by: Thomas Gleixner tglx@linutronix.de Reviewed-by: Ingo Molnar mingo@kernel.org Acked-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: linux-doc@vger.kernel.org Link: https://lkml.kernel.org/r/7b23d2d26d717b8e14ba137c94b70943f1ae4b5c.155776931... Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- Documentation/x86/topology.txt | 4 ++ arch/x86/include/asm/processor.h | 4 +- arch/x86/kernel/cpu/topology.c | 85 +++++++++++++++++++++++++------- arch/x86/kernel/smpboot.c | 2 + 4 files changed, 75 insertions(+), 20 deletions(-)

diff --git a/Documentation/x86/topology.txt b/Documentation/x86/topology.txt index 2953e3ec9a025..e49487f51a25d 100644 --- a/Documentation/x86/topology.txt +++ b/Documentation/x86/topology.txt @@ -46,6 +46,10 @@ The topology of a system is described in the units of:

The number of cores in a package. This information is retrieved via CPUID.

+ - cpuinfo_x86.x86_max_dies: + + The number of dies in a package. This information is retrieved via CPUID. + - cpuinfo_x86.phys_proc_id:

The physical ID of the package. This information is retrieved via CPUID diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index 7a4b529881ff4..c147b2ad30299 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -117,7 +117,8 @@ struct cpuinfo_x86 { int x86_power; unsigned long loops_per_jiffy; /* cpuid returned max cores value: */ - u16 x86_max_cores; + u16 x86_max_cores; + u16 x86_max_dies; u16 apicid; u16 initial_apicid; u16 x86_clflush_size; @@ -129,6 +130,7 @@ struct cpuinfo_x86 { u16 logical_proc_id; /* Core id: */ u16 cpu_core_id; + u16 cpu_die_id; /* Index into per_cpu list: */ u16 cpu_index; u32 microcode; diff --git a/arch/x86/kernel/cpu/topology.c b/arch/x86/kernel/cpu/topology.c index 71ca064e37948..3f7141dd4ce88 100644 --- a/arch/x86/kernel/cpu/topology.c +++ b/arch/x86/kernel/cpu/topology.c @@ -13,33 +13,63 @@ /* leaf 0xb SMT level */ #define SMT_LEVEL 0

-/* leaf 0xb sub-leaf types */ +/* extended topology sub-leaf types */ #define INVALID_TYPE 0 #define SMT_TYPE 1 #define CORE_TYPE 2 +#define DIE_TYPE 5

#define LEAFB_SUBTYPE(ecx) (((ecx) >> 8) & 0xff) #define BITS_SHIFT_NEXT_LEVEL(eax) ((eax) & 0x1f) #define LEVEL_MAX_SIBLINGS(ebx) ((ebx) & 0xffff)

-int detect_extended_topology_early(struct cpuinfo_x86 *c) -{ #ifdef CONFIG_SMP +/* + * Check if given CPUID extended toplogy "leaf" is implemented + */ +static int check_extended_topology_leaf(int leaf) +{ unsigned int eax, ebx, ecx, edx;

- if (c->cpuid_level < 0xb) + cpuid_count(leaf, SMT_LEVEL, &eax, &ebx, &ecx, &edx); + + if (ebx == 0 || (LEAFB_SUBTYPE(ecx) != SMT_TYPE)) return -1;

- cpuid_count(0xb, SMT_LEVEL, &eax, &ebx, &ecx, &edx); + return 0; +} +/* + * Return best CPUID Extended Toplogy Leaf supported + */ +static int detect_extended_topology_leaf(struct cpuinfo_x86 *c) +{ + if (c->cpuid_level >= 0x1f) { + if (check_extended_topology_leaf(0x1f) == 0) + return 0x1f; + }

- /* - * check if the cpuid leaf 0xb is actually implemented. - */ - if (ebx == 0 || (LEAFB_SUBTYPE(ecx) != SMT_TYPE)) + if (c->cpuid_level >= 0xb) { + if (check_extended_topology_leaf(0xb) == 0) + return 0xb; + } + + return -1; +} +#endif + +int detect_extended_topology_early(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned int eax, ebx, ecx, edx; + int leaf; + + leaf = detect_extended_topology_leaf(c); + if (leaf < 0) return -1;

set_cpu_cap(c, X86_FEATURE_XTOPOLOGY);

+ cpuid_count(leaf, SMT_LEVEL, &eax, &ebx, &ecx, &edx); /* * initial apic id, which also represents 32-bit extended x2apic id. */ @@ -50,7 +80,7 @@ int detect_extended_topology_early(struct cpuinfo_x86 *c) }

/* - * Check for extended topology enumeration cpuid leaf 0xb and if it + * Check for extended topology enumeration cpuid leaf, and if it * exists, use it for populating initial_apicid and cpu topology * detection. */ @@ -58,22 +88,28 @@ int detect_extended_topology(struct cpuinfo_x86 *c) { #ifdef CONFIG_SMP unsigned int eax, ebx, ecx, edx, sub_index; - unsigned int ht_mask_width, core_plus_mask_width; + unsigned int ht_mask_width, core_plus_mask_width, die_plus_mask_width; unsigned int core_select_mask, core_level_siblings; + unsigned int die_select_mask, die_level_siblings; + int leaf;

- if (detect_extended_topology_early(c) < 0) + leaf = detect_extended_topology_leaf(c); + if (leaf < 0) return -1;

/* * Populate HT related information from sub-leaf level 0. */ - cpuid_count(0xb, SMT_LEVEL, &eax, &ebx, &ecx, &edx); + cpuid_count(leaf, SMT_LEVEL, &eax, &ebx, &ecx, &edx); + c->initial_apicid = edx; core_level_siblings = smp_num_siblings = LEVEL_MAX_SIBLINGS(ebx); core_plus_mask_width = ht_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + die_level_siblings = LEVEL_MAX_SIBLINGS(ebx); + die_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax);

sub_index = 1; do { - cpuid_count(0xb, sub_index, &eax, &ebx, &ecx, &edx); + cpuid_count(leaf, sub_index, &eax, &ebx, &ecx, &edx);

/* * Check for the Core type in the implemented sub leaves. @@ -81,23 +117,34 @@ int detect_extended_topology(struct cpuinfo_x86 *c) if (LEAFB_SUBTYPE(ecx) == CORE_TYPE) { core_level_siblings = LEVEL_MAX_SIBLINGS(ebx); core_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); - break; + die_level_siblings = core_level_siblings; + die_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + } + if (LEAFB_SUBTYPE(ecx) == DIE_TYPE) { + die_level_siblings = LEVEL_MAX_SIBLINGS(ebx); + die_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); }

sub_index++; } while (LEAFB_SUBTYPE(ecx) != INVALID_TYPE);

core_select_mask = (~(-1 << core_plus_mask_width)) >> ht_mask_width; - - c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, ht_mask_width) - & core_select_mask; - c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, core_plus_mask_width); + die_select_mask = (~(-1 << die_plus_mask_width)) >> + core_plus_mask_width; + + c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, + ht_mask_width) & core_select_mask; + c->cpu_die_id = apic->phys_pkg_id(c->initial_apicid, + core_plus_mask_width) & die_select_mask; + c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, + die_plus_mask_width); /* * Reinit the apicid, now that we have extended initial_apicid. */ c->apicid = apic->phys_pkg_id(c->initial_apicid, 0);

c->x86_max_cores = (core_level_siblings / smp_num_siblings); + c->x86_max_dies = (die_level_siblings / core_level_siblings); #endif return 0; } diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 4755a60bb0a66..42fc95021b998 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -401,6 +401,7 @@ static bool match_smt(struct cpuinfo_x86 *c, struct cpuinfo_x86 *o) int cpu1 = c->cpu_index, cpu2 = o->cpu_index;

if (c->phys_proc_id == o->phys_proc_id && + c->cpu_die_id == o->cpu_die_id && per_cpu(cpu_llc_id, cpu1) == per_cpu(cpu_llc_id, cpu2)) { if (c->cpu_core_id == o->cpu_core_id) return topology_sane(c, o, "smt"); @@ -412,6 +413,7 @@ static bool match_smt(struct cpuinfo_x86 *c, struct cpuinfo_x86 *o) }

} else if (c->phys_proc_id == o->phys_proc_id && + c->cpu_die_id == o->cpu_die_id && c->cpu_core_id == o->cpu_core_id) { return topology_sane(c, o, "smt"); }

From: Len Brown len.brown@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 0e344d8c709fe01d882fc0fb5452bedfe5eba67a category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 0e344d8c709fe01d882fc0fb5452bedfe5eba67a upstream.

Export die_id in cpu topology, for the benefit of hardware that has multiple-die/package.

Signed-off-by: Len Brown len.brown@intel.com Signed-off-by: Thomas Gleixner tglx@linutronix.de Reviewed-by: Ingo Molnar mingo@kernel.org Acked-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: linux-doc@vger.kernel.org Link: https://lkml.kernel.org/r/e7d1caaf4fbd24ee40db6d557ab28d7d83298900.155776931... Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- Documentation/cputopology.txt | 15 ++++++++++++--- drivers/base/topology.c | 4 ++++ include/linux/topology.h | 3 +++ 3 files changed, 19 insertions(+), 3 deletions(-)

diff --git a/Documentation/cputopology.txt b/Documentation/cputopology.txt index cb61277e23082..2ff8a1e9a2db0 100644 --- a/Documentation/cputopology.txt +++ b/Documentation/cputopology.txt @@ -12,6 +12,12 @@ physical_package_id: socket number, but the actual value is architecture and platform dependent.

+die_id: + + the CPU die ID of cpuX. Typically it is the hardware platform's + identifier (rather than the kernel's). The actual value is + architecture and platform dependent. + core_id:

the CPU core ID of cpuX. Typically it is the hardware platform's @@ -81,6 +87,7 @@ For an architecture to support this feature, it must define some of these macros in include/asm-XXX/topology.h::

#define topology_physical_package_id(cpu) + #define topology_die_id(cpu) #define topology_core_id(cpu) #define topology_book_id(cpu) #define topology_drawer_id(cpu) @@ -99,9 +106,11 @@ provides default definitions for any of the above macros that are not defined by include/asm-XXX/topology.h:

1) topology_physical_package_id: -1 -2) topology_core_id: 0 -3) topology_sibling_cpumask: just the given CPU -4) topology_core_cpumask: just the given CPU +2) topology_die_id: -1 +3) topology_core_id: 0 +4) topology_sibling_cpumask: just the given CPU +5) topology_core_cpumask: just the given CPU +6) topology_die_cpumask: just the given CPU

For architectures that don't support books (CONFIG_SCHED_BOOK) there are no default definitions for topology_book_id() and topology_book_cpumask(). diff --git a/drivers/base/topology.c b/drivers/base/topology.c index 5fd9f167ecc10..50352cf96f853 100644 --- a/drivers/base/topology.c +++ b/drivers/base/topology.c @@ -43,6 +43,9 @@ static ssize_t name##_list_show(struct device *dev, \ define_id_show_func(physical_package_id); static DEVICE_ATTR_RO(physical_package_id);

+define_id_show_func(die_id); +static DEVICE_ATTR_RO(die_id); + define_id_show_func(core_id); static DEVICE_ATTR_RO(core_id);

@@ -72,6 +75,7 @@ static DEVICE_ATTR_RO(drawer_siblings_list);

static struct attribute *default_attrs[] = { &dev_attr_physical_package_id.attr, + &dev_attr_die_id.attr, &dev_attr_core_id.attr, &dev_attr_thread_siblings.attr, &dev_attr_thread_siblings_list.attr, diff --git a/include/linux/topology.h b/include/linux/topology.h index cb0775e1ee4bd..5cc8595dd0e4e 100644 --- a/include/linux/topology.h +++ b/include/linux/topology.h @@ -184,6 +184,9 @@ static inline int cpu_to_mem(int cpu) #ifndef topology_physical_package_id #define topology_physical_package_id(cpu) ((void)(cpu), -1) #endif +#ifndef topology_die_id +#define topology_die_id(cpu) ((void)(cpu), -1) +#endif #ifndef topology_core_id #define topology_core_id(cpu) ((void)(cpu), 0) #endif

From: Len Brown len.brown@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 212bf4fdb7f9eeeb99afd97ebad677d43e7b55ac category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 212bf4fdb7f9eeeb99afd97ebad677d43e7b55ac upstream.

Define topology_logical_die_id() ala existing topology_logical_package_id()

Signed-off-by: Len Brown len.brown@intel.com Signed-off-by: Thomas Gleixner tglx@linutronix.de Tested-by: Zhang Rui rui.zhang@intel.com Reviewed-by: Ingo Molnar mingo@kernel.org Acked-by: Peter Zijlstra (Intel) peterz@infradead.org Link: https://lkml.kernel.org/r/2f3526e25ae14fbeff26fb26e877d159df8946d9.155776931... Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/include/asm/processor.h | 1 + arch/x86/include/asm/topology.h | 5 ++++ arch/x86/kernel/cpu/common.c | 1 + arch/x86/kernel/smpboot.c | 45 ++++++++++++++++++++++++++++++++ 4 files changed, 52 insertions(+)

diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index f6bfff5c5b214..123c4769c809b 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -130,6 +130,7 @@ struct cpuinfo_x86 { /* Core id: */ u16 cpu_core_id; u16 cpu_die_id; + u16 logical_die_id; /* Index into per_cpu list: */ u16 cpu_index; u32 microcode; diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index 3777dbe9c0ffd..9de16b4f60234 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -106,6 +106,7 @@ extern const struct cpumask *cpu_coregroup_mask(int cpu);

#define topology_logical_package_id(cpu) (cpu_data(cpu).logical_proc_id) #define topology_physical_package_id(cpu) (cpu_data(cpu).phys_proc_id) +#define topology_logical_die_id(cpu) (cpu_data(cpu).logical_die_id) #define topology_die_id(cpu) (cpu_data(cpu).cpu_die_id) #define topology_core_id(cpu) (cpu_data(cpu).cpu_core_id)

@@ -131,13 +132,17 @@ static inline int topology_max_smt_threads(void) }

int topology_update_package_map(unsigned int apicid, unsigned int cpu); +int topology_update_die_map(unsigned int dieid, unsigned int cpu); int topology_phys_to_logical_pkg(unsigned int pkg); +int topology_phys_to_logical_die(unsigned int die, unsigned int cpu); bool topology_is_primary_thread(unsigned int cpu); bool topology_smt_supported(void); #else #define topology_max_packages() (1) static inline int topology_update_package_map(unsigned int apicid, unsigned int cpu) { return 0; } +static inline int +topology_update_die_map(unsigned int dieid, unsigned int cpu) { return 0; } static inline int topology_phys_to_logical_pkg(unsigned int pkg) { return 0; } static inline int topology_phys_to_logical_die(unsigned int die, unsigned int cpu) { return 0; } diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 6e52e5a706e0b..f49aa9200db7e 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1391,6 +1391,7 @@ static void validate_apic_and_package_id(struct cpuinfo_x86 *c) cpu, apicid, c->initial_apicid); } BUG_ON(topology_update_package_map(c->phys_proc_id, cpu)); + BUG_ON(topology_update_die_map(c->cpu_die_id, cpu)); #else c->logical_proc_id = 0; #endif diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 42fc95021b998..e5aa5964a90d7 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -100,6 +100,7 @@ EXPORT_PER_CPU_SYMBOL(cpu_info); unsigned int __max_logical_packages __read_mostly; EXPORT_SYMBOL(__max_logical_packages); static unsigned int logical_packages __read_mostly; +static unsigned int logical_die __read_mostly;

/* Maximum number of SMT threads on any online core */ int __read_mostly __max_smt_threads = 1; @@ -314,6 +315,26 @@ int topology_phys_to_logical_pkg(unsigned int phys_pkg) return -1; } EXPORT_SYMBOL(topology_phys_to_logical_pkg); +/** + * topology_phys_to_logical_die - Map a physical die id to logical + * + * Returns logical die id or -1 if not found + */ +int topology_phys_to_logical_die(unsigned int die_id, unsigned int cur_cpu) +{ + int cpu; + int proc_id = cpu_data(cur_cpu).phys_proc_id; + + for_each_possible_cpu(cpu) { + struct cpuinfo_x86 *c = &cpu_data(cpu); + + if (c->initialized && c->cpu_die_id == die_id && + c->phys_proc_id == proc_id) + return c->logical_die_id; + } + return -1; +} +EXPORT_SYMBOL(topology_phys_to_logical_die);

/** * topology_update_package_map - Update the physical to logical package map @@ -338,6 +359,29 @@ int topology_update_package_map(unsigned int pkg, unsigned int cpu) cpu_data(cpu).logical_proc_id = new; return 0; } +/** + * topology_update_die_map - Update the physical to logical die map + * @die: The die id as retrieved via CPUID + * @cpu: The cpu for which this is updated + */ +int topology_update_die_map(unsigned int die, unsigned int cpu) +{ + int new; + + /* Already available somewhere? */ + new = topology_phys_to_logical_die(die, cpu); + if (new >= 0) + goto found; + + new = logical_die++; + if (new != die) { + pr_info("CPU %u Converting physical %u to logical die %u\n", + cpu, die, new); + } +found: + cpu_data(cpu).logical_die_id = new; + return 0; +}

void __init smp_store_boot_cpu_info(void) { @@ -347,6 +391,7 @@ void __init smp_store_boot_cpu_info(void) *c = boot_cpu_data; c->cpu_index = id; topology_update_package_map(c->phys_proc_id, id); + topology_update_die_map(c->cpu_die_id, id); c->initialized = true; }

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 32fb480e0a2cf1f71e4174d6477198c94dbc746c category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 32fb480e0a2cf1f71e4174d6477198c94dbc746c upstream.

RAPL "package" domains are actually implemented in hardware per-die. Thus, the new multi-die/package systems have mulitple domains within each physical package.

Update the intel_rapl driver to be "die aware" -- exporting multiple domains within a single package, when present. No change on single die/package systems.

Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Len Brown len.brown@intel.com Signed-off-by: Thomas Gleixner tglx@linutronix.de Reviewed-by: Ingo Molnar mingo@kernel.org Acked-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Acked-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: linux-pm@vger.kernel.org Link: https://lkml.kernel.org/r/9fcb4719aeb7efccf3bc75ed8dd559e46121649f.155776931... Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/powercap/intel_rapl.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/powercap/intel_rapl.c b/drivers/powercap/intel_rapl.c index 4d753445e28b8..55f75025d2470 100644 --- a/drivers/powercap/intel_rapl.c +++ b/drivers/powercap/intel_rapl.c @@ -266,7 +266,7 @@ static struct rapl_domain *platform_rapl_domain; /* Platform (PSys) domain */ /* caller to ensure CPU hotplug lock is held */ static struct rapl_package *rapl_find_package_domain(int cpu) { - int id = topology_physical_package_id(cpu); + int id = topology_logical_die_id(cpu); struct rapl_package *rp;

list_for_each_entry(rp, &rapl_packages, plist) { @@ -1464,7 +1464,7 @@ static void rapl_remove_package(struct rapl_package *rp) /* called from CPU hotplug notifier, hotplug lock held */ static struct rapl_package *rapl_add_package(int cpu) { - int id = topology_physical_package_id(cpu); + int id = topology_logical_die_id(cpu); struct rapl_package *rp; int ret;

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit f7c4e0c89bbd0f008b33d9dce02e207d9dea9f54 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit f7c4e0c89bbd0f008b33d9dce02e207d9dea9f54 upstream.

To support both MSR and MMIO Interface, use 'reg' to discribe RAPL registers instead of 'msr'.

Reviewed-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Tested-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/powercap/intel_rapl.c | 98 +++++++++++++++++------------------ 1 file changed, 49 insertions(+), 49 deletions(-)

diff --git a/drivers/powercap/intel_rapl.c b/drivers/powercap/intel_rapl.c index 0c861567d37bd..fb7d475604919 100644 --- a/drivers/powercap/intel_rapl.c +++ b/drivers/powercap/intel_rapl.c @@ -95,13 +95,13 @@ enum rapl_domain_type { RAPL_DOMAIN_MAX, };

-enum rapl_domain_msr_id { - RAPL_DOMAIN_MSR_LIMIT, - RAPL_DOMAIN_MSR_STATUS, - RAPL_DOMAIN_MSR_PERF, - RAPL_DOMAIN_MSR_POLICY, - RAPL_DOMAIN_MSR_INFO, - RAPL_DOMAIN_MSR_MAX, +enum rapl_domain_reg_id { + RAPL_DOMAIN_REG_LIMIT, + RAPL_DOMAIN_REG_STATUS, + RAPL_DOMAIN_REG_PERF, + RAPL_DOMAIN_REG_POLICY, + RAPL_DOMAIN_REG_INFO, + RAPL_DOMAIN_REG_MAX, };

/* per domain data, some are optional */ @@ -166,7 +166,7 @@ struct rapl_package; struct rapl_domain { const char *name; enum rapl_domain_type id; - int msrs[RAPL_DOMAIN_MSR_MAX]; + int regs[RAPL_DOMAIN_REG_MAX]; struct powercap_zone power_zone; struct rapl_domain_data rdd; struct rapl_power_limit rpl[NR_POWER_LIMITS]; @@ -228,7 +228,7 @@ struct rapl_primitive_info { const char *name; u64 mask; int shift; - enum rapl_domain_msr_id id; + enum rapl_domain_reg_id id; enum unit_type unit; u32 flag; }; @@ -654,11 +654,11 @@ static void rapl_init_domains(struct rapl_package *rp) case BIT(RAPL_DOMAIN_PACKAGE): rd->name = rapl_domain_names[RAPL_DOMAIN_PACKAGE]; rd->id = RAPL_DOMAIN_PACKAGE; - rd->msrs[0] = MSR_PKG_POWER_LIMIT; - rd->msrs[1] = MSR_PKG_ENERGY_STATUS; - rd->msrs[2] = MSR_PKG_PERF_STATUS; - rd->msrs[3] = 0; - rd->msrs[4] = MSR_PKG_POWER_INFO; + rd->regs[0] = MSR_PKG_POWER_LIMIT; + rd->regs[1] = MSR_PKG_ENERGY_STATUS; + rd->regs[2] = MSR_PKG_PERF_STATUS; + rd->regs[3] = 0; + rd->regs[4] = MSR_PKG_POWER_INFO; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->rpl[1].prim_id = PL2_ENABLE; @@ -667,33 +667,33 @@ static void rapl_init_domains(struct rapl_package *rp) case BIT(RAPL_DOMAIN_PP0): rd->name = rapl_domain_names[RAPL_DOMAIN_PP0]; rd->id = RAPL_DOMAIN_PP0; - rd->msrs[0] = MSR_PP0_POWER_LIMIT; - rd->msrs[1] = MSR_PP0_ENERGY_STATUS; - rd->msrs[2] = 0; - rd->msrs[3] = MSR_PP0_POLICY; - rd->msrs[4] = 0; + rd->regs[0] = MSR_PP0_POWER_LIMIT; + rd->regs[1] = MSR_PP0_ENERGY_STATUS; + rd->regs[2] = 0; + rd->regs[3] = MSR_PP0_POLICY; + rd->regs[4] = 0; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; break; case BIT(RAPL_DOMAIN_PP1): rd->name = rapl_domain_names[RAPL_DOMAIN_PP1]; rd->id = RAPL_DOMAIN_PP1; - rd->msrs[0] = MSR_PP1_POWER_LIMIT; - rd->msrs[1] = MSR_PP1_ENERGY_STATUS; - rd->msrs[2] = 0; - rd->msrs[3] = MSR_PP1_POLICY; - rd->msrs[4] = 0; + rd->regs[0] = MSR_PP1_POWER_LIMIT; + rd->regs[1] = MSR_PP1_ENERGY_STATUS; + rd->regs[2] = 0; + rd->regs[3] = MSR_PP1_POLICY; + rd->regs[4] = 0; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; break; case BIT(RAPL_DOMAIN_DRAM): rd->name = rapl_domain_names[RAPL_DOMAIN_DRAM]; rd->id = RAPL_DOMAIN_DRAM; - rd->msrs[0] = MSR_DRAM_POWER_LIMIT; - rd->msrs[1] = MSR_DRAM_ENERGY_STATUS; - rd->msrs[2] = MSR_DRAM_PERF_STATUS; - rd->msrs[3] = 0; - rd->msrs[4] = MSR_DRAM_POWER_INFO; + rd->regs[0] = MSR_DRAM_POWER_LIMIT; + rd->regs[1] = MSR_DRAM_ENERGY_STATUS; + rd->regs[2] = MSR_DRAM_PERF_STATUS; + rd->regs[3] = 0; + rd->regs[4] = MSR_DRAM_POWER_INFO; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->domain_energy_unit = @@ -748,37 +748,37 @@ static u64 rapl_unit_xlate(struct rapl_domain *rd, enum unit_type type, static struct rapl_primitive_info rpi[] = { /* name, mask, shift, msr index, unit divisor */ PRIMITIVE_INFO_INIT(ENERGY_COUNTER, ENERGY_STATUS_MASK, 0, - RAPL_DOMAIN_MSR_STATUS, ENERGY_UNIT, 0), + RAPL_DOMAIN_REG_STATUS, ENERGY_UNIT, 0), PRIMITIVE_INFO_INIT(POWER_LIMIT1, POWER_LIMIT1_MASK, 0, - RAPL_DOMAIN_MSR_LIMIT, POWER_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(POWER_LIMIT2, POWER_LIMIT2_MASK, 32, - RAPL_DOMAIN_MSR_LIMIT, POWER_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(FW_LOCK, POWER_PP_LOCK, 31, - RAPL_DOMAIN_MSR_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL1_ENABLE, POWER_LIMIT1_ENABLE, 15, - RAPL_DOMAIN_MSR_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL1_CLAMP, POWER_LIMIT1_CLAMP, 16, - RAPL_DOMAIN_MSR_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL2_ENABLE, POWER_LIMIT2_ENABLE, 47, - RAPL_DOMAIN_MSR_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL2_CLAMP, POWER_LIMIT2_CLAMP, 48, - RAPL_DOMAIN_MSR_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(TIME_WINDOW1, TIME_WINDOW1_MASK, 17, - RAPL_DOMAIN_MSR_LIMIT, TIME_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, TIME_UNIT, 0), PRIMITIVE_INFO_INIT(TIME_WINDOW2, TIME_WINDOW2_MASK, 49, - RAPL_DOMAIN_MSR_LIMIT, TIME_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, TIME_UNIT, 0), PRIMITIVE_INFO_INIT(THERMAL_SPEC_POWER, POWER_INFO_THERMAL_SPEC_MASK, - 0, RAPL_DOMAIN_MSR_INFO, POWER_UNIT, 0), + 0, RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(MAX_POWER, POWER_INFO_MAX_MASK, 32, - RAPL_DOMAIN_MSR_INFO, POWER_UNIT, 0), + RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(MIN_POWER, POWER_INFO_MIN_MASK, 16, - RAPL_DOMAIN_MSR_INFO, POWER_UNIT, 0), + RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(MAX_TIME_WINDOW, POWER_INFO_MAX_TIME_WIN_MASK, 48, - RAPL_DOMAIN_MSR_INFO, TIME_UNIT, 0), + RAPL_DOMAIN_REG_INFO, TIME_UNIT, 0), PRIMITIVE_INFO_INIT(THROTTLED_TIME, PERF_STATUS_THROTTLE_TIME_MASK, 0, - RAPL_DOMAIN_MSR_PERF, TIME_UNIT, 0), + RAPL_DOMAIN_REG_PERF, TIME_UNIT, 0), PRIMITIVE_INFO_INIT(PRIORITY_LEVEL, PP_POLICY_MASK, 0, - RAPL_DOMAIN_MSR_POLICY, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_POLICY, ARBITRARY_UNIT, 0), /* non-hardware */ PRIMITIVE_INFO_INIT(AVERAGE_POWER, 0, 0, 0, POWER_UNIT, RAPL_PRIMITIVE_DERIVED), @@ -810,7 +810,7 @@ static int rapl_read_data_raw(struct rapl_domain *rd, if (!rp->name || rp->flag & RAPL_PRIMITIVE_DUMMY) return -EINVAL;

- msr = rd->msrs[rp->id]; + msr = rd->regs[rp->id]; if (!msr) return -EINVAL;

@@ -886,7 +886,7 @@ static int rapl_write_data_raw(struct rapl_domain *rd,

memset(&ma, 0, sizeof(ma));

- ma.msr_no = rd->msrs[rp->id]; + ma.msr_no = rd->regs[rp->id]; ma.clear_mask = rp->mask; ma.set_mask = bits;

@@ -1299,8 +1299,8 @@ static int __init rapl_register_psys(void)

rd->name = rapl_domain_names[RAPL_DOMAIN_PLATFORM]; rd->id = RAPL_DOMAIN_PLATFORM; - rd->msrs[0] = MSR_PLATFORM_POWER_LIMIT; - rd->msrs[1] = MSR_PLATFORM_ENERGY_STATUS; + rd->regs[0] = MSR_PLATFORM_POWER_LIMIT; + rd->regs[1] = MSR_PLATFORM_ENERGY_STATUS; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->rpl[1].prim_id = PL2_ENABLE;

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit ff956826a403f5cf189978d5ff6b3eb53aa11610 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit ff956826a403f5cf189978d5ff6b3eb53aa11610 upstream.

Create a new header file for the common definitions that might be used by different RAPL Interface.

Reviewed-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Tested-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- MAINTAINERS | 1 + drivers/powercap/intel_rapl.c | 101 +----------------------------- include/linux/intel_rapl.h | 113 ++++++++++++++++++++++++++++++++++ 3 files changed, 116 insertions(+), 99 deletions(-) create mode 100644 include/linux/intel_rapl.h

diff --git a/MAINTAINERS b/MAINTAINERS index 6634d287526c5..674616dc5143c 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -11702,6 +11702,7 @@ F: drivers/base/power/ F: include/linux/pm.h F: include/linux/pm_* F: include/linux/powercap.h +F: include/linux/intel_rapl.h F: drivers/powercap/ F: kernel/configs/nopm.config

diff --git a/drivers/powercap/intel_rapl.c b/drivers/powercap/intel_rapl.c index 6cb123678e1cf..a5e4f6b97621b 100644 --- a/drivers/powercap/intel_rapl.c +++ b/drivers/powercap/intel_rapl.c @@ -30,8 +30,9 @@ #include <linux/cpu.h> #include <linux/powercap.h> #include <linux/suspend.h> -#include <asm/iosf_mbi.h> +#include <linux/intel_rapl.h>

+#include <asm/iosf_mbi.h> #include <asm/processor.h> #include <asm/cpu_device_id.h> #include <asm/intel-family.h> @@ -86,59 +87,9 @@ enum unit_type { TIME_UNIT, };

-enum rapl_domain_type { - RAPL_DOMAIN_PACKAGE, /* entire package/socket */ - RAPL_DOMAIN_PP0, /* core power plane */ - RAPL_DOMAIN_PP1, /* graphics uncore */ - RAPL_DOMAIN_DRAM,/* DRAM control_type */ - RAPL_DOMAIN_PLATFORM, /* PSys control_type */ - RAPL_DOMAIN_MAX, -}; - -enum rapl_domain_reg_id { - RAPL_DOMAIN_REG_LIMIT, - RAPL_DOMAIN_REG_STATUS, - RAPL_DOMAIN_REG_PERF, - RAPL_DOMAIN_REG_POLICY, - RAPL_DOMAIN_REG_INFO, - RAPL_DOMAIN_REG_MAX, -}; - /* per domain data, some are optional */ -enum rapl_primitives { - ENERGY_COUNTER, - POWER_LIMIT1, - POWER_LIMIT2, - FW_LOCK, - - PL1_ENABLE, /* power limit 1, aka long term */ - PL1_CLAMP, /* allow frequency to go below OS request */ - PL2_ENABLE, /* power limit 2, aka short term, instantaneous */ - PL2_CLAMP, - - TIME_WINDOW1, /* long term */ - TIME_WINDOW2, /* short term */ - THERMAL_SPEC_POWER, - MAX_POWER, - - MIN_POWER, - MAX_TIME_WINDOW, - THROTTLED_TIME, - PRIORITY_LEVEL, - - /* below are not raw primitive data */ - AVERAGE_POWER, - NR_RAPL_PRIMITIVES, -}; - #define NR_RAW_PRIMITIVES (NR_RAPL_PRIMITIVES - 2)

-/* Can be expanded to include events, etc.*/ -struct rapl_domain_data { - u64 primitives[NR_RAPL_PRIMITIVES]; - unsigned long timestamp; -}; - struct msrl_action { u32 msr_no; u64 clear_mask; @@ -150,60 +101,12 @@ struct msrl_action { #define DOMAIN_STATE_POWER_LIMIT_SET BIT(1) #define DOMAIN_STATE_BIOS_LOCKED BIT(2)

-#define NR_POWER_LIMITS (2) -struct rapl_power_limit { - struct powercap_zone_constraint *constraint; - int prim_id; /* primitive ID used to enable */ - struct rapl_domain *domain; - const char *name; - u64 last_power_limit; -}; - static const char pl1_name[] = "long_term"; static const char pl2_name[] = "short_term";

-struct rapl_package; -struct rapl_domain { - const char *name; - enum rapl_domain_type id; - int regs[RAPL_DOMAIN_REG_MAX]; - struct powercap_zone power_zone; - struct rapl_domain_data rdd; - struct rapl_power_limit rpl[NR_POWER_LIMITS]; - u64 attr_map; /* track capabilities */ - unsigned int state; - unsigned int domain_energy_unit; - struct rapl_package *rp; -}; #define power_zone_to_rapl_domain(_zone) \ container_of(_zone, struct rapl_domain, power_zone)

-/* maximum rapl package domain name: package-%d-die-%d */ -#define PACKAGE_DOMAIN_NAME_LENGTH 30 - - -/* Each rapl package contains multiple domains, these are the common - * data across RAPL domains within a package. - */ -struct rapl_package { - unsigned int id; /* logical die id, equals physical 1-die systems */ - unsigned int nr_domains; - unsigned long domain_map; /* bit map of active domains */ - unsigned int power_unit; - unsigned int energy_unit; - unsigned int time_unit; - struct rapl_domain *domains; /* array of domains, sized at runtime */ - struct powercap_zone *power_zone; /* keep track of parent zone */ - unsigned long power_limit_irq; /* keep track of package power limit - * notify interrupt enable status. - */ - struct list_head plist; - int lead_cpu; /* one active cpu per package for access */ - /* Track active cpus */ - struct cpumask cpumask; - char name[PACKAGE_DOMAIN_NAME_LENGTH]; -}; - struct rapl_defaults { u8 floor_freq_reg_addr; int (*check_unit)(struct rapl_package *rp, int cpu); diff --git a/include/linux/intel_rapl.h b/include/linux/intel_rapl.h new file mode 100644 index 0000000000000..94716036d8294 --- /dev/null +++ b/include/linux/intel_rapl.h @@ -0,0 +1,113 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Data types and headers for RAPL support + * + * Copyright (C) 2019 Intel Corporation. + * + * Author: Zhang Rui rui.zhang@intel.com + */ + +#ifndef __INTEL_RAPL_H__ +#define __INTEL_RAPL_H__ + +#include <linux/types.h> +#include <linux/powercap.h> + +enum rapl_domain_type { + RAPL_DOMAIN_PACKAGE, /* entire package/socket */ + RAPL_DOMAIN_PP0, /* core power plane */ + RAPL_DOMAIN_PP1, /* graphics uncore */ + RAPL_DOMAIN_DRAM, /* DRAM control_type */ + RAPL_DOMAIN_PLATFORM, /* PSys control_type */ + RAPL_DOMAIN_MAX, +}; + +enum rapl_domain_reg_id { + RAPL_DOMAIN_REG_LIMIT, + RAPL_DOMAIN_REG_STATUS, + RAPL_DOMAIN_REG_PERF, + RAPL_DOMAIN_REG_POLICY, + RAPL_DOMAIN_REG_INFO, + RAPL_DOMAIN_REG_MAX, +}; + +struct rapl_package; + +enum rapl_primitives { + ENERGY_COUNTER, + POWER_LIMIT1, + POWER_LIMIT2, + FW_LOCK, + + PL1_ENABLE, /* power limit 1, aka long term */ + PL1_CLAMP, /* allow frequency to go below OS request */ + PL2_ENABLE, /* power limit 2, aka short term, instantaneous */ + PL2_CLAMP, + + TIME_WINDOW1, /* long term */ + TIME_WINDOW2, /* short term */ + THERMAL_SPEC_POWER, + MAX_POWER, + + MIN_POWER, + MAX_TIME_WINDOW, + THROTTLED_TIME, + PRIORITY_LEVEL, + + /* below are not raw primitive data */ + AVERAGE_POWER, + NR_RAPL_PRIMITIVES, +}; + +struct rapl_domain_data { + u64 primitives[NR_RAPL_PRIMITIVES]; + unsigned long timestamp; +}; + +#define NR_POWER_LIMITS (2) +struct rapl_power_limit { + struct powercap_zone_constraint *constraint; + int prim_id; /* primitive ID used to enable */ + struct rapl_domain *domain; + const char *name; + u64 last_power_limit; +}; + +struct rapl_package; + +struct rapl_domain { + const char *name; + enum rapl_domain_type id; + int regs[RAPL_DOMAIN_REG_MAX]; + struct powercap_zone power_zone; + struct rapl_domain_data rdd; + struct rapl_power_limit rpl[NR_POWER_LIMITS]; + u64 attr_map; /* track capabilities */ + unsigned int state; + unsigned int domain_energy_unit; + struct rapl_package *rp; +}; + +/* maximum rapl package domain name: package-%d-die-%d */ +#define PACKAGE_DOMAIN_NAME_LENGTH 30 + +struct rapl_package { + unsigned int id; /* logical die id, equals physical 1-die systems */ + unsigned int nr_domains; + unsigned long domain_map; /* bit map of active domains */ + unsigned int power_unit; + unsigned int energy_unit; + unsigned int time_unit; + struct rapl_domain *domains; /* array of domains, sized at runtime */ + struct powercap_zone *power_zone; /* keep track of parent zone */ + unsigned long power_limit_irq; /* keep track of package power limit + * notify interrupt enable status. + */ + struct list_head plist; + int lead_cpu; /* one active cpu per package for access */ + /* Track active cpus */ + struct cpumask cpumask; + char name[PACKAGE_DOMAIN_NAME_LENGTH]; +}; + +#endif /* __INTEL_RAPL_H__ */

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 7fde2712a7adab721eaabafbd8ff93dff3262d35 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 7fde2712a7adab721eaabafbd8ff93dff3262d35 upstream.

MSR and MMIO RAPL interface have different sets of registers, thus the RAPL register address should be obtained from interface specific structure, i.e. struct rapl_if_private, instead.

Reviewed-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Tested-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/powercap/intel_rapl.c | 73 ++++++++++++++++------------------- include/linux/intel_rapl.h | 4 ++ 2 files changed, 37 insertions(+), 40 deletions(-)

diff --git a/drivers/powercap/intel_rapl.c b/drivers/powercap/intel_rapl.c index 26aa43c9a1a04..a8f1982e2fdad 100644 --- a/drivers/powercap/intel_rapl.c +++ b/drivers/powercap/intel_rapl.c @@ -88,7 +88,19 @@ enum unit_type { };

/* private data for RAPL MSR Interface */ -static struct rapl_if_priv rapl_msr_priv; +static struct rapl_if_priv rapl_msr_priv = { + .reg_unit = MSR_RAPL_POWER_UNIT, + .regs[RAPL_DOMAIN_PACKAGE] = { + MSR_PKG_POWER_LIMIT, MSR_PKG_ENERGY_STATUS, MSR_PKG_PERF_STATUS, 0, MSR_PKG_POWER_INFO }, + .regs[RAPL_DOMAIN_PP0] = { + MSR_PP0_POWER_LIMIT, MSR_PP0_ENERGY_STATUS, 0, MSR_PP0_POLICY, 0 }, + .regs[RAPL_DOMAIN_PP1] = { + MSR_PP1_POWER_LIMIT, MSR_PP1_ENERGY_STATUS, 0, MSR_PP1_POLICY, 0 }, + .regs[RAPL_DOMAIN_DRAM] = { + MSR_DRAM_POWER_LIMIT, MSR_DRAM_ENERGY_STATUS, MSR_DRAM_PERF_STATUS, 0, MSR_DRAM_POWER_INFO }, + .regs[RAPL_DOMAIN_PLATFORM] = { + MSR_PLATFORM_POWER_LIMIT, MSR_PLATFORM_ENERGY_STATUS, 0, 0, 0}, +};

/* per domain data, some are optional */ #define NR_RAW_PRIMITIVES (NR_RAPL_PRIMITIVES - 2) @@ -553,15 +565,17 @@ static void rapl_init_domains(struct rapl_package *rp)

for (i = 0; i < RAPL_DOMAIN_MAX; i++) { unsigned int mask = rp->domain_map & (1 << i); + + rd->regs[RAPL_DOMAIN_REG_LIMIT] = rp->priv->regs[i][RAPL_DOMAIN_REG_LIMIT]; + rd->regs[RAPL_DOMAIN_REG_STATUS] = rp->priv->regs[i][RAPL_DOMAIN_REG_STATUS]; + rd->regs[RAPL_DOMAIN_REG_PERF] = rp->priv->regs[i][RAPL_DOMAIN_REG_PERF]; + rd->regs[RAPL_DOMAIN_REG_POLICY] = rp->priv->regs[i][RAPL_DOMAIN_REG_POLICY]; + rd->regs[RAPL_DOMAIN_REG_INFO] = rp->priv->regs[i][RAPL_DOMAIN_REG_INFO]; + switch (mask) { case BIT(RAPL_DOMAIN_PACKAGE): rd->name = rapl_domain_names[RAPL_DOMAIN_PACKAGE]; rd->id = RAPL_DOMAIN_PACKAGE; - rd->regs[RAPL_DOMAIN_REG_LIMIT] = MSR_PKG_POWER_LIMIT; - rd->regs[RAPL_DOMAIN_REG_STATUS] = MSR_PKG_ENERGY_STATUS; - rd->regs[RAPL_DOMAIN_REG_PERF] = MSR_PKG_PERF_STATUS; - rd->regs[RAPL_DOMAIN_REG_POLICY] = 0; - rd->regs[RAPL_DOMAIN_REG_INFO] = MSR_PKG_POWER_INFO; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->rpl[1].prim_id = PL2_ENABLE; @@ -570,33 +584,18 @@ static void rapl_init_domains(struct rapl_package *rp) case BIT(RAPL_DOMAIN_PP0): rd->name = rapl_domain_names[RAPL_DOMAIN_PP0]; rd->id = RAPL_DOMAIN_PP0; - rd->regs[RAPL_DOMAIN_REG_LIMIT] = MSR_PP0_POWER_LIMIT; - rd->regs[RAPL_DOMAIN_REG_STATUS] = MSR_PP0_ENERGY_STATUS; - rd->regs[RAPL_DOMAIN_REG_PERF] = 0; - rd->regs[RAPL_DOMAIN_REG_POLICY] = MSR_PP0_POLICY; - rd->regs[RAPL_DOMAIN_REG_INFO] = 0; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; break; case BIT(RAPL_DOMAIN_PP1): rd->name = rapl_domain_names[RAPL_DOMAIN_PP1]; rd->id = RAPL_DOMAIN_PP1; - rd->regs[RAPL_DOMAIN_REG_LIMIT] = MSR_PP1_POWER_LIMIT; - rd->regs[RAPL_DOMAIN_REG_STATUS] = MSR_PP1_ENERGY_STATUS; - rd->regs[RAPL_DOMAIN_REG_PERF] = 0; - rd->regs[RAPL_DOMAIN_REG_POLICY] = MSR_PP1_POLICY; - rd->regs[RAPL_DOMAIN_REG_INFO] = 0; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; break; case BIT(RAPL_DOMAIN_DRAM): rd->name = rapl_domain_names[RAPL_DOMAIN_DRAM]; rd->id = RAPL_DOMAIN_DRAM; - rd->regs[RAPL_DOMAIN_REG_LIMIT] = MSR_DRAM_POWER_LIMIT; - rd->regs[RAPL_DOMAIN_REG_STATUS] = MSR_DRAM_ENERGY_STATUS; - rd->regs[RAPL_DOMAIN_REG_PERF] = MSR_DRAM_PERF_STATUS; - rd->regs[RAPL_DOMAIN_REG_POLICY] = 0; - rd->regs[RAPL_DOMAIN_REG_INFO] = MSR_DRAM_POWER_INFO; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->domain_energy_unit = @@ -818,9 +817,9 @@ static int rapl_check_unit_core(struct rapl_package *rp, int cpu) u64 msr_val; u32 value;

- if (rdmsrl_safe_on_cpu(cpu, MSR_RAPL_POWER_UNIT, &msr_val)) { + if (rdmsrl_safe_on_cpu(cpu, rp->priv->reg_unit, &msr_val)) { pr_err("Failed to read power unit MSR 0x%x on CPU %d, exit.\n", - MSR_RAPL_POWER_UNIT, cpu); + rp->priv->reg_unit, cpu); return -ENODEV; }

@@ -844,9 +843,9 @@ static int rapl_check_unit_atom(struct rapl_package *rp, int cpu) u64 msr_val; u32 value;

- if (rdmsrl_safe_on_cpu(cpu, MSR_RAPL_POWER_UNIT, &msr_val)) { + if (rdmsrl_safe_on_cpu(cpu, rp->priv->reg_unit, &msr_val)) { pr_err("Failed to read power unit MSR 0x%x on CPU %d, exit.\n", - MSR_RAPL_POWER_UNIT, cpu); + rp->priv->reg_unit, cpu); return -ENODEV; } value = (msr_val & ENERGY_UNIT_MASK) >> ENERGY_UNIT_OFFSET; @@ -1190,10 +1189,10 @@ static int __init rapl_register_psys(void) struct powercap_zone *power_zone; u64 val;

- if (rdmsrl_safe_on_cpu(0, MSR_PLATFORM_ENERGY_STATUS, &val) || !val) + if (rdmsrl_safe_on_cpu(0, rapl_msr_priv.regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS], &val) || !val) return -ENODEV;

- if (rdmsrl_safe_on_cpu(0, MSR_PLATFORM_POWER_LIMIT, &val) || !val) + if (rdmsrl_safe_on_cpu(0, rapl_msr_priv.regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT], &val) || !val) return -ENODEV;

rd = kzalloc(sizeof(*rd), GFP_KERNEL); @@ -1202,8 +1201,8 @@ static int __init rapl_register_psys(void)

rd->name = rapl_domain_names[RAPL_DOMAIN_PLATFORM]; rd->id = RAPL_DOMAIN_PLATFORM; - rd->regs[RAPL_DOMAIN_REG_LIMIT] = MSR_PLATFORM_POWER_LIMIT; - rd->regs[RAPL_DOMAIN_REG_STATUS] = MSR_PLATFORM_ENERGY_STATUS; + rd->regs[RAPL_DOMAIN_REG_LIMIT] = rapl_msr_priv.regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT]; + rd->regs[RAPL_DOMAIN_REG_STATUS] = rapl_msr_priv.regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS]; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->rpl[1].prim_id = PL2_ENABLE; @@ -1235,23 +1234,17 @@ static int __init rapl_register_powercap(void) return 0; }

-static int rapl_check_domain(int cpu, int domain) +static int rapl_check_domain(int cpu, int domain, struct rapl_package *rp) { - unsigned msr; + u32 reg; u64 val = 0;

switch (domain) { case RAPL_DOMAIN_PACKAGE: - msr = MSR_PKG_ENERGY_STATUS; - break; case RAPL_DOMAIN_PP0: - msr = MSR_PP0_ENERGY_STATUS; - break; case RAPL_DOMAIN_PP1: - msr = MSR_PP1_ENERGY_STATUS; - break; case RAPL_DOMAIN_DRAM: - msr = MSR_DRAM_ENERGY_STATUS; + reg = rp->priv->regs[domain][RAPL_DOMAIN_REG_STATUS]; break; case RAPL_DOMAIN_PLATFORM: /* PSYS(PLATFORM) is not a CPU domain, so avoid printng error */ @@ -1263,7 +1256,7 @@ static int rapl_check_domain(int cpu, int domain) /* make sure domain counters are available and contains non-zero * values, otherwise skip it. */ - if (rdmsrl_safe_on_cpu(cpu, msr, &val) || !val) + if (rdmsrl_safe_on_cpu(cpu, reg, &val) || !val) return -ENODEV;

return 0; @@ -1310,7 +1303,7 @@ static int rapl_detect_domains(struct rapl_package *rp, int cpu)

for (i = 0; i < RAPL_DOMAIN_MAX; i++) { /* use physical package id to read counters */ - if (!rapl_check_domain(cpu, i)) { + if (!rapl_check_domain(cpu, i, rp)) { rp->domain_map |= 1 << i; pr_info("Found RAPL domain %s\n", rapl_domain_names[i]); } diff --git a/include/linux/intel_rapl.h b/include/linux/intel_rapl.h index 7bf1683e4a63f..ec2c9e83274fe 100644 --- a/include/linux/intel_rapl.h +++ b/include/linux/intel_rapl.h @@ -95,11 +95,15 @@ struct rapl_domain { * @platform_rapl_domain: Optional. Some RAPL interface may have platform * level RAPL control. * @pcap_rapl_online: CPU hotplug state for each RAPL interface. + * @reg_unit: Register for getting energy/power/time unit. + * @regs: Register sets for different RAPL Domains. */ struct rapl_if_priv { struct powercap_control_type *control_type; struct rapl_domain *platform_rapl_domain; enum cpuhp_state pcap_rapl_online; + u32 reg_unit; + u32 regs[RAPL_DOMAIN_MAX][RAPL_DOMAIN_REG_MAX]; };

/* maximum rapl package domain name: package-%d-die-%d */

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 8a00676cd690941c9a18bd390c3b2cade631c516 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 8a00676cd690941c9a18bd390c3b2cade631c516 upstream.

Previously, there are three functions: rapl_register_psys(), which registers platform rapl domain. rapl_register_powercap(), which registers powercap control type. rapl_unregsiter_powercap(), which unregisters platform rapl domain and powercap control type.

This is confusing as the function name does not describe what it does clearly.

With this patch, the three functions are removed, and two new functions rapl_register_platform_domain()/rapl_unregister_platform_domain() are introduced instead, and they do exactly what their function name describes.

Plus, as part of the common code, hardcoded MSR accesses in these functions are converted to follow the abstracted register access.

Reviewed-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Tested-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/powercap/intel_rapl.c | 62 +++++++++++++++++------------------ 1 file changed, 31 insertions(+), 31 deletions(-)

diff --git a/drivers/powercap/intel_rapl.c b/drivers/powercap/intel_rapl.c index eb0b8ffa24523..357ea9657ce15 100644 --- a/drivers/powercap/intel_rapl.c +++ b/drivers/powercap/intel_rapl.c @@ -1069,16 +1069,6 @@ static void rapl_update_domain_data(struct rapl_package *rp)

}

-static void rapl_unregister_powercap(void) -{ - if (&rapl_msr_priv.platform_rapl_domain) { - powercap_unregister_zone(rapl_msr_priv.control_type, - &rapl_msr_priv.platform_rapl_domain->power_zone); - kfree(rapl_msr_priv.platform_rapl_domain); - } - powercap_unregister_control_type(rapl_msr_priv.control_type); -} - static int rapl_package_register_powercap(struct rapl_package *rp) { struct rapl_domain *rd; @@ -1148,16 +1138,23 @@ static int rapl_package_register_powercap(struct rapl_package *rp) return ret; }

-static int __init rapl_register_psys(void) +static int __init rapl_add_platform_domain(struct rapl_if_priv *priv) { struct rapl_domain *rd; struct powercap_zone *power_zone; - u64 val; + struct reg_action ra; + int ret;

- if (rdmsrl_safe_on_cpu(0, rapl_msr_priv.regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS], &val) || !val) + ra.reg = priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS]; + ra.mask = ~0; + ret = priv->read_raw(0, &ra); + if (ret || !ra.value) return -ENODEV;

- if (rdmsrl_safe_on_cpu(0, rapl_msr_priv.regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT], &val) || !val) + ra.reg = priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT]; + ra.mask = ~0; + ret = priv->read_raw(0, &ra); + if (ret || !ra.value) return -ENODEV;

rd = kzalloc(sizeof(*rd), GFP_KERNEL); @@ -1166,15 +1163,15 @@ static int __init rapl_register_psys(void)

rd->name = rapl_domain_names[RAPL_DOMAIN_PLATFORM]; rd->id = RAPL_DOMAIN_PLATFORM; - rd->regs[RAPL_DOMAIN_REG_LIMIT] = rapl_msr_priv.regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT]; - rd->regs[RAPL_DOMAIN_REG_STATUS] = rapl_msr_priv.regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS]; + rd->regs[RAPL_DOMAIN_REG_LIMIT] = priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT]; + rd->regs[RAPL_DOMAIN_REG_STATUS] = priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS]; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->rpl[1].prim_id = PL2_ENABLE; rd->rpl[1].name = pl2_name; - rd->rp = rapl_find_package_domain(0, &rapl_msr_priv); + rd->rp = rapl_find_package_domain(0, priv);

- power_zone = powercap_register_zone(&rd->power_zone, rapl_msr_priv.control_type, + power_zone = powercap_register_zone(&rd->power_zone, priv->control_type, "psys", NULL, &zone_ops[RAPL_DOMAIN_PLATFORM], 2, &constraint_ops); @@ -1184,19 +1181,18 @@ static int __init rapl_register_psys(void) return PTR_ERR(power_zone); }

- rapl_msr_priv.platform_rapl_domain = rd; + priv->platform_rapl_domain = rd;

return 0; }

-static int __init rapl_register_powercap(void) +static void rapl_remove_platform_domain(struct rapl_if_priv *priv) { - rapl_msr_priv.control_type = powercap_register_control_type(NULL, "intel-rapl", NULL); - if (IS_ERR(rapl_msr_priv.control_type)) { - pr_debug("failed to register powercap control_type.\n"); - return PTR_ERR(rapl_msr_priv.control_type); + if (priv->platform_rapl_domain) { + powercap_unregister_zone(priv->control_type, + &priv->platform_rapl_domain->power_zone); + kfree(priv->platform_rapl_domain); } - return 0; }

static int rapl_check_domain(int cpu, int domain, struct rapl_package *rp) @@ -1543,9 +1539,12 @@ static int __init rapl_init(void)

rapl_msr_priv.read_raw = rapl_msr_read_raw; rapl_msr_priv.write_raw = rapl_msr_write_raw; - ret = rapl_register_powercap(); - if (ret) - return ret; + + rapl_msr_priv.control_type = powercap_register_control_type(NULL, "intel-rapl", NULL); + if (IS_ERR(rapl_msr_priv.control_type)) { + pr_debug("failed to register powercap control_type.\n"); + return PTR_ERR(rapl_msr_priv.control_type); + }

ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "powercap/rapl:online", rapl_cpu_online, rapl_cpu_down_prep); @@ -1554,7 +1553,7 @@ static int __init rapl_init(void) rapl_msr_priv.pcap_rapl_online = ret;

/* Don't bail out if PSys is not supported */ - rapl_register_psys(); + rapl_add_platform_domain(&rapl_msr_priv);

ret = register_pm_notifier(&rapl_pm_notifier); if (ret) @@ -1566,7 +1565,7 @@ static int __init rapl_init(void) cpuhp_remove_state(rapl_msr_priv.pcap_rapl_online);

err_unreg: - rapl_unregister_powercap(); + powercap_unregister_control_type(rapl_msr_priv.control_type); return ret; }

@@ -1574,7 +1573,8 @@ static void __exit rapl_exit(void) { unregister_pm_notifier(&rapl_pm_notifier); cpuhp_remove_state(rapl_msr_priv.pcap_rapl_online); - rapl_unregister_powercap(); + rapl_remove_platform_domain(&rapl_msr_priv); + powercap_unregister_control_type(rapl_msr_priv.control_type); }

module_init(rapl_init);

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 3382388d714891fc0f575926189f33d22e7c960b category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 3382388d714891fc0f575926189f33d22e7c960b upstream.

Split intel_rapl.c to intel_rapl_common.c and intel_rapl_msr.c, where intel_rapl_common.c contains the common code that can be used by both MSR and MMIO interface. intel_rapl_msr.c contains the implementation of RAPL MSR interface.

Reviewed-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Tested-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Guoqing Jiang jiangguoqing@kylinos.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/powercap/Kconfig | 11 +- drivers/powercap/Makefile | 4 +- .../{intel_rapl.c => intel_rapl_common.c} | 556 +++++++----------- drivers/powercap/intel_rapl_msr.c | 163 +++++ include/linux/intel_rapl.h | 7 + 5 files changed, 397 insertions(+), 344 deletions(-) rename drivers/powercap/{intel_rapl.c => intel_rapl_common.c} (72%) create mode 100644 drivers/powercap/intel_rapl_msr.c

diff --git a/drivers/powercap/Kconfig b/drivers/powercap/Kconfig index 6ac27e5908f59..8ef2332b58630 100644 --- a/drivers/powercap/Kconfig +++ b/drivers/powercap/Kconfig @@ -15,14 +15,17 @@ menuconfig POWERCAP

if POWERCAP # Client driver configurations go here. +config INTEL_RAPL_CORE + tristate + config INTEL_RAPL - tristate "Intel RAPL Support" + tristate "Intel RAPL Support via MSR Interface" depends on X86 && IOSF_MBI - default n + select INTEL_RAPL_CORE ---help--- This enables support for the Intel Running Average Power Limit (RAPL) - technology which allows power limits to be enforced and monitored on - modern Intel processors (Sandy Bridge and later). + technology via MSR interface, which allows power limits to be enforced + and monitored on modern Intel processors (Sandy Bridge and later).

In RAPL, the platform level settings are divided into domains for fine grained control. These domains include processor package, DRAM diff --git a/drivers/powercap/Makefile b/drivers/powercap/Makefile index 1b328854b36e4..7255c94ec61c4 100644 --- a/drivers/powercap/Makefile +++ b/drivers/powercap/Makefile @@ -1,3 +1,5 @@ +# SPDX-License-Identifier: GPL-2.0-only obj-$(CONFIG_POWERCAP) += powercap_sys.o -obj-$(CONFIG_INTEL_RAPL) += intel_rapl.o +obj-$(CONFIG_INTEL_RAPL_CORE) += intel_rapl_common.o +obj-$(CONFIG_INTEL_RAPL) += intel_rapl_msr.o obj-$(CONFIG_IDLE_INJECT) += idle_inject.o diff --git a/drivers/powercap/intel_rapl.c b/drivers/powercap/intel_rapl_common.c similarity index 72% rename from drivers/powercap/intel_rapl.c rename to drivers/powercap/intel_rapl_common.c index b5cad6f9518d6..34a82531a7cf0 100644 --- a/drivers/powercap/intel_rapl.c +++ b/drivers/powercap/intel_rapl_common.c @@ -1,19 +1,7 @@ +// SPDX-License-Identifier: GPL-2.0-only /* - * Intel Running Average Power Limit (RAPL) Driver - * Copyright (c) 2013, Intel Corporation. - * - * This program is free software; you can redistribute it and/or modify it - * under the terms and conditions of the GNU General Public License, - * version 2, as published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for - * more details. - * - * You should have received a copy of the GNU General Public License along with - * this program; if not, write to the Free Software Foundation, Inc. - * + * Common code for Intel Running Average Power Limit (RAPL) support. + * Copyright (c) 2019, Intel Corporation. */ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

@@ -30,10 +18,10 @@ #include <linux/cpu.h> #include <linux/powercap.h> #include <linux/suspend.h> +#include <asm/iosf_mbi.h> #include <linux/intel_rapl.h>

-#include <asm/iosf_mbi.h> -#include <asm/processor.h> +#include <linux/processor.h> #include <asm/cpu_device_id.h> #include <asm/intel-family.h>

@@ -74,34 +62,19 @@ #define PP_POLICY_MASK 0x1F

/* Non HW constants */ -#define RAPL_PRIMITIVE_DERIVED BIT(1) /* not from raw data */ +#define RAPL_PRIMITIVE_DERIVED BIT(1) /* not from raw data */ #define RAPL_PRIMITIVE_DUMMY BIT(2)

#define TIME_WINDOW_MAX_MSEC 40000 #define TIME_WINDOW_MIN_MSEC 250 -#define ENERGY_UNIT_SCALE 1000 /* scale from driver unit to powercap unit */ +#define ENERGY_UNIT_SCALE 1000 /* scale from driver unit to powercap unit */ enum unit_type { - ARBITRARY_UNIT, /* no translation */ + ARBITRARY_UNIT, /* no translation */ POWER_UNIT, ENERGY_UNIT, TIME_UNIT, };

-/* private data for RAPL MSR Interface */ -static struct rapl_if_priv rapl_msr_priv = { - .reg_unit = MSR_RAPL_POWER_UNIT, - .regs[RAPL_DOMAIN_PACKAGE] = { - MSR_PKG_POWER_LIMIT, MSR_PKG_ENERGY_STATUS, MSR_PKG_PERF_STATUS, 0, MSR_PKG_POWER_INFO }, - .regs[RAPL_DOMAIN_PP0] = { - MSR_PP0_POWER_LIMIT, MSR_PP0_ENERGY_STATUS, 0, MSR_PP0_POLICY, 0 }, - .regs[RAPL_DOMAIN_PP1] = { - MSR_PP1_POWER_LIMIT, MSR_PP1_ENERGY_STATUS, 0, MSR_PP1_POLICY, 0 }, - .regs[RAPL_DOMAIN_DRAM] = { - MSR_DRAM_POWER_LIMIT, MSR_DRAM_ENERGY_STATUS, MSR_DRAM_PERF_STATUS, 0, MSR_DRAM_POWER_INFO }, - .regs[RAPL_DOMAIN_PLATFORM] = { - MSR_PLATFORM_POWER_LIMIT, MSR_PLATFORM_ENERGY_STATUS, 0, 0, 0}, -}; - /* per domain data, some are optional */ #define NR_RAW_PRIMITIVES (NR_RAPL_PRIMITIVES - 2)

@@ -120,7 +93,7 @@ struct rapl_defaults { int (*check_unit)(struct rapl_package *rp, int cpu); void (*set_floor_freq)(struct rapl_domain *rd, bool mode); u64 (*compute_time_window)(struct rapl_package *rp, u64 val, - bool to_raw); + bool to_raw); unsigned int dram_domain_energy_unit; }; static struct rapl_defaults *rapl_defaults; @@ -155,19 +128,20 @@ struct rapl_primitive_info {

static void rapl_init_domains(struct rapl_package *rp); static int rapl_read_data_raw(struct rapl_domain *rd, - enum rapl_primitives prim, - bool xlate, u64 *data); + enum rapl_primitives prim, + bool xlate, u64 *data); static int rapl_write_data_raw(struct rapl_domain *rd, - enum rapl_primitives prim, - unsigned long long value); + enum rapl_primitives prim, + unsigned long long value); static u64 rapl_unit_xlate(struct rapl_domain *rd, - enum unit_type type, u64 value, - int to_raw); + enum unit_type type, u64 value, int to_raw); static void package_power_limit_irq_save(struct rapl_package *rp); +static int rapl_init_core(void); +static void rapl_remove_core(void);

-static LIST_HEAD(rapl_packages); /* guarded by CPU hotplug lock */ +static LIST_HEAD(rapl_packages); /* guarded by CPU hotplug lock */

-static const char * const rapl_domain_names[] = { +static const char *const rapl_domain_names[] = { "package", "core", "uncore", @@ -175,21 +149,8 @@ static const char * const rapl_domain_names[] = { "psys", };

-/* caller to ensure CPU hotplug lock is held */ -static struct rapl_package *rapl_find_package_domain(int cpu, struct rapl_if_priv *priv) -{ - int id = topology_logical_die_id(cpu); - struct rapl_package *rp; - - list_for_each_entry(rp, &rapl_packages, plist) { - if (rp->id == id && rp->priv->control_type == priv->control_type) - return rp; - } - - return NULL; -} - -static int get_energy_counter(struct powercap_zone *power_zone, u64 *energy_raw) +static int get_energy_counter(struct powercap_zone *power_zone, + u64 *energy_raw) { struct rapl_domain *rd; u64 energy_now; @@ -288,50 +249,49 @@ static int get_domain_enable(struct powercap_zone *power_zone, bool *mode) static const struct powercap_zone_ops zone_ops[] = { /* RAPL_DOMAIN_PACKAGE */ { - .get_energy_uj = get_energy_counter, - .get_max_energy_range_uj = get_max_energy_counter, - .release = release_zone, - .set_enable = set_domain_enable, - .get_enable = get_domain_enable, - }, + .get_energy_uj = get_energy_counter, + .get_max_energy_range_uj = get_max_energy_counter, + .release = release_zone, + .set_enable = set_domain_enable, + .get_enable = get_domain_enable, + }, /* RAPL_DOMAIN_PP0 */ { - .get_energy_uj = get_energy_counter, - .get_max_energy_range_uj = get_max_energy_counter, - .release = release_zone, - .set_enable = set_domain_enable, - .get_enable = get_domain_enable, - }, + .get_energy_uj = get_energy_counter, + .get_max_energy_range_uj = get_max_energy_counter, + .release = release_zone, + .set_enable = set_domain_enable, + .get_enable = get_domain_enable, + }, /* RAPL_DOMAIN_PP1 */ { - .get_energy_uj = get_energy_counter, - .get_max_energy_range_uj = get_max_energy_counter, - .release = release_zone, - .set_enable = set_domain_enable, - .get_enable = get_domain_enable, - }, + .get_energy_uj = get_energy_counter, + .get_max_energy_range_uj = get_max_energy_counter, + .release = release_zone, + .set_enable = set_domain_enable, + .get_enable = get_domain_enable, + }, /* RAPL_DOMAIN_DRAM */ { - .get_energy_uj = get_energy_counter, - .get_max_energy_range_uj = get_max_energy_counter, - .release = release_zone, - .set_enable = set_domain_enable, - .get_enable = get_domain_enable, - }, + .get_energy_uj = get_energy_counter, + .get_max_energy_range_uj = get_max_energy_counter, + .release = release_zone, + .set_enable = set_domain_enable, + .get_enable = get_domain_enable, + }, /* RAPL_DOMAIN_PLATFORM */ { - .get_energy_uj = get_energy_counter, - .get_max_energy_range_uj = get_max_energy_counter, - .release = release_zone, - .set_enable = set_domain_enable, - .get_enable = get_domain_enable, - }, + .get_energy_uj = get_energy_counter, + .get_max_energy_range_uj = get_max_energy_counter, + .release = release_zone, + .set_enable = set_domain_enable, + .get_enable = get_domain_enable, + }, };

- /* * Constraint index used by powercap can be different than power limit (PL) - * index in that some PLs maybe missing due to non-existant MSRs. So we + * index in that some PLs maybe missing due to non-existent MSRs. So we * need to convert here by finding the valid PLs only (name populated). */ static int contraint_to_pl(struct rapl_domain *rd, int cid) @@ -350,7 +310,7 @@ static int contraint_to_pl(struct rapl_domain *rd, int cid) }

static int set_power_limit(struct powercap_zone *power_zone, int cid, - u64 power_limit) + u64 power_limit) { struct rapl_domain *rd; struct rapl_package *rp; @@ -368,8 +328,8 @@ static int set_power_limit(struct powercap_zone *power_zone, int cid, rp = rd->rp;

if (rd->state & DOMAIN_STATE_BIOS_LOCKED) { - dev_warn(&power_zone->dev, "%s locked by BIOS, monitoring only\n", - rd->name); + dev_warn(&power_zone->dev, + "%s locked by BIOS, monitoring only\n", rd->name); ret = -EACCES; goto set_exit; } @@ -392,7 +352,7 @@ static int set_power_limit(struct powercap_zone *power_zone, int cid, }

static int get_current_power_limit(struct powercap_zone *power_zone, int cid, - u64 *data) + u64 *data) { struct rapl_domain *rd; u64 val; @@ -431,7 +391,7 @@ static int get_current_power_limit(struct powercap_zone *power_zone, int cid, }

static int set_time_window(struct powercap_zone *power_zone, int cid, - u64 window) + u64 window) { struct rapl_domain *rd; int ret = 0; @@ -461,7 +421,8 @@ static int set_time_window(struct powercap_zone *power_zone, int cid, return ret; }

-static int get_time_window(struct powercap_zone *power_zone, int cid, u64 *data) +static int get_time_window(struct powercap_zone *power_zone, int cid, + u64 *data) { struct rapl_domain *rd; u64 val; @@ -496,7 +457,8 @@ static int get_time_window(struct powercap_zone *power_zone, int cid, u64 *data) return ret; }

-static const char *get_constraint_name(struct powercap_zone *power_zone, int cid) +static const char *get_constraint_name(struct powercap_zone *power_zone, + int cid) { struct rapl_domain *rd; int id; @@ -509,9 +471,7 @@ static const char *get_constraint_name(struct powercap_zone *power_zone, int cid return NULL; }

- -static int get_max_power(struct powercap_zone *power_zone, int id, - u64 *data) +static int get_max_power(struct powercap_zone *power_zone, int id, u64 *data) { struct rapl_domain *rd; u64 val; @@ -559,11 +519,16 @@ static void rapl_init_domains(struct rapl_package *rp) for (i = 0; i < RAPL_DOMAIN_MAX; i++) { unsigned int mask = rp->domain_map & (1 << i);

- rd->regs[RAPL_DOMAIN_REG_LIMIT] = rp->priv->regs[i][RAPL_DOMAIN_REG_LIMIT]; - rd->regs[RAPL_DOMAIN_REG_STATUS] = rp->priv->regs[i][RAPL_DOMAIN_REG_STATUS]; - rd->regs[RAPL_DOMAIN_REG_PERF] = rp->priv->regs[i][RAPL_DOMAIN_REG_PERF]; - rd->regs[RAPL_DOMAIN_REG_POLICY] = rp->priv->regs[i][RAPL_DOMAIN_REG_POLICY]; - rd->regs[RAPL_DOMAIN_REG_INFO] = rp->priv->regs[i][RAPL_DOMAIN_REG_INFO]; + rd->regs[RAPL_DOMAIN_REG_LIMIT] = + rp->priv->regs[i][RAPL_DOMAIN_REG_LIMIT]; + rd->regs[RAPL_DOMAIN_REG_STATUS] = + rp->priv->regs[i][RAPL_DOMAIN_REG_STATUS]; + rd->regs[RAPL_DOMAIN_REG_PERF] = + rp->priv->regs[i][RAPL_DOMAIN_REG_PERF]; + rd->regs[RAPL_DOMAIN_REG_POLICY] = + rp->priv->regs[i][RAPL_DOMAIN_REG_POLICY]; + rd->regs[RAPL_DOMAIN_REG_INFO] = + rp->priv->regs[i][RAPL_DOMAIN_REG_INFO];

switch (mask) { case BIT(RAPL_DOMAIN_PACKAGE): @@ -592,7 +557,7 @@ static void rapl_init_domains(struct rapl_package *rp) rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->domain_energy_unit = - rapl_defaults->dram_domain_energy_unit; + rapl_defaults->dram_domain_energy_unit; if (rd->domain_energy_unit) pr_info("DRAM domain energy unit %dpj\n", rd->domain_energy_unit); @@ -606,7 +571,7 @@ static void rapl_init_domains(struct rapl_package *rp) }

static u64 rapl_unit_xlate(struct rapl_domain *rd, enum unit_type type, - u64 value, int to_raw) + u64 value, int to_raw) { u64 units = 1; struct rapl_package *rp = rd->rp; @@ -643,40 +608,40 @@ static u64 rapl_unit_xlate(struct rapl_domain *rd, enum unit_type type, static struct rapl_primitive_info rpi[] = { /* name, mask, shift, msr index, unit divisor */ PRIMITIVE_INFO_INIT(ENERGY_COUNTER, ENERGY_STATUS_MASK, 0, - RAPL_DOMAIN_REG_STATUS, ENERGY_UNIT, 0), + RAPL_DOMAIN_REG_STATUS, ENERGY_UNIT, 0), PRIMITIVE_INFO_INIT(POWER_LIMIT1, POWER_LIMIT1_MASK, 0, - RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(POWER_LIMIT2, POWER_LIMIT2_MASK, 32, - RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(FW_LOCK, POWER_PP_LOCK, 31, - RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL1_ENABLE, POWER_LIMIT1_ENABLE, 15, - RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL1_CLAMP, POWER_LIMIT1_CLAMP, 16, - RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL2_ENABLE, POWER_LIMIT2_ENABLE, 47, - RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL2_CLAMP, POWER_LIMIT2_CLAMP, 48, - RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(TIME_WINDOW1, TIME_WINDOW1_MASK, 17, - RAPL_DOMAIN_REG_LIMIT, TIME_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, TIME_UNIT, 0), PRIMITIVE_INFO_INIT(TIME_WINDOW2, TIME_WINDOW2_MASK, 49, - RAPL_DOMAIN_REG_LIMIT, TIME_UNIT, 0), + RAPL_DOMAIN_REG_LIMIT, TIME_UNIT, 0), PRIMITIVE_INFO_INIT(THERMAL_SPEC_POWER, POWER_INFO_THERMAL_SPEC_MASK, - 0, RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), + 0, RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(MAX_POWER, POWER_INFO_MAX_MASK, 32, - RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), + RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(MIN_POWER, POWER_INFO_MIN_MASK, 16, - RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), + RAPL_DOMAIN_REG_INFO, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(MAX_TIME_WINDOW, POWER_INFO_MAX_TIME_WIN_MASK, 48, - RAPL_DOMAIN_REG_INFO, TIME_UNIT, 0), + RAPL_DOMAIN_REG_INFO, TIME_UNIT, 0), PRIMITIVE_INFO_INIT(THROTTLED_TIME, PERF_STATUS_THROTTLE_TIME_MASK, 0, - RAPL_DOMAIN_REG_PERF, TIME_UNIT, 0), + RAPL_DOMAIN_REG_PERF, TIME_UNIT, 0), PRIMITIVE_INFO_INIT(PRIORITY_LEVEL, PP_POLICY_MASK, 0, - RAPL_DOMAIN_REG_POLICY, ARBITRARY_UNIT, 0), + RAPL_DOMAIN_REG_POLICY, ARBITRARY_UNIT, 0), /* non-hardware */ PRIMITIVE_INFO_INIT(AVERAGE_POWER, 0, 0, 0, POWER_UNIT, - RAPL_PRIMITIVE_DERIVED), + RAPL_PRIMITIVE_DERIVED), {NULL, 0, 0, 0}, };

@@ -694,8 +659,7 @@ static struct rapl_primitive_info rpi[] = { * 63-------------------------- 31--------------------------- 0 */ static int rapl_read_data_raw(struct rapl_domain *rd, - enum rapl_primitives prim, - bool xlate, u64 *data) + enum rapl_primitives prim, bool xlate, u64 *data) { u64 value; struct rapl_primitive_info *rp = &rpi[prim]; @@ -711,7 +675,7 @@ static int rapl_read_data_raw(struct rapl_domain *rd,

cpu = rd->rp->lead_cpu;

- /* special-case package domain, which uses a different bit*/ + /* special-case package domain, which uses a different bit */ if (prim == FW_LOCK && rd->id == RAPL_DOMAIN_PACKAGE) { rp->mask = POWER_PACKAGE_LOCK; rp->shift = 63; @@ -741,8 +705,8 @@ static int rapl_read_data_raw(struct rapl_domain *rd,

/* Similar use of primitive info in the read counterpart */ static int rapl_write_data_raw(struct rapl_domain *rd, - enum rapl_primitives prim, - unsigned long long value) + enum rapl_primitives prim, + unsigned long long value) { struct rapl_primitive_info *rp = &rpi[prim]; int cpu; @@ -786,7 +750,7 @@ static int rapl_check_unit_core(struct rapl_package *rp, int cpu) ra.mask = ~0; if (rp->priv->read_raw(cpu, &ra)) { pr_err("Failed to read power unit REG 0x%x on CPU %d, exit.\n", - rp->priv->reg_unit, cpu); + rp->priv->reg_unit, cpu); return -ENODEV; }

@@ -800,7 +764,7 @@ static int rapl_check_unit_core(struct rapl_package *rp, int cpu) rp->time_unit = 1000000 / (1 << value);

pr_debug("Core CPU %s energy=%dpJ, time=%dus, power=%duW\n", - rp->name, rp->energy_unit, rp->time_unit, rp->power_unit); + rp->name, rp->energy_unit, rp->time_unit, rp->power_unit);

return 0; } @@ -814,7 +778,7 @@ static int rapl_check_unit_atom(struct rapl_package *rp, int cpu) ra.mask = ~0; if (rp->priv->read_raw(cpu, &ra)) { pr_err("Failed to read power unit REG 0x%x on CPU %d, exit.\n", - rp->priv->reg_unit, cpu); + rp->priv->reg_unit, cpu); return -ENODEV; }

@@ -828,7 +792,7 @@ static int rapl_check_unit_atom(struct rapl_package *rp, int cpu) rp->time_unit = 1000000 / (1 << value);

pr_debug("Atom %s energy=%dpJ, time=%dus, power=%duW\n", - rp->name, rp->energy_unit, rp->time_unit, rp->power_unit); + rp->name, rp->energy_unit, rp->time_unit, rp->power_unit);

return 0; } @@ -848,7 +812,6 @@ static void power_limit_irq_save_cpu(void *info) wrmsr_safe(MSR_IA32_PACKAGE_THERM_INTERRUPT, l, h); }

- /* REVISIT: * When package power limit is set artificially low by RAPL, LVT * thermal interrupt for package power limit should be ignored @@ -932,9 +895,9 @@ static void set_floor_freq_atom(struct rapl_domain *rd, bool enable) }

static u64 rapl_compute_time_window_core(struct rapl_package *rp, u64 value, - bool to_raw) + bool to_raw) { - u64 f, y; /* fraction and exp. used for time unit */ + u64 f, y; /* fraction and exp. used for time unit */

/* * Special processing based on 2^Y*(1+F/4), refer @@ -954,7 +917,7 @@ static u64 rapl_compute_time_window_core(struct rapl_package *rp, u64 value, }

static u64 rapl_compute_time_window_atom(struct rapl_package *rp, u64 value, - bool to_raw) + bool to_raw) { /* * Atom time unit encoding is straight forward val * time_unit, @@ -962,8 +925,8 @@ static u64 rapl_compute_time_window_atom(struct rapl_package *rp, u64 value, */ if (!to_raw) return (value) ? value *= rp->time_unit : rp->time_unit; - else - value = div64_u64(value, rp->time_unit); + + value = div64_u64(value, rp->time_unit);

return value; } @@ -1010,49 +973,45 @@ static const struct rapl_defaults rapl_defaults_cht = { .compute_time_window = rapl_compute_time_window_atom, };

-#define RAPL_CPU(_model, _ops) { \ - .vendor = X86_VENDOR_INTEL, \ - .family = 6, \ - .model = _model, \ - .driver_data = (kernel_ulong_t)&_ops, \ - } - static const struct x86_cpu_id rapl_ids[] __initconst = { - RAPL_CPU(INTEL_FAM6_SANDYBRIDGE, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_SANDYBRIDGE_X, rapl_defaults_core), - - RAPL_CPU(INTEL_FAM6_IVYBRIDGE, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_IVYBRIDGE_X, rapl_defaults_core), - - RAPL_CPU(INTEL_FAM6_HASWELL_CORE, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_HASWELL_ULT, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_HASWELL_GT3E, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_HASWELL_X, rapl_defaults_hsw_server), - - RAPL_CPU(INTEL_FAM6_BROADWELL_CORE, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_BROADWELL_GT3E, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_BROADWELL_XEON_D, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_BROADWELL_X, rapl_defaults_hsw_server), - - RAPL_CPU(INTEL_FAM6_SKYLAKE_DESKTOP, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_SKYLAKE_MOBILE, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_SKYLAKE_X, rapl_defaults_hsw_server), - RAPL_CPU(INTEL_FAM6_KABYLAKE_MOBILE, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_KABYLAKE_DESKTOP, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_CANNONLAKE_MOBILE, rapl_defaults_core), - - RAPL_CPU(INTEL_FAM6_ATOM_SILVERMONT, rapl_defaults_byt), - RAPL_CPU(INTEL_FAM6_ATOM_AIRMONT, rapl_defaults_cht), - RAPL_CPU(INTEL_FAM6_ATOM_SILVERMONT_MID, rapl_defaults_tng), - RAPL_CPU(INTEL_FAM6_ATOM_AIRMONT_MID, rapl_defaults_ann), - RAPL_CPU(INTEL_FAM6_ATOM_GOLDMONT, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_ATOM_GOLDMONT_PLUS, rapl_defaults_core), - RAPL_CPU(INTEL_FAM6_ATOM_GOLDMONT_X, rapl_defaults_core), - - RAPL_CPU(INTEL_FAM6_XEON_PHI_KNL, rapl_defaults_hsw_server), - RAPL_CPU(INTEL_FAM6_XEON_PHI_KNM, rapl_defaults_hsw_server), + INTEL_CPU_FAM6(SANDYBRIDGE, rapl_defaults_core), + INTEL_CPU_FAM6(SANDYBRIDGE_X, rapl_defaults_core), + + INTEL_CPU_FAM6(IVYBRIDGE, rapl_defaults_core), + INTEL_CPU_FAM6(IVYBRIDGE_X, rapl_defaults_core), + + INTEL_CPU_FAM6(HASWELL_CORE, rapl_defaults_core), + INTEL_CPU_FAM6(HASWELL_ULT, rapl_defaults_core), + INTEL_CPU_FAM6(HASWELL_GT3E, rapl_defaults_core), + INTEL_CPU_FAM6(HASWELL_X, rapl_defaults_hsw_server), + + INTEL_CPU_FAM6(BROADWELL_CORE, rapl_defaults_core), + INTEL_CPU_FAM6(BROADWELL_GT3E, rapl_defaults_core), + INTEL_CPU_FAM6(BROADWELL_XEON_D, rapl_defaults_core), + INTEL_CPU_FAM6(BROADWELL_X, rapl_defaults_hsw_server), + + INTEL_CPU_FAM6(SKYLAKE_DESKTOP, rapl_defaults_core), + INTEL_CPU_FAM6(SKYLAKE_MOBILE, rapl_defaults_core), + INTEL_CPU_FAM6(SKYLAKE_X, rapl_defaults_hsw_server), + INTEL_CPU_FAM6(KABYLAKE_MOBILE, rapl_defaults_core), + INTEL_CPU_FAM6(KABYLAKE_DESKTOP, rapl_defaults_core), + INTEL_CPU_FAM6(CANNONLAKE_MOBILE, rapl_defaults_core), + INTEL_CPU_FAM6(ICELAKE_MOBILE, rapl_defaults_core), + + INTEL_CPU_FAM6(ATOM_SILVERMONT, rapl_defaults_byt), + INTEL_CPU_FAM6(ATOM_AIRMONT, rapl_defaults_cht), + INTEL_CPU_FAM6(ATOM_SILVERMONT_MID, rapl_defaults_tng), + INTEL_CPU_FAM6(ATOM_AIRMONT_MID, rapl_defaults_ann), + INTEL_CPU_FAM6(ATOM_GOLDMONT, rapl_defaults_core), + INTEL_CPU_FAM6(ATOM_GOLDMONT_PLUS, rapl_defaults_core), + INTEL_CPU_FAM6(ATOM_GOLDMONT_X, rapl_defaults_core), + INTEL_CPU_FAM6(ATOM_TREMONT_X, rapl_defaults_core), + + INTEL_CPU_FAM6(XEON_PHI_KNL, rapl_defaults_hsw_server), + INTEL_CPU_FAM6(XEON_PHI_KNM, rapl_defaults_hsw_server), {} }; + MODULE_DEVICE_TABLE(x86cpu, rapl_ids);

/* Read once for all raw primitive data for domains */ @@ -1068,7 +1027,7 @@ static void rapl_update_domain_data(struct rapl_package *rp) for (prim = 0; prim < NR_RAW_PRIMITIVES; prim++) { if (!rapl_read_data_raw(&rp->domains[dmn], prim, rpi[prim].unit, &val)) - rp->domains[dmn].rdd.primitives[prim] = val; + rp->domains[dmn].rdd.primitives[prim] = val; } }

@@ -1083,20 +1042,18 @@ static int rapl_package_register_powercap(struct rapl_package *rp) /* Update the domain data of the new package */ rapl_update_domain_data(rp);

- /* first we register package domain as the parent zone*/ + /* first we register package domain as the parent zone */ for (rd = rp->domains; rd < rp->domains + rp->nr_domains; rd++) { if (rd->id == RAPL_DOMAIN_PACKAGE) { nr_pl = find_nr_power_limit(rd); pr_debug("register package domain %s\n", rp->name); power_zone = powercap_register_zone(&rd->power_zone, - rp->priv->control_type, - rp->name, NULL, - &zone_ops[rd->id], - nr_pl, - &constraint_ops); + rp->priv->control_type, rp->name, + NULL, &zone_ops[rd->id], nr_pl, + &constraint_ops); if (IS_ERR(power_zone)) { pr_debug("failed to register power zone %s\n", - rp->name); + rp->name); return PTR_ERR(power_zone); } /* track parent zone in per package/socket data */ @@ -1109,21 +1066,21 @@ static int rapl_package_register_powercap(struct rapl_package *rp) pr_err("no package domain found, unknown topology!\n"); return -ENODEV; } - /* now register domains as children of the socket/package*/ + /* now register domains as children of the socket/package */ for (rd = rp->domains; rd < rp->domains + rp->nr_domains; rd++) { if (rd->id == RAPL_DOMAIN_PACKAGE) continue; /* number of power limits per domain varies */ nr_pl = find_nr_power_limit(rd); power_zone = powercap_register_zone(&rd->power_zone, - rp->priv->control_type, rd->name, - rp->power_zone, - &zone_ops[rd->id], nr_pl, - &constraint_ops); + rp->priv->control_type, + rd->name, rp->power_zone, + &zone_ops[rd->id], nr_pl, + &constraint_ops);

if (IS_ERR(power_zone)) { pr_debug("failed to register power_zone, %s:%s\n", - rp->name, rd->name); + rp->name, rd->name); ret = PTR_ERR(power_zone); goto err_cleanup; } @@ -1137,13 +1094,14 @@ static int rapl_package_register_powercap(struct rapl_package *rp) */ while (--rd >= rp->domains) { pr_debug("unregister %s domain %s\n", rp->name, rd->name); - powercap_unregister_zone(rp->priv->control_type, &rd->power_zone); + powercap_unregister_zone(rp->priv->control_type, + &rd->power_zone); }

return ret; }

-static int __init rapl_add_platform_domain(struct rapl_if_priv *priv) +int rapl_add_platform_domain(struct rapl_if_priv *priv) { struct rapl_domain *rd; struct powercap_zone *power_zone; @@ -1168,8 +1126,10 @@ static int __init rapl_add_platform_domain(struct rapl_if_priv *priv)

rd->name = rapl_domain_names[RAPL_DOMAIN_PLATFORM]; rd->id = RAPL_DOMAIN_PLATFORM; - rd->regs[RAPL_DOMAIN_REG_LIMIT] = priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT]; - rd->regs[RAPL_DOMAIN_REG_STATUS] = priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS]; + rd->regs[RAPL_DOMAIN_REG_LIMIT] = + priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_LIMIT]; + rd->regs[RAPL_DOMAIN_REG_STATUS] = + priv->regs[RAPL_DOMAIN_PLATFORM][RAPL_DOMAIN_REG_STATUS]; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; rd->rpl[1].prim_id = PL2_ENABLE; @@ -1190,15 +1150,17 @@ static int __init rapl_add_platform_domain(struct rapl_if_priv *priv)

return 0; } +EXPORT_SYMBOL_GPL(rapl_add_platform_domain);

-static void rapl_remove_platform_domain(struct rapl_if_priv *priv) +void rapl_remove_platform_domain(struct rapl_if_priv *priv) { if (priv->platform_rapl_domain) { powercap_unregister_zone(priv->control_type, - &priv->platform_rapl_domain->power_zone); + &priv->platform_rapl_domain->power_zone); kfree(priv->platform_rapl_domain); } } +EXPORT_SYMBOL_GPL(rapl_remove_platform_domain);

static int rapl_check_domain(int cpu, int domain, struct rapl_package *rp) { @@ -1229,13 +1191,12 @@ static int rapl_check_domain(int cpu, int domain, struct rapl_package *rp) return 0; }

- /* * Check if power limits are available. Two cases when they are not available: * 1. Locked by BIOS, in this case we still provide read-only access so that * users can see what limit is set by the BIOS. * 2. Some CPUs make some domains monitoring only which means PLx MSRs may not - * exist at all. In this case, we do not show the contraints in powercap. + * exist at all. In this case, we do not show the constraints in powercap. * * Called after domains are detected and initialized. */ @@ -1252,9 +1213,10 @@ static void rapl_detect_powerlimit(struct rapl_domain *rd) rd->state |= DOMAIN_STATE_BIOS_LOCKED; } } - /* check if power limit MSRs exists, otherwise domain is monitoring only */ + /* check if power limit MSR exists, otherwise domain is monitoring only */ for (i = 0; i < NR_POWER_LIMITS; i++) { int prim = rd->rpl[i].prim_id; + if (rapl_read_data_raw(rd, prim, false, &val64)) rd->rpl[i].name = NULL; } @@ -1275,7 +1237,7 @@ static int rapl_detect_domains(struct rapl_package *rp, int cpu) pr_info("Found RAPL domain %s\n", rapl_domain_names[i]); } } - rp->nr_domains = bitmap_weight(&rp->domain_map, RAPL_DOMAIN_MAX); + rp->nr_domains = bitmap_weight(&rp->domain_map, RAPL_DOMAIN_MAX); if (!rp->nr_domains) { pr_debug("no valid rapl domains found in %s\n", rp->name); return -ENODEV; @@ -1283,7 +1245,7 @@ static int rapl_detect_domains(struct rapl_package *rp, int cpu) pr_debug("found %d domains on %s\n", rp->nr_domains, rp->name);

rp->domains = kcalloc(rp->nr_domains + 1, sizeof(struct rapl_domain), - GFP_KERNEL); + GFP_KERNEL); if (!rp->domains) return -ENOMEM;

@@ -1296,7 +1258,7 @@ static int rapl_detect_domains(struct rapl_package *rp, int cpu) }

/* called from CPU hotplug notifier, hotplug lock held */ -static void rapl_remove_package(struct rapl_package *rp) +void rapl_remove_package(struct rapl_package *rp) { struct rapl_domain *rd, *rd_package = NULL;

@@ -1315,22 +1277,47 @@ static void rapl_remove_package(struct rapl_package *rp) } pr_debug("remove package, undo power limit on %s: %s\n", rp->name, rd->name); - powercap_unregister_zone(rp->priv->control_type, &rd->power_zone); + powercap_unregister_zone(rp->priv->control_type, + &rd->power_zone); } /* do parent zone last */ - powercap_unregister_zone(rp->priv->control_type, &rd_package->power_zone); + powercap_unregister_zone(rp->priv->control_type, + &rd_package->power_zone); list_del(&rp->plist); + if (list_empty(&rapl_packages)) + rapl_remove_core(); kfree(rp); } +EXPORT_SYMBOL_GPL(rapl_remove_package); + +/* caller to ensure CPU hotplug lock is held */ +struct rapl_package *rapl_find_package_domain(int cpu, struct rapl_if_priv *priv) +{ + int id = topology_logical_die_id(cpu); + struct rapl_package *rp; + + list_for_each_entry(rp, &rapl_packages, plist) { + if (rp->id == id + && rp->priv->control_type == priv->control_type) + return rp; + } + + return NULL; +} +EXPORT_SYMBOL_GPL(rapl_find_package_domain);

/* called from CPU hotplug notifier, hotplug lock held */ -static struct rapl_package *rapl_add_package(int cpu, struct rapl_if_priv *priv) +struct rapl_package *rapl_add_package(int cpu, struct rapl_if_priv *priv) { int id = topology_logical_die_id(cpu); struct rapl_package *rp; struct cpuinfo_x86 *c = &cpu_data(cpu); int ret;

+ ret = rapl_init_core(); + if (ret) + return ERR_PTR(ret); + rp = kzalloc(sizeof(struct rapl_package), GFP_KERNEL); if (!rp) return ERR_PTR(-ENOMEM); @@ -1342,14 +1329,13 @@ static struct rapl_package *rapl_add_package(int cpu, struct rapl_if_priv *priv)

if (topology_max_die_per_package() > 1) snprintf(rp->name, PACKAGE_DOMAIN_NAME_LENGTH, - "package-%d-die-%d", c->phys_proc_id, c->cpu_die_id); + "package-%d-die-%d", c->phys_proc_id, c->cpu_die_id); else snprintf(rp->name, PACKAGE_DOMAIN_NAME_LENGTH, "package-%d", - c->phys_proc_id); + c->phys_proc_id);

/* check if the package contains valid domains */ - if (rapl_detect_domains(rp, cpu) || - rapl_defaults->check_unit(rp, cpu)) { + if (rapl_detect_domains(rp, cpu) || rapl_defaults->check_unit(rp, cpu)) { ret = -ENODEV; goto err_free_package; } @@ -1365,45 +1351,7 @@ static struct rapl_package *rapl_add_package(int cpu, struct rapl_if_priv *priv) kfree(rp); return ERR_PTR(ret); } - -/* Handles CPU hotplug on multi-socket systems. - * If a CPU goes online as the first CPU of the physical package - * we add the RAPL package to the system. Similarly, when the last - * CPU of the package is removed, we remove the RAPL package and its - * associated domains. Cooling devices are handled accordingly at - * per-domain level. - */ -static int rapl_cpu_online(unsigned int cpu) -{ - struct rapl_package *rp; - - rp = rapl_find_package_domain(cpu, &rapl_msr_priv); - if (!rp) { - rp = rapl_add_package(cpu, &rapl_msr_priv); - if (IS_ERR(rp)) - return PTR_ERR(rp); - } - cpumask_set_cpu(cpu, &rp->cpumask); - return 0; -} - -static int rapl_cpu_down_prep(unsigned int cpu) -{ - struct rapl_package *rp; - int lead_cpu; - - rp = rapl_find_package_domain(cpu, &rapl_msr_priv); - if (!rp) - return 0; - - cpumask_clear_cpu(cpu, &rp->cpumask); - lead_cpu = cpumask_first(&rp->cpumask); - if (lead_cpu >= nr_cpu_ids) - rapl_remove_package(rp); - else if (rp->lead_cpu == cpu) - rp->lead_cpu = lead_cpu; - return 0; -} +EXPORT_SYMBOL_GPL(rapl_add_package);

static void power_limit_state_save(void) { @@ -1421,17 +1369,15 @@ static void power_limit_state_save(void) switch (rd->rpl[i].prim_id) { case PL1_ENABLE: ret = rapl_read_data_raw(rd, - POWER_LIMIT1, - true, - &rd->rpl[i].last_power_limit); + POWER_LIMIT1, true, + &rd->rpl[i].last_power_limit); if (ret) rd->rpl[i].last_power_limit = 0; break; case PL2_ENABLE: ret = rapl_read_data_raw(rd, - POWER_LIMIT2, - true, - &rd->rpl[i].last_power_limit); + POWER_LIMIT2, true, + &rd->rpl[i].last_power_limit); if (ret) rd->rpl[i].last_power_limit = 0; break; @@ -1457,15 +1403,13 @@ static void power_limit_state_restore(void) switch (rd->rpl[i].prim_id) { case PL1_ENABLE: if (rd->rpl[i].last_power_limit) - rapl_write_data_raw(rd, - POWER_LIMIT1, - rd->rpl[i].last_power_limit); + rapl_write_data_raw(rd, POWER_LIMIT1, + rd->rpl[i].last_power_limit); break; case PL2_ENABLE: if (rd->rpl[i].last_power_limit) - rapl_write_data_raw(rd, - POWER_LIMIT2, - rd->rpl[i].last_power_limit); + rapl_write_data_raw(rd, POWER_LIMIT2, + rd->rpl[i].last_power_limit); break; } } @@ -1474,7 +1418,7 @@ static void power_limit_state_restore(void) }

static int rapl_pm_callback(struct notifier_block *nb, - unsigned long mode, void *_unused) + unsigned long mode, void *_unused) { switch (mode) { case PM_SUSPEND_PREPARE: @@ -1491,101 +1435,35 @@ static struct notifier_block rapl_pm_notifier = { .notifier_call = rapl_pm_callback, };

-static int rapl_msr_read_raw(int cpu, struct reg_action *ra) -{ - if (rdmsrl_safe_on_cpu(cpu, ra->reg, &ra->value)) { - pr_debug("failed to read msr 0x%x on cpu %d\n", ra->reg, cpu); - return -EIO; - } - ra->value &= ra->mask; - return 0; -} - -static void rapl_msr_update_func(void *info) -{ - struct reg_action *ra = info; - u64 val; - - ra->err = rdmsrl_safe(ra->reg, &val); - if (ra->err) - return; - - val &= ~ra->mask; - val |= ra->value; - - ra->err = wrmsrl_safe(ra->reg, val); -} - - -static int rapl_msr_write_raw(int cpu, struct reg_action *ra) -{ - int ret; - - ret = smp_call_function_single(cpu, rapl_msr_update_func, ra, 1); - if (WARN_ON_ONCE(ret)) - return ret; - - return ra->err; -} - -static int __init rapl_init(void) +static int rapl_init_core(void) { const struct x86_cpu_id *id; int ret;

+ if (rapl_defaults) + return 0; + id = x86_match_cpu(rapl_ids); if (!id) { pr_err("driver does not support CPU family %d model %d\n", - boot_cpu_data.x86, boot_cpu_data.x86_model); + boot_cpu_data.x86, boot_cpu_data.x86_model);

return -ENODEV; }

rapl_defaults = (struct rapl_defaults *)id->driver_data;

- rapl_msr_priv.read_raw = rapl_msr_read_raw; - rapl_msr_priv.write_raw = rapl_msr_write_raw; - - rapl_msr_priv.control_type = powercap_register_control_type(NULL, "intel-rapl", NULL); - if (IS_ERR(rapl_msr_priv.control_type)) { - pr_debug("failed to register powercap control_type.\n"); - return PTR_ERR(rapl_msr_priv.control_type); - } - - ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "powercap/rapl:online", - rapl_cpu_online, rapl_cpu_down_prep); - if (ret < 0) - goto err_unreg; - rapl_msr_priv.pcap_rapl_online = ret; - - /* Don't bail out if PSys is not supported */ - rapl_add_platform_domain(&rapl_msr_priv); - ret = register_pm_notifier(&rapl_pm_notifier); - if (ret) - goto err_unreg_all;

return 0; - -err_unreg_all: - cpuhp_remove_state(rapl_msr_priv.pcap_rapl_online); - -err_unreg: - powercap_unregister_control_type(rapl_msr_priv.control_type); - return ret; }

-static void __exit rapl_exit(void) +static void rapl_remove_core(void) { unregister_pm_notifier(&rapl_pm_notifier); - cpuhp_remove_state(rapl_msr_priv.pcap_rapl_online); - rapl_remove_platform_domain(&rapl_msr_priv); - powercap_unregister_control_type(rapl_msr_priv.control_type); + rapl_defaults = NULL; }

-module_init(rapl_init); -module_exit(rapl_exit); - -MODULE_DESCRIPTION("Driver for Intel RAPL (Running Average Power Limit)"); +MODULE_DESCRIPTION("Intel Runtime Average Power Limit (RAPL) common code"); MODULE_AUTHOR("Jacob Pan jacob.jun.pan@intel.com"); MODULE_LICENSE("GPL v2"); diff --git a/drivers/powercap/intel_rapl_msr.c b/drivers/powercap/intel_rapl_msr.c new file mode 100644 index 0000000000000..89645222e3e0f --- /dev/null +++ b/drivers/powercap/intel_rapl_msr.c @@ -0,0 +1,163 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Intel Running Average Power Limit (RAPL) Driver via MSR interface + * Copyright (c) 2019, Intel Corporation. + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/list.h> +#include <linux/types.h> +#include <linux/device.h> +#include <linux/slab.h> +#include <linux/log2.h> +#include <linux/bitmap.h> +#include <linux/delay.h> +#include <linux/sysfs.h> +#include <linux/cpu.h> +#include <linux/powercap.h> +#include <linux/suspend.h> +#include <linux/intel_rapl.h> +#include <linux/processor.h> + +#include <asm/iosf_mbi.h> +#include <asm/cpu_device_id.h> +#include <asm/intel-family.h> + +/* Local defines */ +#define MSR_PLATFORM_POWER_LIMIT 0x0000065C + +/* private data for RAPL MSR Interface */ +static struct rapl_if_priv rapl_msr_priv = { + .reg_unit = MSR_RAPL_POWER_UNIT, + .regs[RAPL_DOMAIN_PACKAGE] = { + MSR_PKG_POWER_LIMIT, MSR_PKG_ENERGY_STATUS, MSR_PKG_PERF_STATUS, 0, MSR_PKG_POWER_INFO }, + .regs[RAPL_DOMAIN_PP0] = { + MSR_PP0_POWER_LIMIT, MSR_PP0_ENERGY_STATUS, 0, MSR_PP0_POLICY, 0 }, + .regs[RAPL_DOMAIN_PP1] = { + MSR_PP1_POWER_LIMIT, MSR_PP1_ENERGY_STATUS, 0, MSR_PP1_POLICY, 0 }, + .regs[RAPL_DOMAIN_DRAM] = { + MSR_DRAM_POWER_LIMIT, MSR_DRAM_ENERGY_STATUS, MSR_DRAM_PERF_STATUS, 0, MSR_DRAM_POWER_INFO }, + .regs[RAPL_DOMAIN_PLATFORM] = { + MSR_PLATFORM_POWER_LIMIT, MSR_PLATFORM_ENERGY_STATUS, 0, 0, 0}, +}; + +/* Handles CPU hotplug on multi-socket systems. + * If a CPU goes online as the first CPU of the physical package + * we add the RAPL package to the system. Similarly, when the last + * CPU of the package is removed, we remove the RAPL package and its + * associated domains. Cooling devices are handled accordingly at + * per-domain level. + */ +static int rapl_cpu_online(unsigned int cpu) +{ + struct rapl_package *rp; + + rp = rapl_find_package_domain(cpu, &rapl_msr_priv); + if (!rp) { + rp = rapl_add_package(cpu, &rapl_msr_priv); + if (IS_ERR(rp)) + return PTR_ERR(rp); + } + cpumask_set_cpu(cpu, &rp->cpumask); + return 0; +} + +static int rapl_cpu_down_prep(unsigned int cpu) +{ + struct rapl_package *rp; + int lead_cpu; + + rp = rapl_find_package_domain(cpu, &rapl_msr_priv); + if (!rp) + return 0; + + cpumask_clear_cpu(cpu, &rp->cpumask); + lead_cpu = cpumask_first(&rp->cpumask); + if (lead_cpu >= nr_cpu_ids) + rapl_remove_package(rp); + else if (rp->lead_cpu == cpu) + rp->lead_cpu = lead_cpu; + return 0; +} + +static int rapl_msr_read_raw(int cpu, struct reg_action *ra) +{ + if (rdmsrl_safe_on_cpu(cpu, ra->reg, &ra->value)) { + pr_debug("failed to read msr 0x%x on cpu %d\n", ra->reg, cpu); + return -EIO; + } + ra->value &= ra->mask; + return 0; +} + +static void rapl_msr_update_func(void *info) +{ + struct reg_action *ra = info; + u64 val; + + ra->err = rdmsrl_safe(ra->reg, &val); + if (ra->err) + return; + + val &= ~ra->mask; + val |= ra->value; + + ra->err = wrmsrl_safe(ra->reg, val); +} + +static int rapl_msr_write_raw(int cpu, struct reg_action *ra) +{ + int ret; + + ret = smp_call_function_single(cpu, rapl_msr_update_func, ra, 1); + if (WARN_ON_ONCE(ret)) + return ret; + + return ra->err; +} + +static int __init rapl_msr_init(void) +{ + int ret; + + rapl_msr_priv.read_raw = rapl_msr_read_raw; + rapl_msr_priv.write_raw = rapl_msr_write_raw; + + rapl_msr_priv.control_type = powercap_register_control_type(NULL, "intel-rapl", NULL); + if (IS_ERR(rapl_msr_priv.control_type)) { + pr_debug("failed to register powercap control_type.\n"); + return PTR_ERR(rapl_msr_priv.control_type); + } + + ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "powercap/rapl:online", + rapl_cpu_online, rapl_cpu_down_prep); + if (ret < 0) + goto out; + rapl_msr_priv.pcap_rapl_online = ret; + + /* Don't bail out if PSys is not supported */ + rapl_add_platform_domain(&rapl_msr_priv); + + return 0; + +out: + if (ret) + powercap_unregister_control_type(rapl_msr_priv.control_type); + return ret; +} + +static void __exit rapl_msr_exit(void) +{ + cpuhp_remove_state(rapl_msr_priv.pcap_rapl_online); + rapl_remove_platform_domain(&rapl_msr_priv); + powercap_unregister_control_type(rapl_msr_priv.control_type); +} + +module_init(rapl_msr_init); +module_exit(rapl_msr_exit); + +MODULE_DESCRIPTION("Driver for Intel RAPL (Running Average Power Limit) control via MSR interface"); +MODULE_AUTHOR("Zhang Rui rui.zhang@intel.com"); +MODULE_LICENSE("GPL v2"); diff --git a/include/linux/intel_rapl.h b/include/linux/intel_rapl.h index ff215d64d1149..9579f458fe4d3 100644 --- a/include/linux/intel_rapl.h +++ b/include/linux/intel_rapl.h @@ -142,4 +142,11 @@ struct rapl_package { struct rapl_if_priv *priv; };

+struct rapl_package *rapl_find_package_domain(int cpu, struct rapl_if_priv *priv); +struct rapl_package *rapl_add_package(int cpu, struct rapl_if_priv *priv); +void rapl_remove_package(struct rapl_package *rp); + +int rapl_add_platform_domain(struct rapl_if_priv *priv); +void rapl_remove_platform_domain(struct rapl_if_priv *priv); + #endif /* __INTEL_RAPL_H__ */

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 0c2ddedd8bcb88c4100acb9e0fc5ac8752d09501 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 0c2ddedd8bcb88c4100acb9e0fc5ac8752d09501 upstream.

RAPL MSR interface supports 2 power limits for package domain, and 1 power limit for other domains, while RAPL MMIO interface supports 2 power limits for both package and dram domains. And when 2 power limits are supported, the FW_LOCK bit is in bit 63 of the register, instead of bit 31.

Remove the assumption that only pakcage domain supports 2 power limits. And allow the RAPL interface driver to specify the number of power limits supported, for every single RAPL domain it owns..

Reviewed-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Tested-by: Pandruvada, Srinivas srinivas.pandruvada@intel.com Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/powercap/intel_rapl_common.c | 72 ++++++++++------------------ drivers/powercap/intel_rapl_msr.c | 1 + include/linux/intel_rapl.h | 2 + 3 files changed, 28 insertions(+), 47 deletions(-)

diff --git a/drivers/powercap/intel_rapl_common.c b/drivers/powercap/intel_rapl_common.c index 8e4de036f6d04..db8df19d8133f 100644 --- a/drivers/powercap/intel_rapl_common.c +++ b/drivers/powercap/intel_rapl_common.c @@ -38,8 +38,8 @@ #define POWER_LIMIT2_MASK (0x7FFFULL<<32) #define POWER_LIMIT2_ENABLE BIT_ULL(47) #define POWER_LIMIT2_CLAMP BIT_ULL(48) -#define POWER_PACKAGE_LOCK BIT_ULL(63) -#define POWER_PP_LOCK BIT(31) +#define POWER_HIGH_LOCK BIT_ULL(63) +#define POWER_LOW_LOCK BIT(31)

#define TIME_WINDOW1_MASK (0x7FULL<<17) #define TIME_WINDOW2_MASK (0x7FULL<<49) @@ -513,60 +513,38 @@ static const struct powercap_zone_constraint_ops constraint_ops = { /* called after domain detection and package level data are set */ static void rapl_init_domains(struct rapl_package *rp) { - int i; + enum rapl_domain_type i; + enum rapl_domain_reg_id j; struct rapl_domain *rd = rp->domains;

for (i = 0; i < RAPL_DOMAIN_MAX; i++) { unsigned int mask = rp->domain_map & (1 << i);

- rd->regs[RAPL_DOMAIN_REG_LIMIT] = - rp->priv->regs[i][RAPL_DOMAIN_REG_LIMIT]; - rd->regs[RAPL_DOMAIN_REG_STATUS] = - rp->priv->regs[i][RAPL_DOMAIN_REG_STATUS]; - rd->regs[RAPL_DOMAIN_REG_PERF] = - rp->priv->regs[i][RAPL_DOMAIN_REG_PERF]; - rd->regs[RAPL_DOMAIN_REG_POLICY] = - rp->priv->regs[i][RAPL_DOMAIN_REG_POLICY]; - rd->regs[RAPL_DOMAIN_REG_INFO] = - rp->priv->regs[i][RAPL_DOMAIN_REG_INFO]; - - switch (mask) { - case BIT(RAPL_DOMAIN_PACKAGE): - rd->name = rapl_domain_names[RAPL_DOMAIN_PACKAGE]; - rd->id = RAPL_DOMAIN_PACKAGE; - rd->rpl[0].prim_id = PL1_ENABLE; - rd->rpl[0].name = pl1_name; + if (!mask) + continue; + + rd->rp = rp; + rd->name = rapl_domain_names[i]; + rd->id = i; + rd->rpl[0].prim_id = PL1_ENABLE; + rd->rpl[0].name = pl1_name; + /* some domain may support two power limits */ + if (rp->priv->limits[i] == 2) { rd->rpl[1].prim_id = PL2_ENABLE; rd->rpl[1].name = pl2_name; - break; - case BIT(RAPL_DOMAIN_PP0): - rd->name = rapl_domain_names[RAPL_DOMAIN_PP0]; - rd->id = RAPL_DOMAIN_PP0; - rd->rpl[0].prim_id = PL1_ENABLE; - rd->rpl[0].name = pl1_name; - break; - case BIT(RAPL_DOMAIN_PP1): - rd->name = rapl_domain_names[RAPL_DOMAIN_PP1]; - rd->id = RAPL_DOMAIN_PP1; - rd->rpl[0].prim_id = PL1_ENABLE; - rd->rpl[0].name = pl1_name; - break; - case BIT(RAPL_DOMAIN_DRAM): - rd->name = rapl_domain_names[RAPL_DOMAIN_DRAM]; - rd->id = RAPL_DOMAIN_DRAM; - rd->rpl[0].prim_id = PL1_ENABLE; - rd->rpl[0].name = pl1_name; + } + + for (j = 0; j < RAPL_DOMAIN_REG_MAX; j++) + rd->regs[j] = rp->priv->regs[i][j]; + + if (i == RAPL_DOMAIN_DRAM) { rd->domain_energy_unit = rapl_defaults->dram_domain_energy_unit; if (rd->domain_energy_unit) pr_info("DRAM domain energy unit %dpj\n", rd->domain_energy_unit); - break; - } - if (mask) { - rd->rp = rp; - rd++; } + rd++; } }

@@ -613,7 +591,7 @@ static struct rapl_primitive_info rpi[] = { RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0), PRIMITIVE_INFO_INIT(POWER_LIMIT2, POWER_LIMIT2_MASK, 32, RAPL_DOMAIN_REG_LIMIT, POWER_UNIT, 0), - PRIMITIVE_INFO_INIT(FW_LOCK, POWER_PP_LOCK, 31, + PRIMITIVE_INFO_INIT(FW_LOCK, POWER_LOW_LOCK, 31, RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), PRIMITIVE_INFO_INIT(PL1_ENABLE, POWER_LIMIT1_ENABLE, 15, RAPL_DOMAIN_REG_LIMIT, ARBITRARY_UNIT, 0), @@ -675,9 +653,9 @@ static int rapl_read_data_raw(struct rapl_domain *rd,

cpu = rd->rp->lead_cpu;

- /* special-case package domain, which uses a different bit */ - if (prim == FW_LOCK && rd->id == RAPL_DOMAIN_PACKAGE) { - rp->mask = POWER_PACKAGE_LOCK; + /* domain with 2 limits has different bit */ + if (prim == FW_LOCK && rd->rp->priv->limits[rd->id] == 2) { + rp->mask = POWER_HIGH_LOCK; rp->shift = 63; } /* non-hardware data are collected by the polling thread */ diff --git a/drivers/powercap/intel_rapl_msr.c b/drivers/powercap/intel_rapl_msr.c index 6cd8a8fb9238b..bc14a4579acb8 100644 --- a/drivers/powercap/intel_rapl_msr.c +++ b/drivers/powercap/intel_rapl_msr.c @@ -41,6 +41,7 @@ static struct rapl_if_priv rapl_msr_priv = { MSR_DRAM_POWER_LIMIT, MSR_DRAM_ENERGY_STATUS, MSR_DRAM_PERF_STATUS, 0, MSR_DRAM_POWER_INFO }, .regs[RAPL_DOMAIN_PLATFORM] = { MSR_PLATFORM_POWER_LIMIT, MSR_PLATFORM_ENERGY_STATUS, 0, 0, 0}, + .limits[RAPL_DOMAIN_PACKAGE] = 2, };

/* Handles CPU hotplug on multi-socket systems. diff --git a/include/linux/intel_rapl.h b/include/linux/intel_rapl.h index 649e19981eb0f..0c179d92d1106 100644 --- a/include/linux/intel_rapl.h +++ b/include/linux/intel_rapl.h @@ -104,6 +104,7 @@ struct reg_action { * @pcap_rapl_online: CPU hotplug state for each RAPL interface. * @reg_unit: Register for getting energy/power/time unit. * @regs: Register sets for different RAPL Domains. + * @limits: Number of power limits supported by each domain. * @read_raw: Callback for reading RAPL interface specific * registers. * @write_raw: Callback for writing RAPL interface specific @@ -115,6 +116,7 @@ struct rapl_if_priv { enum cpuhp_state pcap_rapl_online; u64 reg_unit; u64 regs[RAPL_DOMAIN_MAX][RAPL_DOMAIN_REG_MAX]; + int limits[RAPL_DOMAIN_MAX]; int (*read_raw)(int cpu, struct reg_action *ra); int (*write_raw)(int cpu, struct reg_action *ra); };

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 0ab74bcd1b50821391b264150d26b7f03ba6740b category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 0ab74bcd1b50821391b264150d26b7f03ba6740b upstream.

Add IceLake desktop support in intel_rapl driver

Signed-off-by: Gayatri Kammela gayatri.kammela@intel.com Signed-off-by: Joe Konno joe.konno@intel.com Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/powercap/intel_rapl_common.c | 1 + 1 file changed, 1 insertion(+)

diff --git a/drivers/powercap/intel_rapl_common.c b/drivers/powercap/intel_rapl_common.c index f1b7bcc328919..e9e2342616c67 100644 --- a/drivers/powercap/intel_rapl_common.c +++ b/drivers/powercap/intel_rapl_common.c @@ -974,6 +974,7 @@ static const struct x86_cpu_id rapl_ids[] __initconst = { INTEL_CPU_FAM6(KABYLAKE_DESKTOP, rapl_defaults_core), INTEL_CPU_FAM6(CANNONLAKE_MOBILE, rapl_defaults_core), INTEL_CPU_FAM6(ICELAKE_MOBILE, rapl_defaults_core), + INTEL_CPU_FAM6(ICELAKE_DESKTOP, rapl_defaults_core),

INTEL_CPU_FAM6(ATOM_SILVERMONT, rapl_defaults_byt), INTEL_CPU_FAM6(ATOM_AIRMONT, rapl_defaults_cht),

From: Zhang Rui rui.zhang@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 3231a21d5ca6f6baea95588406775304f35a203e category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 3231a21d5ca6f6baea95588406775304f35a203e upstream.

Add ICX-D support in intel_rapl driver

Signed-off-by: Zhang Rui rui.zhang@intel.com Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/powercap/intel_rapl_common.c | 1 + 1 file changed, 1 insertion(+)

diff --git a/drivers/powercap/intel_rapl_common.c b/drivers/powercap/intel_rapl_common.c index 3a5440d90017f..b624a88b2c250 100644 --- a/drivers/powercap/intel_rapl_common.c +++ b/drivers/powercap/intel_rapl_common.c @@ -976,6 +976,7 @@ static const struct x86_cpu_id rapl_ids[] __initconst = { INTEL_CPU_FAM6(ICELAKE_MOBILE, rapl_defaults_core), INTEL_CPU_FAM6(ICELAKE_DESKTOP, rapl_defaults_core), INTEL_CPU_FAM6(ICELAKE_X, rapl_defaults_hsw_server), + INTEL_CPU_FAM6(ICELAKE_XEON_D, rapl_defaults_hsw_server),

INTEL_CPU_FAM6(ATOM_SILVERMONT, rapl_defaults_byt), INTEL_CPU_FAM6(ATOM_AIRMONT, rapl_defaults_cht),

From: Qiuxu Zhuo qiuxu.zhuo@intel.com

mainline inclusion from mainline-v5.1-rc1 commit 98f2fc829e3b55a0ff1b82753fc1e10941535cb9 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

Delete the duplicated code from skx_edac.c and rename skx_edac.c to skx_base.c. Update the Makefile to build the skx_edac driver from skx_base.c and skx_common.c.

Add SPDX to skx_base.c and clean out unnecessary #include lines.

[ bp: Drop the license boilerplate - there's an SPDX identifier now. ]

Co-developed-by: Tony Luck tony.luck@intel.com Signed-off-by: Qiuxu Zhuo qiuxu.zhuo@intel.com Signed-off-by: Tony Luck tony.luck@intel.com Signed-off-by: Borislav Petkov bp@suse.de Cc: James Morse james.morse@arm.com Cc: Mauro Carvalho Chehab mchehab@kernel.org Cc: linux-edac linux-edac@vger.kernel.org Link: https://lkml.kernel.org/r/20190130191519.15393-4-tony.luck@intel.com Signed-off-by: Guoqing Jiang jiangguoqing@kylinos.cn Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/edac/Makefile | 4 +- drivers/edac/skx_base.c | 650 +++++++++++++++++++ drivers/edac/skx_edac.c | 1357 --------------------------------------- 3 files changed, 653 insertions(+), 1358 deletions(-) create mode 100644 drivers/edac/skx_base.c delete mode 100644 drivers/edac/skx_edac.c

diff --git a/drivers/edac/Makefile b/drivers/edac/Makefile index 02b43a7d8c3ee..b250d59910ba1 100644 --- a/drivers/edac/Makefile +++ b/drivers/edac/Makefile @@ -30,7 +30,6 @@ obj-$(CONFIG_EDAC_I5400) += i5400_edac.o obj-$(CONFIG_EDAC_I7300) += i7300_edac.o obj-$(CONFIG_EDAC_I7CORE) += i7core_edac.o obj-$(CONFIG_EDAC_SBRIDGE) += sb_edac.o -obj-$(CONFIG_EDAC_SKX) += skx_edac.o obj-$(CONFIG_EDAC_PND2) += pnd2_edac.o obj-$(CONFIG_EDAC_E7XXX) += e7xxx_edac.o obj-$(CONFIG_EDAC_E752X) += e752x_edac.o @@ -58,6 +57,9 @@ obj-$(CONFIG_EDAC_MPC85XX) += mpc85xx_edac_mod.o layerscape_edac_mod-y := fsl_ddr_edac.o layerscape_edac.o obj-$(CONFIG_EDAC_LAYERSCAPE) += layerscape_edac_mod.o

+skx_edac-y := skx_common.o skx_base.o +obj-$(CONFIG_EDAC_SKX) += skx_edac.o + obj-$(CONFIG_EDAC_MV64X60) += mv64x60_edac.o obj-$(CONFIG_EDAC_CELL) += cell_edac.o obj-$(CONFIG_EDAC_PPC4XX) += ppc4xx_edac.o diff --git a/drivers/edac/skx_base.c b/drivers/edac/skx_base.c new file mode 100644 index 0000000000000..adae4c848ca1e --- /dev/null +++ b/drivers/edac/skx_base.c @@ -0,0 +1,650 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * EDAC driver for Intel(R) Xeon(R) Skylake processors + * Copyright (c) 2016, Intel Corporation. + */ + +#include <linux/kernel.h> +#include <linux/processor.h> +#include <asm/cpu_device_id.h> +#include <asm/intel-family.h> +#include <asm/mce.h> + +#include "edac_module.h" +#include "skx_common.h" + +#define EDAC_MOD_STR "skx_edac" + +/* + * Debug macros + */ +#define skx_printk(level, fmt, arg...) \ + edac_printk(level, "skx", fmt, ##arg) + +#define skx_mc_printk(mci, level, fmt, arg...) \ + edac_mc_chipset_printk(mci, level, "skx", fmt, ##arg) + +static struct list_head *skx_edac_list; + +static u64 skx_tolm, skx_tohm; +static int skx_num_sockets; +static unsigned int nvdimm_count; + +#define MASK26 0x3FFFFFF /* Mask for 2^26 */ +#define MASK29 0x1FFFFFFF /* Mask for 2^29 */ + +static struct skx_dev *get_skx_dev(struct pci_bus *bus, u8 idx) +{ + struct skx_dev *d; + + list_for_each_entry(d, skx_edac_list, list) { + if (d->seg == pci_domain_nr(bus) && d->bus[idx] == bus->number) + return d; + } + + return NULL; +} + +enum munittype { + CHAN0, CHAN1, CHAN2, SAD_ALL, UTIL_ALL, SAD +}; + +struct munit { + u16 did; + u16 devfn[SKX_NUM_IMC]; + u8 busidx; + u8 per_socket; + enum munittype mtype; +}; + +/* + * List of PCI device ids that we need together with some device + * number and function numbers to tell which memory controller the + * device belongs to. + */ +static const struct munit skx_all_munits[] = { + { 0x2054, { }, 1, 1, SAD_ALL }, + { 0x2055, { }, 1, 1, UTIL_ALL }, + { 0x2040, { PCI_DEVFN(10, 0), PCI_DEVFN(12, 0) }, 2, 2, CHAN0 }, + { 0x2044, { PCI_DEVFN(10, 4), PCI_DEVFN(12, 4) }, 2, 2, CHAN1 }, + { 0x2048, { PCI_DEVFN(11, 0), PCI_DEVFN(13, 0) }, 2, 2, CHAN2 }, + { 0x208e, { }, 1, 0, SAD }, + { } +}; + +static int get_all_munits(const struct munit *m) +{ + struct pci_dev *pdev, *prev; + struct skx_dev *d; + u32 reg; + int i = 0, ndev = 0; + + prev = NULL; + for (;;) { + pdev = pci_get_device(PCI_VENDOR_ID_INTEL, m->did, prev); + if (!pdev) + break; + ndev++; + if (m->per_socket == SKX_NUM_IMC) { + for (i = 0; i < SKX_NUM_IMC; i++) + if (m->devfn[i] == pdev->devfn) + break; + if (i == SKX_NUM_IMC) + goto fail; + } + d = get_skx_dev(pdev->bus, m->busidx); + if (!d) + goto fail; + + /* Be sure that the device is enabled */ + if (unlikely(pci_enable_device(pdev) < 0)) { + skx_printk(KERN_ERR, "Couldn't enable device %04x:%04x\n", + PCI_VENDOR_ID_INTEL, m->did); + goto fail; + } + + switch (m->mtype) { + case CHAN0: case CHAN1: case CHAN2: + pci_dev_get(pdev); + d->imc[i].chan[m->mtype].cdev = pdev; + break; + case SAD_ALL: + pci_dev_get(pdev); + d->sad_all = pdev; + break; + case UTIL_ALL: + pci_dev_get(pdev); + d->util_all = pdev; + break; + case SAD: + /* + * one of these devices per core, including cores + * that don't exist on this SKU. Ignore any that + * read a route table of zero, make sure all the + * non-zero values match. + */ + pci_read_config_dword(pdev, 0xB4, &reg); + if (reg != 0) { + if (d->mcroute == 0) { + d->mcroute = reg; + } else if (d->mcroute != reg) { + skx_printk(KERN_ERR, "mcroute mismatch\n"); + goto fail; + } + } + ndev--; + break; + } + + prev = pdev; + } + + return ndev; +fail: + pci_dev_put(pdev); + return -ENODEV; +} + +static const struct x86_cpu_id skx_cpuids[] = { + { X86_VENDOR_INTEL, 6, INTEL_FAM6_SKYLAKE_X, 0, 0 }, + { } +}; +MODULE_DEVICE_TABLE(x86cpu, skx_cpuids); + +#define SKX_GET_MTMTR(dev, reg) \ + pci_read_config_dword((dev), 0x87c, &(reg)) + +static bool skx_check_ecc(struct pci_dev *pdev) +{ + u32 mtmtr; + + SKX_GET_MTMTR(pdev, mtmtr); + + return !!GET_BITFIELD(mtmtr, 2, 2); +} + +static int skx_get_dimm_config(struct mem_ctl_info *mci) +{ + struct skx_pvt *pvt = mci->pvt_info; + struct skx_imc *imc = pvt->imc; + u32 mtr, amap, mcddrtcfg; + struct dimm_info *dimm; + int i, j; + int ndimms; + + for (i = 0; i < SKX_NUM_CHANNELS; i++) { + ndimms = 0; + pci_read_config_dword(imc->chan[i].cdev, 0x8C, &amap); + pci_read_config_dword(imc->chan[i].cdev, 0x400, &mcddrtcfg); + for (j = 0; j < SKX_NUM_DIMMS; j++) { + dimm = EDAC_DIMM_PTR(mci->layers, mci->dimms, + mci->n_layers, i, j, 0); + pci_read_config_dword(imc->chan[i].cdev, + 0x80 + 4 * j, &mtr); + if (IS_DIMM_PRESENT(mtr)) { + ndimms += skx_get_dimm_info(mtr, amap, dimm, imc, i, j); + } else if (IS_NVDIMM_PRESENT(mcddrtcfg, j)) { + ndimms += skx_get_nvdimm_info(dimm, imc, i, j, + EDAC_MOD_STR); + nvdimm_count++; + } + } + if (ndimms && !skx_check_ecc(imc->chan[0].cdev)) { + skx_printk(KERN_ERR, "ECC is disabled on imc %d\n", imc->mc); + return -ENODEV; + } + } + + return 0; +} + +#define SKX_MAX_SAD 24 + +#define SKX_GET_SAD(d, i, reg) \ + pci_read_config_dword((d)->sad_all, 0x60 + 8 * (i), &(reg)) +#define SKX_GET_ILV(d, i, reg) \ + pci_read_config_dword((d)->sad_all, 0x64 + 8 * (i), &(reg)) + +#define SKX_SAD_MOD3MODE(sad) GET_BITFIELD((sad), 30, 31) +#define SKX_SAD_MOD3(sad) GET_BITFIELD((sad), 27, 27) +#define SKX_SAD_LIMIT(sad) (((u64)GET_BITFIELD((sad), 7, 26) << 26) | MASK26) +#define SKX_SAD_MOD3ASMOD2(sad) GET_BITFIELD((sad), 5, 6) +#define SKX_SAD_ATTR(sad) GET_BITFIELD((sad), 3, 4) +#define SKX_SAD_INTERLEAVE(sad) GET_BITFIELD((sad), 1, 2) +#define SKX_SAD_ENABLE(sad) GET_BITFIELD((sad), 0, 0) + +#define SKX_ILV_REMOTE(tgt) (((tgt) & 8) == 0) +#define SKX_ILV_TARGET(tgt) ((tgt) & 7) + +static bool skx_sad_decode(struct decoded_addr *res) +{ + struct skx_dev *d = list_first_entry(skx_edac_list, typeof(*d), list); + u64 addr = res->addr; + int i, idx, tgt, lchan, shift; + u32 sad, ilv; + u64 limit, prev_limit; + int remote = 0; + + /* Simple sanity check for I/O space or out of range */ + if (addr >= skx_tohm || (addr >= skx_tolm && addr < BIT_ULL(32))) { + edac_dbg(0, "Address 0x%llx out of range\n", addr); + return false; + } + +restart: + prev_limit = 0; + for (i = 0; i < SKX_MAX_SAD; i++) { + SKX_GET_SAD(d, i, sad); + limit = SKX_SAD_LIMIT(sad); + if (SKX_SAD_ENABLE(sad)) { + if (addr >= prev_limit && addr <= limit) + goto sad_found; + } + prev_limit = limit + 1; + } + edac_dbg(0, "No SAD entry for 0x%llx\n", addr); + return false; + +sad_found: + SKX_GET_ILV(d, i, ilv); + + switch (SKX_SAD_INTERLEAVE(sad)) { + case 0: + idx = GET_BITFIELD(addr, 6, 8); + break; + case 1: + idx = GET_BITFIELD(addr, 8, 10); + break; + case 2: + idx = GET_BITFIELD(addr, 12, 14); + break; + case 3: + idx = GET_BITFIELD(addr, 30, 32); + break; + } + + tgt = GET_BITFIELD(ilv, 4 * idx, 4 * idx + 3); + + /* If point to another node, find it and start over */ + if (SKX_ILV_REMOTE(tgt)) { + if (remote) { + edac_dbg(0, "Double remote!\n"); + return false; + } + remote = 1; + list_for_each_entry(d, skx_edac_list, list) { + if (d->imc[0].src_id == SKX_ILV_TARGET(tgt)) + goto restart; + } + edac_dbg(0, "Can't find node %d\n", SKX_ILV_TARGET(tgt)); + return false; + } + + if (SKX_SAD_MOD3(sad) == 0) { + lchan = SKX_ILV_TARGET(tgt); + } else { + switch (SKX_SAD_MOD3MODE(sad)) { + case 0: + shift = 6; + break; + case 1: + shift = 8; + break; + case 2: + shift = 12; + break; + default: + edac_dbg(0, "illegal mod3mode\n"); + return false; + } + switch (SKX_SAD_MOD3ASMOD2(sad)) { + case 0: + lchan = (addr >> shift) % 3; + break; + case 1: + lchan = (addr >> shift) % 2; + break; + case 2: + lchan = (addr >> shift) % 2; + lchan = (lchan << 1) | !lchan; + break; + case 3: + lchan = ((addr >> shift) % 2) << 1; + break; + } + lchan = (lchan << 1) | (SKX_ILV_TARGET(tgt) & 1); + } + + res->dev = d; + res->socket = d->imc[0].src_id; + res->imc = GET_BITFIELD(d->mcroute, lchan * 3, lchan * 3 + 2); + res->channel = GET_BITFIELD(d->mcroute, lchan * 2 + 18, lchan * 2 + 19); + + edac_dbg(2, "0x%llx: socket=%d imc=%d channel=%d\n", + res->addr, res->socket, res->imc, res->channel); + return true; +} + +#define SKX_MAX_TAD 8 + +#define SKX_GET_TADBASE(d, mc, i, reg) \ + pci_read_config_dword((d)->imc[mc].chan[0].cdev, 0x850 + 4 * (i), &(reg)) +#define SKX_GET_TADWAYNESS(d, mc, i, reg) \ + pci_read_config_dword((d)->imc[mc].chan[0].cdev, 0x880 + 4 * (i), &(reg)) +#define SKX_GET_TADCHNILVOFFSET(d, mc, ch, i, reg) \ + pci_read_config_dword((d)->imc[mc].chan[ch].cdev, 0x90 + 4 * (i), &(reg)) + +#define SKX_TAD_BASE(b) ((u64)GET_BITFIELD((b), 12, 31) << 26) +#define SKX_TAD_SKT_GRAN(b) GET_BITFIELD((b), 4, 5) +#define SKX_TAD_CHN_GRAN(b) GET_BITFIELD((b), 6, 7) +#define SKX_TAD_LIMIT(b) (((u64)GET_BITFIELD((b), 12, 31) << 26) | MASK26) +#define SKX_TAD_OFFSET(b) ((u64)GET_BITFIELD((b), 4, 23) << 26) +#define SKX_TAD_SKTWAYS(b) (1 << GET_BITFIELD((b), 10, 11)) +#define SKX_TAD_CHNWAYS(b) (GET_BITFIELD((b), 8, 9) + 1) + +/* which bit used for both socket and channel interleave */ +static int skx_granularity[] = { 6, 8, 12, 30 }; + +static u64 skx_do_interleave(u64 addr, int shift, int ways, u64 lowbits) +{ + addr >>= shift; + addr /= ways; + addr <<= shift; + + return addr | (lowbits & ((1ull << shift) - 1)); +} + +static bool skx_tad_decode(struct decoded_addr *res) +{ + int i; + u32 base, wayness, chnilvoffset; + int skt_interleave_bit, chn_interleave_bit; + u64 channel_addr; + + for (i = 0; i < SKX_MAX_TAD; i++) { + SKX_GET_TADBASE(res->dev, res->imc, i, base); + SKX_GET_TADWAYNESS(res->dev, res->imc, i, wayness); + if (SKX_TAD_BASE(base) <= res->addr && res->addr <= SKX_TAD_LIMIT(wayness)) + goto tad_found; + } + edac_dbg(0, "No TAD entry for 0x%llx\n", res->addr); + return false; + +tad_found: + res->sktways = SKX_TAD_SKTWAYS(wayness); + res->chanways = SKX_TAD_CHNWAYS(wayness); + skt_interleave_bit = skx_granularity[SKX_TAD_SKT_GRAN(base)]; + chn_interleave_bit = skx_granularity[SKX_TAD_CHN_GRAN(base)]; + + SKX_GET_TADCHNILVOFFSET(res->dev, res->imc, res->channel, i, chnilvoffset); + channel_addr = res->addr - SKX_TAD_OFFSET(chnilvoffset); + + if (res->chanways == 3 && skt_interleave_bit > chn_interleave_bit) { + /* Must handle channel first, then socket */ + channel_addr = skx_do_interleave(channel_addr, chn_interleave_bit, + res->chanways, channel_addr); + channel_addr = skx_do_interleave(channel_addr, skt_interleave_bit, + res->sktways, channel_addr); + } else { + /* Handle socket then channel. Preserve low bits from original address */ + channel_addr = skx_do_interleave(channel_addr, skt_interleave_bit, + res->sktways, res->addr); + channel_addr = skx_do_interleave(channel_addr, chn_interleave_bit, + res->chanways, res->addr); + } + + res->chan_addr = channel_addr; + + edac_dbg(2, "0x%llx: chan_addr=0x%llx sktways=%d chanways=%d\n", + res->addr, res->chan_addr, res->sktways, res->chanways); + return true; +} + +#define SKX_MAX_RIR 4 + +#define SKX_GET_RIRWAYNESS(d, mc, ch, i, reg) \ + pci_read_config_dword((d)->imc[mc].chan[ch].cdev, \ + 0x108 + 4 * (i), &(reg)) +#define SKX_GET_RIRILV(d, mc, ch, idx, i, reg) \ + pci_read_config_dword((d)->imc[mc].chan[ch].cdev, \ + 0x120 + 16 * (idx) + 4 * (i), &(reg)) + +#define SKX_RIR_VALID(b) GET_BITFIELD((b), 31, 31) +#define SKX_RIR_LIMIT(b) (((u64)GET_BITFIELD((b), 1, 11) << 29) | MASK29) +#define SKX_RIR_WAYS(b) (1 << GET_BITFIELD((b), 28, 29)) +#define SKX_RIR_CHAN_RANK(b) GET_BITFIELD((b), 16, 19) +#define SKX_RIR_OFFSET(b) ((u64)(GET_BITFIELD((b), 2, 15) << 26)) + +static bool skx_rir_decode(struct decoded_addr *res) +{ + int i, idx, chan_rank; + int shift; + u32 rirway, rirlv; + u64 rank_addr, prev_limit = 0, limit; + + if (res->dev->imc[res->imc].chan[res->channel].dimms[0].close_pg) + shift = 6; + else + shift = 13; + + for (i = 0; i < SKX_MAX_RIR; i++) { + SKX_GET_RIRWAYNESS(res->dev, res->imc, res->channel, i, rirway); + limit = SKX_RIR_LIMIT(rirway); + if (SKX_RIR_VALID(rirway)) { + if (prev_limit <= res->chan_addr && + res->chan_addr <= limit) + goto rir_found; + } + prev_limit = limit; + } + edac_dbg(0, "No RIR entry for 0x%llx\n", res->addr); + return false; + +rir_found: + rank_addr = res->chan_addr >> shift; + rank_addr /= SKX_RIR_WAYS(rirway); + rank_addr <<= shift; + rank_addr |= res->chan_addr & GENMASK_ULL(shift - 1, 0); + + res->rank_address = rank_addr; + idx = (res->chan_addr >> shift) % SKX_RIR_WAYS(rirway); + + SKX_GET_RIRILV(res->dev, res->imc, res->channel, idx, i, rirlv); + res->rank_address = rank_addr - SKX_RIR_OFFSET(rirlv); + chan_rank = SKX_RIR_CHAN_RANK(rirlv); + res->channel_rank = chan_rank; + res->dimm = chan_rank / 4; + res->rank = chan_rank % 4; + + edac_dbg(2, "0x%llx: dimm=%d rank=%d chan_rank=%d rank_addr=0x%llx\n", + res->addr, res->dimm, res->rank, + res->channel_rank, res->rank_address); + return true; +} + +static u8 skx_close_row[] = { + 15, 16, 17, 18, 20, 21, 22, 28, 10, 11, 12, 13, 29, 30, 31, 32, 33 +}; + +static u8 skx_close_column[] = { + 3, 4, 5, 14, 19, 23, 24, 25, 26, 27 +}; + +static u8 skx_open_row[] = { + 14, 15, 16, 20, 28, 21, 22, 23, 24, 25, 26, 27, 29, 30, 31, 32, 33 +}; + +static u8 skx_open_column[] = { + 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 +}; + +static u8 skx_open_fine_column[] = { + 3, 4, 5, 7, 8, 9, 10, 11, 12, 13 +}; + +static int skx_bits(u64 addr, int nbits, u8 *bits) +{ + int i, res = 0; + + for (i = 0; i < nbits; i++) + res |= ((addr >> bits[i]) & 1) << i; + return res; +} + +static int skx_bank_bits(u64 addr, int b0, int b1, int do_xor, int x0, int x1) +{ + int ret = GET_BITFIELD(addr, b0, b0) | (GET_BITFIELD(addr, b1, b1) << 1); + + if (do_xor) + ret ^= GET_BITFIELD(addr, x0, x0) | (GET_BITFIELD(addr, x1, x1) << 1); + + return ret; +} + +static bool skx_mad_decode(struct decoded_addr *r) +{ + struct skx_dimm *dimm = &r->dev->imc[r->imc].chan[r->channel].dimms[r->dimm]; + int bg0 = dimm->fine_grain_bank ? 6 : 13; + + if (dimm->close_pg) { + r->row = skx_bits(r->rank_address, dimm->rowbits, skx_close_row); + r->column = skx_bits(r->rank_address, dimm->colbits, skx_close_column); + r->column |= 0x400; /* C10 is autoprecharge, always set */ + r->bank_address = skx_bank_bits(r->rank_address, 8, 9, dimm->bank_xor_enable, 22, 28); + r->bank_group = skx_bank_bits(r->rank_address, 6, 7, dimm->bank_xor_enable, 20, 21); + } else { + r->row = skx_bits(r->rank_address, dimm->rowbits, skx_open_row); + if (dimm->fine_grain_bank) + r->column = skx_bits(r->rank_address, dimm->colbits, skx_open_fine_column); + else + r->column = skx_bits(r->rank_address, dimm->colbits, skx_open_column); + r->bank_address = skx_bank_bits(r->rank_address, 18, 19, dimm->bank_xor_enable, 22, 23); + r->bank_group = skx_bank_bits(r->rank_address, bg0, 17, dimm->bank_xor_enable, 20, 21); + } + r->row &= (1u << dimm->rowbits) - 1; + + edac_dbg(2, "0x%llx: row=0x%x col=0x%x bank_addr=%d bank_group=%d\n", + r->addr, r->row, r->column, r->bank_address, + r->bank_group); + return true; +} + +static bool skx_decode(struct decoded_addr *res) +{ + return skx_sad_decode(res) && skx_tad_decode(res) && + skx_rir_decode(res) && skx_mad_decode(res); +} + +static struct notifier_block skx_mce_dec = { + .notifier_call = skx_mce_check_error, + .priority = MCE_PRIO_EDAC, +}; + +/* + * skx_init: + * make sure we are running on the correct cpu model + * search for all the devices we need + * check which DIMMs are present. + */ +static int __init skx_init(void) +{ + const struct x86_cpu_id *id; + const struct munit *m; + const char *owner; + int rc = 0, i, off[3] = {0xd0, 0xd4, 0xd8}; + u8 mc = 0, src_id, node_id; + struct skx_dev *d; + + edac_dbg(2, "\n"); + + owner = edac_get_owner(); + if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR))) + return -EBUSY; + + id = x86_match_cpu(skx_cpuids); + if (!id) + return -ENODEV; + + rc = skx_get_hi_lo(0x2034, off, &skx_tolm, &skx_tohm); + if (rc) + return rc; + + rc = skx_get_all_bus_mappings(0x2016, 0xcc, SKX, &skx_edac_list); + if (rc < 0) + goto fail; + if (rc == 0) { + edac_dbg(2, "No memory controllers found\n"); + return -ENODEV; + } + skx_num_sockets = rc; + + for (m = skx_all_munits; m->did; m++) { + rc = get_all_munits(m); + if (rc < 0) + goto fail; + if (rc != m->per_socket * skx_num_sockets) { + edac_dbg(2, "Expected %d, got %d of 0x%x\n", + m->per_socket * skx_num_sockets, rc, m->did); + rc = -ENODEV; + goto fail; + } + } + + list_for_each_entry(d, skx_edac_list, list) { + rc = skx_get_src_id(d, &src_id); + if (rc < 0) + goto fail; + rc = skx_get_node_id(d, &node_id); + if (rc < 0) + goto fail; + edac_dbg(2, "src_id=%d node_id=%d\n", src_id, node_id); + for (i = 0; i < SKX_NUM_IMC; i++) { + d->imc[i].mc = mc++; + d->imc[i].lmc = i; + d->imc[i].src_id = src_id; + d->imc[i].node_id = node_id; + rc = skx_register_mci(&d->imc[i], d->imc[i].chan[0].cdev, + "Skylake Socket", EDAC_MOD_STR, + skx_get_dimm_config); + if (rc < 0) + goto fail; + } + } + + skx_set_decode(skx_decode); + + if (nvdimm_count && skx_adxl_get() == -ENODEV) + skx_printk(KERN_NOTICE, "Only decoding DDR4 address!\n"); + + /* Ensure that the OPSTATE is set correctly for POLL or NMI */ + opstate_init(); + + setup_skx_debug("skx_test"); + + mce_register_decode_chain(&skx_mce_dec); + + return 0; +fail: + skx_remove(); + return rc; +} + +static void __exit skx_exit(void) +{ + edac_dbg(2, "\n"); + mce_unregister_decode_chain(&skx_mce_dec); + teardown_skx_debug(); + if (nvdimm_count) + skx_adxl_put(); + skx_remove(); +} + +module_init(skx_init); +module_exit(skx_exit); + +module_param(edac_op_state, int, 0444); +MODULE_PARM_DESC(edac_op_state, "EDAC Error Reporting state: 0=Poll,1=NMI"); + +MODULE_LICENSE("GPL v2"); +MODULE_AUTHOR("Tony Luck"); +MODULE_DESCRIPTION("MC Driver for Intel Skylake server processors"); diff --git a/drivers/edac/skx_edac.c b/drivers/edac/skx_edac.c deleted file mode 100644 index a99ea61dad321..0000000000000 --- a/drivers/edac/skx_edac.c +++ /dev/null @@ -1,1357 +0,0 @@ -/* - * EDAC driver for Intel(R) Xeon(R) Skylake processors - * Copyright (c) 2016, Intel Corporation. - * - * This program is free software; you can redistribute it and/or modify it - * under the terms and conditions of the GNU General Public License, - * version 2, as published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for - * more details. - */ - -#include <linux/module.h> -#include <linux/init.h> -#include <linux/acpi.h> -#include <linux/dmi.h> -#include <linux/pci.h> -#include <linux/pci_ids.h> -#include <linux/slab.h> -#include <linux/delay.h> -#include <linux/edac.h> -#include <linux/mmzone.h> -#include <linux/smp.h> -#include <linux/bitmap.h> -#include <linux/math64.h> -#include <linux/mod_devicetable.h> -#include <linux/adxl.h> -#include <acpi/nfit.h> -#include <asm/cpu_device_id.h> -#include <asm/intel-family.h> -#include <asm/processor.h> -#include <asm/mce.h> - -#include "edac_module.h" - -#define EDAC_MOD_STR "skx_edac" -#define MSG_SIZE 1024 - -/* - * Debug macros - */ -#define skx_printk(level, fmt, arg...) \ - edac_printk(level, "skx", fmt, ##arg) - -#define skx_mc_printk(mci, level, fmt, arg...) \ - edac_mc_chipset_printk(mci, level, "skx", fmt, ##arg) - -/* - * Get a bit field at register value <v>, from bit <lo> to bit <hi> - */ -#define GET_BITFIELD(v, lo, hi) \ - (((v) & GENMASK_ULL((hi), (lo))) >> (lo)) - -static LIST_HEAD(skx_edac_list); - -static u64 skx_tolm, skx_tohm; -static char *skx_msg; -static unsigned int nvdimm_count; - -enum { - INDEX_SOCKET, - INDEX_MEMCTRL, - INDEX_CHANNEL, - INDEX_DIMM, - INDEX_MAX -}; - -static const char * const component_names[] = { - [INDEX_SOCKET] = "ProcessorSocketId", - [INDEX_MEMCTRL] = "MemoryControllerId", - [INDEX_CHANNEL] = "ChannelId", - [INDEX_DIMM] = "DimmSlotId", -}; - -static int component_indices[ARRAY_SIZE(component_names)]; -static int adxl_component_count; -static const char * const *adxl_component_names; -static u64 *adxl_values; -static char *adxl_msg; - -#define NUM_IMC 2 /* memory controllers per socket */ -#define NUM_CHANNELS 3 /* channels per memory controller */ -#define NUM_DIMMS 2 /* Max DIMMS per channel */ - -#define MASK26 0x3FFFFFF /* Mask for 2^26 */ -#define MASK29 0x1FFFFFFF /* Mask for 2^29 */ - -/* - * Each cpu socket contains some pci devices that provide global - * information, and also some that are local to each of the two - * memory controllers on the die. - */ -struct skx_dev { - struct list_head list; - u8 bus[4]; - int seg; - struct pci_dev *sad_all; - struct pci_dev *util_all; - u32 mcroute; - struct skx_imc { - struct mem_ctl_info *mci; - u8 mc; /* system wide mc# */ - u8 lmc; /* socket relative mc# */ - u8 src_id, node_id; - struct skx_channel { - struct pci_dev *cdev; - struct skx_dimm { - u8 close_pg; - u8 bank_xor_enable; - u8 fine_grain_bank; - u8 rowbits; - u8 colbits; - } dimms[NUM_DIMMS]; - } chan[NUM_CHANNELS]; - } imc[NUM_IMC]; -}; -static int skx_num_sockets; - -struct skx_pvt { - struct skx_imc *imc; -}; - -struct decoded_addr { - struct skx_dev *dev; - u64 addr; - int socket; - int imc; - int channel; - u64 chan_addr; - int sktways; - int chanways; - int dimm; - int rank; - int channel_rank; - u64 rank_address; - int row; - int column; - int bank_address; - int bank_group; -}; - -static struct skx_dev *get_skx_dev(struct pci_bus *bus, u8 idx) -{ - struct skx_dev *d; - - list_for_each_entry(d, &skx_edac_list, list) { - if (d->seg == pci_domain_nr(bus) && d->bus[idx] == bus->number) - return d; - } - - return NULL; -} - -enum munittype { - CHAN0, CHAN1, CHAN2, SAD_ALL, UTIL_ALL, SAD -}; - -struct munit { - u16 did; - u16 devfn[NUM_IMC]; - u8 busidx; - u8 per_socket; - enum munittype mtype; -}; - -/* - * List of PCI device ids that we need together with some device - * number and function numbers to tell which memory controller the - * device belongs to. - */ -static const struct munit skx_all_munits[] = { - { 0x2054, { }, 1, 1, SAD_ALL }, - { 0x2055, { }, 1, 1, UTIL_ALL }, - { 0x2040, { PCI_DEVFN(10, 0), PCI_DEVFN(12, 0) }, 2, 2, CHAN0 }, - { 0x2044, { PCI_DEVFN(10, 4), PCI_DEVFN(12, 4) }, 2, 2, CHAN1 }, - { 0x2048, { PCI_DEVFN(11, 0), PCI_DEVFN(13, 0) }, 2, 2, CHAN2 }, - { 0x208e, { }, 1, 0, SAD }, - { } -}; - -/* - * We use the per-socket device 0x2016 to count how many sockets are present, - * and to detemine which PCI buses are associated with each socket. Allocate - * and build the full list of all the skx_dev structures that we need here. - */ -static int get_all_bus_mappings(void) -{ - struct pci_dev *pdev, *prev; - struct skx_dev *d; - u32 reg; - int ndev = 0; - - prev = NULL; - for (;;) { - pdev = pci_get_device(PCI_VENDOR_ID_INTEL, 0x2016, prev); - if (!pdev) - break; - ndev++; - d = kzalloc(sizeof(*d), GFP_KERNEL); - if (!d) { - pci_dev_put(pdev); - return -ENOMEM; - } - d->seg = pci_domain_nr(pdev->bus); - pci_read_config_dword(pdev, 0xCC, &reg); - d->bus[0] = GET_BITFIELD(reg, 0, 7); - d->bus[1] = GET_BITFIELD(reg, 8, 15); - d->bus[2] = GET_BITFIELD(reg, 16, 23); - d->bus[3] = GET_BITFIELD(reg, 24, 31); - edac_dbg(2, "busses: %x, %x, %x, %x\n", - d->bus[0], d->bus[1], d->bus[2], d->bus[3]); - list_add_tail(&d->list, &skx_edac_list); - skx_num_sockets++; - prev = pdev; - } - - return ndev; -} - -static int get_all_munits(const struct munit *m) -{ - struct pci_dev *pdev, *prev; - struct skx_dev *d; - u32 reg; - int i = 0, ndev = 0; - - prev = NULL; - for (;;) { - pdev = pci_get_device(PCI_VENDOR_ID_INTEL, m->did, prev); - if (!pdev) - break; - ndev++; - if (m->per_socket == NUM_IMC) { - for (i = 0; i < NUM_IMC; i++) - if (m->devfn[i] == pdev->devfn) - break; - if (i == NUM_IMC) - goto fail; - } - d = get_skx_dev(pdev->bus, m->busidx); - if (!d) - goto fail; - - /* Be sure that the device is enabled */ - if (unlikely(pci_enable_device(pdev) < 0)) { - skx_printk(KERN_ERR, - "Couldn't enable %04x:%04x\n", PCI_VENDOR_ID_INTEL, m->did); - goto fail; - } - - switch (m->mtype) { - case CHAN0: case CHAN1: case CHAN2: - pci_dev_get(pdev); - d->imc[i].chan[m->mtype].cdev = pdev; - break; - case SAD_ALL: - pci_dev_get(pdev); - d->sad_all = pdev; - break; - case UTIL_ALL: - pci_dev_get(pdev); - d->util_all = pdev; - break; - case SAD: - /* - * one of these devices per core, including cores - * that don't exist on this SKU. Ignore any that - * read a route table of zero, make sure all the - * non-zero values match. - */ - pci_read_config_dword(pdev, 0xB4, &reg); - if (reg != 0) { - if (d->mcroute == 0) - d->mcroute = reg; - else if (d->mcroute != reg) { - skx_printk(KERN_ERR, - "mcroute mismatch\n"); - goto fail; - } - } - ndev--; - break; - } - - prev = pdev; - } - - return ndev; -fail: - pci_dev_put(pdev); - return -ENODEV; -} - -static const struct x86_cpu_id skx_cpuids[] = { - { X86_VENDOR_INTEL, 6, INTEL_FAM6_SKYLAKE_X, 0, 0 }, - { } -}; -MODULE_DEVICE_TABLE(x86cpu, skx_cpuids); - -static u8 get_src_id(struct skx_dev *d) -{ - u32 reg; - - pci_read_config_dword(d->util_all, 0xF0, &reg); - - return GET_BITFIELD(reg, 12, 14); -} - -static u8 skx_get_node_id(struct skx_dev *d) -{ - u32 reg; - - pci_read_config_dword(d->util_all, 0xF4, &reg); - - return GET_BITFIELD(reg, 0, 2); -} - -static int get_dimm_attr(u32 reg, int lobit, int hibit, int add, int minval, - int maxval, char *name) -{ - u32 val = GET_BITFIELD(reg, lobit, hibit); - - if (val < minval || val > maxval) { - edac_dbg(2, "bad %s = %d (raw=%x)\n", name, val, reg); - return -EINVAL; - } - return val + add; -} - -#define IS_DIMM_PRESENT(mtr) GET_BITFIELD((mtr), 15, 15) -#define IS_NVDIMM_PRESENT(mcddrtcfg, i) GET_BITFIELD((mcddrtcfg), (i), (i)) - -#define numrank(reg) get_dimm_attr((reg), 12, 13, 0, 0, 2, "ranks") -#define numrow(reg) get_dimm_attr((reg), 2, 4, 12, 1, 6, "rows") -#define numcol(reg) get_dimm_attr((reg), 0, 1, 10, 0, 2, "cols") - -static int get_width(u32 mtr) -{ - switch (GET_BITFIELD(mtr, 8, 9)) { - case 0: - return DEV_X4; - case 1: - return DEV_X8; - case 2: - return DEV_X16; - } - return DEV_UNKNOWN; -} - -static int skx_get_hi_lo(void) -{ - struct pci_dev *pdev; - u32 reg; - - pdev = pci_get_device(PCI_VENDOR_ID_INTEL, 0x2034, NULL); - if (!pdev) { - edac_dbg(0, "Can't get tolm/tohm\n"); - return -ENODEV; - } - - pci_read_config_dword(pdev, 0xD0, &reg); - skx_tolm = reg; - pci_read_config_dword(pdev, 0xD4, &reg); - skx_tohm = reg; - pci_read_config_dword(pdev, 0xD8, &reg); - skx_tohm |= (u64)reg << 32; - - pci_dev_put(pdev); - edac_dbg(2, "tolm=%llx tohm=%llx\n", skx_tolm, skx_tohm); - - return 0; -} - -static int get_dimm_info(u32 mtr, u32 amap, struct dimm_info *dimm, - struct skx_imc *imc, int chan, int dimmno) -{ - int banks = 16, ranks, rows, cols, npages; - u64 size; - - ranks = numrank(mtr); - rows = numrow(mtr); - cols = numcol(mtr); - - /* - * Compute size in 8-byte (2^3) words, then shift to MiB (2^20) - */ - size = ((1ull << (rows + cols + ranks)) * banks) >> (20 - 3); - npages = MiB_TO_PAGES(size); - - edac_dbg(0, "mc#%d: channel %d, dimm %d, %lld MiB (%d pages) bank: %d, rank: %d, row: %#x, col: %#x\n", - imc->mc, chan, dimmno, size, npages, - banks, 1 << ranks, rows, cols); - - imc->chan[chan].dimms[dimmno].close_pg = GET_BITFIELD(mtr, 0, 0); - imc->chan[chan].dimms[dimmno].bank_xor_enable = GET_BITFIELD(mtr, 9, 9); - imc->chan[chan].dimms[dimmno].fine_grain_bank = GET_BITFIELD(amap, 0, 0); - imc->chan[chan].dimms[dimmno].rowbits = rows; - imc->chan[chan].dimms[dimmno].colbits = cols; - - dimm->nr_pages = npages; - dimm->grain = 32; - dimm->dtype = get_width(mtr); - dimm->mtype = MEM_DDR4; - dimm->edac_mode = EDAC_SECDED; /* likely better than this */ - snprintf(dimm->label, sizeof(dimm->label), "CPU_SrcID#%u_MC#%u_Chan#%u_DIMM#%u", - imc->src_id, imc->lmc, chan, dimmno); - - return 1; -} - -static int get_nvdimm_info(struct dimm_info *dimm, struct skx_imc *imc, - int chan, int dimmno) -{ - int smbios_handle; - u32 dev_handle; - u16 flags; - u64 size = 0; - - nvdimm_count++; - - dev_handle = ACPI_NFIT_BUILD_DEVICE_HANDLE(dimmno, chan, imc->lmc, - imc->src_id, 0); - - smbios_handle = nfit_get_smbios_id(dev_handle, &flags); - if (smbios_handle == -EOPNOTSUPP) { - pr_warn_once(EDAC_MOD_STR ": Can't find size of NVDIMM. Try enabling CONFIG_ACPI_NFIT\n"); - goto unknown_size; - } - - if (smbios_handle < 0) { - skx_printk(KERN_ERR, "Can't find handle for NVDIMM ADR=%x\n", dev_handle); - goto unknown_size; - } - - if (flags & ACPI_NFIT_MEM_MAP_FAILED) { - skx_printk(KERN_ERR, "NVDIMM ADR=%x is not mapped\n", dev_handle); - goto unknown_size; - } - - size = dmi_memdev_size(smbios_handle); - if (size == ~0ull) - skx_printk(KERN_ERR, "Can't find size for NVDIMM ADR=%x/SMBIOS=%x\n", - dev_handle, smbios_handle); - -unknown_size: - dimm->nr_pages = size >> PAGE_SHIFT; - dimm->grain = 32; - dimm->dtype = DEV_UNKNOWN; - dimm->mtype = MEM_NVDIMM; - dimm->edac_mode = EDAC_SECDED; /* likely better than this */ - - edac_dbg(0, "mc#%d: channel %d, dimm %d, %llu MiB (%u pages)\n", - imc->mc, chan, dimmno, size >> 20, dimm->nr_pages); - - snprintf(dimm->label, sizeof(dimm->label), "CPU_SrcID#%u_MC#%u_Chan#%u_DIMM#%u", - imc->src_id, imc->lmc, chan, dimmno); - - return (size == 0 || size == ~0ull) ? 0 : 1; -} - -#define SKX_GET_MTMTR(dev, reg) \ - pci_read_config_dword((dev), 0x87c, &reg) - -static bool skx_check_ecc(struct pci_dev *pdev) -{ - u32 mtmtr; - - SKX_GET_MTMTR(pdev, mtmtr); - - return !!GET_BITFIELD(mtmtr, 2, 2); -} - -static int skx_get_dimm_config(struct mem_ctl_info *mci) -{ - struct skx_pvt *pvt = mci->pvt_info; - struct skx_imc *imc = pvt->imc; - u32 mtr, amap, mcddrtcfg; - struct dimm_info *dimm; - int i, j; - int ndimms; - - for (i = 0; i < NUM_CHANNELS; i++) { - ndimms = 0; - pci_read_config_dword(imc->chan[i].cdev, 0x8C, &amap); - pci_read_config_dword(imc->chan[i].cdev, 0x400, &mcddrtcfg); - for (j = 0; j < NUM_DIMMS; j++) { - dimm = EDAC_DIMM_PTR(mci->layers, mci->dimms, - mci->n_layers, i, j, 0); - pci_read_config_dword(imc->chan[i].cdev, - 0x80 + 4*j, &mtr); - if (IS_DIMM_PRESENT(mtr)) - ndimms += get_dimm_info(mtr, amap, dimm, imc, i, j); - else if (IS_NVDIMM_PRESENT(mcddrtcfg, j)) - ndimms += get_nvdimm_info(dimm, imc, i, j); - } - if (ndimms && !skx_check_ecc(imc->chan[0].cdev)) { - skx_printk(KERN_ERR, "ECC is disabled on imc %d\n", imc->mc); - return -ENODEV; - } - } - - return 0; -} - -static void skx_unregister_mci(struct skx_imc *imc) -{ - struct mem_ctl_info *mci = imc->mci; - - if (!mci) - return; - - edac_dbg(0, "MC%d: mci = %p\n", imc->mc, mci); - - /* Remove MC sysfs nodes */ - edac_mc_del_mc(mci->pdev); - - edac_dbg(1, "%s: free mci struct\n", mci->ctl_name); - kfree(mci->ctl_name); - edac_mc_free(mci); -} - -static int skx_register_mci(struct skx_imc *imc) -{ - struct mem_ctl_info *mci; - struct edac_mc_layer layers[2]; - struct pci_dev *pdev = imc->chan[0].cdev; - struct skx_pvt *pvt; - int rc; - - /* allocate a new MC control structure */ - layers[0].type = EDAC_MC_LAYER_CHANNEL; - layers[0].size = NUM_CHANNELS; - layers[0].is_virt_csrow = false; - layers[1].type = EDAC_MC_LAYER_SLOT; - layers[1].size = NUM_DIMMS; - layers[1].is_virt_csrow = true; - mci = edac_mc_alloc(imc->mc, ARRAY_SIZE(layers), layers, - sizeof(struct skx_pvt)); - - if (unlikely(!mci)) - return -ENOMEM; - - edac_dbg(0, "MC#%d: mci = %p\n", imc->mc, mci); - - /* Associate skx_dev and mci for future usage */ - imc->mci = mci; - pvt = mci->pvt_info; - pvt->imc = imc; - - mci->ctl_name = kasprintf(GFP_KERNEL, "Skylake Socket#%d IMC#%d", - imc->node_id, imc->lmc); - if (!mci->ctl_name) { - rc = -ENOMEM; - goto fail0; - } - - mci->mtype_cap = MEM_FLAG_DDR4 | MEM_FLAG_NVDIMM; - mci->edac_ctl_cap = EDAC_FLAG_NONE; - mci->edac_cap = EDAC_FLAG_NONE; - mci->mod_name = EDAC_MOD_STR; - mci->dev_name = pci_name(imc->chan[0].cdev); - mci->ctl_page_to_phys = NULL; - - rc = skx_get_dimm_config(mci); - if (rc < 0) - goto fail; - - /* record ptr to the generic device */ - mci->pdev = &pdev->dev; - - /* add this new MC control structure to EDAC's list of MCs */ - if (unlikely(edac_mc_add_mc(mci))) { - edac_dbg(0, "MC: failed edac_mc_add_mc()\n"); - rc = -EINVAL; - goto fail; - } - - return 0; - -fail: - kfree(mci->ctl_name); -fail0: - edac_mc_free(mci); - imc->mci = NULL; - return rc; -} - -#define SKX_MAX_SAD 24 - -#define SKX_GET_SAD(d, i, reg) \ - pci_read_config_dword((d)->sad_all, 0x60 + 8 * (i), &reg) -#define SKX_GET_ILV(d, i, reg) \ - pci_read_config_dword((d)->sad_all, 0x64 + 8 * (i), &reg) - -#define SKX_SAD_MOD3MODE(sad) GET_BITFIELD((sad), 30, 31) -#define SKX_SAD_MOD3(sad) GET_BITFIELD((sad), 27, 27) -#define SKX_SAD_LIMIT(sad) (((u64)GET_BITFIELD((sad), 7, 26) << 26) | MASK26) -#define SKX_SAD_MOD3ASMOD2(sad) GET_BITFIELD((sad), 5, 6) -#define SKX_SAD_ATTR(sad) GET_BITFIELD((sad), 3, 4) -#define SKX_SAD_INTERLEAVE(sad) GET_BITFIELD((sad), 1, 2) -#define SKX_SAD_ENABLE(sad) GET_BITFIELD((sad), 0, 0) - -#define SKX_ILV_REMOTE(tgt) (((tgt) & 8) == 0) -#define SKX_ILV_TARGET(tgt) ((tgt) & 7) - -static bool skx_sad_decode(struct decoded_addr *res) -{ - struct skx_dev *d = list_first_entry(&skx_edac_list, typeof(*d), list); - u64 addr = res->addr; - int i, idx, tgt, lchan, shift; - u32 sad, ilv; - u64 limit, prev_limit; - int remote = 0; - - /* Simple sanity check for I/O space or out of range */ - if (addr >= skx_tohm || (addr >= skx_tolm && addr < BIT_ULL(32))) { - edac_dbg(0, "Address %llx out of range\n", addr); - return false; - } - -restart: - prev_limit = 0; - for (i = 0; i < SKX_MAX_SAD; i++) { - SKX_GET_SAD(d, i, sad); - limit = SKX_SAD_LIMIT(sad); - if (SKX_SAD_ENABLE(sad)) { - if (addr >= prev_limit && addr <= limit) - goto sad_found; - } - prev_limit = limit + 1; - } - edac_dbg(0, "No SAD entry for %llx\n", addr); - return false; - -sad_found: - SKX_GET_ILV(d, i, ilv); - - switch (SKX_SAD_INTERLEAVE(sad)) { - case 0: - idx = GET_BITFIELD(addr, 6, 8); - break; - case 1: - idx = GET_BITFIELD(addr, 8, 10); - break; - case 2: - idx = GET_BITFIELD(addr, 12, 14); - break; - case 3: - idx = GET_BITFIELD(addr, 30, 32); - break; - } - - tgt = GET_BITFIELD(ilv, 4 * idx, 4 * idx + 3); - - /* If point to another node, find it and start over */ - if (SKX_ILV_REMOTE(tgt)) { - if (remote) { - edac_dbg(0, "Double remote!\n"); - return false; - } - remote = 1; - list_for_each_entry(d, &skx_edac_list, list) { - if (d->imc[0].src_id == SKX_ILV_TARGET(tgt)) - goto restart; - } - edac_dbg(0, "Can't find node %d\n", SKX_ILV_TARGET(tgt)); - return false; - } - - if (SKX_SAD_MOD3(sad) == 0) - lchan = SKX_ILV_TARGET(tgt); - else { - switch (SKX_SAD_MOD3MODE(sad)) { - case 0: - shift = 6; - break; - case 1: - shift = 8; - break; - case 2: - shift = 12; - break; - default: - edac_dbg(0, "illegal mod3mode\n"); - return false; - } - switch (SKX_SAD_MOD3ASMOD2(sad)) { - case 0: - lchan = (addr >> shift) % 3; - break; - case 1: - lchan = (addr >> shift) % 2; - break; - case 2: - lchan = (addr >> shift) % 2; - lchan = (lchan << 1) | !lchan; - break; - case 3: - lchan = ((addr >> shift) % 2) << 1; - break; - } - lchan = (lchan << 1) | (SKX_ILV_TARGET(tgt) & 1); - } - - res->dev = d; - res->socket = d->imc[0].src_id; - res->imc = GET_BITFIELD(d->mcroute, lchan * 3, lchan * 3 + 2); - res->channel = GET_BITFIELD(d->mcroute, lchan * 2 + 18, lchan * 2 + 19); - - edac_dbg(2, "%llx: socket=%d imc=%d channel=%d\n", - res->addr, res->socket, res->imc, res->channel); - return true; -} - -#define SKX_MAX_TAD 8 - -#define SKX_GET_TADBASE(d, mc, i, reg) \ - pci_read_config_dword((d)->imc[mc].chan[0].cdev, 0x850 + 4 * (i), &reg) -#define SKX_GET_TADWAYNESS(d, mc, i, reg) \ - pci_read_config_dword((d)->imc[mc].chan[0].cdev, 0x880 + 4 * (i), &reg) -#define SKX_GET_TADCHNILVOFFSET(d, mc, ch, i, reg) \ - pci_read_config_dword((d)->imc[mc].chan[ch].cdev, 0x90 + 4 * (i), &reg) - -#define SKX_TAD_BASE(b) ((u64)GET_BITFIELD((b), 12, 31) << 26) -#define SKX_TAD_SKT_GRAN(b) GET_BITFIELD((b), 4, 5) -#define SKX_TAD_CHN_GRAN(b) GET_BITFIELD((b), 6, 7) -#define SKX_TAD_LIMIT(b) (((u64)GET_BITFIELD((b), 12, 31) << 26) | MASK26) -#define SKX_TAD_OFFSET(b) ((u64)GET_BITFIELD((b), 4, 23) << 26) -#define SKX_TAD_SKTWAYS(b) (1 << GET_BITFIELD((b), 10, 11)) -#define SKX_TAD_CHNWAYS(b) (GET_BITFIELD((b), 8, 9) + 1) - -/* which bit used for both socket and channel interleave */ -static int skx_granularity[] = { 6, 8, 12, 30 }; - -static u64 skx_do_interleave(u64 addr, int shift, int ways, u64 lowbits) -{ - addr >>= shift; - addr /= ways; - addr <<= shift; - - return addr | (lowbits & ((1ull << shift) - 1)); -} - -static bool skx_tad_decode(struct decoded_addr *res) -{ - int i; - u32 base, wayness, chnilvoffset; - int skt_interleave_bit, chn_interleave_bit; - u64 channel_addr; - - for (i = 0; i < SKX_MAX_TAD; i++) { - SKX_GET_TADBASE(res->dev, res->imc, i, base); - SKX_GET_TADWAYNESS(res->dev, res->imc, i, wayness); - if (SKX_TAD_BASE(base) <= res->addr && res->addr <= SKX_TAD_LIMIT(wayness)) - goto tad_found; - } - edac_dbg(0, "No TAD entry for %llx\n", res->addr); - return false; - -tad_found: - res->sktways = SKX_TAD_SKTWAYS(wayness); - res->chanways = SKX_TAD_CHNWAYS(wayness); - skt_interleave_bit = skx_granularity[SKX_TAD_SKT_GRAN(base)]; - chn_interleave_bit = skx_granularity[SKX_TAD_CHN_GRAN(base)]; - - SKX_GET_TADCHNILVOFFSET(res->dev, res->imc, res->channel, i, chnilvoffset); - channel_addr = res->addr - SKX_TAD_OFFSET(chnilvoffset); - - if (res->chanways == 3 && skt_interleave_bit > chn_interleave_bit) { - /* Must handle channel first, then socket */ - channel_addr = skx_do_interleave(channel_addr, chn_interleave_bit, - res->chanways, channel_addr); - channel_addr = skx_do_interleave(channel_addr, skt_interleave_bit, - res->sktways, channel_addr); - } else { - /* Handle socket then channel. Preserve low bits from original address */ - channel_addr = skx_do_interleave(channel_addr, skt_interleave_bit, - res->sktways, res->addr); - channel_addr = skx_do_interleave(channel_addr, chn_interleave_bit, - res->chanways, res->addr); - } - - res->chan_addr = channel_addr; - - edac_dbg(2, "%llx: chan_addr=%llx sktways=%d chanways=%d\n", - res->addr, res->chan_addr, res->sktways, res->chanways); - return true; -} - -#define SKX_MAX_RIR 4 - -#define SKX_GET_RIRWAYNESS(d, mc, ch, i, reg) \ - pci_read_config_dword((d)->imc[mc].chan[ch].cdev, \ - 0x108 + 4 * (i), &reg) -#define SKX_GET_RIRILV(d, mc, ch, idx, i, reg) \ - pci_read_config_dword((d)->imc[mc].chan[ch].cdev, \ - 0x120 + 16 * idx + 4 * (i), &reg) - -#define SKX_RIR_VALID(b) GET_BITFIELD((b), 31, 31) -#define SKX_RIR_LIMIT(b) (((u64)GET_BITFIELD((b), 1, 11) << 29) | MASK29) -#define SKX_RIR_WAYS(b) (1 << GET_BITFIELD((b), 28, 29)) -#define SKX_RIR_CHAN_RANK(b) GET_BITFIELD((b), 16, 19) -#define SKX_RIR_OFFSET(b) ((u64)(GET_BITFIELD((b), 2, 15) << 26)) - -static bool skx_rir_decode(struct decoded_addr *res) -{ - int i, idx, chan_rank; - int shift; - u32 rirway, rirlv; - u64 rank_addr, prev_limit = 0, limit; - - if (res->dev->imc[res->imc].chan[res->channel].dimms[0].close_pg) - shift = 6; - else - shift = 13; - - for (i = 0; i < SKX_MAX_RIR; i++) { - SKX_GET_RIRWAYNESS(res->dev, res->imc, res->channel, i, rirway); - limit = SKX_RIR_LIMIT(rirway); - if (SKX_RIR_VALID(rirway)) { - if (prev_limit <= res->chan_addr && - res->chan_addr <= limit) - goto rir_found; - } - prev_limit = limit; - } - edac_dbg(0, "No RIR entry for %llx\n", res->addr); - return false; - -rir_found: - rank_addr = res->chan_addr >> shift; - rank_addr /= SKX_RIR_WAYS(rirway); - rank_addr <<= shift; - rank_addr |= res->chan_addr & GENMASK_ULL(shift - 1, 0); - - res->rank_address = rank_addr; - idx = (res->chan_addr >> shift) % SKX_RIR_WAYS(rirway); - - SKX_GET_RIRILV(res->dev, res->imc, res->channel, idx, i, rirlv); - res->rank_address = rank_addr - SKX_RIR_OFFSET(rirlv); - chan_rank = SKX_RIR_CHAN_RANK(rirlv); - res->channel_rank = chan_rank; - res->dimm = chan_rank / 4; - res->rank = chan_rank % 4; - - edac_dbg(2, "%llx: dimm=%d rank=%d chan_rank=%d rank_addr=%llx\n", - res->addr, res->dimm, res->rank, - res->channel_rank, res->rank_address); - return true; -} - -static u8 skx_close_row[] = { - 15, 16, 17, 18, 20, 21, 22, 28, 10, 11, 12, 13, 29, 30, 31, 32, 33 -}; -static u8 skx_close_column[] = { - 3, 4, 5, 14, 19, 23, 24, 25, 26, 27 -}; -static u8 skx_open_row[] = { - 14, 15, 16, 20, 28, 21, 22, 23, 24, 25, 26, 27, 29, 30, 31, 32, 33 -}; -static u8 skx_open_column[] = { - 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 -}; -static u8 skx_open_fine_column[] = { - 3, 4, 5, 7, 8, 9, 10, 11, 12, 13 -}; - -static int skx_bits(u64 addr, int nbits, u8 *bits) -{ - int i, res = 0; - - for (i = 0; i < nbits; i++) - res |= ((addr >> bits[i]) & 1) << i; - return res; -} - -static int skx_bank_bits(u64 addr, int b0, int b1, int do_xor, int x0, int x1) -{ - int ret = GET_BITFIELD(addr, b0, b0) | (GET_BITFIELD(addr, b1, b1) << 1); - - if (do_xor) - ret ^= GET_BITFIELD(addr, x0, x0) | (GET_BITFIELD(addr, x1, x1) << 1); - - return ret; -} - -static bool skx_mad_decode(struct decoded_addr *r) -{ - struct skx_dimm *dimm = &r->dev->imc[r->imc].chan[r->channel].dimms[r->dimm]; - int bg0 = dimm->fine_grain_bank ? 6 : 13; - - if (dimm->close_pg) { - r->row = skx_bits(r->rank_address, dimm->rowbits, skx_close_row); - r->column = skx_bits(r->rank_address, dimm->colbits, skx_close_column); - r->column |= 0x400; /* C10 is autoprecharge, always set */ - r->bank_address = skx_bank_bits(r->rank_address, 8, 9, dimm->bank_xor_enable, 22, 28); - r->bank_group = skx_bank_bits(r->rank_address, 6, 7, dimm->bank_xor_enable, 20, 21); - } else { - r->row = skx_bits(r->rank_address, dimm->rowbits, skx_open_row); - if (dimm->fine_grain_bank) - r->column = skx_bits(r->rank_address, dimm->colbits, skx_open_fine_column); - else - r->column = skx_bits(r->rank_address, dimm->colbits, skx_open_column); - r->bank_address = skx_bank_bits(r->rank_address, 18, 19, dimm->bank_xor_enable, 22, 23); - r->bank_group = skx_bank_bits(r->rank_address, bg0, 17, dimm->bank_xor_enable, 20, 21); - } - r->row &= (1u << dimm->rowbits) - 1; - - edac_dbg(2, "%llx: row=%x col=%x bank_addr=%d bank_group=%d\n", - r->addr, r->row, r->column, r->bank_address, - r->bank_group); - return true; -} - -static bool skx_decode(struct decoded_addr *res) -{ - - return skx_sad_decode(res) && skx_tad_decode(res) && - skx_rir_decode(res) && skx_mad_decode(res); -} - -#ifdef CONFIG_EDAC_DEBUG -/* - * Debug feature. Make /sys/kernel/debug/skx_edac_test/addr. - * Write an address to this file to exercise the address decode - * logic in this driver. - */ -static struct dentry *skx_test; -static u64 skx_fake_addr; - -static int debugfs_u64_set(void *data, u64 val) -{ - struct decoded_addr res; - - res.addr = val; - skx_decode(&res); - - return 0; -} - -DEFINE_SIMPLE_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n"); - -static struct dentry *mydebugfs_create(const char *name, umode_t mode, - struct dentry *parent, u64 *value) -{ - return debugfs_create_file(name, mode, parent, value, &fops_u64_wo); -} - -static void setup_skx_debug(void) -{ - skx_test = debugfs_create_dir("skx_edac_test", NULL); - mydebugfs_create("addr", S_IWUSR, skx_test, &skx_fake_addr); -} - -static void teardown_skx_debug(void) -{ - debugfs_remove_recursive(skx_test); -} -#else -static void setup_skx_debug(void) -{ -} - -static void teardown_skx_debug(void) -{ -} -#endif /*CONFIG_EDAC_DEBUG*/ - -static bool skx_adxl_decode(struct decoded_addr *res) - -{ - int i, len = 0; - - if (res->addr >= skx_tohm || (res->addr >= skx_tolm && - res->addr < BIT_ULL(32))) { - edac_dbg(0, "Address 0x%llx out of range\n", res->addr); - return false; - } - - if (adxl_decode(res->addr, adxl_values)) { - edac_dbg(0, "Failed to decode 0x%llx\n", res->addr); - return false; - } - - res->socket = (int)adxl_values[component_indices[INDEX_SOCKET]]; - res->imc = (int)adxl_values[component_indices[INDEX_MEMCTRL]]; - res->channel = (int)adxl_values[component_indices[INDEX_CHANNEL]]; - res->dimm = (int)adxl_values[component_indices[INDEX_DIMM]]; - - for (i = 0; i < adxl_component_count; i++) { - if (adxl_values[i] == ~0x0ull) - continue; - - len += snprintf(adxl_msg + len, MSG_SIZE - len, " %s:0x%llx", - adxl_component_names[i], adxl_values[i]); - if (MSG_SIZE - len <= 0) - break; - } - - return true; -} - -static void skx_mce_output_error(struct mem_ctl_info *mci, - const struct mce *m, - struct decoded_addr *res) -{ - enum hw_event_mc_err_type tp_event; - char *type, *optype; - bool ripv = GET_BITFIELD(m->mcgstatus, 0, 0); - bool overflow = GET_BITFIELD(m->status, 62, 62); - bool uncorrected_error = GET_BITFIELD(m->status, 61, 61); - bool recoverable; - u32 core_err_cnt = GET_BITFIELD(m->status, 38, 52); - u32 mscod = GET_BITFIELD(m->status, 16, 31); - u32 errcode = GET_BITFIELD(m->status, 0, 15); - u32 optypenum = GET_BITFIELD(m->status, 4, 6); - - recoverable = GET_BITFIELD(m->status, 56, 56); - - if (uncorrected_error) { - core_err_cnt = 1; - if (ripv) { - type = "FATAL"; - tp_event = HW_EVENT_ERR_FATAL; - } else { - type = "NON_FATAL"; - tp_event = HW_EVENT_ERR_UNCORRECTED; - } - } else { - type = "CORRECTED"; - tp_event = HW_EVENT_ERR_CORRECTED; - } - - /* - * According with Table 15-9 of the Intel Architecture spec vol 3A, - * memory errors should fit in this mask: - * 000f 0000 1mmm cccc (binary) - * where: - * f = Correction Report Filtering Bit. If 1, subsequent errors - * won't be shown - * mmm = error type - * cccc = channel - * If the mask doesn't match, report an error to the parsing logic - */ - if (!((errcode & 0xef80) == 0x80)) { - optype = "Can't parse: it is not a mem"; - } else { - switch (optypenum) { - case 0: - optype = "generic undef request error"; - break; - case 1: - optype = "memory read error"; - break; - case 2: - optype = "memory write error"; - break; - case 3: - optype = "addr/cmd error"; - break; - case 4: - optype = "memory scrubbing error"; - break; - default: - optype = "reserved"; - break; - } - } - if (adxl_component_count) { - snprintf(skx_msg, MSG_SIZE, "%s%s err_code:%04x:%04x %s", - overflow ? " OVERFLOW" : "", - (uncorrected_error && recoverable) ? " recoverable" : "", - mscod, errcode, adxl_msg); - } else { - snprintf(skx_msg, MSG_SIZE, - "%s%s err_code:%04x:%04x socket:%d imc:%d rank:%d bg:%d ba:%d row:%x col:%x", - overflow ? " OVERFLOW" : "", - (uncorrected_error && recoverable) ? " recoverable" : "", - mscod, errcode, - res->socket, res->imc, res->rank, - res->bank_group, res->bank_address, res->row, res->column); - } - - edac_dbg(0, "%s\n", skx_msg); - - /* Call the helper to output message */ - edac_mc_handle_error(tp_event, mci, core_err_cnt, - m->addr >> PAGE_SHIFT, m->addr & ~PAGE_MASK, 0, - res->channel, res->dimm, -1, - optype, skx_msg); -} - -static struct mem_ctl_info *get_mci(int src_id, int lmc) -{ - struct skx_dev *d; - - if (lmc > NUM_IMC - 1) { - skx_printk(KERN_ERR, "Bad lmc %d\n", lmc); - return NULL; - } - - list_for_each_entry(d, &skx_edac_list, list) { - if (d->imc[0].src_id == src_id) - return d->imc[lmc].mci; - } - - skx_printk(KERN_ERR, "No mci for src_id %d lmc %d\n", src_id, lmc); - - return NULL; -} - -static int skx_mce_check_error(struct notifier_block *nb, unsigned long val, - void *data) -{ - struct mce *mce = (struct mce *)data; - struct decoded_addr res; - struct mem_ctl_info *mci; - char *type; - - if (edac_get_report_status() == EDAC_REPORTING_DISABLED) - return NOTIFY_DONE; - - /* ignore unless this is memory related with an address */ - if ((mce->status & 0xefff) >> 7 != 1 || !(mce->status & MCI_STATUS_ADDRV)) - return NOTIFY_DONE; - - memset(&res, 0, sizeof(res)); - res.addr = mce->addr; - - if (adxl_component_count) { - if (!skx_adxl_decode(&res)) - return NOTIFY_DONE; - - mci = get_mci(res.socket, res.imc); - } else { - if (!skx_decode(&res)) - return NOTIFY_DONE; - - mci = res.dev->imc[res.imc].mci; - } - - if (!mci) - return NOTIFY_DONE; - - if (mce->mcgstatus & MCG_STATUS_MCIP) - type = "Exception"; - else - type = "Event"; - - skx_mc_printk(mci, KERN_DEBUG, "HANDLING MCE MEMORY ERROR\n"); - - skx_mc_printk(mci, KERN_DEBUG, "CPU %d: Machine Check %s: %Lx " - "Bank %d: %016Lx\n", mce->extcpu, type, - mce->mcgstatus, mce->bank, mce->status); - skx_mc_printk(mci, KERN_DEBUG, "TSC %llx ", mce->tsc); - skx_mc_printk(mci, KERN_DEBUG, "ADDR %llx ", mce->addr); - skx_mc_printk(mci, KERN_DEBUG, "MISC %llx ", mce->misc); - - skx_mc_printk(mci, KERN_DEBUG, "PROCESSOR %u:%x TIME %llu SOCKET " - "%u APIC %x\n", mce->cpuvendor, mce->cpuid, - mce->time, mce->socketid, mce->apicid); - - skx_mce_output_error(mci, mce, &res); - - return NOTIFY_DONE; -} - -static struct notifier_block skx_mce_dec = { - .notifier_call = skx_mce_check_error, - .priority = MCE_PRIO_EDAC, -}; - -static void skx_remove(void) -{ - int i, j; - struct skx_dev *d, *tmp; - - edac_dbg(0, "\n"); - - list_for_each_entry_safe(d, tmp, &skx_edac_list, list) { - list_del(&d->list); - for (i = 0; i < NUM_IMC; i++) { - skx_unregister_mci(&d->imc[i]); - for (j = 0; j < NUM_CHANNELS; j++) - pci_dev_put(d->imc[i].chan[j].cdev); - } - pci_dev_put(d->util_all); - pci_dev_put(d->sad_all); - - kfree(d); - } -} - -static void __init skx_adxl_get(void) -{ - const char * const *names; - int i, j; - - names = adxl_get_component_names(); - if (!names) { - skx_printk(KERN_NOTICE, "No firmware support for address translation."); - skx_printk(KERN_CONT, " Only decoding DDR4 address!\n"); - return; - } - - for (i = 0; i < INDEX_MAX; i++) { - for (j = 0; names[j]; j++) { - if (!strcmp(component_names[i], names[j])) { - component_indices[i] = j; - break; - } - } - - if (!names[j]) - goto err; - } - - adxl_component_names = names; - while (*names++) - adxl_component_count++; - - adxl_values = kcalloc(adxl_component_count, sizeof(*adxl_values), - GFP_KERNEL); - if (!adxl_values) { - adxl_component_count = 0; - return; - } - - adxl_msg = kzalloc(MSG_SIZE, GFP_KERNEL); - if (!adxl_msg) { - adxl_component_count = 0; - kfree(adxl_values); - } - - return; -err: - skx_printk(KERN_ERR, "'%s' is not matched from DSM parameters: ", - component_names[i]); - for (j = 0; names[j]; j++) - skx_printk(KERN_CONT, "%s ", names[j]); - skx_printk(KERN_CONT, "\n"); -} - -static void __exit skx_adxl_put(void) -{ - kfree(adxl_values); - kfree(adxl_msg); -} - -/* - * skx_init: - * make sure we are running on the correct cpu model - * search for all the devices we need - * check which DIMMs are present. - */ -static int __init skx_init(void) -{ - const struct x86_cpu_id *id; - const struct munit *m; - const char *owner; - int rc = 0, i; - u8 mc = 0, src_id, node_id; - struct skx_dev *d; - - edac_dbg(2, "\n"); - - owner = edac_get_owner(); - if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR))) - return -EBUSY; - - id = x86_match_cpu(skx_cpuids); - if (!id) - return -ENODEV; - - rc = skx_get_hi_lo(); - if (rc) - return rc; - - rc = get_all_bus_mappings(); - if (rc < 0) - goto fail; - if (rc == 0) { - edac_dbg(2, "No memory controllers found\n"); - return -ENODEV; - } - - for (m = skx_all_munits; m->did; m++) { - rc = get_all_munits(m); - if (rc < 0) - goto fail; - if (rc != m->per_socket * skx_num_sockets) { - edac_dbg(2, "Expected %d, got %d of %x\n", - m->per_socket * skx_num_sockets, rc, m->did); - rc = -ENODEV; - goto fail; - } - } - - list_for_each_entry(d, &skx_edac_list, list) { - src_id = get_src_id(d); - node_id = skx_get_node_id(d); - edac_dbg(2, "src_id=%d node_id=%d\n", src_id, node_id); - for (i = 0; i < NUM_IMC; i++) { - d->imc[i].mc = mc++; - d->imc[i].lmc = i; - d->imc[i].src_id = src_id; - d->imc[i].node_id = node_id; - rc = skx_register_mci(&d->imc[i]); - if (rc < 0) - goto fail; - } - } - - skx_msg = kzalloc(MSG_SIZE, GFP_KERNEL); - if (!skx_msg) { - rc = -ENOMEM; - goto fail; - } - - if (nvdimm_count) - skx_adxl_get(); - - /* Ensure that the OPSTATE is set correctly for POLL or NMI */ - opstate_init(); - - setup_skx_debug(); - - mce_register_decode_chain(&skx_mce_dec); - - return 0; -fail: - skx_remove(); - return rc; -} - -static void __exit skx_exit(void) -{ - edac_dbg(2, "\n"); - mce_unregister_decode_chain(&skx_mce_dec); - skx_remove(); - if (nvdimm_count) - skx_adxl_put(); - kfree(skx_msg); - teardown_skx_debug(); -} - -module_init(skx_init); -module_exit(skx_exit); - -module_param(edac_op_state, int, 0444); -MODULE_PARM_DESC(edac_op_state, "EDAC Error Reporting state: 0=Poll,1=NMI"); - -MODULE_LICENSE("GPL v2"); -MODULE_AUTHOR("Tony Luck"); -MODULE_DESCRIPTION("MC Driver for Intel Skylake server processors");

From: Tony Luck tony.luck@intel.com

mainline inclusion from mainline-v5.1-rc1 commit cbfa482f7e2becbb774dd30117efac48819252f8 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit cbfa482f7e2becbb774dd30117efac48819252f8 upstream.

A new error code for systems that use DRAM as an extra level of cache looks like:

000F 0010 1MMM CCCC

where the MMM and CCCC bits are used for the same purpose as the original code. For this new class of errors the ADXL translation will provide details of both the DIMM used as cache for the error location and the component that is being cached.

Note: This new error code is first supported in Skylake. Older EDAC drivers do not need to be updated.

Signed-off-by: Tony Luck tony.luck@intel.com Signed-off-by: Borislav Petkov bp@suse.de Cc: Aristeu Rozanski aris@redhat.com Cc: James Morse james.morse@arm.com Cc: Mauro Carvalho Chehab mchehab@kernel.org Cc: Qiuxu Zhuo qiuxu.zhuo@intel.com Cc: linux-edac linux-edac@vger.kernel.org Link: https://lkml.kernel.org/r/20190205182109.27828-1-tony.luck@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/edac/skx_common.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/edac/skx_common.c b/drivers/edac/skx_common.c index fb3d9a563ba8c..0821e74513c48 100644 --- a/drivers/edac/skx_common.c +++ b/drivers/edac/skx_common.c @@ -494,9 +494,11 @@ static void skx_mce_output_error(struct mem_ctl_info *mci, }

/* - * According with Table 15-9 of the Intel Architecture spec vol 3A, - * memory errors should fit in this mask: + * According to Intel Architecture spec vol 3B, + * Table 15-10 "IA32_MCi_Status [15:0] Compound Error Code Encoding" + * memory errors should fit one of these masks: * 000f 0000 1mmm cccc (binary) + * 000f 0010 1mmm cccc (binary) [RAM used as cache] * where: * f = Correction Report Filtering Bit. If 1, subsequent errors * won't be shown @@ -504,7 +506,7 @@ static void skx_mce_output_error(struct mem_ctl_info *mci, * cccc = channel * If the mask doesn't match, report an error to the parsing logic */ - if (!((errcode & 0xef80) == 0x80)) { + if (!((errcode & 0xef80) == 0x80 || (errcode & 0xef80) == 0x280)) { optype = "Can't parse: it is not a mem"; } else { switch (optypenum) {

From: Qiuxu Zhuo qiuxu.zhuo@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 5c5d3ac2064ae2466c81d40186bcc09b2d5b7892 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 5c5d3ac2064ae2466c81d40186bcc09b2d5b7892 upstream.

Two new CPU models share the same memory controller architecture with Jacobsville/Tremont, so can use the same i10nm EDAC driver.

Add ICX and ICX-D CPU model numbers for EDAC support.

Signed-off-by: Qiuxu Zhuo qiuxu.zhuo@intel.com Signed-off-by: Tony Luck tony.luck@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/edac/i10nm_base.c | 2 ++ 1 file changed, 2 insertions(+)

diff --git a/drivers/edac/i10nm_base.c b/drivers/edac/i10nm_base.c index 87baed8e0807d..eb13f07cdaa5a 100644 --- a/drivers/edac/i10nm_base.c +++ b/drivers/edac/i10nm_base.c @@ -124,6 +124,8 @@ static int i10nm_get_all_munits(void)

static const struct x86_cpu_id i10nm_cpuids[] = { { X86_VENDOR_INTEL, 6, INTEL_FAM6_ATOM_TREMONT_X, 0, 0 }, + { X86_VENDOR_INTEL, 6, INTEL_FAM6_ICELAKE_X, 0, 0 }, + { X86_VENDOR_INTEL, 6, INTEL_FAM6_ICELAKE_XEON_D, 0, 0 }, { } }; MODULE_DEVICE_TABLE(x86cpu, i10nm_cpuids);

From: Qiuxu Zhuo qiuxu.zhuo@intel.com

mainline inclusion from mainline-v5.3-rc1 commit 1dc78f1ffa3a386b986b659884952d816021f38f category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 1dc78f1ffa3a386b986b659884952d816021f38f upstream.

The source ID register offset for Skylake server is 0xf0, while for Icelake server is 0xf8. Pass the correct offset to get the source ID.

Signed-off-by: Qiuxu Zhuo qiuxu.zhuo@intel.com Signed-off-by: Tony Luck tony.luck@intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/edac/i10nm_base.c | 2 +- drivers/edac/skx_base.c | 2 +- drivers/edac/skx_common.c | 4 ++-- drivers/edac/skx_common.h | 2 +- 4 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/edac/i10nm_base.c b/drivers/edac/i10nm_base.c index 6c435db0bcd9c..216c36a0505e9 100644 --- a/drivers/edac/i10nm_base.c +++ b/drivers/edac/i10nm_base.c @@ -267,7 +267,7 @@ static int __init i10nm_init(void) goto fail;

list_for_each_entry(d, i10nm_edac_list, list) { - rc = skx_get_src_id(d, &src_id); + rc = skx_get_src_id(d, 0xf8, &src_id); if (rc < 0) goto fail;

diff --git a/drivers/edac/skx_base.c b/drivers/edac/skx_base.c index c6c8be2814ec9..c0ff22b96ec4e 100644 --- a/drivers/edac/skx_base.c +++ b/drivers/edac/skx_base.c @@ -639,7 +639,7 @@ static int __init skx_init(void) }

list_for_each_entry(d, skx_edac_list, list) { - rc = skx_get_src_id(d, &src_id); + rc = skx_get_src_id(d, 0xf0, &src_id); if (rc < 0) goto fail; rc = skx_get_node_id(d, &node_id); diff --git a/drivers/edac/skx_common.c b/drivers/edac/skx_common.c index b0dddcfa9baa2..d8ff63d91b860 100644 --- a/drivers/edac/skx_common.c +++ b/drivers/edac/skx_common.c @@ -136,11 +136,11 @@ void skx_set_decode(skx_decode_f decode) skx_decode = decode; }

-int skx_get_src_id(struct skx_dev *d, u8 *id) +int skx_get_src_id(struct skx_dev *d, int off, u8 *id) { u32 reg;

- if (pci_read_config_dword(d->util_all, 0xf0, &reg)) { + if (pci_read_config_dword(d->util_all, off, &reg)) { skx_printk(KERN_ERR, "Failed to read src id\n"); return -ENODEV; } diff --git a/drivers/edac/skx_common.h b/drivers/edac/skx_common.h index d18fa98669af1..08cc971a50ea2 100644 --- a/drivers/edac/skx_common.h +++ b/drivers/edac/skx_common.h @@ -118,7 +118,7 @@ int __init skx_adxl_get(void); void __exit skx_adxl_put(void); void skx_set_decode(skx_decode_f decode);

-int skx_get_src_id(struct skx_dev *d, u8 *id); +int skx_get_src_id(struct skx_dev *d, int off, u8 *id); int skx_get_node_id(struct skx_dev *d, u8 *id);

int skx_get_all_bus_mappings(unsigned int did, int off, enum type,

From: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com

mainline inclusion from mainline-v5.3-rc1 commit 35f2c14d2a076b063a76c5bf275c46c0743ba3a0 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 35f2c14d2a076b063a76c5bf275c46c0743ba3a0 upstream.

Encapsulate common functions which all Intel Speed Select Technology interface drivers can use. This creates API to register misc device for user kernel communication and handle all common IOCTLs. As part of the registry it allows a callback which is to handle domain specific ioctl processing.

There can be multiple drivers register for services, which can be built as modules. So this driver handle contention during registry and as well as during removal. Once user space opened the misc device, the registered driver will be prevented from removal. Also once misc device is opened by the user space new client driver can't register, till the misc device is closed.

There are two types of client drivers, one to handle mail box interface and the other is to allow direct read/write to some specific MMIO space.

This common driver implements IOCTL ISST_IF_GET_PLATFORM_INFO.

Signed-off-by: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com Signed-off-by: Andy Shevchenko andriy.shevchenko@linux.intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/platform/x86/Kconfig | 2 + drivers/platform/x86/Makefile | 1 + .../x86/intel_speed_select_if/Kconfig | 17 ++ .../x86/intel_speed_select_if/Makefile | 7 + .../intel_speed_select_if/isst_if_common.c | 182 ++++++++++++++++++ .../intel_speed_select_if/isst_if_common.h | 60 ++++++ include/uapi/linux/isst_if.h | 41 ++++ 7 files changed, 310 insertions(+) create mode 100644 drivers/platform/x86/intel_speed_select_if/Kconfig create mode 100644 drivers/platform/x86/intel_speed_select_if/Makefile create mode 100644 drivers/platform/x86/intel_speed_select_if/isst_if_common.c create mode 100644 drivers/platform/x86/intel_speed_select_if/isst_if_common.h create mode 100644 include/uapi/linux/isst_if.h

diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig index a13bb4ddd0cf1..2f30b80dc0afe 100644 --- a/drivers/platform/x86/Kconfig +++ b/drivers/platform/x86/Kconfig @@ -1244,6 +1244,8 @@ config INTEL_ATOMISP2_PM To compile this driver as a module, choose M here: the module will be called intel_atomisp2_pm.

+source "drivers/platform/x86/intel_speed_select_if/Kconfig" + endif # X86_PLATFORM_DEVICES

config PMC_ATOM diff --git a/drivers/platform/x86/Makefile b/drivers/platform/x86/Makefile index dc29af4d8e2fa..fb426466817bf 100644 --- a/drivers/platform/x86/Makefile +++ b/drivers/platform/x86/Makefile @@ -93,3 +93,4 @@ obj-$(CONFIG_INTEL_TURBO_MAX_3) += intel_turbo_max_3.o obj-$(CONFIG_INTEL_CHTDC_TI_PWRBTN) += intel_chtdc_ti_pwrbtn.o obj-$(CONFIG_I2C_MULTI_INSTANTIATE) += i2c-multi-instantiate.o obj-$(CONFIG_INTEL_ATOMISP2_PM) += intel_atomisp2_pm.o +obj-$(CONFIG_INTEL_SPEED_SELECT_INTERFACE) += intel_speed_select_if/ diff --git a/drivers/platform/x86/intel_speed_select_if/Kconfig b/drivers/platform/x86/intel_speed_select_if/Kconfig new file mode 100644 index 0000000000000..ce3e3dc076d23 --- /dev/null +++ b/drivers/platform/x86/intel_speed_select_if/Kconfig @@ -0,0 +1,17 @@ +menu "Intel Speed Select Technology interface support" + depends on PCI + depends on X86_64 || COMPILE_TEST + +config INTEL_SPEED_SELECT_INTERFACE + tristate "Intel(R) Speed Select Technology interface drivers" + help + This config enables the Intel(R) Speed Select Technology interface + drivers. The Intel(R) speed select technology features are non + architectural and only supported on specific Xeon(R) servers. + These drivers provide interface to directly communicate with hardware + via MMIO and Mail boxes to enumerate and control all the speed select + features. + + Enable this config, if there is a need to enable and control the + Intel(R) Speed Select Technology features from the user space. +endmenu diff --git a/drivers/platform/x86/intel_speed_select_if/Makefile b/drivers/platform/x86/intel_speed_select_if/Makefile new file mode 100644 index 0000000000000..c12687672fc90 --- /dev/null +++ b/drivers/platform/x86/intel_speed_select_if/Makefile @@ -0,0 +1,7 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# Makefile - Intel Speed Select Interface drivers +# Copyright (c) 2019, Intel Corporation. +# + +obj-$(CONFIG_INTEL_SPEED_SELECT_INTERFACE) += isst_if_common.o diff --git a/drivers/platform/x86/intel_speed_select_if/isst_if_common.c b/drivers/platform/x86/intel_speed_select_if/isst_if_common.c new file mode 100644 index 0000000000000..ab2bb4862dc8a --- /dev/null +++ b/drivers/platform/x86/intel_speed_select_if/isst_if_common.c @@ -0,0 +1,182 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel Speed Select Interface: Common functions + * Copyright (c) 2019, Intel Corporation. + * All rights reserved. + * + * Author: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com + */ + +#include <linux/fs.h> +#include <linux/miscdevice.h> +#include <linux/module.h> +#include <linux/uaccess.h> +#include <uapi/linux/isst_if.h> + +#include "isst_if_common.h" + +static struct isst_if_cmd_cb punit_callbacks[ISST_IF_DEV_MAX]; + +static int isst_if_get_platform_info(void __user *argp) +{ + struct isst_if_platform_info info; + + info.api_version = ISST_IF_API_VERSION, + info.driver_version = ISST_IF_DRIVER_VERSION, + info.max_cmds_per_ioctl = ISST_IF_CMD_LIMIT, + info.mbox_supported = punit_callbacks[ISST_IF_DEV_MBOX].registered; + info.mmio_supported = punit_callbacks[ISST_IF_DEV_MMIO].registered; + + if (copy_to_user(argp, &info, sizeof(info))) + return -EFAULT; + + return 0; +} + +static long isst_if_def_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + void __user *argp = (void __user *)arg; + long ret = -ENOTTY; + + switch (cmd) { + case ISST_IF_GET_PLATFORM_INFO: + ret = isst_if_get_platform_info(argp); + break; + default: + break; + } + + return ret; +} + +static DEFINE_MUTEX(punit_misc_dev_lock); +static int misc_usage_count; +static int misc_device_ret; +static int misc_device_open; + +static int isst_if_open(struct inode *inode, struct file *file) +{ + int i, ret = 0; + + /* Fail open, if a module is going away */ + mutex_lock(&punit_misc_dev_lock); + for (i = 0; i < ISST_IF_DEV_MAX; ++i) { + struct isst_if_cmd_cb *cb = &punit_callbacks[i]; + + if (cb->registered && !try_module_get(cb->owner)) { + ret = -ENODEV; + break; + } + } + if (ret) { + int j; + + for (j = 0; j < i; ++j) { + struct isst_if_cmd_cb *cb; + + cb = &punit_callbacks[j]; + if (cb->registered) + module_put(cb->owner); + } + } else { + misc_device_open++; + } + mutex_unlock(&punit_misc_dev_lock); + + return ret; +} + +static int isst_if_relase(struct inode *inode, struct file *f) +{ + int i; + + mutex_lock(&punit_misc_dev_lock); + misc_device_open--; + for (i = 0; i < ISST_IF_DEV_MAX; ++i) { + struct isst_if_cmd_cb *cb = &punit_callbacks[i]; + + if (cb->registered) + module_put(cb->owner); + } + mutex_unlock(&punit_misc_dev_lock); + + return 0; +} + +static const struct file_operations isst_if_char_driver_ops = { + .open = isst_if_open, + .unlocked_ioctl = isst_if_def_ioctl, + .release = isst_if_relase, +}; + +static struct miscdevice isst_if_char_driver = { + .minor = MISC_DYNAMIC_MINOR, + .name = "isst_interface", + .fops = &isst_if_char_driver_ops, +}; + +/** + * isst_if_cdev_register() - Register callback for IOCTL + * @device_type: The device type this callback handling. + * @cb: Callback structure. + * + * This function registers a callback to device type. On very first call + * it will register a misc device, which is used for user kernel interface. + * Other calls simply increment ref count. Registry will fail, if the user + * already opened misc device for operation. Also if the misc device + * creation failed, then it will not try again and all callers will get + * failure code. + * + * Return: Return the return value from the misc creation device or -EINVAL + * for unsupported device type. + */ +int isst_if_cdev_register(int device_type, struct isst_if_cmd_cb *cb) +{ + if (misc_device_ret) + return misc_device_ret; + + if (device_type >= ISST_IF_DEV_MAX) + return -EINVAL; + + mutex_lock(&punit_misc_dev_lock); + if (misc_device_open) { + mutex_unlock(&punit_misc_dev_lock); + return -EAGAIN; + } + if (!misc_usage_count) { + misc_device_ret = misc_register(&isst_if_char_driver); + if (misc_device_ret) + goto unlock_exit; + } + memcpy(&punit_callbacks[device_type], cb, sizeof(*cb)); + punit_callbacks[device_type].registered = 1; + misc_usage_count++; +unlock_exit: + mutex_unlock(&punit_misc_dev_lock); + + return misc_device_ret; +} +EXPORT_SYMBOL_GPL(isst_if_cdev_register); + +/** + * isst_if_cdev_unregister() - Unregister callback for IOCTL + * @device_type: The device type to unregister. + * + * This function unregisters the previously registered callback. If this + * is the last callback unregistering, then misc device is removed. + * + * Return: None. + */ +void isst_if_cdev_unregister(int device_type) +{ + mutex_lock(&punit_misc_dev_lock); + misc_usage_count--; + punit_callbacks[device_type].registered = 0; + if (!misc_usage_count && !misc_device_ret) + misc_deregister(&isst_if_char_driver); + mutex_unlock(&punit_misc_dev_lock); +} +EXPORT_SYMBOL_GPL(isst_if_cdev_unregister); + +MODULE_LICENSE("GPL v2"); diff --git a/drivers/platform/x86/intel_speed_select_if/isst_if_common.h b/drivers/platform/x86/intel_speed_select_if/isst_if_common.h new file mode 100644 index 0000000000000..11f339226fb40 --- /dev/null +++ b/drivers/platform/x86/intel_speed_select_if/isst_if_common.h @@ -0,0 +1,60 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Intel Speed Select Interface: Drivers Internal defines + * Copyright (c) 2019, Intel Corporation. + * All rights reserved. + * + * Author: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com + */ + +#ifndef __ISST_IF_COMMON_H +#define __ISST_IF_COMMON_H + +/* + * Validate maximum commands in a single request. + * This is enough to handle command to every core in one ioctl, or all + * possible message id to one CPU. Limit is also helpful for resonse time + * per IOCTL request, as PUNIT may take different times to process each + * request and may hold for long for too many commands. + */ +#define ISST_IF_CMD_LIMIT 64 + +#define ISST_IF_API_VERSION 0x01 +#define ISST_IF_DRIVER_VERSION 0x01 + +#define ISST_IF_DEV_MBOX 0 +#define ISST_IF_DEV_MMIO 1 +#define ISST_IF_DEV_MAX 2 + +/** + * struct isst_if_cmd_cb - Used to register a IOCTL handler + * @registered: Used by the common code to store registry. Caller don't + * to touch this field + * @cmd_size: The command size of the individual command in IOCTL + * @offset: Offset to the first valid member in command structure. + * This will be the offset of the start of the command + * after command count field + * @cmd_callback: Callback function to handle IOCTL. The callback has the + * command pointer with data for command. There is a pointer + * called write_only, which when set, will not copy the + * response to user ioctl buffer. The "resume" argument + * can be used to avoid storing the command for replay + * during system resume + * + * This structure is used to register an handler for IOCTL. To avoid + * code duplication common code handles all the IOCTL command read/write + * including handling multiple command in single IOCTL. The caller just + * need to execute a command via the registered callback. + */ +struct isst_if_cmd_cb { + int registered; + int cmd_size; + int offset; + struct module *owner; + long (*cmd_callback)(u8 *ptr, int *write_only, int resume); +}; + +/* Internal interface functions */ +int isst_if_cdev_register(int type, struct isst_if_cmd_cb *cb); +void isst_if_cdev_unregister(int type); +#endif diff --git a/include/uapi/linux/isst_if.h b/include/uapi/linux/isst_if.h new file mode 100644 index 0000000000000..fa94480b5f748 --- /dev/null +++ b/include/uapi/linux/isst_if.h @@ -0,0 +1,41 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Intel Speed Select Interface: OS to hardware Interface + * Copyright (c) 2019, Intel Corporation. + * All rights reserved. + * + * Author: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com + */ + +#ifndef __ISST_IF_H +#define __ISST_IF_H + +#include <linux/types.h> + +/** + * struct isst_if_platform_info - Define platform information + * @api_version: Version of the firmware document, which this driver + * can communicate + * @driver_version: Driver version, which will help user to send right + * commands. Even if the firmware is capable, driver may + * not be ready + * @max_cmds_per_ioctl: Returns the maximum number of commands driver will + * accept in a single ioctl + * @mbox_supported: Support of mail box interface + * @mmio_supported: Support of mmio interface for core-power feature + * + * Used to return output of IOCTL ISST_IF_GET_PLATFORM_INFO. This + * information can be used by the user space, to get the driver, firmware + * support and also number of commands to send in a single IOCTL request. + */ +struct isst_if_platform_info { + __u16 api_version; + __u16 driver_version; + __u16 max_cmds_per_ioctl; + __u8 mbox_supported; + __u8 mmio_supported; +}; + +#define ISST_IF_MAGIC 0xFE +#define ISST_IF_GET_PLATFORM_INFO _IOR(ISST_IF_MAGIC, 0, struct isst_if_platform_info *) +#endif

From: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com

mainline inclusion from mainline-v5.3-rc1 commit fb5b36a413b9f30fba573fc2a596ab7142dfaf12 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit fb5b36a413b9f30fba573fc2a596ab7142dfaf12 upstream.

Add processing for IOCTL command ISST_IF_GET_PHY_ID. This converts from the Linux logical CPU to PUNIT CPU numbering scheme.

Signed-off-by: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com Signed-off-by: Andy Shevchenko andriy.shevchenko@linux.intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- .../intel_speed_select_if/isst_if_common.c | 74 +++++++++++++++++++ include/uapi/linux/isst_if.h | 28 +++++++ 2 files changed, 102 insertions(+)

diff --git a/drivers/platform/x86/intel_speed_select_if/isst_if_common.c b/drivers/platform/x86/intel_speed_select_if/isst_if_common.c index 0e16cbf685d0a..72e74d72724b3 100644 --- a/drivers/platform/x86/intel_speed_select_if/isst_if_common.c +++ b/drivers/platform/x86/intel_speed_select_if/isst_if_common.c @@ -134,16 +134,90 @@ static void isst_if_cpu_info_exit(void) kfree(isst_cpu_info); };

+static long isst_if_proc_phyid_req(u8 *cmd_ptr, int *write_only, int resume) +{ + struct isst_if_cpu_map *cpu_map; + + cpu_map = (struct isst_if_cpu_map *)cmd_ptr; + if (cpu_map->logical_cpu >= nr_cpu_ids || + cpu_map->logical_cpu >= num_possible_cpus()) + return -EINVAL; + + *write_only = 0; + cpu_map->physical_cpu = isst_cpu_info[cpu_map->logical_cpu].punit_cpu_id; + + return 0; +} + +static long isst_if_exec_multi_cmd(void __user *argp, struct isst_if_cmd_cb *cb) +{ + unsigned char __user *ptr; + u32 cmd_count; + u8 *cmd_ptr; + long ret; + int i; + + /* Each multi command has u32 command count as the first field */ + if (copy_from_user(&cmd_count, argp, sizeof(cmd_count))) + return -EFAULT; + + if (!cmd_count || cmd_count > ISST_IF_CMD_LIMIT) + return -EINVAL; + + cmd_ptr = kmalloc(cb->cmd_size, GFP_KERNEL); + if (!cmd_ptr) + return -ENOMEM; + + /* cb->offset points to start of the command after the command count */ + ptr = argp + cb->offset; + + for (i = 0; i < cmd_count; ++i) { + int wr_only; + + if (signal_pending(current)) { + ret = -EINTR; + break; + } + + if (copy_from_user(cmd_ptr, ptr, cb->cmd_size)) { + ret = -EFAULT; + break; + } + + ret = cb->cmd_callback(cmd_ptr, &wr_only, 0); + if (ret) + break; + + if (!wr_only && copy_to_user(ptr, cmd_ptr, cb->cmd_size)) { + ret = -EFAULT; + break; + } + + ptr += cb->cmd_size; + } + + kfree(cmd_ptr); + + return i ? i : ret; +} + static long isst_if_def_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { void __user *argp = (void __user *)arg; + struct isst_if_cmd_cb cmd_cb; long ret = -ENOTTY;

switch (cmd) { case ISST_IF_GET_PLATFORM_INFO: ret = isst_if_get_platform_info(argp); break; + case ISST_IF_GET_PHY_ID: + cmd_cb.cmd_size = sizeof(struct isst_if_cpu_map); + cmd_cb.offset = offsetof(struct isst_if_cpu_maps, cpu_map); + cmd_cb.cmd_callback = isst_if_proc_phyid_req; + ret = isst_if_exec_multi_cmd(argp, &cmd_cb); + break; default: break; } diff --git a/include/uapi/linux/isst_if.h b/include/uapi/linux/isst_if.h index fa94480b5f748..15d1f286a8301 100644 --- a/include/uapi/linux/isst_if.h +++ b/include/uapi/linux/isst_if.h @@ -36,6 +36,34 @@ struct isst_if_platform_info { __u8 mmio_supported; };

+/** + * struct isst_if_cpu_map - CPU mapping between logical and physical CPU + * @logical_cpu: Linux logical CPU number + * @physical_cpu: PUNIT CPU number + * + * Used to convert from Linux logical CPU to PUNIT CPU numbering scheme. + * The PUNIT CPU number is different than APIC ID based CPU numbering. + */ +struct isst_if_cpu_map { + __u32 logical_cpu; + __u32 physical_cpu; +}; + +/** + * struct isst_if_cpu_maps - structure for CPU map IOCTL + * @cmd_count: Number of CPU mapping command in cpu_map[] + * @cpu_map[]: Holds one or more CPU map data structure + * + * This structure used with ioctl ISST_IF_GET_PHY_ID to send + * one or more CPU mapping commands. Here IOCTL return value indicates + * number of commands sent or error number if no commands have been sent. + */ +struct isst_if_cpu_maps { + __u32 cmd_count; + struct isst_if_cpu_map cpu_map[1]; +}; + #define ISST_IF_MAGIC 0xFE #define ISST_IF_GET_PLATFORM_INFO _IOR(ISST_IF_MAGIC, 0, struct isst_if_platform_info *) +#define ISST_IF_GET_PHY_ID _IOWR(ISST_IF_MAGIC, 1, struct isst_if_cpu_map *) #endif

From: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com

mainline inclusion from mainline-v5.3-rc1 commit 31a166fe9c269af17977e650846ee4ea50361c07 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 31a166fe9c269af17977e650846ee4ea50361c07 upstream.

Add an IOCTL to send mailbox commands to PUNIT using PUNIT PCI device. A limited set of mailbox commands can be sent to PUNIT.

This MMIO interface is used by the intel-speed-select tool under tools/x86/power to enumerate and control Intel Speed Select features. The MBOX commands ids and semantics of the message can be checked from the source code of the tool.

Signed-off-by: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com Signed-off-by: Andy Shevchenko andriy.shevchenko@linux.intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- .../x86/intel_speed_select_if/Makefile | 1 + .../intel_speed_select_if/isst_if_common.c | 85 ++++++++ .../intel_speed_select_if/isst_if_common.h | 3 + .../intel_speed_select_if/isst_if_mbox_pci.c | 199 ++++++++++++++++++ include/uapi/linux/isst_if.h | 38 ++++ 5 files changed, 326 insertions(+) create mode 100644 drivers/platform/x86/intel_speed_select_if/isst_if_mbox_pci.c

diff --git a/drivers/platform/x86/intel_speed_select_if/Makefile b/drivers/platform/x86/intel_speed_select_if/Makefile index 7e94919208d3e..8dec8c858649e 100644 --- a/drivers/platform/x86/intel_speed_select_if/Makefile +++ b/drivers/platform/x86/intel_speed_select_if/Makefile @@ -6,3 +6,4 @@

obj-$(CONFIG_INTEL_SPEED_SELECT_INTERFACE) += isst_if_common.o obj-$(CONFIG_INTEL_SPEED_SELECT_INTERFACE) += isst_if_mmio.o +obj-$(CONFIG_INTEL_SPEED_SELECT_INTERFACE) += isst_if_mbox_pci.o diff --git a/drivers/platform/x86/intel_speed_select_if/isst_if_common.c b/drivers/platform/x86/intel_speed_select_if/isst_if_common.c index 3f96a3925bc6b..391fc3f121616 100644 --- a/drivers/platform/x86/intel_speed_select_if/isst_if_common.c +++ b/drivers/platform/x86/intel_speed_select_if/isst_if_common.c @@ -25,6 +25,86 @@

static struct isst_if_cmd_cb punit_callbacks[ISST_IF_DEV_MAX];

+struct isst_valid_cmd_ranges { + u16 cmd; + u16 sub_cmd_beg; + u16 sub_cmd_end; +}; + +struct isst_cmd_set_req_type { + u16 cmd; + u16 sub_cmd; + u16 param; +}; + +static const struct isst_valid_cmd_ranges isst_valid_cmds[] = { + {0xD0, 0x00, 0x03}, + {0x7F, 0x00, 0x0B}, + {0x7F, 0x10, 0x12}, + {0x7F, 0x20, 0x23}, +}; + +static const struct isst_cmd_set_req_type isst_cmd_set_reqs[] = { + {0xD0, 0x00, 0x08}, + {0xD0, 0x01, 0x08}, + {0xD0, 0x02, 0x08}, + {0xD0, 0x03, 0x08}, + {0x7F, 0x02, 0x00}, + {0x7F, 0x08, 0x00}, +}; + +/** + * isst_if_mbox_cmd_invalid() - Check invalid mailbox commands + * @cmd: Pointer to the command structure to verify. + * + * Invalid command to PUNIT to may result in instability of the platform. + * This function has a whitelist of commands, which are allowed. + * + * Return: Return true if the command is invalid, else false. + */ +bool isst_if_mbox_cmd_invalid(struct isst_if_mbox_cmd *cmd) +{ + int i; + + if (cmd->logical_cpu >= nr_cpu_ids) + return true; + + for (i = 0; i < ARRAY_SIZE(isst_valid_cmds); ++i) { + if (cmd->command == isst_valid_cmds[i].cmd && + (cmd->sub_command >= isst_valid_cmds[i].sub_cmd_beg && + cmd->sub_command <= isst_valid_cmds[i].sub_cmd_end)) { + return false; + } + } + + return true; +} +EXPORT_SYMBOL_GPL(isst_if_mbox_cmd_invalid); + +/** + * isst_if_mbox_cmd_set_req() - Check mailbox command is a set request + * @cmd: Pointer to the command structure to verify. + * + * Check if the given mail box level is set request and not a get request. + * + * Return: Return true if the command is set_req, else false. + */ +bool isst_if_mbox_cmd_set_req(struct isst_if_mbox_cmd *cmd) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(isst_cmd_set_reqs); ++i) { + if (cmd->command == isst_cmd_set_reqs[i].cmd && + cmd->sub_command == isst_cmd_set_reqs[i].sub_cmd && + cmd->parameter == isst_cmd_set_reqs[i].param) { + return true; + } + } + + return false; +} +EXPORT_SYMBOL_GPL(isst_if_mbox_cmd_set_req); + static int isst_if_get_platform_info(void __user *argp) { struct isst_if_platform_info info; @@ -224,6 +304,11 @@ static long isst_if_def_ioctl(struct file *file, unsigned int cmd, if (cb->registered) ret = isst_if_exec_multi_cmd(argp, cb); break; + case ISST_IF_MBOX_COMMAND: + cb = &punit_callbacks[ISST_IF_DEV_MBOX]; + if (cb->registered) + ret = isst_if_exec_multi_cmd(argp, cb); + break; default: break; } diff --git a/drivers/platform/x86/intel_speed_select_if/isst_if_common.h b/drivers/platform/x86/intel_speed_select_if/isst_if_common.h index cdc7d019748a5..7c0f71221da70 100644 --- a/drivers/platform/x86/intel_speed_select_if/isst_if_common.h +++ b/drivers/platform/x86/intel_speed_select_if/isst_if_common.h @@ -11,6 +11,7 @@ #define __ISST_IF_COMMON_H

#define INTEL_RAPL_PRIO_DEVID_0 0x3451 +#define INTEL_CFG_MBOX_DEVID_0 0x3459

/* * Validate maximum commands in a single request. @@ -60,4 +61,6 @@ struct isst_if_cmd_cb { int isst_if_cdev_register(int type, struct isst_if_cmd_cb *cb); void isst_if_cdev_unregister(int type); struct pci_dev *isst_if_get_pci_dev(int cpu, int bus, int dev, int fn); +bool isst_if_mbox_cmd_set_req(struct isst_if_mbox_cmd *mbox_cmd); +bool isst_if_mbox_cmd_invalid(struct isst_if_mbox_cmd *cmd); #endif diff --git a/drivers/platform/x86/intel_speed_select_if/isst_if_mbox_pci.c b/drivers/platform/x86/intel_speed_select_if/isst_if_mbox_pci.c new file mode 100644 index 0000000000000..1c4f2893cd800 --- /dev/null +++ b/drivers/platform/x86/intel_speed_select_if/isst_if_mbox_pci.c @@ -0,0 +1,199 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel Speed Select Interface: Mbox via PCI Interface + * Copyright (c) 2019, Intel Corporation. + * All rights reserved. + * + * Author: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com + */ + +#include <linux/cpufeature.h> +#include <linux/module.h> +#include <linux/pci.h> +#include <linux/sched/signal.h> +#include <linux/uaccess.h> +#include <uapi/linux/isst_if.h> + +#include "isst_if_common.h" + +#define PUNIT_MAILBOX_DATA 0xA0 +#define PUNIT_MAILBOX_INTERFACE 0xA4 +#define PUNIT_MAILBOX_BUSY_BIT 31 + +/* + * Commands has variable amount of processing time. Most of the commands will + * be done in 0-3 tries, but some takes up to 50. + * The real processing time was observed as 25us for the most of the commands + * at 2GHz. It is possible to optimize this count taking samples on customer + * systems. + */ +#define OS_MAILBOX_RETRY_COUNT 50 + +struct isst_if_device { + struct mutex mutex; +}; + +static int isst_if_mbox_cmd(struct pci_dev *pdev, + struct isst_if_mbox_cmd *mbox_cmd) +{ + u32 retries, data; + int ret; + + /* Poll for rb bit == 0 */ + retries = OS_MAILBOX_RETRY_COUNT; + do { + ret = pci_read_config_dword(pdev, PUNIT_MAILBOX_INTERFACE, + &data); + if (ret) + return ret; + + if (data & BIT_ULL(PUNIT_MAILBOX_BUSY_BIT)) { + ret = -EBUSY; + continue; + } + ret = 0; + break; + } while (--retries); + + if (ret) + return ret; + + /* Write DATA register */ + ret = pci_write_config_dword(pdev, PUNIT_MAILBOX_DATA, + mbox_cmd->req_data); + if (ret) + return ret; + + /* Write command register */ + data = BIT_ULL(PUNIT_MAILBOX_BUSY_BIT) | + (mbox_cmd->parameter & GENMASK_ULL(13, 0)) << 16 | + (mbox_cmd->sub_command << 8) | + mbox_cmd->command; + + ret = pci_write_config_dword(pdev, PUNIT_MAILBOX_INTERFACE, data); + if (ret) + return ret; + + /* Poll for rb bit == 0 */ + retries = OS_MAILBOX_RETRY_COUNT; + do { + ret = pci_read_config_dword(pdev, PUNIT_MAILBOX_INTERFACE, + &data); + if (ret) + return ret; + + if (data & BIT_ULL(PUNIT_MAILBOX_BUSY_BIT)) { + ret = -EBUSY; + continue; + } + + if (data & 0xff) + return -ENXIO; + + ret = pci_read_config_dword(pdev, PUNIT_MAILBOX_DATA, &data); + if (ret) + return ret; + + mbox_cmd->resp_data = data; + ret = 0; + break; + } while (--retries); + + return ret; +} + +static long isst_if_mbox_proc_cmd(u8 *cmd_ptr, int *write_only, int resume) +{ + struct isst_if_mbox_cmd *mbox_cmd; + struct isst_if_device *punit_dev; + struct pci_dev *pdev; + int ret; + + mbox_cmd = (struct isst_if_mbox_cmd *)cmd_ptr; + + if (isst_if_mbox_cmd_invalid(mbox_cmd)) + return -EINVAL; + + if (isst_if_mbox_cmd_set_req(mbox_cmd) && !capable(CAP_SYS_ADMIN)) + return -EPERM; + + pdev = isst_if_get_pci_dev(mbox_cmd->logical_cpu, 1, 30, 1); + if (!pdev) + return -EINVAL; + + punit_dev = pci_get_drvdata(pdev); + if (!punit_dev) + return -EINVAL; + + /* + * Basically we are allowing one complete mailbox transaction on + * a mapped PCI device at a time. + */ + mutex_lock(&punit_dev->mutex); + ret = isst_if_mbox_cmd(pdev, mbox_cmd); + mutex_unlock(&punit_dev->mutex); + if (ret) + return ret; + + *write_only = 0; + + return 0; +} + +static const struct pci_device_id isst_if_mbox_ids[] = { + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, INTEL_CFG_MBOX_DEVID_0)}, + { 0 }, +}; +MODULE_DEVICE_TABLE(pci, isst_if_mbox_ids); + +static int isst_if_mbox_probe(struct pci_dev *pdev, + const struct pci_device_id *ent) +{ + struct isst_if_device *punit_dev; + struct isst_if_cmd_cb cb; + int ret; + + punit_dev = devm_kzalloc(&pdev->dev, sizeof(*punit_dev), GFP_KERNEL); + if (!punit_dev) + return -ENOMEM; + + ret = pcim_enable_device(pdev); + if (ret) + return ret; + + mutex_init(&punit_dev->mutex); + pci_set_drvdata(pdev, punit_dev); + + memset(&cb, 0, sizeof(cb)); + cb.cmd_size = sizeof(struct isst_if_mbox_cmd); + cb.offset = offsetof(struct isst_if_mbox_cmds, mbox_cmd); + cb.cmd_callback = isst_if_mbox_proc_cmd; + cb.owner = THIS_MODULE; + ret = isst_if_cdev_register(ISST_IF_DEV_MBOX, &cb); + + if (ret) + mutex_destroy(&punit_dev->mutex); + + return ret; +} + +static void isst_if_mbox_remove(struct pci_dev *pdev) +{ + struct isst_if_device *punit_dev; + + punit_dev = pci_get_drvdata(pdev); + isst_if_cdev_unregister(ISST_IF_DEV_MBOX); + mutex_destroy(&punit_dev->mutex); +} + +static struct pci_driver isst_if_pci_driver = { + .name = "isst_if_mbox_pci", + .id_table = isst_if_mbox_ids, + .probe = isst_if_mbox_probe, + .remove = isst_if_mbox_remove, +}; + +module_pci_driver(isst_if_pci_driver); + +MODULE_LICENSE("GPL v2"); +MODULE_DESCRIPTION("Intel speed select interface pci mailbox driver"); diff --git a/include/uapi/linux/isst_if.h b/include/uapi/linux/isst_if.h index fe2492ade078e..e4b1c2ec32792 100644 --- a/include/uapi/linux/isst_if.h +++ b/include/uapi/linux/isst_if.h @@ -95,8 +95,46 @@ struct isst_if_io_regs { struct isst_if_io_reg io_reg[1]; };

+/** + * struct isst_if_mbox_cmd - Structure to define mail box command + * @logical_cpu: Logical CPU number to get target PCI device + * @parameter: Mailbox parameter value + * @req_data: Request data for the mailbox + * @resp_data: Response data for mailbox command response + * @command: Mailbox command value + * @sub_command: Mailbox sub command value + * @reserved: Unused, set to 0 + * + * Structure to specify mailbox command to be sent to PUNIT. + */ +struct isst_if_mbox_cmd { + __u32 logical_cpu; + __u32 parameter; + __u32 req_data; + __u32 resp_data; + __u16 command; + __u16 sub_command; + __u32 reserved; +}; + +/** + * struct isst_if_mbox_cmds - structure for mailbox commands + * @cmd_count: Number of mailbox commands in mbox_cmd[] + * @mbox_cmd[]: Holds one or more mbox commands + * + * This structure used with ioctl ISST_IF_MBOX_COMMAND to send + * one or more mailbox commands to PUNIT. Here IOCTL return value + * indicates number of commands sent or error number if no commands have + * been sent. + */ +struct isst_if_mbox_cmds { + __u32 cmd_count; + struct isst_if_mbox_cmd mbox_cmd[1]; +}; + #define ISST_IF_MAGIC 0xFE #define ISST_IF_GET_PLATFORM_INFO _IOR(ISST_IF_MAGIC, 0, struct isst_if_platform_info *) #define ISST_IF_GET_PHY_ID _IOWR(ISST_IF_MAGIC, 1, struct isst_if_cpu_map *) #define ISST_IF_IO_CMD _IOW(ISST_IF_MAGIC, 2, struct isst_if_io_regs *) +#define ISST_IF_MBOX_COMMAND _IOWR(ISST_IF_MAGIC, 3, struct isst_if_mbox_cmds *) #endif

From: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com

mainline inclusion from mainline-v5.3-rc1 commit e765f37b9b8b4fa65682e9a78a2ca2b11d3d9096 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit e765f37b9b8b4fa65682e9a78a2ca2b11d3d9096 upstream.

While using new non arhitectural features using PUNIT Mailbox and MMIO read/write interface, still there is need to operate using MSRs to control PUNIT. User space could have used user user-space MSR interface for this, but when user space MSR access is disabled, then it can't. Here only limited number of MSRs are allowed using this new interface.

Signed-off-by: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com Signed-off-by: Andy Shevchenko andriy.shevchenko@linux.intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- .../intel_speed_select_if/isst_if_common.c | 59 +++++++++++++++++++ include/uapi/linux/isst_if.h | 32 ++++++++++ 2 files changed, 91 insertions(+)

diff --git a/drivers/platform/x86/intel_speed_select_if/isst_if_common.c b/drivers/platform/x86/intel_speed_select_if/isst_if_common.c index 391fc3f121616..de2fb5292f1cd 100644 --- a/drivers/platform/x86/intel_speed_select_if/isst_if_common.c +++ b/drivers/platform/x86/intel_speed_select_if/isst_if_common.c @@ -25,6 +25,11 @@

static struct isst_if_cmd_cb punit_callbacks[ISST_IF_DEV_MAX];

+static int punit_msr_white_list[] = { + MSR_TURBO_RATIO_LIMIT, + MSR_CONFIG_TDP_CONTROL, +}; + struct isst_valid_cmd_ranges { u16 cmd; u16 sub_cmd_beg; @@ -229,6 +234,54 @@ static long isst_if_proc_phyid_req(u8 *cmd_ptr, int *write_only, int resume) return 0; }

+static bool match_punit_msr_white_list(int msr) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(punit_msr_white_list); ++i) { + if (punit_msr_white_list[i] == msr) + return true; + } + + return false; +} + +static long isst_if_msr_cmd_req(u8 *cmd_ptr, int *write_only, int resume) +{ + struct isst_if_msr_cmd *msr_cmd; + int ret; + + msr_cmd = (struct isst_if_msr_cmd *)cmd_ptr; + + if (!match_punit_msr_white_list(msr_cmd->msr)) + return -EINVAL; + + if (msr_cmd->logical_cpu >= nr_cpu_ids) + return -EINVAL; + + if (msr_cmd->read_write) { + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + ret = wrmsrl_safe_on_cpu(msr_cmd->logical_cpu, + msr_cmd->msr, + msr_cmd->data); + *write_only = 1; + } else { + u64 data; + + ret = rdmsrl_safe_on_cpu(msr_cmd->logical_cpu, + msr_cmd->msr, &data); + if (!ret) { + msr_cmd->data = data; + *write_only = 0; + } + } + + + return ret; +} + static long isst_if_exec_multi_cmd(void __user *argp, struct isst_if_cmd_cb *cb) { unsigned char __user *ptr; @@ -309,6 +362,12 @@ static long isst_if_def_ioctl(struct file *file, unsigned int cmd, if (cb->registered) ret = isst_if_exec_multi_cmd(argp, cb); break; + case ISST_IF_MSR_COMMAND: + cmd_cb.cmd_size = sizeof(struct isst_if_msr_cmd); + cmd_cb.offset = offsetof(struct isst_if_msr_cmds, msr_cmd); + cmd_cb.cmd_callback = isst_if_msr_cmd_req; + ret = isst_if_exec_multi_cmd(argp, &cmd_cb); + break; default: break; } diff --git a/include/uapi/linux/isst_if.h b/include/uapi/linux/isst_if.h index e4b1c2ec32792..d10b832c58c58 100644 --- a/include/uapi/linux/isst_if.h +++ b/include/uapi/linux/isst_if.h @@ -132,9 +132,41 @@ struct isst_if_mbox_cmds { struct isst_if_mbox_cmd mbox_cmd[1]; };

+/** + * struct isst_if_msr_cmd - Structure to define msr command + * @read_write: Value 0: Read, 1: Write + * @logical_cpu: Logical CPU number + * @msr: MSR number + * @data: For write operation, data to write, for read + * place holder + * + * Structure to specify MSR command related to PUNIT. + */ +struct isst_if_msr_cmd { + __u32 read_write; /* Read:0, Write:1 */ + __u32 logical_cpu; + __u64 msr; + __u64 data; +}; + +/** + * struct isst_if_msr_cmds - structure for msr commands + * @cmd_count: Number of mailbox commands in msr_cmd[] + * @msr_cmd[]: Holds one or more msr commands + * + * This structure used with ioctl ISST_IF_MSR_COMMAND to send + * one or more MSR commands. IOCTL return value indicates number of + * commands sent or error number if no commands have been sent. + */ +struct isst_if_msr_cmds { + __u32 cmd_count; + struct isst_if_msr_cmd msr_cmd[1]; +}; + #define ISST_IF_MAGIC 0xFE #define ISST_IF_GET_PLATFORM_INFO _IOR(ISST_IF_MAGIC, 0, struct isst_if_platform_info *) #define ISST_IF_GET_PHY_ID _IOWR(ISST_IF_MAGIC, 1, struct isst_if_cpu_map *) #define ISST_IF_IO_CMD _IOW(ISST_IF_MAGIC, 2, struct isst_if_io_regs *) #define ISST_IF_MBOX_COMMAND _IOWR(ISST_IF_MAGIC, 3, struct isst_if_mbox_cmds *) +#define ISST_IF_MSR_COMMAND _IOWR(ISST_IF_MAGIC, 4, struct isst_if_msr_cmds *) #endif

From: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com

mainline inclusion from mainline-v5.3-rc1 commit 3fb4f7cd472c7f5905c91508e988f6b28372210d category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 3fb4f7cd472c7f5905c91508e988f6b28372210d upstream.

The Intel(R) Speed select technologies contains four features.

Performance profile:An non architectural mechanism that allows multiple optimized performance profiles per system via static and/or dynamic adjustment of core count, workload, Tjmax, and TDP, etc. aka ISS in the documentation.

Base Frequency: Enables users to increase guaranteed base frequency on certain cores (high priority cores) in exchange for lower base frequency on remaining cores (low priority cores). aka PBF in the documenation.

Turbo frequency: Enables the ability to set different turbo ratio limits to cores based on priority. aka FACT in the documentation.

Core power: An Interface that allows user to define per core/tile priority.

There is a multi level help for commands and options. This can be used to check required arguments for each feature and commands for the feature.

To start navigating the features start with

$sudo intel-speed-select --help

For help on a specific feature for example $sudo intel-speed-select perf-profile --help

To get help for a command for a feature for example $sudo intel-speed-select perf-profile get-lock-status --help

Signed-off-by: Srinivas Pandruvada srinivas.pandruvada@linux.intel.com Acked-by: Len Brown len.brown@intel.com Acked-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Andy Shevchenko andriy.shevchenko@linux.intel.com Signed-off-by: Youquan Song youquan.song@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- tools/Makefile | 11 +- tools/power/x86/intel-speed-select/Build | 1 + tools/power/x86/intel-speed-select/Makefile | 56 + .../x86/intel-speed-select/isst-config.c | 1607 +++++++++++++++++ .../power/x86/intel-speed-select/isst-core.c | 721 ++++++++ .../x86/intel-speed-select/isst-display.c | 479 +++++ tools/power/x86/intel-speed-select/isst.h | 231 +++ 7 files changed, 3101 insertions(+), 5 deletions(-) create mode 100644 tools/power/x86/intel-speed-select/Build create mode 100644 tools/power/x86/intel-speed-select/Makefile create mode 100644 tools/power/x86/intel-speed-select/isst-config.c create mode 100644 tools/power/x86/intel-speed-select/isst-core.c create mode 100644 tools/power/x86/intel-speed-select/isst-display.c create mode 100644 tools/power/x86/intel-speed-select/isst.h

diff --git a/tools/Makefile b/tools/Makefile index be02c8b904dba..a078ba8e69c82 100644 --- a/tools/Makefile +++ b/tools/Makefile @@ -17,6 +17,7 @@ help: @echo ' gpio - GPIO tools' @echo ' hv - tools used when in Hyper-V clients' @echo ' iio - IIO tools' + @echo ' intel-speed-select - Intel Speed Select tool' @echo ' kvm_stat - top-like utility for displaying kvm statistics' @echo ' leds - LEDs tools' @echo ' liblockdep - user-space wrapper for kernel locking-validator' @@ -79,7 +80,7 @@ perf: FORCE selftests: FORCE $(call descend,testing/$@)

-turbostat x86_energy_perf_policy: FORCE +turbostat x86_energy_perf_policy intel-speed-select: FORCE $(call descend,power/x86/$@)

tmon: FORCE @@ -111,7 +112,7 @@ liblockdep_install: selftests_install: $(call descend,testing/$(@:_install=),install)

-turbostat_install x86_energy_perf_policy_install: +turbostat_install x86_energy_perf_policy_install intel-speed-select_install: $(call descend,power/x86/$(@:_install=),install)

tmon_install: @@ -128,7 +129,7 @@ install: acpi_install cgroup_install cpupower_install gpio_install \ perf_install selftests_install turbostat_install usb_install \ virtio_install vm_install bpf_install x86_energy_perf_policy_install \ tmon_install freefall_install objtool_install kvm_stat_install \ - wmi_install + wmi_install intel-speed-select_install

acpi_clean: $(call descend,power/acpi,clean) @@ -158,7 +159,7 @@ perf_clean: selftests_clean: $(call descend,testing/$(@:_clean=),clean)

-turbostat_clean x86_energy_perf_policy_clean: +turbostat_clean x86_energy_perf_policy_clean intel-speed-select_clean: $(call descend,power/x86/$(@:_clean=),clean)

tmon_clean: @@ -174,6 +175,6 @@ clean: acpi_clean cgroup_clean cpupower_clean hv_clean firewire_clean \ perf_clean selftests_clean turbostat_clean spi_clean usb_clean virtio_clean \ vm_clean bpf_clean iio_clean x86_energy_perf_policy_clean tmon_clean \ freefall_clean build_clean libbpf_clean libsubcmd_clean liblockdep_clean \ - gpio_clean objtool_clean leds_clean wmi_clean + gpio_clean objtool_clean leds_clean wmi_clean intel-speed-select_clean

.PHONY: FORCE diff --git a/tools/power/x86/intel-speed-select/Build b/tools/power/x86/intel-speed-select/Build new file mode 100644 index 0000000000000..b61456d75190f --- /dev/null +++ b/tools/power/x86/intel-speed-select/Build @@ -0,0 +1 @@ +intel-speed-select-y += isst-config.o isst-core.o isst-display.o diff --git a/tools/power/x86/intel-speed-select/Makefile b/tools/power/x86/intel-speed-select/Makefile new file mode 100644 index 0000000000000..12c6939dca2a4 --- /dev/null +++ b/tools/power/x86/intel-speed-select/Makefile @@ -0,0 +1,56 @@ +# SPDX-License-Identifier: GPL-2.0 +include ../../../scripts/Makefile.include + +bindir ?= /usr/bin + +ifeq ($(srctree),) +srctree := $(patsubst %/,%,$(dir $(CURDIR))) +srctree := $(patsubst %/,%,$(dir $(srctree))) +srctree := $(patsubst %/,%,$(dir $(srctree))) +srctree := $(patsubst %/,%,$(dir $(srctree))) +endif + +# Do not use make's built-in rules +# (this improves performance and avoids hard-to-debug behaviour); +MAKEFLAGS += -r + +override CFLAGS += -O2 -Wall -g -D_GNU_SOURCE -I$(OUTPUT)include + +ALL_TARGETS := intel-speed-select +ALL_PROGRAMS := $(patsubst %,$(OUTPUT)%,$(ALL_TARGETS)) + +all: $(ALL_PROGRAMS) + +export srctree OUTPUT CC LD CFLAGS +include $(srctree)/tools/build/Makefile.include + +# +# We need the following to be outside of kernel tree +# +$(OUTPUT)include/linux/isst_if.h: ../../../../include/uapi/linux/isst_if.h + mkdir -p $(OUTPUT)include/linux 2>&1 || true + ln -sf $(CURDIR)/../../../../include/uapi/linux/isst_if.h $@ + +prepare: $(OUTPUT)include/linux/isst_if.h + +ISST_IN := $(OUTPUT)intel-speed-select-in.o + +$(ISST_IN): prepare FORCE + $(Q)$(MAKE) $(build)=intel-speed-select +$(OUTPUT)intel-speed-select: $(ISST_IN) + $(QUIET_LINK)$(CC) $(CFLAGS) $(LDFLAGS) $< -o $@ + +clean: + rm -f $(ALL_PROGRAMS) + rm -rf $(OUTPUT)include/linux/isst_if.h + find $(if $(OUTPUT),$(OUTPUT),.) -name '*.o' -delete -o -name '.*.d' -delete + +install: $(ALL_PROGRAMS) + install -d -m 755 $(DESTDIR)$(bindir); \ + for program in $(ALL_PROGRAMS); do \ + install $$program $(DESTDIR)$(bindir); \ + done + +FORCE: + +.PHONY: all install clean FORCE prepare diff --git a/tools/power/x86/intel-speed-select/isst-config.c b/tools/power/x86/intel-speed-select/isst-config.c new file mode 100644 index 0000000000000..91c5ad1685a15 --- /dev/null +++ b/tools/power/x86/intel-speed-select/isst-config.c @@ -0,0 +1,1607 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel Speed Select -- Enumerate and control features + * Copyright (c) 2019 Intel Corporation. + */ + +#include <linux/isst_if.h> + +#include "isst.h" + +struct process_cmd_struct { + char *feature; + char *command; + void (*process_fn)(void); +}; + +static const char *version_str = "v1.0"; +static const int supported_api_ver = 1; +static struct isst_if_platform_info isst_platform_info; +static char *progname; +static int debug_flag; +static FILE *outf; + +static int cpu_model; + +#define MAX_CPUS_IN_ONE_REQ 64 +static short max_target_cpus; +static unsigned short target_cpus[MAX_CPUS_IN_ONE_REQ]; + +static int topo_max_cpus; +static size_t present_cpumask_size; +static cpu_set_t *present_cpumask; +static size_t target_cpumask_size; +static cpu_set_t *target_cpumask; +static int tdp_level = 0xFF; +static int fact_bucket = 0xFF; +static int fact_avx = 0xFF; +static unsigned long long fact_trl; +static int out_format_json; +static int cmd_help; + +/* clos related */ +static int current_clos = -1; +static int clos_epp = -1; +static int clos_prop_prio = -1; +static int clos_min = -1; +static int clos_max = -1; +static int clos_desired = -1; +static int clos_priority_type; + +struct _cpu_map { + unsigned short core_id; + unsigned short pkg_id; + unsigned short die_id; + unsigned short punit_cpu; + unsigned short punit_cpu_core; +}; +struct _cpu_map *cpu_map; + +void debug_printf(const char *format, ...) +{ + va_list args; + + va_start(args, format); + + if (debug_flag) + vprintf(format, args); + + va_end(args); +} + +static void update_cpu_model(void) +{ + unsigned int ebx, ecx, edx; + unsigned int fms, family; + + __cpuid(1, fms, ebx, ecx, edx); + family = (fms >> 8) & 0xf; + cpu_model = (fms >> 4) & 0xf; + if (family == 6 || family == 0xf) + cpu_model += ((fms >> 16) & 0xf) << 4; +} + +/* Open a file, and exit on failure */ +static FILE *fopen_or_exit(const char *path, const char *mode) +{ + FILE *filep = fopen(path, mode); + + if (!filep) + err(1, "%s: open failed", path); + + return filep; +} + +/* Parse a file containing a single int */ +static int parse_int_file(int fatal, const char *fmt, ...) +{ + va_list args; + char path[PATH_MAX]; + FILE *filep; + int value; + + va_start(args, fmt); + vsnprintf(path, sizeof(path), fmt, args); + va_end(args); + if (fatal) { + filep = fopen_or_exit(path, "r"); + } else { + filep = fopen(path, "r"); + if (!filep) + return -1; + } + if (fscanf(filep, "%d", &value) != 1) + err(1, "%s: failed to parse number from file", path); + fclose(filep); + + return value; +} + +int cpufreq_sysfs_present(void) +{ + DIR *dir; + + dir = opendir("/sys/devices/system/cpu/cpu0/cpufreq"); + if (dir) { + closedir(dir); + return 1; + } + + return 0; +} + +int out_format_is_json(void) +{ + return out_format_json; +} + +int get_physical_package_id(int cpu) +{ + return parse_int_file( + 1, "/sys/devices/system/cpu/cpu%d/topology/physical_package_id", + cpu); +} + +int get_physical_core_id(int cpu) +{ + return parse_int_file( + 1, "/sys/devices/system/cpu/cpu%d/topology/core_id", cpu); +} + +int get_physical_die_id(int cpu) +{ + int ret; + + ret = parse_int_file(0, "/sys/devices/system/cpu/cpu%d/topology/die_id", + cpu); + if (ret < 0) + ret = 0; + + return ret; +} + +int get_topo_max_cpus(void) +{ + return topo_max_cpus; +} + +#define MAX_PACKAGE_COUNT 8 +#define MAX_DIE_PER_PACKAGE 2 +static void for_each_online_package_in_set(void (*callback)(int, void *, void *, + void *, void *), + void *arg1, void *arg2, void *arg3, + void *arg4) +{ + int max_packages[MAX_PACKAGE_COUNT * MAX_PACKAGE_COUNT]; + int pkg_index = 0, i; + + memset(max_packages, 0xff, sizeof(max_packages)); + for (i = 0; i < topo_max_cpus; ++i) { + int j, online, pkg_id, die_id = 0, skip = 0; + + if (!CPU_ISSET_S(i, present_cpumask_size, present_cpumask)) + continue; + if (i) + online = parse_int_file( + 1, "/sys/devices/system/cpu/cpu%d/online", i); + else + online = + 1; /* online entry for CPU 0 needs some special configs */ + + die_id = get_physical_die_id(i); + if (die_id < 0) + die_id = 0; + pkg_id = get_physical_package_id(i); + /* Create an unique id for package, die combination to store */ + pkg_id = (MAX_PACKAGE_COUNT * pkg_id + die_id); + + for (j = 0; j < pkg_index; ++j) { + if (max_packages[j] == pkg_id) { + skip = 1; + break; + } + } + + if (!skip && online && callback) { + callback(i, arg1, arg2, arg3, arg4); + max_packages[pkg_index++] = pkg_id; + } + } +} + +static void for_each_online_target_cpu_in_set( + void (*callback)(int, void *, void *, void *, void *), void *arg1, + void *arg2, void *arg3, void *arg4) +{ + int i; + + for (i = 0; i < topo_max_cpus; ++i) { + int online; + + if (!CPU_ISSET_S(i, target_cpumask_size, target_cpumask)) + continue; + if (i) + online = parse_int_file( + 1, "/sys/devices/system/cpu/cpu%d/online", i); + else + online = + 1; /* online entry for CPU 0 needs some special configs */ + + if (online && callback) + callback(i, arg1, arg2, arg3, arg4); + } +} + +#define BITMASK_SIZE 32 +static void set_max_cpu_num(void) +{ + FILE *filep; + unsigned long dummy; + + topo_max_cpus = 0; + filep = fopen_or_exit( + "/sys/devices/system/cpu/cpu0/topology/thread_siblings", "r"); + while (fscanf(filep, "%lx,", &dummy) == 1) + topo_max_cpus += BITMASK_SIZE; + fclose(filep); + topo_max_cpus--; /* 0 based */ + + debug_printf("max cpus %d\n", topo_max_cpus); +} + +size_t alloc_cpu_set(cpu_set_t **cpu_set) +{ + cpu_set_t *_cpu_set; + size_t size; + + _cpu_set = CPU_ALLOC((topo_max_cpus + 1)); + if (_cpu_set == NULL) + err(3, "CPU_ALLOC"); + size = CPU_ALLOC_SIZE((topo_max_cpus + 1)); + CPU_ZERO_S(size, _cpu_set); + + *cpu_set = _cpu_set; + return size; +} + +void free_cpu_set(cpu_set_t *cpu_set) +{ + CPU_FREE(cpu_set); +} + +static int cpu_cnt[MAX_PACKAGE_COUNT][MAX_DIE_PER_PACKAGE]; +static void set_cpu_present_cpu_mask(void) +{ + size_t size; + DIR *dir; + int i; + + size = alloc_cpu_set(&present_cpumask); + present_cpumask_size = size; + for (i = 0; i < topo_max_cpus; ++i) { + char buffer[256]; + + snprintf(buffer, sizeof(buffer), + "/sys/devices/system/cpu/cpu%d", i); + dir = opendir(buffer); + if (dir) { + int pkg_id, die_id; + + CPU_SET_S(i, size, present_cpumask); + die_id = get_physical_die_id(i); + if (die_id < 0) + die_id = 0; + + pkg_id = get_physical_package_id(i); + if (pkg_id < MAX_PACKAGE_COUNT && + die_id < MAX_DIE_PER_PACKAGE) + cpu_cnt[pkg_id][die_id]++; + } + closedir(dir); + } +} + +int get_cpu_count(int pkg_id, int die_id) +{ + if (pkg_id < MAX_PACKAGE_COUNT && die_id < MAX_DIE_PER_PACKAGE) + return cpu_cnt[pkg_id][die_id] + 1; + + return 0; +} + +static void set_cpu_target_cpu_mask(void) +{ + size_t size; + int i; + + size = alloc_cpu_set(&target_cpumask); + target_cpumask_size = size; + for (i = 0; i < max_target_cpus; ++i) { + if (!CPU_ISSET_S(target_cpus[i], present_cpumask_size, + present_cpumask)) + continue; + + CPU_SET_S(target_cpus[i], size, target_cpumask); + } +} + +static void create_cpu_map(void) +{ + const char *pathname = "/dev/isst_interface"; + int i, fd = 0; + struct isst_if_cpu_maps map; + + cpu_map = malloc(sizeof(*cpu_map) * topo_max_cpus); + if (!cpu_map) + err(3, "cpumap"); + + fd = open(pathname, O_RDWR); + if (fd < 0) + err(-1, "%s open failed", pathname); + + for (i = 0; i < topo_max_cpus; ++i) { + if (!CPU_ISSET_S(i, present_cpumask_size, present_cpumask)) + continue; + + map.cmd_count = 1; + map.cpu_map[0].logical_cpu = i; + + debug_printf(" map logical_cpu:%d\n", + map.cpu_map[0].logical_cpu); + if (ioctl(fd, ISST_IF_GET_PHY_ID, &map) == -1) { + perror("ISST_IF_GET_PHY_ID"); + fprintf(outf, "Error: map logical_cpu:%d\n", + map.cpu_map[0].logical_cpu); + continue; + } + cpu_map[i].core_id = get_physical_core_id(i); + cpu_map[i].pkg_id = get_physical_package_id(i); + cpu_map[i].die_id = get_physical_die_id(i); + cpu_map[i].punit_cpu = map.cpu_map[0].physical_cpu; + cpu_map[i].punit_cpu_core = (map.cpu_map[0].physical_cpu >> + 1); // shift to get core id + + debug_printf( + "map logical_cpu:%d core: %d die:%d pkg:%d punit_cpu:%d punit_core:%d\n", + i, cpu_map[i].core_id, cpu_map[i].die_id, + cpu_map[i].pkg_id, cpu_map[i].punit_cpu, + cpu_map[i].punit_cpu_core); + } + + if (fd) + close(fd); +} + +int find_logical_cpu(int pkg_id, int die_id, int punit_core_id) +{ + int i; + + for (i = 0; i < topo_max_cpus; ++i) { + if (cpu_map[i].pkg_id == pkg_id && + cpu_map[i].die_id == die_id && + cpu_map[i].punit_cpu_core == punit_core_id) + return i; + } + + return -EINVAL; +} + +void set_cpu_mask_from_punit_coremask(int cpu, unsigned long long core_mask, + size_t core_cpumask_size, + cpu_set_t *core_cpumask, int *cpu_cnt) +{ + int i, cnt = 0; + int die_id, pkg_id; + + *cpu_cnt = 0; + die_id = get_physical_die_id(cpu); + pkg_id = get_physical_package_id(cpu); + + for (i = 0; i < 64; ++i) { + if (core_mask & BIT(i)) { + int j; + + for (j = 0; j < topo_max_cpus; ++j) { + if (cpu_map[j].pkg_id == pkg_id && + cpu_map[j].die_id == die_id && + cpu_map[j].punit_cpu_core == i) { + CPU_SET_S(j, core_cpumask_size, + core_cpumask); + ++cnt; + } + } + } + } + + *cpu_cnt = cnt; +} + +int find_phy_core_num(int logical_cpu) +{ + if (logical_cpu < topo_max_cpus) + return cpu_map[logical_cpu].punit_cpu_core; + + return -EINVAL; +} + +static int isst_send_mmio_command(unsigned int cpu, unsigned int reg, int write, + unsigned int *value) +{ + struct isst_if_io_regs io_regs; + const char *pathname = "/dev/isst_interface"; + int cmd; + int fd; + + debug_printf("mmio_cmd cpu:%d reg:%d write:%d\n", cpu, reg, write); + + fd = open(pathname, O_RDWR); + if (fd < 0) + err(-1, "%s open failed", pathname); + + io_regs.req_count = 1; + io_regs.io_reg[0].logical_cpu = cpu; + io_regs.io_reg[0].reg = reg; + cmd = ISST_IF_IO_CMD; + if (write) { + io_regs.io_reg[0].read_write = 1; + io_regs.io_reg[0].value = *value; + } else { + io_regs.io_reg[0].read_write = 0; + } + + if (ioctl(fd, cmd, &io_regs) == -1) { + perror("ISST_IF_IO_CMD"); + fprintf(outf, "Error: mmio_cmd cpu:%d reg:%x read_write:%x\n", + cpu, reg, write); + } else { + if (!write) + *value = io_regs.io_reg[0].value; + + debug_printf( + "mmio_cmd response: cpu:%d reg:%x rd_write:%x resp:%x\n", + cpu, reg, write, *value); + } + + close(fd); + + return 0; +} + +int isst_send_mbox_command(unsigned int cpu, unsigned char command, + unsigned char sub_command, unsigned int parameter, + unsigned int req_data, unsigned int *resp) +{ + const char *pathname = "/dev/isst_interface"; + int fd; + struct isst_if_mbox_cmds mbox_cmds = { 0 }; + + debug_printf( + "mbox_send: cpu:%d command:%x sub_command:%x parameter:%x req_data:%x\n", + cpu, command, sub_command, parameter, req_data); + + if (isst_platform_info.mmio_supported && command == CONFIG_CLOS) { + unsigned int value; + int write = 0; + int clos_id, core_id, ret = 0; + + debug_printf("CLOS %d\n", cpu); + + if (parameter & BIT(MBOX_CMD_WRITE_BIT)) { + value = req_data; + write = 1; + } + + switch (sub_command) { + case CLOS_PQR_ASSOC: + core_id = parameter & 0xff; + ret = isst_send_mmio_command( + cpu, PQR_ASSOC_OFFSET + core_id * 4, write, + &value); + if (!ret && !write) + *resp = value; + break; + case CLOS_PM_CLOS: + clos_id = parameter & 0x03; + ret = isst_send_mmio_command( + cpu, PM_CLOS_OFFSET + clos_id * 4, write, + &value); + if (!ret && !write) + *resp = value; + break; + case CLOS_PM_QOS_CONFIG: + ret = isst_send_mmio_command(cpu, PM_QOS_CONFIG_OFFSET, + write, &value); + if (!ret && !write) + *resp = value; + break; + case CLOS_STATUS: + break; + default: + break; + } + return ret; + } + + mbox_cmds.cmd_count = 1; + mbox_cmds.mbox_cmd[0].logical_cpu = cpu; + mbox_cmds.mbox_cmd[0].command = command; + mbox_cmds.mbox_cmd[0].sub_command = sub_command; + mbox_cmds.mbox_cmd[0].parameter = parameter; + mbox_cmds.mbox_cmd[0].req_data = req_data; + + fd = open(pathname, O_RDWR); + if (fd < 0) + err(-1, "%s open failed", pathname); + + if (ioctl(fd, ISST_IF_MBOX_COMMAND, &mbox_cmds) == -1) { + perror("ISST_IF_MBOX_COMMAND"); + fprintf(outf, + "Error: mbox_cmd cpu:%d command:%x sub_command:%x parameter:%x req_data:%x\n", + cpu, command, sub_command, parameter, req_data); + } else { + *resp = mbox_cmds.mbox_cmd[0].resp_data; + debug_printf( + "mbox_cmd response: cpu:%d command:%x sub_command:%x parameter:%x req_data:%x resp:%x\n", + cpu, command, sub_command, parameter, req_data, *resp); + } + + close(fd); + + return 0; +} + +int isst_send_msr_command(unsigned int cpu, unsigned int msr, int write, + unsigned long long *req_resp) +{ + struct isst_if_msr_cmds msr_cmds; + const char *pathname = "/dev/isst_interface"; + int fd; + + fd = open(pathname, O_RDWR); + if (fd < 0) + err(-1, "%s open failed", pathname); + + msr_cmds.cmd_count = 1; + msr_cmds.msr_cmd[0].logical_cpu = cpu; + msr_cmds.msr_cmd[0].msr = msr; + msr_cmds.msr_cmd[0].read_write = write; + if (write) + msr_cmds.msr_cmd[0].data = *req_resp; + + if (ioctl(fd, ISST_IF_MSR_COMMAND, &msr_cmds) == -1) { + perror("ISST_IF_MSR_COMMAD"); + fprintf(outf, "Error: msr_cmd cpu:%d msr:%x read_write:%d\n", + cpu, msr, write); + } else { + if (!write) + *req_resp = msr_cmds.msr_cmd[0].data; + + debug_printf( + "msr_cmd response: cpu:%d msr:%x rd_write:%x resp:%llx %llx\n", + cpu, msr, write, *req_resp, msr_cmds.msr_cmd[0].data); + } + + close(fd); + + return 0; +} + +static int isst_fill_platform_info(void) +{ + const char *pathname = "/dev/isst_interface"; + int fd; + + fd = open(pathname, O_RDWR); + if (fd < 0) + err(-1, "%s open failed", pathname); + + if (ioctl(fd, ISST_IF_GET_PLATFORM_INFO, &isst_platform_info) == -1) { + perror("ISST_IF_GET_PLATFORM_INFO"); + close(fd); + return -1; + } + + close(fd); + + return 0; +} + +static void isst_print_platform_information(void) +{ + struct isst_if_platform_info platform_info; + const char *pathname = "/dev/isst_interface"; + int fd; + + fd = open(pathname, O_RDWR); + if (fd < 0) + err(-1, "%s open failed", pathname); + + if (ioctl(fd, ISST_IF_GET_PLATFORM_INFO, &platform_info) == -1) { + perror("ISST_IF_GET_PLATFORM_INFO"); + } else { + fprintf(outf, "Platform: API version : %d\n", + platform_info.api_version); + fprintf(outf, "Platform: Driver version : %d\n", + platform_info.driver_version); + fprintf(outf, "Platform: mbox supported : %d\n", + platform_info.mbox_supported); + fprintf(outf, "Platform: mmio supported : %d\n", + platform_info.mmio_supported); + } + + close(fd); + + exit(0); +} + +static void exec_on_get_ctdp_cpu(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + int (*fn_ptr)(int cpu, void *arg); + int ret; + + fn_ptr = arg1; + ret = fn_ptr(cpu, arg2); + if (ret) + perror("get_tdp_*"); + else + isst_display_result(cpu, outf, "perf-profile", (char *)arg3, + *(unsigned int *)arg4); +} + +#define _get_tdp_level(desc, suffix, object, help) \ + static void get_tdp_##object(void) \ + { \ + struct isst_pkg_ctdp ctdp; \ +\ + if (cmd_help) { \ + fprintf(stderr, \ + "Print %s [No command arguments are required]\n", \ + help); \ + exit(0); \ + } \ + isst_ctdp_display_information_start(outf); \ + if (max_target_cpus) \ + for_each_online_target_cpu_in_set( \ + exec_on_get_ctdp_cpu, isst_get_ctdp_##suffix, \ + &ctdp, desc, &ctdp.object); \ + else \ + for_each_online_package_in_set(exec_on_get_ctdp_cpu, \ + isst_get_ctdp_##suffix, \ + &ctdp, desc, \ + &ctdp.object); \ + isst_ctdp_display_information_end(outf); \ + } + +_get_tdp_level("get-config-levels", levels, levels, "TDP levels"); +_get_tdp_level("get-config-version", levels, version, "TDP version"); +_get_tdp_level("get-config-enabled", levels, enabled, "TDP enable status"); +_get_tdp_level("get-config-current_level", levels, current_level, + "Current TDP Level"); +_get_tdp_level("get-lock-status", levels, locked, "TDP lock status"); + +static void dump_isst_config_for_cpu(int cpu, void *arg1, void *arg2, + void *arg3, void *arg4) +{ + struct isst_pkg_ctdp pkg_dev; + int ret; + + memset(&pkg_dev, 0, sizeof(pkg_dev)); + ret = isst_get_process_ctdp(cpu, tdp_level, &pkg_dev); + if (ret) { + perror("isst_get_process_ctdp"); + } else { + isst_ctdp_display_information(cpu, outf, tdp_level, &pkg_dev); + isst_get_process_ctdp_complete(cpu, &pkg_dev); + } +} + +static void dump_isst_config(void) +{ + if (cmd_help) { + fprintf(stderr, + "Print Intel(R) Speed Select Technology Performance profile configuration\n"); + fprintf(stderr, + "including base frequency and turbo frequency configurations\n"); + fprintf(stderr, "Optional: -l|--level : Specify tdp level\n"); + fprintf(stderr, + "\tIf no arguments, dump information for all TDP levels\n"); + exit(0); + } + + isst_ctdp_display_information_start(outf); + + if (max_target_cpus) + for_each_online_target_cpu_in_set(dump_isst_config_for_cpu, + NULL, NULL, NULL, NULL); + else + for_each_online_package_in_set(dump_isst_config_for_cpu, NULL, + NULL, NULL, NULL); + + isst_ctdp_display_information_end(outf); +} + +static void set_tdp_level_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + int ret; + + ret = isst_set_tdp_level(cpu, tdp_level); + if (ret) + perror("set_tdp_level_for_cpu"); + else + isst_display_result(cpu, outf, "perf-profile", "set_tdp_level", + ret); +} + +static void set_tdp_level(void) +{ + if (cmd_help) { + fprintf(stderr, "Set Config TDP level\n"); + fprintf(stderr, + "\t Arguments: -l|--level : Specify tdp level\n"); + exit(0); + } + + if (tdp_level == 0xff) { + fprintf(outf, "Invalid command: specify tdp_level\n"); + exit(1); + } + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(set_tdp_level_for_cpu, NULL, + NULL, NULL, NULL); + else + for_each_online_package_in_set(set_tdp_level_for_cpu, NULL, + NULL, NULL, NULL); + isst_ctdp_display_information_end(outf); +} + +static void dump_pbf_config_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + struct isst_pbf_info pbf_info; + int ret; + + ret = isst_get_pbf_info(cpu, tdp_level, &pbf_info); + if (ret) { + perror("isst_get_pbf_info"); + } else { + isst_pbf_display_information(cpu, outf, tdp_level, &pbf_info); + isst_get_pbf_info_complete(&pbf_info); + } +} + +static void dump_pbf_config(void) +{ + if (cmd_help) { + fprintf(stderr, + "Print Intel(R) Speed Select Technology base frequency configuration for a TDP level\n"); + fprintf(stderr, + "\tArguments: -l|--level : Specify tdp level\n"); + exit(0); + } + + if (tdp_level == 0xff) { + fprintf(outf, "Invalid command: specify tdp_level\n"); + exit(1); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(dump_pbf_config_for_cpu, NULL, + NULL, NULL, NULL); + else + for_each_online_package_in_set(dump_pbf_config_for_cpu, NULL, + NULL, NULL, NULL); + isst_ctdp_display_information_end(outf); +} + +static void set_pbf_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + int ret; + int status = *(int *)arg4; + + ret = isst_set_pbf_fact_status(cpu, 1, status); + if (ret) { + perror("isst_set_pbf"); + } else { + if (status) + isst_display_result(cpu, outf, "base-freq", "enable", + ret); + else + isst_display_result(cpu, outf, "base-freq", "disable", + ret); + } +} + +static void set_pbf_enable(void) +{ + int status = 1; + + if (cmd_help) { + fprintf(stderr, + "Enable Intel Speed Select Technology base frequency feature [No command arguments are required]\n"); + exit(0); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(set_pbf_for_cpu, NULL, NULL, + NULL, &status); + else + for_each_online_package_in_set(set_pbf_for_cpu, NULL, NULL, + NULL, &status); + isst_ctdp_display_information_end(outf); +} + +static void set_pbf_disable(void) +{ + int status = 0; + + if (cmd_help) { + fprintf(stderr, + "Disable Intel Speed Select Technology base frequency feature [No command arguments are required]\n"); + exit(0); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(set_pbf_for_cpu, NULL, NULL, + NULL, &status); + else + for_each_online_package_in_set(set_pbf_for_cpu, NULL, NULL, + NULL, &status); + isst_ctdp_display_information_end(outf); +} + +static void dump_fact_config_for_cpu(int cpu, void *arg1, void *arg2, + void *arg3, void *arg4) +{ + struct isst_fact_info fact_info; + int ret; + + ret = isst_get_fact_info(cpu, tdp_level, &fact_info); + if (ret) + perror("isst_get_fact_bucket_info"); + else + isst_fact_display_information(cpu, outf, tdp_level, fact_bucket, + fact_avx, &fact_info); +} + +static void dump_fact_config(void) +{ + if (cmd_help) { + fprintf(stderr, + "Print complete Intel Speed Select Technology turbo frequency configuration for a TDP level. Other arguments are optional.\n"); + fprintf(stderr, + "\tArguments: -l|--level : Specify tdp level\n"); + fprintf(stderr, + "\tArguments: -b|--bucket : Bucket index to dump\n"); + fprintf(stderr, + "\tArguments: -r|--trl-type : Specify trl type: sse|avx2|avx512\n"); + exit(0); + } + + if (tdp_level == 0xff) { + fprintf(outf, "Invalid command: specify tdp_level\n"); + exit(1); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(dump_fact_config_for_cpu, + NULL, NULL, NULL, NULL); + else + for_each_online_package_in_set(dump_fact_config_for_cpu, NULL, + NULL, NULL, NULL); + isst_ctdp_display_information_end(outf); +} + +static void set_fact_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + int ret; + int status = *(int *)arg4; + + ret = isst_set_pbf_fact_status(cpu, 0, status); + if (ret) + perror("isst_set_fact"); + else { + if (status) { + struct isst_pkg_ctdp pkg_dev; + + ret = isst_get_ctdp_levels(cpu, &pkg_dev); + if (ret) { + isst_display_result(cpu, outf, "turbo-freq", + "enable", ret); + return; + } + ret = isst_set_trl(cpu, fact_trl); + isst_display_result(cpu, outf, "turbo-freq", "enable", + ret); + } else { + /* Since we modified TRL during Fact enable, restore it */ + isst_set_trl_from_current_tdp(cpu, fact_trl); + isst_display_result(cpu, outf, "turbo-freq", "disable", + ret); + } + } +} + +static void set_fact_enable(void) +{ + int status = 1; + + if (cmd_help) { + fprintf(stderr, + "Enable Intel Speed Select Technology Turbo frequency feature\n"); + fprintf(stderr, + "Optional: -t|--trl : Specify turbo ratio limit\n"); + exit(0); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(set_fact_for_cpu, NULL, NULL, + NULL, &status); + else + for_each_online_package_in_set(set_fact_for_cpu, NULL, NULL, + NULL, &status); + isst_ctdp_display_information_end(outf); +} + +static void set_fact_disable(void) +{ + int status = 0; + + if (cmd_help) { + fprintf(stderr, + "Disable Intel Speed Select Technology turbo frequency feature\n"); + fprintf(stderr, + "Optional: -t|--trl : Specify turbo ratio limit\n"); + exit(0); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(set_fact_for_cpu, NULL, NULL, + NULL, &status); + else + for_each_online_package_in_set(set_fact_for_cpu, NULL, NULL, + NULL, &status); + isst_ctdp_display_information_end(outf); +} + +static void enable_clos_qos_config(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + int ret; + int status = *(int *)arg4; + + ret = isst_pm_qos_config(cpu, status, clos_priority_type); + if (ret) { + perror("isst_pm_qos_config"); + } else { + if (status) + isst_display_result(cpu, outf, "core-power", "enable", + ret); + else + isst_display_result(cpu, outf, "core-power", "disable", + ret); + } +} + +static void set_clos_enable(void) +{ + int status = 1; + + if (cmd_help) { + fprintf(stderr, "Enable core-power for a package/die\n"); + fprintf(stderr, + "\tClos Enable: Specify priority type with [--priority|-p]\n"); + fprintf(stderr, "\t\t 0: Proportional, 1: Ordered\n"); + exit(0); + } + + if (cpufreq_sysfs_present()) { + fprintf(stderr, + "cpufreq subsystem and core-power enable will interfere with each other!\n"); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(enable_clos_qos_config, NULL, + NULL, NULL, &status); + else + for_each_online_package_in_set(enable_clos_qos_config, NULL, + NULL, NULL, &status); + isst_ctdp_display_information_end(outf); +} + +static void set_clos_disable(void) +{ + int status = 0; + + if (cmd_help) { + fprintf(stderr, + "Disable core-power: [No command arguments are required]\n"); + exit(0); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(enable_clos_qos_config, NULL, + NULL, NULL, &status); + else + for_each_online_package_in_set(enable_clos_qos_config, NULL, + NULL, NULL, &status); + isst_ctdp_display_information_end(outf); +} + +static void dump_clos_config_for_cpu(int cpu, void *arg1, void *arg2, + void *arg3, void *arg4) +{ + struct isst_clos_config clos_config; + int ret; + + ret = isst_pm_get_clos(cpu, current_clos, &clos_config); + if (ret) + perror("isst_pm_get_clos"); + else + isst_clos_display_information(cpu, outf, current_clos, + &clos_config); +} + +static void dump_clos_config(void) +{ + if (cmd_help) { + fprintf(stderr, + "Print Intel Speed Select Technology core power configuration\n"); + fprintf(stderr, + "\tArguments: [-c | --clos]: Specify clos id\n"); + exit(0); + } + if (current_clos < 0 || current_clos > 3) { + fprintf(stderr, "Invalid clos id\n"); + exit(0); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(dump_clos_config_for_cpu, + NULL, NULL, NULL, NULL); + else + for_each_online_package_in_set(dump_clos_config_for_cpu, NULL, + NULL, NULL, NULL); + isst_ctdp_display_information_end(outf); +} + +static void set_clos_config_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + struct isst_clos_config clos_config; + int ret; + + clos_config.pkg_id = get_physical_package_id(cpu); + clos_config.die_id = get_physical_die_id(cpu); + + clos_config.epp = clos_epp; + clos_config.clos_prop_prio = clos_prop_prio; + clos_config.clos_min = clos_min; + clos_config.clos_max = clos_max; + clos_config.clos_desired = clos_desired; + ret = isst_set_clos(cpu, current_clos, &clos_config); + if (ret) + perror("isst_set_clos"); + else + isst_display_result(cpu, outf, "core-power", "config", ret); +} + +static void set_clos_config(void) +{ + if (cmd_help) { + fprintf(stderr, + "Set core-power configuration for one of the four clos ids\n"); + fprintf(stderr, + "\tSpecify targeted clos id with [--clos|-c]\n"); + fprintf(stderr, "\tSpecify clos EPP with [--epp|-e]\n"); + fprintf(stderr, + "\tSpecify clos Proportional Priority [--weight|-w]\n"); + fprintf(stderr, "\tSpecify clos min with [--min|-n]\n"); + fprintf(stderr, "\tSpecify clos max with [--max|-m]\n"); + fprintf(stderr, "\tSpecify clos desired with [--desired|-d]\n"); + exit(0); + } + + if (current_clos < 0 || current_clos > 3) { + fprintf(stderr, "Invalid clos id\n"); + exit(0); + } + if (clos_epp < 0 || clos_epp > 0x0F) { + fprintf(stderr, "clos epp is not specified, default: 0\n"); + clos_epp = 0; + } + if (clos_prop_prio < 0 || clos_prop_prio > 0x0F) { + fprintf(stderr, + "clos frequency weight is not specified, default: 0\n"); + clos_prop_prio = 0; + } + if (clos_min < 0) { + fprintf(stderr, "clos min is not specified, default: 0\n"); + clos_min = 0; + } + if (clos_max < 0) { + fprintf(stderr, "clos max is not specified, default: 0xff\n"); + clos_max = 0xff; + } + if (clos_desired < 0) { + fprintf(stderr, "clos desired is not specified, default: 0\n"); + clos_desired = 0x00; + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(set_clos_config_for_cpu, NULL, + NULL, NULL, NULL); + else + for_each_online_package_in_set(set_clos_config_for_cpu, NULL, + NULL, NULL, NULL); + isst_ctdp_display_information_end(outf); +} + +static void set_clos_assoc_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + int ret; + + ret = isst_clos_associate(cpu, current_clos); + if (ret) + perror("isst_clos_associate"); + else + isst_display_result(cpu, outf, "core-power", "assoc", ret); +} + +static void set_clos_assoc(void) +{ + if (cmd_help) { + fprintf(stderr, "Associate a clos id to a CPU\n"); + fprintf(stderr, + "\tSpecify targeted clos id with [--clos|-c]\n"); + exit(0); + } + + if (current_clos < 0 || current_clos > 3) { + fprintf(stderr, "Invalid clos id\n"); + exit(0); + } + if (max_target_cpus) + for_each_online_target_cpu_in_set(set_clos_assoc_for_cpu, NULL, + NULL, NULL, NULL); + else { + fprintf(stderr, + "Invalid target cpu. Specify with [-c|--cpu]\n"); + } +} + +static void get_clos_assoc_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + int clos, ret; + + ret = isst_clos_get_assoc_status(cpu, &clos); + if (ret) + perror("isst_clos_get_assoc_status"); + else + isst_display_result(cpu, outf, "core-power", "get-assoc", clos); +} + +static void get_clos_assoc(void) +{ + if (cmd_help) { + fprintf(stderr, "Get associate clos id to a CPU\n"); + fprintf(stderr, "\tSpecify targeted cpu id with [--cpu|-c]\n"); + exit(0); + } + if (max_target_cpus) + for_each_online_target_cpu_in_set(get_clos_assoc_for_cpu, NULL, + NULL, NULL, NULL); + else { + fprintf(stderr, + "Invalid target cpu. Specify with [-c|--cpu]\n"); + } +} + +static struct process_cmd_struct isst_cmds[] = { + { "perf-profile", "get-lock-status", get_tdp_locked }, + { "perf-profile", "get-config-levels", get_tdp_levels }, + { "perf-profile", "get-config-version", get_tdp_version }, + { "perf-profile", "get-config-enabled", get_tdp_enabled }, + { "perf-profile", "get-config-current-level", get_tdp_current_level }, + { "perf-profile", "set-config-level", set_tdp_level }, + { "perf-profile", "info", dump_isst_config }, + { "base-freq", "info", dump_pbf_config }, + { "base-freq", "enable", set_pbf_enable }, + { "base-freq", "disable", set_pbf_disable }, + { "turbo-freq", "info", dump_fact_config }, + { "turbo-freq", "enable", set_fact_enable }, + { "turbo-freq", "disable", set_fact_disable }, + { "core-power", "info", dump_clos_config }, + { "core-power", "enable", set_clos_enable }, + { "core-power", "disable", set_clos_disable }, + { "core-power", "config", set_clos_config }, + { "core-power", "assoc", set_clos_assoc }, + { "core-power", "get-assoc", get_clos_assoc }, + { NULL, NULL, NULL } +}; + +/* + * parse cpuset with following syntax + * 1,2,4..6,8-10 and set bits in cpu_subset + */ +void parse_cpu_command(char *optarg) +{ + unsigned int start, end; + char *next; + + next = optarg; + + while (next && *next) { + if (*next == '-') /* no negative cpu numbers */ + goto error; + + start = strtoul(next, &next, 10); + + if (max_target_cpus < MAX_CPUS_IN_ONE_REQ) + target_cpus[max_target_cpus++] = start; + + if (*next == '\0') + break; + + if (*next == ',') { + next += 1; + continue; + } + + if (*next == '-') { + next += 1; /* start range */ + } else if (*next == '.') { + next += 1; + if (*next == '.') + next += 1; /* start range */ + else + goto error; + } + + end = strtoul(next, &next, 10); + if (end <= start) + goto error; + + while (++start <= end) { + if (max_target_cpus < MAX_CPUS_IN_ONE_REQ) + target_cpus[max_target_cpus++] = start; + } + + if (*next == ',') + next += 1; + else if (*next != '\0') + goto error; + } + +#ifdef DEBUG + { + int i; + + for (i = 0; i < max_target_cpus; ++i) + printf("cpu [%d] in arg\n", target_cpus[i]); + } +#endif + return; + +error: + fprintf(stderr, ""--cpu %s" malformed\n", optarg); + exit(-1); +} + +static void parse_cmd_args(int argc, int start, char **argv) +{ + int opt; + int option_index; + + static struct option long_options[] = { + { "bucket", required_argument, 0, 'b' }, + { "level", required_argument, 0, 'l' }, + { "trl-type", required_argument, 0, 'r' }, + { "trl", required_argument, 0, 't' }, + { "help", no_argument, 0, 'h' }, + { "clos", required_argument, 0, 'c' }, + { "desired", required_argument, 0, 'd' }, + { "epp", required_argument, 0, 'e' }, + { "min", required_argument, 0, 'n' }, + { "max", required_argument, 0, 'm' }, + { "priority", required_argument, 0, 'p' }, + { "weight", required_argument, 0, 'w' }, + { 0, 0, 0, 0 } + }; + + option_index = start; + + optind = start + 1; + while ((opt = getopt_long(argc, argv, "b:l:t:c:d:e:n:m:p:w:h", + long_options, &option_index)) != -1) { + switch (opt) { + case 'b': + fact_bucket = atoi(optarg); + break; + case 'h': + cmd_help = 1; + break; + case 'l': + tdp_level = atoi(optarg); + break; + case 't': + sscanf(optarg, "0x%llx", &fact_trl); + break; + case 'r': + if (!strncmp(optarg, "sse", 3)) { + fact_avx = 0x01; + } else if (!strncmp(optarg, "avx2", 4)) { + fact_avx = 0x02; + } else if (!strncmp(optarg, "avx512", 4)) { + fact_avx = 0x04; + } else { + fprintf(outf, "Invalid sse,avx options\n"); + exit(1); + } + break; + /* CLOS related */ + case 'c': + current_clos = atoi(optarg); + printf("clos %d\n", current_clos); + break; + case 'd': + clos_desired = atoi(optarg); + break; + case 'e': + clos_epp = atoi(optarg); + break; + case 'n': + clos_min = atoi(optarg); + break; + case 'm': + clos_max = atoi(optarg); + break; + case 'p': + clos_priority_type = atoi(optarg); + break; + case 'w': + clos_prop_prio = atoi(optarg); + break; + default: + printf("no match\n"); + } + } +} + +static void isst_help(void) +{ + printf("perf-profile:\tAn architectural mechanism that allows multiple optimized \n\ + performance profiles per system via static and/or dynamic\n\ + adjustment of core count, workload, Tjmax, and\n\ + TDP, etc.\n"); + printf("\nCommands : For feature=perf-profile\n"); + printf("\tinfo\n"); + printf("\tget-lock-status\n"); + printf("\tget-config-levels\n"); + printf("\tget-config-version\n"); + printf("\tget-config-enabled\n"); + printf("\tget-config-current-level\n"); + printf("\tset-config-level\n"); +} + +static void pbf_help(void) +{ + printf("base-freq:\tEnables users to increase guaranteed base frequency\n\ + on certain cores (high priority cores) in exchange for lower\n\ + base frequency on remaining cores (low priority cores).\n"); + printf("\tcommand : info\n"); + printf("\tcommand : enable\n"); + printf("\tcommand : disable\n"); +} + +static void fact_help(void) +{ + printf("turbo-freq:\tEnables the ability to set different turbo ratio\n\ + limits to cores based on priority.\n"); + printf("\nCommand: For feature=turbo-freq\n"); + printf("\tcommand : info\n"); + printf("\tcommand : enable\n"); + printf("\tcommand : disable\n"); +} + +static void core_power_help(void) +{ + printf("core-power:\tInterface that allows user to define per core/tile\n\ + priority.\n"); + printf("\nCommands : For feature=core-power\n"); + printf("\tinfo\n"); + printf("\tenable\n"); + printf("\tdisable\n"); + printf("\tconfig\n"); + printf("\tassoc\n"); + printf("\tget-assoc\n"); +} + +struct process_cmd_help_struct { + char *feature; + void (*process_fn)(void); +}; + +static struct process_cmd_help_struct isst_help_cmds[] = { + { "perf-profile", isst_help }, + { "base-freq", pbf_help }, + { "turbo-freq", fact_help }, + { "core-power", core_power_help }, + { NULL, NULL } +}; + +void process_command(int argc, char **argv) +{ + int i = 0, matched = 0; + char *feature = argv[optind]; + char *cmd = argv[optind + 1]; + + if (!feature || !cmd) + return; + + debug_printf("feature name [%s] command [%s]\n", feature, cmd); + if (!strcmp(cmd, "-h") || !strcmp(cmd, "--help")) { + while (isst_help_cmds[i].feature) { + if (!strcmp(isst_help_cmds[i].feature, feature)) { + isst_help_cmds[i].process_fn(); + exit(0); + } + ++i; + } + } + + create_cpu_map(); + + i = 0; + while (isst_cmds[i].feature) { + if (!strcmp(isst_cmds[i].feature, feature) && + !strcmp(isst_cmds[i].command, cmd)) { + parse_cmd_args(argc, optind + 1, argv); + isst_cmds[i].process_fn(); + matched = 1; + break; + } + ++i; + } + + if (!matched) + fprintf(stderr, "Invalid command\n"); +} + +static void usage(void) +{ + printf("Intel(R) Speed Select Technology\n"); + printf("\nUsage:\n"); + printf("intel-speed-select [OPTIONS] FEATURE COMMAND COMMAND_ARGUMENTS\n"); + printf("\nUse this tool to enumerate and control the Intel Speed Select Technology features,\n"); + printf("\nFEATURE : [perf-profile|base-freq|turbo-freq|core-power]\n"); + printf("\nFor help on each feature, use --h|--help\n"); + printf("\tFor example: intel-speed-select perf-profile -h\n"); + + printf("\nFor additional help on each command for a feature, use --h|--help\n"); + printf("\tFor example: intel-speed-select perf-profile get-lock-status -h\n"); + printf("\t\t This will print help for the command "get-lock-status" for the feature "perf-profile"\n"); + + printf("\nOPTIONS\n"); + printf("\t[-c|--cpu] : logical cpu number\n"); + printf("\t\tDefault: Die scoped for all dies in the system with multiple dies/package\n"); + printf("\t\t\t Or Package scoped for all Packages when each package contains one die\n"); + printf("\t[-d|--debug] : Debug mode\n"); + printf("\t[-h|--help] : Print help\n"); + printf("\t[-i|--info] : Print platform information\n"); + printf("\t[-o|--out] : Output file\n"); + printf("\t\t\tDefault : stderr\n"); + printf("\t[-f|--format] : output format [json|text]. Default: text\n"); + printf("\t[-v|--version] : Print version\n"); + + printf("\nResult format\n"); + printf("\tResult display uses a common format for each command:\n"); + printf("\tResults are formatted in text/JSON with\n"); + printf("\t\tPackage, Die, CPU, and command specific results.\n"); + printf("\t\t\tFor Set commands, status is 0 for success and rest for failures\n"); + exit(1); +} + +static void print_version(void) +{ + fprintf(outf, "Version %s\n", version_str); + fprintf(outf, "Build date %s time %s\n", __DATE__, __TIME__); + exit(0); +} + +static void cmdline(int argc, char **argv) +{ + int opt; + int option_index = 0; + + static struct option long_options[] = { + { "cpu", required_argument, 0, 'c' }, + { "debug", no_argument, 0, 'd' }, + { "format", required_argument, 0, 'f' }, + { "help", no_argument, 0, 'h' }, + { "info", no_argument, 0, 'i' }, + { "out", required_argument, 0, 'o' }, + { "version", no_argument, 0, 'v' }, + { 0, 0, 0, 0 } + }; + + progname = argv[0]; + while ((opt = getopt_long_only(argc, argv, "+c:df:hio:v", long_options, + &option_index)) != -1) { + switch (opt) { + case 'c': + parse_cpu_command(optarg); + break; + case 'd': + debug_flag = 1; + printf("Debug Mode ON\n"); + break; + case 'f': + if (!strncmp(optarg, "json", 4)) + out_format_json = 1; + break; + case 'h': + usage(); + break; + case 'i': + isst_print_platform_information(); + break; + case 'o': + if (outf) + fclose(outf); + outf = fopen_or_exit(optarg, "w"); + break; + case 'v': + print_version(); + break; + default: + usage(); + } + } + + if (geteuid() != 0) { + fprintf(stderr, "Must run as root\n"); + exit(0); + } + + if (optind > (argc - 2)) { + fprintf(stderr, "Feature name and|or command not specified\n"); + exit(0); + } + update_cpu_model(); + printf("Intel(R) Speed Select Technology\n"); + printf("Executing on CPU model:%d[0x%x]\n", cpu_model, cpu_model); + set_max_cpu_num(); + set_cpu_present_cpu_mask(); + set_cpu_target_cpu_mask(); + isst_fill_platform_info(); + if (isst_platform_info.api_version > supported_api_ver) { + printf("Incompatible API versions; Upgrade of tool is required\n"); + exit(0); + } + + process_command(argc, argv); +} + +int main(int argc, char **argv) +{ + outf = stderr; + cmdline(argc, argv); + return 0; +} diff --git a/tools/power/x86/intel-speed-select/isst-core.c b/tools/power/x86/intel-speed-select/isst-core.c new file mode 100644 index 0000000000000..8de4ac39a0081 --- /dev/null +++ b/tools/power/x86/intel-speed-select/isst-core.c @@ -0,0 +1,721 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel Speed Select -- Enumerate and control features + * Copyright (c) 2019 Intel Corporation. + */ + +#include "isst.h" + +int isst_get_ctdp_levels(int cpu, struct isst_pkg_ctdp *pkg_dev) +{ + unsigned int resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_GET_LEVELS_INFO, 0, 0, &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CONFIG_TDP_GET_LEVELS_INFO resp:%x\n", cpu, resp); + + pkg_dev->version = resp & 0xff; + pkg_dev->levels = (resp >> 8) & 0xff; + pkg_dev->current_level = (resp >> 16) & 0xff; + pkg_dev->locked = !!(resp & BIT(24)); + pkg_dev->enabled = !!(resp & BIT(31)); + + return 0; +} + +int isst_get_ctdp_control(int cpu, int config_index, + struct isst_pkg_ctdp_level_info *ctdp_level) +{ + unsigned int resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_GET_TDP_CONTROL, 0, + config_index, &resp); + if (ret) + return ret; + + ctdp_level->fact_support = resp & BIT(0); + ctdp_level->pbf_support = !!(resp & BIT(1)); + ctdp_level->fact_enabled = !!(resp & BIT(16)); + ctdp_level->pbf_enabled = !!(resp & BIT(17)); + + debug_printf( + "cpu:%d CONFIG_TDP_GET_TDP_CONTROL resp:%x fact_support:%d pbf_support: %d fact_enabled:%d pbf_enabled:%d\n", + cpu, resp, ctdp_level->fact_support, ctdp_level->pbf_support, + ctdp_level->fact_enabled, ctdp_level->pbf_enabled); + + return 0; +} + +int isst_get_tdp_info(int cpu, int config_index, + struct isst_pkg_ctdp_level_info *ctdp_level) +{ + unsigned int resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_TDP_INFO, + 0, config_index, &resp); + if (ret) + return ret; + + ctdp_level->pkg_tdp = resp & GENMASK(14, 0); + ctdp_level->tdp_ratio = (resp & GENMASK(23, 16)) >> 16; + + debug_printf( + "cpu:%d ctdp:%d CONFIG_TDP_GET_TDP_INFO resp:%x tdp_ratio:%d pkg_tdp:%d\n", + cpu, config_index, resp, ctdp_level->tdp_ratio, + ctdp_level->pkg_tdp); + return 0; +} + +int isst_get_pwr_info(int cpu, int config_index, + struct isst_pkg_ctdp_level_info *ctdp_level) +{ + unsigned int resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_PWR_INFO, + 0, config_index, &resp); + if (ret) + return ret; + + ctdp_level->pkg_max_power = resp & GENMASK(14, 0); + ctdp_level->pkg_min_power = (resp & GENMASK(30, 16)) >> 16; + + debug_printf( + "cpu:%d ctdp:%d CONFIG_TDP_GET_PWR_INFO resp:%x pkg_max_power:%d pkg_min_power:%d\n", + cpu, config_index, resp, ctdp_level->pkg_max_power, + ctdp_level->pkg_min_power); + + return 0; +} + +int isst_get_tjmax_info(int cpu, int config_index, + struct isst_pkg_ctdp_level_info *ctdp_level) +{ + unsigned int resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_TJMAX_INFO, + 0, config_index, &resp); + if (ret) + return ret; + + ctdp_level->t_proc_hot = resp & GENMASK(7, 0); + + debug_printf( + "cpu:%d ctdp:%d CONFIG_TDP_GET_TJMAX_INFO resp:%x t_proc_hot:%d\n", + cpu, config_index, resp, ctdp_level->t_proc_hot); + + return 0; +} + +int isst_get_coremask_info(int cpu, int config_index, + struct isst_pkg_ctdp_level_info *ctdp_level) +{ + unsigned int resp; + int i, ret; + + ctdp_level->cpu_count = 0; + for (i = 0; i < 2; ++i) { + unsigned long long mask; + int cpu_count = 0; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_GET_CORE_MASK, 0, + (i << 8) | config_index, &resp); + if (ret) + return ret; + + debug_printf( + "cpu:%d ctdp:%d mask:%d CONFIG_TDP_GET_CORE_MASK resp:%x\n", + cpu, config_index, i, resp); + + mask = (unsigned long long)resp << (32 * i); + set_cpu_mask_from_punit_coremask(cpu, mask, + ctdp_level->core_cpumask_size, + ctdp_level->core_cpumask, + &cpu_count); + ctdp_level->cpu_count += cpu_count; + debug_printf("cpu:%d ctdp:%d mask:%d cpu count:%d\n", cpu, + config_index, i, ctdp_level->cpu_count); + } + + return 0; +} + +int isst_get_get_trl(int cpu, int level, int avx_level, int *trl) +{ + unsigned int req, resp; + int ret; + + req = level | (avx_level << 16); + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_GET_TURBO_LIMIT_RATIOS, 0, req, + &resp); + if (ret) + return ret; + + debug_printf( + "cpu:%d CONFIG_TDP_GET_TURBO_LIMIT_RATIOS req:%x resp:%x\n", + cpu, req, resp); + + trl[0] = resp & GENMASK(7, 0); + trl[1] = (resp & GENMASK(15, 8)) >> 8; + trl[2] = (resp & GENMASK(23, 16)) >> 16; + trl[3] = (resp & GENMASK(31, 24)) >> 24; + + req = level | BIT(8) | (avx_level << 16); + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_GET_TURBO_LIMIT_RATIOS, 0, req, + &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CONFIG_TDP_GET_TURBO_LIMIT req:%x resp:%x\n", cpu, + req, resp); + + trl[4] = resp & GENMASK(7, 0); + trl[5] = (resp & GENMASK(15, 8)) >> 8; + trl[6] = (resp & GENMASK(23, 16)) >> 16; + trl[7] = (resp & GENMASK(31, 24)) >> 24; + + return 0; +} + +int isst_set_tdp_level_msr(int cpu, int tdp_level) +{ + int ret; + + debug_printf("cpu: tdp_level via MSR %d\n", cpu, tdp_level); + + if (isst_get_config_tdp_lock_status(cpu)) { + debug_printf("cpu: tdp_locked %d\n", cpu); + return -1; + } + + if (tdp_level > 2) + return -1; /* invalid value */ + + ret = isst_send_msr_command(cpu, 0x64b, 1, + (unsigned long long *)&tdp_level); + if (ret) + return ret; + + debug_printf("cpu: tdp_level via MSR successful %d\n", cpu, tdp_level); + + return 0; +} + +int isst_set_tdp_level(int cpu, int tdp_level) +{ + unsigned int resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_SET_LEVEL, 0, + tdp_level, &resp); + if (ret) + return isst_set_tdp_level_msr(cpu, tdp_level); + + return 0; +} + +int isst_get_pbf_info(int cpu, int level, struct isst_pbf_info *pbf_info) +{ + unsigned int req, resp; + int i, ret; + + pbf_info->core_cpumask_size = alloc_cpu_set(&pbf_info->core_cpumask); + + for (i = 0; i < 2; ++i) { + unsigned long long mask; + int count; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_PBF_GET_CORE_MASK_INFO, + 0, (i << 8) | level, &resp); + if (ret) + return ret; + + debug_printf( + "cpu:%d CONFIG_TDP_PBF_GET_CORE_MASK_INFO resp:%x\n", + cpu, resp); + + mask = (unsigned long long)resp << (32 * i); + set_cpu_mask_from_punit_coremask(cpu, mask, + pbf_info->core_cpumask_size, + pbf_info->core_cpumask, + &count); + } + + req = level; + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_PBF_GET_P1HI_P1LO_INFO, 0, req, + &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CONFIG_TDP_PBF_GET_P1HI_P1LO_INFO resp:%x\n", cpu, + resp); + + pbf_info->p1_low = resp & 0xff; + pbf_info->p1_high = (resp & GENMASK(15, 8)) >> 8; + + req = level; + ret = isst_send_mbox_command( + cpu, CONFIG_TDP, CONFIG_TDP_PBF_GET_TDP_INFO, 0, req, &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CONFIG_TDP_PBF_GET_TDP_INFO resp:%x\n", cpu, resp); + + pbf_info->tdp = resp & 0xffff; + + req = level; + ret = isst_send_mbox_command( + cpu, CONFIG_TDP, CONFIG_TDP_PBF_GET_TJ_MAX_INFO, 0, req, &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CONFIG_TDP_PBF_GET_TJ_MAX_INFO resp:%x\n", cpu, + resp); + pbf_info->t_control = (resp >> 8) & 0xff; + pbf_info->t_prochot = resp & 0xff; + + return 0; +} + +void isst_get_pbf_info_complete(struct isst_pbf_info *pbf_info) +{ + free_cpu_set(pbf_info->core_cpumask); +} + +int isst_set_pbf_fact_status(int cpu, int pbf, int enable) +{ + struct isst_pkg_ctdp pkg_dev; + struct isst_pkg_ctdp_level_info ctdp_level; + int current_level; + unsigned int req = 0, resp; + int ret; + + ret = isst_get_ctdp_levels(cpu, &pkg_dev); + if (ret) + return ret; + + current_level = pkg_dev.current_level; + + ret = isst_get_ctdp_control(cpu, current_level, &ctdp_level); + if (ret) + return ret; + + if (pbf) { + if (ctdp_level.fact_enabled) + req = BIT(16); + + if (enable) + req |= BIT(17); + else + req &= ~BIT(17); + } else { + if (ctdp_level.pbf_enabled) + req = BIT(17); + + if (enable) + req |= BIT(16); + else + req &= ~BIT(16); + } + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_SET_TDP_CONTROL, 0, req, &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CONFIG_TDP_SET_TDP_CONTROL pbf/fact:%d req:%x\n", + cpu, pbf, req); + + return 0; +} + +int isst_get_fact_bucket_info(int cpu, int level, + struct isst_fact_bucket_info *bucket_info) +{ + unsigned int resp; + int i, k, ret; + + for (i = 0; i < 2; ++i) { + int j; + + ret = isst_send_mbox_command( + cpu, CONFIG_TDP, + CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_NUMCORES, 0, + (i << 8) | level, &resp); + if (ret) + return ret; + + debug_printf( + "cpu:%d CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_NUMCORES index:%d level:%d resp:%x\n", + cpu, i, level, resp); + + for (j = 0; j < 4; ++j) { + bucket_info[j + (i * 4)].high_priority_cores_count = + (resp >> (j * 8)) & 0xff; + } + } + + for (k = 0; k < 3; ++k) { + for (i = 0; i < 2; ++i) { + int j; + + ret = isst_send_mbox_command( + cpu, CONFIG_TDP, + CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_RATIOS, 0, + (k << 16) | (i << 8) | level, &resp); + if (ret) + return ret; + + debug_printf( + "cpu:%d CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_RATIOS index:%d level:%d avx:%d resp:%x\n", + cpu, i, level, k, resp); + + for (j = 0; j < 4; ++j) { + switch (k) { + case 0: + bucket_info[j + (i * 4)].sse_trl = + (resp >> (j * 8)) & 0xff; + break; + case 1: + bucket_info[j + (i * 4)].avx_trl = + (resp >> (j * 8)) & 0xff; + break; + case 2: + bucket_info[j + (i * 4)].avx512_trl = + (resp >> (j * 8)) & 0xff; + break; + default: + break; + } + } + } + } + + return 0; +} + +int isst_get_fact_info(int cpu, int level, struct isst_fact_info *fact_info) +{ + unsigned int resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_TDP, + CONFIG_TDP_GET_FACT_LP_CLIPPING_RATIO, 0, + level, &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CONFIG_TDP_GET_FACT_LP_CLIPPING_RATIO resp:%x\n", + cpu, resp); + + fact_info->lp_clipping_ratio_license_sse = resp & 0xff; + fact_info->lp_clipping_ratio_license_avx2 = (resp >> 8) & 0xff; + fact_info->lp_clipping_ratio_license_avx512 = (resp >> 16) & 0xff; + + ret = isst_get_fact_bucket_info(cpu, level, fact_info->bucket_info); + + return ret; +} + +int isst_set_trl(int cpu, unsigned long long trl) +{ + int ret; + + if (!trl) + trl = 0xFFFFFFFFFFFFFFFFULL; + + ret = isst_send_msr_command(cpu, 0x1AD, 1, &trl); + if (ret) + return ret; + + return 0; +} + +int isst_set_trl_from_current_tdp(int cpu, unsigned long long trl) +{ + unsigned long long msr_trl; + int ret; + + if (trl) { + msr_trl = trl; + } else { + struct isst_pkg_ctdp pkg_dev; + int trl[8]; + int i; + + ret = isst_get_ctdp_levels(cpu, &pkg_dev); + if (ret) + return ret; + + ret = isst_get_get_trl(cpu, pkg_dev.current_level, 0, trl); + if (ret) + return ret; + + msr_trl = 0; + for (i = 0; i < 8; ++i) { + unsigned long long _trl = trl[i]; + + msr_trl |= (_trl << (i * 8)); + } + } + ret = isst_send_msr_command(cpu, 0x1AD, 1, &msr_trl); + if (ret) + return ret; + + return 0; +} + +/* Return 1 if locked */ +int isst_get_config_tdp_lock_status(int cpu) +{ + unsigned long long tdp_control = 0; + int ret; + + ret = isst_send_msr_command(cpu, 0x64b, 0, &tdp_control); + if (ret) + return ret; + + ret = !!(tdp_control & BIT(31)); + + return ret; +} + +void isst_get_process_ctdp_complete(int cpu, struct isst_pkg_ctdp *pkg_dev) +{ + int i; + + if (!pkg_dev->processed) + return; + + for (i = 0; i < pkg_dev->levels; ++i) { + struct isst_pkg_ctdp_level_info *ctdp_level; + + ctdp_level = &pkg_dev->ctdp_level[i]; + if (ctdp_level->pbf_support) + free_cpu_set(ctdp_level->pbf_info.core_cpumask); + free_cpu_set(ctdp_level->core_cpumask); + } +} + +int isst_get_process_ctdp(int cpu, int tdp_level, struct isst_pkg_ctdp *pkg_dev) +{ + int i, ret; + + if (pkg_dev->processed) + return 0; + + ret = isst_get_ctdp_levels(cpu, pkg_dev); + if (ret) + return ret; + + debug_printf("cpu: %d ctdp enable:%d current level: %d levels:%d\n", + cpu, pkg_dev->enabled, pkg_dev->current_level, + pkg_dev->levels); + + for (i = 0; i <= pkg_dev->levels; ++i) { + struct isst_pkg_ctdp_level_info *ctdp_level; + + if (tdp_level != 0xff && i != tdp_level) + continue; + + debug_printf("cpu:%d Get Information for TDP level:%d\n", cpu, + i); + ctdp_level = &pkg_dev->ctdp_level[i]; + + ctdp_level->processed = 1; + ctdp_level->level = i; + ctdp_level->control_cpu = cpu; + ctdp_level->pkg_id = get_physical_package_id(cpu); + ctdp_level->die_id = get_physical_die_id(cpu); + + ret = isst_get_ctdp_control(cpu, i, ctdp_level); + if (ret) + return ret; + + ret = isst_get_tdp_info(cpu, i, ctdp_level); + if (ret) + return ret; + + ret = isst_get_pwr_info(cpu, i, ctdp_level); + if (ret) + return ret; + + ret = isst_get_tjmax_info(cpu, i, ctdp_level); + if (ret) + return ret; + + ctdp_level->core_cpumask_size = + alloc_cpu_set(&ctdp_level->core_cpumask); + ret = isst_get_coremask_info(cpu, i, ctdp_level); + if (ret) + return ret; + + ret = isst_get_get_trl(cpu, i, 0, + ctdp_level->trl_sse_active_cores); + if (ret) + return ret; + + ret = isst_get_get_trl(cpu, i, 1, + ctdp_level->trl_avx_active_cores); + if (ret) + return ret; + + ret = isst_get_get_trl(cpu, i, 2, + ctdp_level->trl_avx_512_active_cores); + if (ret) + return ret; + + if (ctdp_level->pbf_support) { + ret = isst_get_pbf_info(cpu, i, &ctdp_level->pbf_info); + if (!ret) + ctdp_level->pbf_found = 1; + } + + if (ctdp_level->fact_support) { + ret = isst_get_fact_info(cpu, i, + &ctdp_level->fact_info); + if (ret) + return ret; + } + } + + pkg_dev->processed = 1; + + return 0; +} + +int isst_pm_qos_config(int cpu, int enable_clos, int priority_type) +{ + unsigned int req, resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_QOS_CONFIG, 0, 0, + &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CLOS_PM_QOS_CONFIG resp:%x\n", cpu, resp); + + req = resp; + + if (enable_clos) + req = req | BIT(1); + else + req = req & ~BIT(1); + + if (priority_type) + req = req | BIT(2); + else + req = req & ~BIT(2); + + ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_QOS_CONFIG, + BIT(MBOX_CMD_WRITE_BIT), req, &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CLOS_PM_QOS_CONFIG priority type:%d req:%x\n", cpu, + priority_type, req); + + return 0; +} + +int isst_pm_get_clos(int cpu, int clos, struct isst_clos_config *clos_config) +{ + unsigned int resp; + int ret; + + ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_CLOS, clos, 0, + &resp); + if (ret) + return ret; + + clos_config->pkg_id = get_physical_package_id(cpu); + clos_config->die_id = get_physical_die_id(cpu); + + clos_config->epp = resp & 0x0f; + clos_config->clos_prop_prio = (resp >> 4) & 0x0f; + clos_config->clos_min = (resp >> 8) & 0xff; + clos_config->clos_max = (resp >> 16) & 0xff; + clos_config->clos_desired = (resp >> 24) & 0xff; + + return 0; +} + +int isst_set_clos(int cpu, int clos, struct isst_clos_config *clos_config) +{ + unsigned int req, resp; + unsigned int param; + int ret; + + req = clos_config->epp & 0x0f; + req |= (clos_config->clos_prop_prio & 0x0f) << 4; + req |= (clos_config->clos_min & 0xff) << 8; + req |= (clos_config->clos_max & 0xff) << 16; + req |= (clos_config->clos_desired & 0xff) << 24; + + param = BIT(MBOX_CMD_WRITE_BIT) | clos; + + ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_CLOS, param, req, + &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CLOS_PM_CLOS param:%x req:%x\n", cpu, param, req); + + return 0; +} + +int isst_clos_get_assoc_status(int cpu, int *clos_id) +{ + unsigned int resp; + unsigned int param; + int core_id, ret; + + core_id = find_phy_core_num(cpu); + param = core_id; + + ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PQR_ASSOC, param, 0, + &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CLOS_PQR_ASSOC param:%x resp:%x\n", cpu, param, + resp); + *clos_id = (resp >> 16) & 0x03; + + return 0; +} + +int isst_clos_associate(int cpu, int clos_id) +{ + unsigned int req, resp; + unsigned int param; + int core_id, ret; + + req = (clos_id & 0x03) << 16; + core_id = find_phy_core_num(cpu); + param = BIT(MBOX_CMD_WRITE_BIT) | core_id; + + ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PQR_ASSOC, param, + req, &resp); + if (ret) + return ret; + + debug_printf("cpu:%d CLOS_PQR_ASSOC param:%x req:%x\n", cpu, param, + req); + + return 0; +} diff --git a/tools/power/x86/intel-speed-select/isst-display.c b/tools/power/x86/intel-speed-select/isst-display.c new file mode 100644 index 0000000000000..f368b83237427 --- /dev/null +++ b/tools/power/x86/intel-speed-select/isst-display.c @@ -0,0 +1,479 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel dynamic_speed_select -- Enumerate and control features + * Copyright (c) 2019 Intel Corporation. + */ + +#include "isst.h" + +#define DISP_FREQ_MULTIPLIER 100000 + +static void printcpumask(int str_len, char *str, int mask_size, + cpu_set_t *cpu_mask) +{ + int i, max_cpus = get_topo_max_cpus(); + unsigned int *mask; + int size, index, curr_index; + + size = max_cpus / (sizeof(unsigned int) * 8); + if (max_cpus % (sizeof(unsigned int) * 8)) + size++; + + mask = calloc(size, sizeof(unsigned int)); + if (!mask) + return; + + for (i = 0; i < max_cpus; ++i) { + int mask_index, bit_index; + + if (!CPU_ISSET_S(i, mask_size, cpu_mask)) + continue; + + mask_index = i / (sizeof(unsigned int) * 8); + bit_index = i % (sizeof(unsigned int) * 8); + mask[mask_index] |= BIT(bit_index); + } + + curr_index = 0; + for (i = size - 1; i >= 0; --i) { + index = snprintf(&str[curr_index], str_len - curr_index, "%08x", + mask[i]); + curr_index += index; + if (i) { + strncat(&str[curr_index], ",", str_len - curr_index); + curr_index++; + } + } + + free(mask); +} + +static void format_and_print_txt(FILE *outf, int level, char *header, + char *value) +{ + char *spaces = " "; + static char delimiters[256]; + int i, j = 0; + + if (!level) + return; + + if (level == 1) { + strcpy(delimiters, " "); + } else { + for (i = 0; i < level - 1; ++i) + j += snprintf(&delimiters[j], sizeof(delimiters) - j, + "%s", spaces); + } + + if (header && value) { + fprintf(outf, "%s", delimiters); + fprintf(outf, "%s:%s\n", header, value); + } else if (header) { + fprintf(outf, "%s", delimiters); + fprintf(outf, "%s\n", header); + } +} + +static int last_level; +static void format_and_print(FILE *outf, int level, char *header, char *value) +{ + char *spaces = " "; + static char delimiters[256]; + int i; + + if (!out_format_is_json()) { + format_and_print_txt(outf, level, header, value); + return; + } + + if (level == 0) { + if (header) + fprintf(outf, "{"); + else + fprintf(outf, "\n}\n"); + + } else { + int j = 0; + + for (i = 0; i < level; ++i) + j += snprintf(&delimiters[j], sizeof(delimiters) - j, + "%s", spaces); + + if (last_level == level) + fprintf(outf, ",\n"); + + if (value) { + if (last_level != level) + fprintf(outf, "\n"); + + fprintf(outf, "%s"%s": ", delimiters, header); + fprintf(outf, ""%s"", value); + } else { + for (i = last_level - 1; i >= level; --i) { + int k = 0; + + for (j = i; j > 0; --j) + k += snprintf(&delimiters[k], + sizeof(delimiters) - k, + "%s", spaces); + if (i == level && header) + fprintf(outf, "\n%s},", delimiters); + else + fprintf(outf, "\n%s}", delimiters); + } + if (abs(last_level - level) < 3) + fprintf(outf, "\n"); + if (header) + fprintf(outf, "%s"%s": {", delimiters, + header); + } + } + + last_level = level; +} + +static void print_packag_info(int cpu, FILE *outf) +{ + char header[256]; + + snprintf(header, sizeof(header), "package-%d", + get_physical_package_id(cpu)); + format_and_print(outf, 1, header, NULL); + snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); + format_and_print(outf, 2, header, NULL); + snprintf(header, sizeof(header), "cpu-%d", cpu); + format_and_print(outf, 3, header, NULL); +} + +static void _isst_pbf_display_information(int cpu, FILE *outf, int level, + struct isst_pbf_info *pbf_info, + int disp_level) +{ + char header[256]; + char value[256]; + + snprintf(header, sizeof(header), "speed-select-base-freq"); + format_and_print(outf, disp_level, header, NULL); + + snprintf(header, sizeof(header), "high-priority-base-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + pbf_info->p1_high * DISP_FREQ_MULTIPLIER); + format_and_print(outf, disp_level + 1, header, value); + + snprintf(header, sizeof(header), "high-priority-cpu-mask"); + printcpumask(sizeof(value), value, pbf_info->core_cpumask_size, + pbf_info->core_cpumask); + format_and_print(outf, disp_level + 1, header, value); + + snprintf(header, sizeof(header), "low-priority-base-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + pbf_info->p1_low * DISP_FREQ_MULTIPLIER); + format_and_print(outf, disp_level + 1, header, value); + + snprintf(header, sizeof(header), "tjunction-temperature(C)"); + snprintf(value, sizeof(value), "%d", pbf_info->t_prochot); + format_and_print(outf, disp_level + 1, header, value); + + snprintf(header, sizeof(header), "thermal-design-power(W)"); + snprintf(value, sizeof(value), "%d", pbf_info->tdp); + format_and_print(outf, disp_level + 1, header, value); +} + +static void _isst_fact_display_information(int cpu, FILE *outf, int level, + int fact_bucket, int fact_avx, + struct isst_fact_info *fact_info, + int base_level) +{ + struct isst_fact_bucket_info *bucket_info = fact_info->bucket_info; + char header[256]; + char value[256]; + int j; + + snprintf(header, sizeof(header), "speed-select-turbo-freq"); + format_and_print(outf, base_level, header, NULL); + for (j = 0; j < ISST_FACT_MAX_BUCKETS; ++j) { + if (fact_bucket != 0xff && fact_bucket != j) + continue; + + if (!bucket_info[j].high_priority_cores_count) + break; + + snprintf(header, sizeof(header), "bucket-%d", j); + format_and_print(outf, base_level + 1, header, NULL); + + snprintf(header, sizeof(header), "high-priority-cores-count"); + snprintf(value, sizeof(value), "%d", + bucket_info[j].high_priority_cores_count); + format_and_print(outf, base_level + 2, header, value); + + if (fact_avx & 0x01) { + snprintf(header, sizeof(header), + "high-priority-max-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + bucket_info[j].sse_trl * DISP_FREQ_MULTIPLIER); + format_and_print(outf, base_level + 2, header, value); + } + + if (fact_avx & 0x02) { + snprintf(header, sizeof(header), + "high-priority-max-avx2-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + bucket_info[j].avx_trl * DISP_FREQ_MULTIPLIER); + format_and_print(outf, base_level + 2, header, value); + } + + if (fact_avx & 0x04) { + snprintf(header, sizeof(header), + "high-priority-max-avx512-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + bucket_info[j].avx512_trl * + DISP_FREQ_MULTIPLIER); + format_and_print(outf, base_level + 2, header, value); + } + } + snprintf(header, sizeof(header), + "speed-select-turbo-freq-clip-frequencies"); + format_and_print(outf, base_level + 1, header, NULL); + snprintf(header, sizeof(header), "low-priority-max-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + fact_info->lp_clipping_ratio_license_sse * + DISP_FREQ_MULTIPLIER); + format_and_print(outf, base_level + 2, header, value); + snprintf(header, sizeof(header), + "low-priority-max-avx2-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + fact_info->lp_clipping_ratio_license_avx2 * + DISP_FREQ_MULTIPLIER); + format_and_print(outf, base_level + 2, header, value); + snprintf(header, sizeof(header), + "low-priority-max-avx512-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + fact_info->lp_clipping_ratio_license_avx512 * + DISP_FREQ_MULTIPLIER); + format_and_print(outf, base_level + 2, header, value); +} + +void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level, + struct isst_pkg_ctdp *pkg_dev) +{ + char header[256]; + char value[256]; + int i, base_level = 1; + + print_packag_info(cpu, outf); + + for (i = 0; i <= pkg_dev->levels; ++i) { + struct isst_pkg_ctdp_level_info *ctdp_level; + int j; + + ctdp_level = &pkg_dev->ctdp_level[i]; + if (!ctdp_level->processed) + continue; + + snprintf(header, sizeof(header), "perf-profile-level-%d", + ctdp_level->level); + format_and_print(outf, base_level + 3, header, NULL); + + snprintf(header, sizeof(header), "cpu-count"); + j = get_cpu_count(get_physical_die_id(cpu), + get_physical_die_id(cpu)); + snprintf(value, sizeof(value), "%d", j); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), "enable-cpu-mask"); + printcpumask(sizeof(value), value, + ctdp_level->core_cpumask_size, + ctdp_level->core_cpumask); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), "thermal-design-power-ratio"); + snprintf(value, sizeof(value), "%d", ctdp_level->tdp_ratio); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), "base-frequency(KHz)"); + snprintf(value, sizeof(value), "%d", + ctdp_level->tdp_ratio * DISP_FREQ_MULTIPLIER); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), + "speed-select-turbo-freq-support"); + snprintf(value, sizeof(value), "%d", ctdp_level->fact_support); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), + "speed-select-base-freq-support"); + snprintf(value, sizeof(value), "%d", ctdp_level->pbf_support); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), + "speed-select-base-freq-enabled"); + snprintf(value, sizeof(value), "%d", ctdp_level->pbf_enabled); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), + "speed-select-turbo-freq-enabled"); + snprintf(value, sizeof(value), "%d", ctdp_level->fact_enabled); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), "thermal-design-power(W)"); + snprintf(value, sizeof(value), "%d", ctdp_level->pkg_tdp); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), "tjunction-max(C)"); + snprintf(value, sizeof(value), "%d", ctdp_level->t_proc_hot); + format_and_print(outf, base_level + 4, header, value); + + snprintf(header, sizeof(header), "turbo-ratio-limits-sse"); + format_and_print(outf, base_level + 4, header, NULL); + for (j = 0; j < 8; ++j) { + snprintf(header, sizeof(header), "bucket-%d", j); + format_and_print(outf, base_level + 5, header, NULL); + + snprintf(header, sizeof(header), "core-count"); + snprintf(value, sizeof(value), "%d", j); + format_and_print(outf, base_level + 6, header, value); + + snprintf(header, sizeof(header), "turbo-ratio"); + snprintf(value, sizeof(value), "%d", + ctdp_level->trl_sse_active_cores[j]); + format_and_print(outf, base_level + 6, header, value); + } + snprintf(header, sizeof(header), "turbo-ratio-limits-avx"); + format_and_print(outf, base_level + 4, header, NULL); + for (j = 0; j < 8; ++j) { + snprintf(header, sizeof(header), "bucket-%d", j); + format_and_print(outf, base_level + 5, header, NULL); + + snprintf(header, sizeof(header), "core-count"); + snprintf(value, sizeof(value), "%d", j); + format_and_print(outf, base_level + 6, header, value); + + snprintf(header, sizeof(header), "turbo-ratio"); + snprintf(value, sizeof(value), "%d", + ctdp_level->trl_avx_active_cores[j]); + format_and_print(outf, base_level + 6, header, value); + } + + snprintf(header, sizeof(header), "turbo-ratio-limits-avx512"); + format_and_print(outf, base_level + 4, header, NULL); + for (j = 0; j < 8; ++j) { + snprintf(header, sizeof(header), "bucket-%d", j); + format_and_print(outf, base_level + 5, header, NULL); + + snprintf(header, sizeof(header), "core-count"); + snprintf(value, sizeof(value), "%d", j); + format_and_print(outf, base_level + 6, header, value); + + snprintf(header, sizeof(header), "turbo-ratio"); + snprintf(value, sizeof(value), "%d", + ctdp_level->trl_avx_512_active_cores[j]); + format_and_print(outf, base_level + 6, header, value); + } + if (ctdp_level->pbf_support) + _isst_pbf_display_information(cpu, outf, i, + &ctdp_level->pbf_info, + base_level + 4); + if (ctdp_level->fact_support) + _isst_fact_display_information(cpu, outf, i, 0xff, 0xff, + &ctdp_level->fact_info, + base_level + 4); + } + + format_and_print(outf, 1, NULL, NULL); +} + +void isst_ctdp_display_information_start(FILE *outf) +{ + last_level = 0; + format_and_print(outf, 0, "start", NULL); +} + +void isst_ctdp_display_information_end(FILE *outf) +{ + format_and_print(outf, 0, NULL, NULL); +} + +void isst_pbf_display_information(int cpu, FILE *outf, int level, + struct isst_pbf_info *pbf_info) +{ + print_packag_info(cpu, outf); + _isst_pbf_display_information(cpu, outf, level, pbf_info, 4); + format_and_print(outf, 1, NULL, NULL); +} + +void isst_fact_display_information(int cpu, FILE *outf, int level, + int fact_bucket, int fact_avx, + struct isst_fact_info *fact_info) +{ + print_packag_info(cpu, outf); + _isst_fact_display_information(cpu, outf, level, fact_bucket, fact_avx, + fact_info, 4); + format_and_print(outf, 1, NULL, NULL); +} + +void isst_clos_display_information(int cpu, FILE *outf, int clos, + struct isst_clos_config *clos_config) +{ + char header[256]; + char value[256]; + + snprintf(header, sizeof(header), "package-%d", + get_physical_package_id(cpu)); + format_and_print(outf, 1, header, NULL); + snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); + format_and_print(outf, 2, header, NULL); + snprintf(header, sizeof(header), "cpu-%d", cpu); + format_and_print(outf, 3, header, NULL); + + snprintf(header, sizeof(header), "core-power"); + format_and_print(outf, 4, header, NULL); + + snprintf(header, sizeof(header), "clos"); + snprintf(value, sizeof(value), "%d", clos); + format_and_print(outf, 5, header, value); + + snprintf(header, sizeof(header), "epp"); + snprintf(value, sizeof(value), "%d", clos_config->epp); + format_and_print(outf, 5, header, value); + + snprintf(header, sizeof(header), "clos-proportional-priority"); + snprintf(value, sizeof(value), "%d", clos_config->clos_prop_prio); + format_and_print(outf, 5, header, value); + + snprintf(header, sizeof(header), "clos-min"); + snprintf(value, sizeof(value), "%d", clos_config->clos_min); + format_and_print(outf, 5, header, value); + + snprintf(header, sizeof(header), "clos-max"); + snprintf(value, sizeof(value), "%d", clos_config->clos_max); + format_and_print(outf, 5, header, value); + + snprintf(header, sizeof(header), "clos-desired"); + snprintf(value, sizeof(value), "%d", clos_config->clos_desired); + format_and_print(outf, 5, header, value); + + format_and_print(outf, 1, NULL, NULL); +} + +void isst_display_result(int cpu, FILE *outf, char *feature, char *cmd, + int result) +{ + char header[256]; + char value[256]; + + snprintf(header, sizeof(header), "package-%d", + get_physical_package_id(cpu)); + format_and_print(outf, 1, header, NULL); + snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); + format_and_print(outf, 2, header, NULL); + snprintf(header, sizeof(header), "cpu-%d", cpu); + format_and_print(outf, 3, header, NULL); + snprintf(header, sizeof(header), "%s", feature); + format_and_print(outf, 4, header, NULL); + snprintf(header, sizeof(header), "%s", cmd); + snprintf(value, sizeof(value), "%d", result); + format_and_print(outf, 5, header, value); + + format_and_print(outf, 1, NULL, NULL); +} diff --git a/tools/power/x86/intel-speed-select/isst.h b/tools/power/x86/intel-speed-select/isst.h new file mode 100644 index 0000000000000..2218817616093 --- /dev/null +++ b/tools/power/x86/intel-speed-select/isst.h @@ -0,0 +1,231 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Intel Speed Select -- Enumerate and control features + * Copyright (c) 2019 Intel Corporation. + */ + +#ifndef _ISST_H_ +#define _ISST_H_ + +#include <stdio.h> +#include <unistd.h> +#include <sys/types.h> +#include <sched.h> +#include <sys/stat.h> +#include <sys/resource.h> +#include <getopt.h> +#include <err.h> +#include <fcntl.h> +#include <signal.h> +#include <sys/time.h> +#include <limits.h> +#include <stdlib.h> +#include <string.h> +#include <cpuid.h> +#include <dirent.h> +#include <errno.h> + +#include <stdarg.h> +#include <sys/ioctl.h> + +#define BIT(x) (1 << (x)) +#define GENMASK(h, l) (((~0UL) << (l)) & (~0UL >> (sizeof(long) * 8 - 1 - (h)))) +#define GENMASK_ULL(h, l) \ + (((~0ULL) << (l)) & (~0ULL >> (sizeof(long long) * 8 - 1 - (h)))) + +#define CONFIG_TDP 0x7f +#define CONFIG_TDP_GET_LEVELS_INFO 0x00 +#define CONFIG_TDP_GET_TDP_CONTROL 0x01 +#define CONFIG_TDP_SET_TDP_CONTROL 0x02 +#define CONFIG_TDP_GET_TDP_INFO 0x03 +#define CONFIG_TDP_GET_PWR_INFO 0x04 +#define CONFIG_TDP_GET_TJMAX_INFO 0x05 +#define CONFIG_TDP_GET_CORE_MASK 0x06 +#define CONFIG_TDP_GET_TURBO_LIMIT_RATIOS 0x07 +#define CONFIG_TDP_SET_LEVEL 0x08 +#define CONFIG_TDP_GET_UNCORE_P0_P1_INFO 0X09 +#define CONFIG_TDP_GET_P1_INFO 0x0a +#define CONFIG_TDP_GET_MEM_FREQ 0x0b + +#define CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_NUMCORES 0x10 +#define CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_RATIOS 0x11 +#define CONFIG_TDP_GET_FACT_LP_CLIPPING_RATIO 0x12 + +#define CONFIG_TDP_PBF_GET_CORE_MASK_INFO 0x20 +#define CONFIG_TDP_PBF_GET_P1HI_P1LO_INFO 0x21 +#define CONFIG_TDP_PBF_GET_TJ_MAX_INFO 0x22 +#define CONFIG_TDP_PBF_GET_TDP_INFO 0X23 + +#define CONFIG_CLOS 0xd0 +#define CLOS_PQR_ASSOC 0x00 +#define CLOS_PM_CLOS 0x01 +#define CLOS_PM_QOS_CONFIG 0x02 +#define CLOS_STATUS 0x03 + +#define MBOX_CMD_WRITE_BIT 0x08 + +#define PM_QOS_INFO_OFFSET 0x00 +#define PM_QOS_CONFIG_OFFSET 0x04 +#define PM_CLOS_OFFSET 0x08 +#define PQR_ASSOC_OFFSET 0x20 + +struct isst_clos_config { + int pkg_id; + int die_id; + unsigned char epp; + unsigned char clos_prop_prio; + unsigned char clos_min; + unsigned char clos_max; + unsigned char clos_desired; +}; + +struct isst_fact_bucket_info { + int high_priority_cores_count; + int sse_trl; + int avx_trl; + int avx512_trl; +}; + +struct isst_pbf_info { + int pbf_acticated; + int pbf_available; + size_t core_cpumask_size; + cpu_set_t *core_cpumask; + int p1_high; + int p1_low; + int t_control; + int t_prochot; + int tdp; +}; + +#define ISST_TRL_MAX_ACTIVE_CORES 8 +#define ISST_FACT_MAX_BUCKETS 8 +struct isst_fact_info { + int lp_clipping_ratio_license_sse; + int lp_clipping_ratio_license_avx2; + int lp_clipping_ratio_license_avx512; + struct isst_fact_bucket_info bucket_info[ISST_FACT_MAX_BUCKETS]; +}; + +struct isst_pkg_ctdp_level_info { + int processed; + int control_cpu; + int pkg_id; + int die_id; + int level; + int fact_support; + int pbf_support; + int fact_enabled; + int pbf_enabled; + int tdp_ratio; + int active; + int tdp_control; + int pkg_tdp; + int pkg_min_power; + int pkg_max_power; + int fact; + int t_proc_hot; + int uncore_p0; + int uncore_p1; + int sse_p1; + int avx2_p1; + int avx512_p1; + int mem_freq; + size_t core_cpumask_size; + cpu_set_t *core_cpumask; + int cpu_count; + int trl_sse_active_cores[ISST_TRL_MAX_ACTIVE_CORES]; + int trl_avx_active_cores[ISST_TRL_MAX_ACTIVE_CORES]; + int trl_avx_512_active_cores[ISST_TRL_MAX_ACTIVE_CORES]; + int kobj_bucket_index; + int active_bucket; + int fact_max_index; + int fact_max_config; + int pbf_found; + int pbf_active; + struct isst_pbf_info pbf_info; + struct isst_fact_info fact_info; +}; + +#define ISST_MAX_TDP_LEVELS (4 + 1) /* +1 for base config */ +struct isst_pkg_ctdp { + int locked; + int version; + int processed; + int levels; + int current_level; + int enabled; + struct isst_pkg_ctdp_level_info ctdp_level[ISST_MAX_TDP_LEVELS]; +}; + +extern int get_topo_max_cpus(void); +extern int get_cpu_count(int pkg_id, int die_id); + +/* Common interfaces */ +extern void debug_printf(const char *format, ...); +extern int out_format_is_json(void); +extern int get_physical_package_id(int cpu); +extern int get_physical_die_id(int cpu); +extern size_t alloc_cpu_set(cpu_set_t **cpu_set); +extern void free_cpu_set(cpu_set_t *cpu_set); +extern int find_logical_cpu(int pkg_id, int die_id, int phy_cpu); +extern int find_phy_cpu_num(int logical_cpu); +extern int find_phy_core_num(int logical_cpu); +extern void set_cpu_mask_from_punit_coremask(int cpu, + unsigned long long core_mask, + size_t core_cpumask_size, + cpu_set_t *core_cpumask, + int *cpu_cnt); + +extern int isst_send_mbox_command(unsigned int cpu, unsigned char command, + unsigned char sub_command, + unsigned int write, + unsigned int req_data, unsigned int *resp); + +extern int isst_send_msr_command(unsigned int cpu, unsigned int command, + int write, unsigned long long *req_resp); + +extern int isst_get_ctdp_levels(int cpu, struct isst_pkg_ctdp *pkg_dev); +extern int isst_get_process_ctdp(int cpu, int tdp_level, + struct isst_pkg_ctdp *pkg_dev); +extern void isst_get_process_ctdp_complete(int cpu, + struct isst_pkg_ctdp *pkg_dev); +extern void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level, + struct isst_pkg_ctdp *pkg_dev); +extern void isst_ctdp_display_information_start(FILE *outf); +extern void isst_ctdp_display_information_end(FILE *outf); +extern void isst_pbf_display_information(int cpu, FILE *outf, int level, + struct isst_pbf_info *info); +extern int isst_set_tdp_level(int cpu, int tdp_level); +extern int isst_set_tdp_level_msr(int cpu, int tdp_level); +extern int isst_set_pbf_fact_status(int cpu, int pbf, int enable); +extern int isst_get_pbf_info(int cpu, int level, + struct isst_pbf_info *pbf_info); +extern void isst_get_pbf_info_complete(struct isst_pbf_info *pbf_info); +extern int isst_get_fact_info(int cpu, int level, + struct isst_fact_info *fact_info); +extern int isst_get_fact_bucket_info(int cpu, int level, + struct isst_fact_bucket_info *bucket_info); +extern void isst_fact_display_information(int cpu, FILE *outf, int level, + int fact_bucket, int fact_avx, + struct isst_fact_info *fact_info); +extern int isst_set_trl(int cpu, unsigned long long trl); +extern int isst_set_trl_from_current_tdp(int cpu, unsigned long long trl); +extern int isst_get_config_tdp_lock_status(int cpu); + +extern int isst_pm_qos_config(int cpu, int enable_clos, int priority_type); +extern int isst_pm_get_clos(int cpu, int clos, + struct isst_clos_config *clos_config); +extern int isst_set_clos(int cpu, int clos, + struct isst_clos_config *clos_config); +extern int isst_clos_associate(int cpu, int clos); +extern int isst_clos_get_assoc_status(int cpu, int *clos_id); +extern void isst_clos_display_information(int cpu, FILE *outf, int clos, + struct isst_clos_config *clos_config); + +extern int isst_read_reg(unsigned short reg, unsigned int *val); +extern int isst_write_reg(int reg, unsigned int val); + +extern void isst_display_result(int cpu, FILE *outf, char *feature, char *cmd, + int result); +#endif

From: "Rafael J. Wysocki" rafael.j.wysocki@intel.com

mainline inclusion from mainline-v5.6-rc1 commit 987c785319b99e32602f7f86cfae3cf9b81e402b category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 987c785319b99e32602f7f86cfae3cf9b81e402b upstream

In order to separate the ACPI _CST evaluation from checks specific to the ACPI processor driver, move the majority of the acpi_processor_get_power_info_cst() function body to a new function, acpi_processor_evaluate_cst(), that will extract the C-states information from _CST output, and redefine acpi_processor_get_power_info_cst() as a wrapper around it.

No intentional functional impact.

Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: yjia yingbao.jia@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/acpi/processor_idle.c | 52 +++++++++++++++++++++-------------- 1 file changed, 32 insertions(+), 20 deletions(-)

diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 3e398ada86412..2bcc532811643 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -304,21 +304,17 @@ static int acpi_processor_get_power_info_default(struct acpi_processor *pr) return 0; }

-static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) +static int acpi_processor_evaluate_cst(acpi_handle handle, u32 cpu, + struct acpi_processor_power *info) { + struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; + union acpi_object *cst; acpi_status status; u64 count; - int current_count; + int current_count = 0; int i, ret = 0; - struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; - union acpi_object *cst; - - if (nocst) - return -ENODEV;

- current_count = 0; - - status = acpi_evaluate_object(pr->handle, "_CST", NULL, &buffer); + status = acpi_evaluate_object(handle, "_CST", NULL, &buffer); if (ACPI_FAILURE(status)) { ACPI_DEBUG_PRINT((ACPI_DB_INFO, "No _CST, giving up\n")); return -ENODEV; @@ -342,9 +338,6 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) goto end; }

- /* Tell driver that at least _CST is supported. */ - pr->flags.has_cst = 1; - for (i = 1; i <= count; i++) { union acpi_object *element; union acpi_object *obj; @@ -390,7 +383,7 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) cx.entry_method = ACPI_CSTATE_SYSTEMIO; if (reg->space_id == ACPI_ADR_SPACE_FIXED_HARDWARE) { if (acpi_processor_ffh_cstate_probe - (pr->id, &cx, reg) == 0) { + (cpu, &cx, reg) == 0) { cx.entry_method = ACPI_CSTATE_FFH; } else if (cx.type == ACPI_STATE_C1) { /* @@ -439,7 +432,7 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) continue;

current_count++; - memcpy(&(pr->power.states[current_count]), &cx, sizeof(cx)); + memcpy(&info->states[current_count], &cx, sizeof(cx));

/* * We support total ACPI_PROCESSOR_MAX_POWER - 1 @@ -453,12 +446,9 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) } }

- ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Found %d power states\n", - current_count)); + acpi_handle_info(handle, "Found %d idle states\n", current_count);

- /* Validate number of power states discovered */ - if (current_count < 2) - ret = -EFAULT; + info->count = current_count;

end: kfree(buffer.pointer); @@ -466,6 +456,28 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) return ret; }

+static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) +{ + int ret; + + if (nocst) + return -ENODEV; + + ret = acpi_processor_evaluate_cst(pr->handle, pr->id, &pr->power); + if (ret) + return ret; + + /* + * It is expected that there will be at least 2 states, C1 and + * something else (C2 or C3), so fail if that is not the case. + */ + if (pr->power.count < 2) + return -EFAULT; + + pr->flags.has_cst = 1; + return 0; +} + static void acpi_processor_power_verify_c3(struct acpi_processor *pr, struct acpi_processor_cx *cx) {

From: "Rafael J. Wysocki" rafael.j.wysocki@intel.com

mainline inclusion from mainline-v5.6-rc1 commit 77fb4e0a559a960eb36d0b2c50c781c5492577eb category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 77fb4e0a559a960eb36d0b2c50c781c5492577eb upsteam

The intel_idle driver will be modified to use ACPI _CST subsequently and it will need to call acpi_processor_evaluate_cst(), so move that function to acpi_processor.c so that it is always present (which is required by intel_idle) and export it to modules to allow the ACPI processor driver (which is modular) to call it.

No intentional functional impact.

Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: yjia yingbao.jia@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/acpi/acpi_processor.c | 157 ++++++++++++++++++++++++++++++++++ drivers/acpi/processor_idle.c | 142 ------------------------------ include/linux/acpi.h | 9 ++ 3 files changed, 166 insertions(+), 142 deletions(-)

diff --git a/drivers/acpi/acpi_processor.c b/drivers/acpi/acpi_processor.c index 6de4e6b9f8b76..cfdf7cf6d8f1f 100644 --- a/drivers/acpi/acpi_processor.c +++ b/drivers/acpi/acpi_processor.c @@ -732,4 +732,161 @@ bool acpi_processor_claim_cst_control(void) return true; } EXPORT_SYMBOL_GPL(acpi_processor_claim_cst_control); + +/** + * acpi_processor_evaluate_cst - Evaluate the processor _CST control method. + * @handle: ACPI handle of the processor object containing the _CST. + * @cpu: The numeric ID of the target CPU. + * @info: Object write the C-states information into. + * + * Extract the C-state information for the given CPU from the output of the _CST + * control method under the corresponding ACPI processor object (or processor + * device object) and populate @info with it. + * + * If any ACPI_ADR_SPACE_FIXED_HARDWARE C-states are found, invoke + * acpi_processor_ffh_cstate_probe() to verify them and update the + * cpu_cstate_entry data for @cpu. + */ +int acpi_processor_evaluate_cst(acpi_handle handle, u32 cpu, + struct acpi_processor_power *info) +{ + struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; + union acpi_object *cst; + acpi_status status; + u64 count; + int last_index = 0; + int i, ret = 0; + + status = acpi_evaluate_object(handle, "_CST", NULL, &buffer); + if (ACPI_FAILURE(status)) { + acpi_handle_debug(handle, "No _CST\n"); + return -ENODEV; + } + + cst = buffer.pointer; + + /* There must be at least 2 elements. */ + if (!cst || cst->type != ACPI_TYPE_PACKAGE || cst->package.count < 2) { + acpi_handle_warn(handle, "Invalid _CST output\n"); + ret = -EFAULT; + goto end; + } + + count = cst->package.elements[0].integer.value; + + /* Validate the number of C-states. */ + if (count < 1 || count != cst->package.count - 1) { + acpi_handle_warn(handle, "Inconsistent _CST data\n"); + ret = -EFAULT; + goto end; + } + + for (i = 1; i <= count; i++) { + union acpi_object *element; + union acpi_object *obj; + struct acpi_power_register *reg; + struct acpi_processor_cx cx; + + /* + * If there is not enough space for all C-states, skip the + * excess ones and log a warning. + */ + if (last_index >= ACPI_PROCESSOR_MAX_POWER - 1) { + acpi_handle_warn(handle, + "No room for more idle states (limit: %d)\n", + ACPI_PROCESSOR_MAX_POWER - 1); + break; + } + + memset(&cx, 0, sizeof(cx)); + + element = &cst->package.elements[i]; + if (element->type != ACPI_TYPE_PACKAGE) + continue; + + if (element->package.count != 4) + continue; + + obj = &element->package.elements[0]; + + if (obj->type != ACPI_TYPE_BUFFER) + continue; + + reg = (struct acpi_power_register *)obj->buffer.pointer; + + obj = &element->package.elements[1]; + if (obj->type != ACPI_TYPE_INTEGER) + continue; + + cx.type = obj->integer.value; + /* + * There are known cases in which the _CST output does not + * contain C1, so if the type of the first state found is not + * C1, leave an empty slot for C1 to be filled in later. + */ + if (i == 1 && cx.type != ACPI_STATE_C1) + last_index = 1; + + cx.address = reg->address; + cx.index = last_index + 1; + + if (reg->space_id == ACPI_ADR_SPACE_FIXED_HARDWARE) { + if (!acpi_processor_ffh_cstate_probe(cpu, &cx, reg)) { + /* + * In the majority of cases _CST describes C1 as + * a FIXED_HARDWARE C-state, but if the command + * line forbids using MWAIT, use CSTATE_HALT for + * C1 regardless. + */ + if (cx.type == ACPI_STATE_C1 && + boot_option_idle_override == IDLE_NOMWAIT) { + cx.entry_method = ACPI_CSTATE_HALT; + snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI HLT"); + } else { + cx.entry_method = ACPI_CSTATE_FFH; + } + } else if (cx.type == ACPI_STATE_C1) { + /* + * In the special case of C1, FIXED_HARDWARE can + * be handled by executing the HLT instruction. + */ + cx.entry_method = ACPI_CSTATE_HALT; + snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI HLT"); + } else { + continue; + } + } else if (reg->space_id == ACPI_ADR_SPACE_SYSTEM_IO) { + cx.entry_method = ACPI_CSTATE_SYSTEMIO; + snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI IOPORT 0x%x", + cx.address); + } else { + continue; + } + + if (cx.type == ACPI_STATE_C1) + cx.valid = 1; + + obj = &element->package.elements[2]; + if (obj->type != ACPI_TYPE_INTEGER) + continue; + + cx.latency = obj->integer.value; + + obj = &element->package.elements[3]; + if (obj->type != ACPI_TYPE_INTEGER) + continue; + + memcpy(&info->states[++last_index], &cx, sizeof(cx)); + } + + acpi_handle_info(handle, "Found %d idle states\n", last_index); + + info->count = last_index; + + end: + kfree(buffer.pointer); + + return ret; +} +EXPORT_SYMBOL_GPL(acpi_processor_evaluate_cst); #endif /* CONFIG_ACPI_PROCESSOR_CSTATE */ diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 5641d5c8eedef..b97860ec75b62 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -304,148 +304,6 @@ static int acpi_processor_get_power_info_default(struct acpi_processor *pr) return 0; }

-static int acpi_processor_evaluate_cst(acpi_handle handle, u32 cpu, - struct acpi_processor_power *info) -{ - struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; - union acpi_object *cst; - acpi_status status; - u64 count; - int last_index = 0; - int i, ret = 0; - - status = acpi_evaluate_object(handle, "_CST", NULL, &buffer); - if (ACPI_FAILURE(status)) { - acpi_handle_debug(handle, "No _CST\n"); - return -ENODEV; - } - - cst = buffer.pointer; - - /* There must be at least 2 elements. */ - if (!cst || cst->type != ACPI_TYPE_PACKAGE || cst->package.count < 2) { - acpi_handle_warn(handle, "Invalid _CST output\n"); - ret = -EFAULT; - goto end; - } - - count = cst->package.elements[0].integer.value; - - /* Validate the number of C-states. */ - if (count < 1 || count != cst->package.count - 1) { - acpi_handle_warn(handle, "Inconsistent _CST data\n"); - ret = -EFAULT; - goto end; - } - - for (i = 1; i <= count; i++) { - union acpi_object *element; - union acpi_object *obj; - struct acpi_power_register *reg; - struct acpi_processor_cx cx; - - /* - * If there is not enough space for all C-states, skip the - * excess ones and log a warning. - */ - if (last_index >= ACPI_PROCESSOR_MAX_POWER - 1) { - acpi_handle_warn(handle, - "No room for more idle states (limit: %d)\n", - ACPI_PROCESSOR_MAX_POWER - 1); - break; - } - - memset(&cx, 0, sizeof(cx)); - - element = &cst->package.elements[i]; - if (element->type != ACPI_TYPE_PACKAGE) - continue; - - if (element->package.count != 4) - continue; - - obj = &element->package.elements[0]; - - if (obj->type != ACPI_TYPE_BUFFER) - continue; - - reg = (struct acpi_power_register *)obj->buffer.pointer; - - obj = &element->package.elements[1]; - if (obj->type != ACPI_TYPE_INTEGER) - continue; - - cx.type = obj->integer.value; - /* - * There are known cases in which the _CST output does not - * contain C1, so if the type of the first state found is not - * C1, leave an empty slot for C1 to be filled in later. - */ - if (i == 1 && cx.type != ACPI_STATE_C1) - last_index = 1; - - cx.address = reg->address; - cx.index = last_index + 1; - - if (reg->space_id == ACPI_ADR_SPACE_FIXED_HARDWARE) { - if (!acpi_processor_ffh_cstate_probe(cpu, &cx, reg)) { - /* - * In the majority of cases _CST describes C1 as - * a FIXED_HARDWARE C-state, but if the command - * line forbids using MWAIT, use CSTATE_HALT for - * C1 regardless. - */ - if (cx.type == ACPI_STATE_C1 && - boot_option_idle_override == IDLE_NOMWAIT) { - cx.entry_method = ACPI_CSTATE_HALT; - snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI HLT"); - } else { - cx.entry_method = ACPI_CSTATE_FFH; - } - } else if (cx.type == ACPI_STATE_C1) { - /* - * In the special case of C1, FIXED_HARDWARE can - * be handled by executing the HLT instruction. - */ - cx.entry_method = ACPI_CSTATE_HALT; - snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI HLT"); - } else { - continue; - } - } else if (reg->space_id == ACPI_ADR_SPACE_SYSTEM_IO) { - cx.entry_method = ACPI_CSTATE_SYSTEMIO; - snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI IOPORT 0x%x", - cx.address); - } else { - continue; - } - - if (cx.type == ACPI_STATE_C1) - cx.valid = 1; - - obj = &element->package.elements[2]; - if (obj->type != ACPI_TYPE_INTEGER) - continue; - - cx.latency = obj->integer.value; - - obj = &element->package.elements[3]; - if (obj->type != ACPI_TYPE_INTEGER) - continue; - - memcpy(&info->states[++last_index], &cx, sizeof(cx)); - } - - acpi_handle_info(handle, "Found %d idle states\n", last_index); - - info->count = last_index; - - end: - kfree(buffer.pointer); - - return ret; -} - static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) { int ret; diff --git a/include/linux/acpi.h b/include/linux/acpi.h index f7383264c1c44..964d07f180c46 100644 --- a/include/linux/acpi.h +++ b/include/linux/acpi.h @@ -296,10 +296,19 @@ static inline bool invalid_phys_cpuid(phys_cpuid_t phys_id) /* Validate the processor object's proc_id */ bool acpi_duplicate_processor_id(int proc_id); /* Processor _CTS control */ +struct acpi_processor_power; + #ifdef CONFIG_ACPI_PROCESSOR_CSTATE bool acpi_processor_claim_cst_control(void); +int acpi_processor_evaluate_cst(acpi_handle handle, u32 cpu, + struct acpi_processor_power *info); #else static inline bool acpi_processor_claim_cst_control(void) { return false; } +static inline int acpi_processor_evaluate_cst(acpi_handle handle, u32 cpu, + struct acpi_processor_power *info) +{ + return -ENODEV; +} #endif

#ifdef CONFIG_ACPI_HOTPLUG_CPU

From: "Rafael J. Wysocki" rafael.j.wysocki@intel.com

mainline inclusion from mainline-v5.6-rc1 commit 18734958e9bfbc055805d110a38dc76307eba742 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 18734958e9bfbc055805d110a38dc76307eba742 upstream

Modify the intel_idle driver to get the C-states information from ACPI _CST if the processor model is not recognized by it.

The processor is still required to support MWAIT and the information from ACPI _CST will only be used if all of the C-states listed by _CST are of the ACPI_CSTATE_FFH type (which means that they are expected to be entered via MWAIT).

Moreover, the driver assumes that the _CST information is the same for all CPUs in the system, so it is sufficient to evaluate _CST for one of them and extract the common list of C-states from there. Also _CST is evaluated once at the system initialization time and the driver does not respond to _CST change notifications (that can be changed in the future).

The main functional difference between intel_idle with this change and the ACPI processor driver is that the former sets the target residency to be equal to the exit latency (provided by _CST) for C1-type C-states and to 3 times the exit latency value for the other C-state types, whereas the latter obtains the target residency by multiplying the exit latency by the same number (2 by default) for all C-state types. Therefore it is expected that in general using the former instead of the latter on the same system will lead to improved energy-efficiency.

Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: yjia yingbao.jia@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/idle/intel_idle.c | 190 ++++++++++++++++++++++++++++++++------ 1 file changed, 162 insertions(+), 28 deletions(-)

diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 4bb1fc966c958..22fffcbebbf5b 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -53,6 +53,7 @@

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

+#include <linux/acpi.h> #include <linux/kernel.h> #include <linux/cpuidle.h> #include <linux/tick.h> @@ -1125,6 +1126,129 @@ static const struct x86_cpu_id intel_idle_ids[] __initconst = { {} };

+#define INTEL_CPU_FAM6_MWAIT \ + { X86_VENDOR_INTEL, 6, X86_MODEL_ANY, X86_FEATURE_MWAIT, 0 } + +static const struct x86_cpu_id intel_mwait_ids[] __initconst = { + INTEL_CPU_FAM6_MWAIT, + {} +}; + +static bool intel_idle_max_cstate_reached(int cstate) +{ + if (cstate + 1 > max_cstate) { + pr_info("max_cstate %d reached\n", max_cstate); + return true; + } + return false; +} + +#ifdef CONFIG_ACPI_PROCESSOR_CSTATE +#include <acpi/processor.h> + +static struct acpi_processor_power acpi_state_table; + +/** + * intel_idle_cst_usable - Check if the _CST information can be used. + * + * Check if all of the C-states listed by _CST in the max_cstate range are + * ACPI_CSTATE_FFH, which means that they should be entered via MWAIT. + */ +static bool intel_idle_cst_usable(void) +{ + int cstate, limit; + + limit = min_t(int, min_t(int, CPUIDLE_STATE_MAX, max_cstate + 1), + acpi_state_table.count); + + for (cstate = 1; cstate < limit; cstate++) { + struct acpi_processor_cx *cx = &acpi_state_table.states[cstate]; + + if (cx->entry_method != ACPI_CSTATE_FFH) + return false; + } + + return true; +} + +static bool intel_idle_acpi_cst_extract(void) +{ + unsigned int cpu; + + for_each_possible_cpu(cpu) { + struct acpi_processor *pr = per_cpu(processors, cpu); + + if (!pr) + continue; + + if (acpi_processor_evaluate_cst(pr->handle, cpu, &acpi_state_table)) + continue; + + acpi_state_table.count++; + + if (!intel_idle_cst_usable()) + continue; + + if (!acpi_processor_claim_cst_control()) { + acpi_state_table.count = 0; + return false; + } + + return true; + } + + pr_debug("ACPI _CST not found or not usable\n"); + return false; +} + +static void intel_idle_init_cstates_acpi(struct cpuidle_driver *drv) +{ + int cstate, limit = min_t(int, CPUIDLE_STATE_MAX, acpi_state_table.count); + + /* + * If limit > 0, intel_idle_cst_usable() has returned 'true', so all of + * the interesting states are ACPI_CSTATE_FFH. + */ + for (cstate = 1; cstate < limit; cstate++) { + struct acpi_processor_cx *cx; + struct cpuidle_state *state; + + if (intel_idle_max_cstate_reached(cstate)) + break; + + cx = &acpi_state_table.states[cstate]; + + state = &drv->states[drv->state_count++]; + + snprintf(state->name, CPUIDLE_NAME_LEN, "C%d_ACPI", cstate); + strlcpy(state->desc, cx->desc, CPUIDLE_DESC_LEN); + state->exit_latency = cx->latency; + /* + * For C1-type C-states use the same number for both the exit + * latency and target residency, because that is the case for + * C1 in the majority of the static C-states tables above. + * For the other types of C-states, however, set the target + * residency to 3 times the exit latency which should lead to + * a reasonable balance between energy-efficiency and + * performance in the majority of interesting cases. + */ + state->target_residency = cx->latency; + if (cx->type > ACPI_STATE_C1) + state->target_residency *= 3; + + state->flags = MWAIT2flg(cx->address); + if (cx->type > ACPI_STATE_C2) + state->flags |= CPUIDLE_FLAG_TLB_FLUSHED; + + state->enter = intel_idle; + state->enter_s2idle = intel_idle_s2idle; + } +} +#else /* !CONFIG_ACPI_PROCESSOR_CSTATE */ +static inline bool intel_idle_acpi_cst_extract(void) { return false; } +static inline void intel_idle_init_cstates_acpi(struct cpuidle_driver *drv) { } +#endif /* !CONFIG_ACPI_PROCESSOR_CSTATE */ + /* * intel_idle_probe() */ @@ -1139,17 +1263,15 @@ static int __init intel_idle_probe(void) }

id = x86_match_cpu(intel_idle_ids); - if (!id) { - if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && - boot_cpu_data.x86 == 6) - pr_debug("does not run on family %d model %d\n", - boot_cpu_data.x86, boot_cpu_data.x86_model); - return -ENODEV; - } - - if (!boot_cpu_has(X86_FEATURE_MWAIT)) { - pr_debug("Please enable MWAIT in BIOS SETUP\n"); - return -ENODEV; + if (id) { + if (!boot_cpu_has(X86_FEATURE_MWAIT)) { + pr_debug("Please enable MWAIT in BIOS SETUP\n"); + return -ENODEV; + } + } else { + id = x86_match_cpu(intel_mwait_ids); + if (!id) + return -ENODEV; }

if (boot_cpu_data.cpuid_level < CPUID_MWAIT_LEAF) @@ -1165,7 +1287,10 @@ static int __init intel_idle_probe(void) pr_debug("MWAIT substates: 0x%x\n", mwait_substates);

icpu = (const struct idle_cpu *)id->driver_data; - cpuidle_state_table = icpu->state_table; + if (icpu) + cpuidle_state_table = icpu->state_table; + else if (!intel_idle_acpi_cst_extract()) + return -ENODEV;

pr_debug("v" INTEL_IDLE_VERSION " model 0x%X\n", boot_cpu_data.x86_model); @@ -1347,31 +1472,19 @@ static void intel_idle_state_table_update(void) } }

-/* - * intel_idle_cpuidle_driver_init() - * allocate, initialize cpuidle_states - */ -static void __init intel_idle_cpuidle_driver_init(void) +static void intel_idle_init_cstates_icpu(struct cpuidle_driver *drv) { int cstate; - struct cpuidle_driver *drv = &intel_idle_driver; - - intel_idle_state_table_update(); - - cpuidle_poll_state_init(drv); - drv->state_count = 1;

for (cstate = 0; cstate < CPUIDLE_STATE_MAX; ++cstate) { unsigned int mwait_hint;

- if (!cpuidle_state_table[cstate].enter && - !cpuidle_state_table[cstate].enter_s2idle) + if (intel_idle_max_cstate_reached(cstate)) break;

- if (cstate + 1 > max_cstate) { - pr_info("max_cstate %d reached\n", max_cstate); + if (!cpuidle_state_table[cstate].enter && + !cpuidle_state_table[cstate].enter_s2idle) break; - }

/* If marked as unusable, skip this state. */ if (cpuidle_state_table[cstate].disabled != 0) { @@ -1394,6 +1507,24 @@ static void __init intel_idle_cpuidle_driver_init(void) } }

+/* + * intel_idle_cpuidle_driver_init() + * allocate, initialize cpuidle_states + */ +static void __init intel_idle_cpuidle_driver_init(void) +{ + struct cpuidle_driver *drv = &intel_idle_driver; + + intel_idle_state_table_update(); + + cpuidle_poll_state_init(drv); + drv->state_count = 1; + + if (icpu) + intel_idle_init_cstates_icpu(drv); + else + intel_idle_init_cstates_acpi(drv); +}

/* * intel_idle_cpu_init() @@ -1412,6 +1543,9 @@ static int intel_idle_cpu_init(unsigned int cpu) return -EIO; }

+ if (!icpu) + return 0; + if (icpu->auto_demotion_disable_flags) auto_demotion_disable();

From: "Rafael J. Wysocki" rafael.j.wysocki@intel.com

mainline inclusion from mainline-v5.0-rc1 commit aa5eee355b466cb33f97f79bed9740a472c4ab73 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit aa5eee355b466cb33f97f79bed9740a472c4ab73 upstream

Important information is missing from user/admin cpuidle documentation available today, so add a new user/admin document for cpuidle containing current and comprehensive information to admin-guide and drop the old .txt documents it is replacing.

Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Reviewed-by: Viresh Kumar viresh.kumar@linaro.org Reviewed-by: Ulf Hansson ulf.hansson@linaro.org Signed-off-by: yjia yingbao.jia@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- .../admin-guide/pm/working-state.rst | 1 + Documentation/cpuidle/core.txt | 23 ----- Documentation/cpuidle/sysfs.txt | 98 ------------------- 3 files changed, 1 insertion(+), 121 deletions(-) delete mode 100644 Documentation/cpuidle/core.txt delete mode 100644 Documentation/cpuidle/sysfs.txt

diff --git a/Documentation/admin-guide/pm/working-state.rst b/Documentation/admin-guide/pm/working-state.rst index fa01bf083dfe6..b6cef9b5e961e 100644 --- a/Documentation/admin-guide/pm/working-state.rst +++ b/Documentation/admin-guide/pm/working-state.rst @@ -5,5 +5,6 @@ Working-State Power Management .. toctree:: :maxdepth: 2

+ cpuidle cpufreq intel_pstate diff --git a/Documentation/cpuidle/core.txt b/Documentation/cpuidle/core.txt deleted file mode 100644 index 63ecc5dc9d8a6..0000000000000 --- a/Documentation/cpuidle/core.txt +++ /dev/null @@ -1,23 +0,0 @@ - - Supporting multiple CPU idle levels in kernel - - cpuidle - -General Information: - -Various CPUs today support multiple idle levels that are differentiated -by varying exit latencies and power consumption during idle. -cpuidle is a generic in-kernel infrastructure that separates -idle policy (governor) from idle mechanism (driver) and provides a -standardized infrastructure to support independent development of -governors and drivers. - -cpuidle resides under drivers/cpuidle. - -Boot options: -"cpuidle_sysfs_switch" -enables current_governor interface in /sys/devices/system/cpu/cpuidle/, -which can be used to switch governors at run time. This boot option -is meant for developer testing only. In normal usage, kernel picks the -best governor based on governor ratings. -SEE ALSO: sysfs.txt in this directory. diff --git a/Documentation/cpuidle/sysfs.txt b/Documentation/cpuidle/sysfs.txt deleted file mode 100644 index d1587f434e7bb..0000000000000 --- a/Documentation/cpuidle/sysfs.txt +++ /dev/null @@ -1,98 +0,0 @@ - - - Supporting multiple CPU idle levels in kernel - - cpuidle sysfs - -System global cpuidle related information and tunables are under -/sys/devices/system/cpu/cpuidle - -The current interfaces in this directory has self-explanatory names: -* current_driver -* current_governor_ro - -With cpuidle_sysfs_switch boot option (meant for developer testing) -following objects are visible instead. -* current_driver -* available_governors -* current_governor -In this case users can switch the governor at run time by writing -to current_governor. - - -Per logical CPU specific cpuidle information are under -/sys/devices/system/cpu/cpuX/cpuidle -for each online cpu X - --------------------------------------------------------------------------------- -# ls -lR /sys/devices/system/cpu/cpu0/cpuidle/ -/sys/devices/system/cpu/cpu0/cpuidle/: -total 0 -drwxr-xr-x 2 root root 0 Feb 8 10:42 state0 -drwxr-xr-x 2 root root 0 Feb 8 10:42 state1 -drwxr-xr-x 2 root root 0 Feb 8 10:42 state2 -drwxr-xr-x 2 root root 0 Feb 8 10:42 state3 - -/sys/devices/system/cpu/cpu0/cpuidle/state0: -total 0 --r--r--r-- 1 root root 4096 Feb 8 10:42 desc --rw-r--r-- 1 root root 4096 Feb 8 10:42 disable --r--r--r-- 1 root root 4096 Feb 8 10:42 latency --r--r--r-- 1 root root 4096 Feb 8 10:42 name --r--r--r-- 1 root root 4096 Feb 8 10:42 power --r--r--r-- 1 root root 4096 Feb 8 10:42 residency --r--r--r-- 1 root root 4096 Feb 8 10:42 time --r--r--r-- 1 root root 4096 Feb 8 10:42 usage - -/sys/devices/system/cpu/cpu0/cpuidle/state1: -total 0 --r--r--r-- 1 root root 4096 Feb 8 10:42 desc --rw-r--r-- 1 root root 4096 Feb 8 10:42 disable --r--r--r-- 1 root root 4096 Feb 8 10:42 latency --r--r--r-- 1 root root 4096 Feb 8 10:42 name --r--r--r-- 1 root root 4096 Feb 8 10:42 power --r--r--r-- 1 root root 4096 Feb 8 10:42 residency --r--r--r-- 1 root root 4096 Feb 8 10:42 time --r--r--r-- 1 root root 4096 Feb 8 10:42 usage - -/sys/devices/system/cpu/cpu0/cpuidle/state2: -total 0 --r--r--r-- 1 root root 4096 Feb 8 10:42 desc --rw-r--r-- 1 root root 4096 Feb 8 10:42 disable --r--r--r-- 1 root root 4096 Feb 8 10:42 latency --r--r--r-- 1 root root 4096 Feb 8 10:42 name --r--r--r-- 1 root root 4096 Feb 8 10:42 power --r--r--r-- 1 root root 4096 Feb 8 10:42 residency --r--r--r-- 1 root root 4096 Feb 8 10:42 time --r--r--r-- 1 root root 4096 Feb 8 10:42 usage - -/sys/devices/system/cpu/cpu0/cpuidle/state3: -total 0 --r--r--r-- 1 root root 4096 Feb 8 10:42 desc --rw-r--r-- 1 root root 4096 Feb 8 10:42 disable --r--r--r-- 1 root root 4096 Feb 8 10:42 latency --r--r--r-- 1 root root 4096 Feb 8 10:42 name --r--r--r-- 1 root root 4096 Feb 8 10:42 power --r--r--r-- 1 root root 4096 Feb 8 10:42 residency --r--r--r-- 1 root root 4096 Feb 8 10:42 time --r--r--r-- 1 root root 4096 Feb 8 10:42 usage --------------------------------------------------------------------------------- - - -* desc : Small description about the idle state (string) -* disable : Option to disable this idle state (bool) -> see note below -* latency : Latency to exit out of this idle state (in microseconds) -* residency : Time after which a state becomes more effecient than any - shallower state (in microseconds) -* name : Name of the idle state (string) -* power : Power consumed while in this idle state (in milliwatts) -* time : Total time spent in this idle state (in microseconds) -* usage : Number of times this state was entered (count) - -Note: -The behavior and the effect of the disable variable depends on the -implementation of a particular governor. In the ladder governor, for -example, it is not coherent, i.e. if one is disabling a light state, -then all deeper states are disabled as well, but the disable variable -does not reflect it. Likewise, if one enables a deep state but a lighter -state still is disabled, then this has no effect.

From: "Rafael J. Wysocki" rafael.j.wysocki@intel.com

mainline inclusion from mainline-v5.6-rc1 commit bff8e60a86f4960133f90ad9add9adeb082b8154 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit bff8e60a86f4960133f90ad9add9adeb082b8154 upstream

Update the intel_idle driver to get the C-states information from ACPI _CST in some cases in which the processor is known to the driver, as long as that information is available and the new use_acpi flag is set in the profile of the processor in question.

In the cases when there is a specific table of C-states for the given processor in the driver, that table is used as the primary source of information on the available C-states, but if ACPI _CST is present, the C-states that are not listed by it will not be enabled by default (they still can be enabled later by user space via sysfs, though).

The new CPUIDLE_FLAG_ALWAYS_ENABLE flag can be used for marking C-states that should be enabled by default even if they are not listed by ACPI _CST.

Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: yjia yingbao.jia@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/idle/intel_idle.c | 45 ++++++++++++++++++++++++++++++++++++--- 1 file changed, 42 insertions(+), 3 deletions(-)

diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 22fffcbebbf5b..487edbad098d3 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -92,6 +92,7 @@ struct idle_cpu { unsigned long auto_demotion_disable_flags; bool byt_auto_demotion_disable_flag; bool disable_promotion_to_c1e; + bool use_acpi; };

static const struct idle_cpu *icpu; @@ -102,6 +103,11 @@ static void intel_idle_s2idle(struct cpuidle_device *dev, struct cpuidle_driver *drv, int index); static struct cpuidle_state *cpuidle_state_table;

+/* + * Enable this state by default even if the ACPI _CST does not list it. + */ +#define CPUIDLE_FLAG_ALWAYS_ENABLE BIT(15) + /* * Set this flag for states where the HW flushes the TLB for us * and so we don't need cross-calls to keep it consistent. @@ -1244,9 +1250,33 @@ static void intel_idle_init_cstates_acpi(struct cpuidle_driver *drv) state->enter_s2idle = intel_idle_s2idle; } } + +static bool intel_idle_off_by_default(u32 mwait_hint) +{ + int cstate, limit; + + /* + * If there are no _CST C-states, do not disable any C-states by + * default. + */ + if (!acpi_state_table.count) + return false; + + limit = min_t(int, CPUIDLE_STATE_MAX, acpi_state_table.count); + /* + * If limit > 0, intel_idle_cst_usable() has returned 'true', so all of + * the interesting states are ACPI_CSTATE_FFH. + */ + for (cstate = 1; cstate < limit; cstate++) { + if (acpi_state_table.states[cstate].address == mwait_hint) + return false; + } + return true; +} #else /* !CONFIG_ACPI_PROCESSOR_CSTATE */ static inline bool intel_idle_acpi_cst_extract(void) { return false; } static inline void intel_idle_init_cstates_acpi(struct cpuidle_driver *drv) { } +static inline bool intel_idle_off_by_default(u32 mwait_hint) { return false; } #endif /* !CONFIG_ACPI_PROCESSOR_CSTATE */

/* @@ -1287,10 +1317,13 @@ static int __init intel_idle_probe(void) pr_debug("MWAIT substates: 0x%x\n", mwait_substates);

icpu = (const struct idle_cpu *)id->driver_data; - if (icpu) + if (icpu) { cpuidle_state_table = icpu->state_table; - else if (!intel_idle_acpi_cst_extract()) + if (icpu->use_acpi) + intel_idle_acpi_cst_extract(); + } else if (!intel_idle_acpi_cst_extract()) { return -ENODEV; + }

pr_debug("v" INTEL_IDLE_VERSION " model 0x%X\n", boot_cpu_data.x86_model); @@ -1498,7 +1531,13 @@ static void intel_idle_init_cstates_icpu(struct cpuidle_driver *drv) continue;

/* Structure copy. */ - drv->states[drv->state_count++] = cpuidle_state_table[cstate]; + drv->states[drv->state_count] = cpuidle_state_table[cstate]; + + if (icpu->use_acpi && intel_idle_off_by_default(mwait_hint) && + !(cpuidle_state_table[cstate].flags & CPUIDLE_FLAG_ALWAYS_ENABLE)) + drv->states[drv->state_count].flags |= CPUIDLE_FLAG_OFF; + + drv->state_count++; }

if (icpu->byt_auto_demotion_disable_flag) {

From: Guoqing Jiang jiangguoqing@kylinos.cn

mainline inclusion from mainline-v5.6-rc1 commit e6d4f08a677654385869ba0c39d7c9ceec47e5c5 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit e6d4f08a677654385869ba0c39d7c9ceec47e5c5 upstream

In many cases, especially on server systems, it is desirable to avoid enabling C-states that have been disabled in the platform firmware (BIOS) setup, except for C1E.

As a rule, the C-states disabled this way are not listed by ACPI _CST, so if that is used by intel_idle along with the specific table of C-states that it has for the given processor, the C-states disabled through the platform firmware will not be enabled by default by intel_idle.

Accordingly, set the use_acpi flag (introduced previously) in all server processor profiles defined in intel_idle (so as to make it use ACPI _CST to decide which C-states to enable by default) and set the CPUIDLE_FLAG_ALWAYS_ENABLE flag (also introduced previously) for C1E in all C-states tables in intel_idle that contain C1 too (so that C1E is enabled regardless of whether or not it is listed by ACPI _CST).

Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: yjia yingbao.jia@intel.com Signed-off-by: Guoqing Jiang jiangguoqing@kylinos.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- drivers/idle/intel_idle.c | 70 ++++++++++++++++++++++++++++----------- 1 file changed, 50 insertions(+), 20 deletions(-)

diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index cfd3ecafed7b0..a921070184a00 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -143,7 +143,7 @@ static struct cpuidle_state nehalem_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -180,7 +180,7 @@ static struct cpuidle_state snb_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -315,7 +315,7 @@ static struct cpuidle_state ivb_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -360,7 +360,7 @@ static struct cpuidle_state ivt_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 80, .enter = &intel_idle, @@ -397,7 +397,7 @@ static struct cpuidle_state ivt_cstates_4s[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 250, .enter = &intel_idle, @@ -434,7 +434,7 @@ static struct cpuidle_state ivt_cstates_8s[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 500, .enter = &intel_idle, @@ -471,7 +471,7 @@ static struct cpuidle_state hsw_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -539,7 +539,7 @@ static struct cpuidle_state bdw_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -608,7 +608,7 @@ static struct cpuidle_state skl_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -677,7 +677,7 @@ static struct cpuidle_state skx_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -827,7 +827,7 @@ static struct cpuidle_state bxt_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -888,7 +888,7 @@ static struct cpuidle_state dnv_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -1010,6 +1010,13 @@ static const struct idle_cpu idle_cpu_nehalem = { .disable_promotion_to_c1e = true, };

+static const struct idle_cpu idle_cpu_nhx = { + .state_table = nehalem_cstates, + .auto_demotion_disable_flags = NHM_C1_AUTO_DEMOTE | NHM_C3_AUTO_DEMOTE, + .disable_promotion_to_c1e = true, + .use_acpi = true, +}; + static const struct idle_cpu idle_cpu_atom = { .state_table = atom_cstates, }; @@ -1028,6 +1035,12 @@ static const struct idle_cpu idle_cpu_snb = { .disable_promotion_to_c1e = true, };

+static const struct idle_cpu idle_cpu_snx = { + .state_table = snb_cstates, + .disable_promotion_to_c1e = true, + .use_acpi = true, +}; + static const struct idle_cpu idle_cpu_byt = { .state_table = byt_cstates, .disable_promotion_to_c1e = true, @@ -1048,6 +1061,7 @@ static const struct idle_cpu idle_cpu_ivb = { static const struct idle_cpu idle_cpu_ivt = { .state_table = ivt_cstates, .disable_promotion_to_c1e = true, + .use_acpi = true, };

static const struct idle_cpu idle_cpu_hsw = { @@ -1055,11 +1069,23 @@ static const struct idle_cpu idle_cpu_hsw = { .disable_promotion_to_c1e = true, };

+static const struct idle_cpu idle_cpu_hsx = { + .state_table = hsw_cstates, + .disable_promotion_to_c1e = true, + .use_acpi = true, +}; + static const struct idle_cpu idle_cpu_bdw = { .state_table = bdw_cstates, .disable_promotion_to_c1e = true, };

+static const struct idle_cpu idle_cpu_bdx = { + .state_table = bdw_cstates, + .disable_promotion_to_c1e = true, + .use_acpi = true, +}; + static const struct idle_cpu idle_cpu_skl = { .state_table = skl_cstates, .disable_promotion_to_c1e = true, @@ -1068,15 +1094,18 @@ static const struct idle_cpu idle_cpu_skl = { static const struct idle_cpu idle_cpu_skx = { .state_table = skx_cstates, .disable_promotion_to_c1e = true, + .use_acpi = true, };

static const struct idle_cpu idle_cpu_avn = { .state_table = avn_cstates, .disable_promotion_to_c1e = true, + .use_acpi = true, };

static const struct idle_cpu idle_cpu_knl = { .state_table = knl_cstates, + .use_acpi = true, };

static const struct idle_cpu idle_cpu_bxt = { @@ -1087,6 +1116,7 @@ static const struct idle_cpu idle_cpu_bxt = { static const struct idle_cpu idle_cpu_dnv = { .state_table = dnv_cstates, .disable_promotion_to_c1e = true, + .use_acpi = true, };

#define ICPU(model, cpu) \ @@ -1094,16 +1124,16 @@ static const struct idle_cpu idle_cpu_dnv = {

static const struct x86_cpu_id intel_idle_ids[] __initconst = { ICPU(INTEL_FAM6_NEHALEM_EP, idle_cpu_nehalem), - ICPU(INTEL_FAM6_NEHALEM, idle_cpu_nehalem), + ICPU(INTEL_FAM6_NEHALEM, idle_cpu_nhx), ICPU(INTEL_FAM6_NEHALEM_G, idle_cpu_nehalem), ICPU(INTEL_FAM6_WESTMERE, idle_cpu_nehalem), - ICPU(INTEL_FAM6_WESTMERE_EP, idle_cpu_nehalem), - ICPU(INTEL_FAM6_NEHALEM_EX, idle_cpu_nehalem), + ICPU(INTEL_FAM6_WESTMERE_EP, idle_cpu_nhx), + ICPU(INTEL_FAM6_NEHALEM_EX, idle_cpu_nhx), ICPU(INTEL_FAM6_ATOM_BONNELL, idle_cpu_atom), ICPU(INTEL_FAM6_ATOM_BONNELL_MID, idle_cpu_lincroft), - ICPU(INTEL_FAM6_WESTMERE_EX, idle_cpu_nehalem), + ICPU(INTEL_FAM6_WESTMERE_EX, idle_cpu_nhx), ICPU(INTEL_FAM6_SANDYBRIDGE, idle_cpu_snb), - ICPU(INTEL_FAM6_SANDYBRIDGE_X, idle_cpu_snb), + ICPU(INTEL_FAM6_SANDYBRIDGE_X, idle_cpu_snx), ICPU(INTEL_FAM6_ATOM_SALTWELL, idle_cpu_atom), ICPU(INTEL_FAM6_ATOM_SILVERMONT, idle_cpu_byt), ICPU(INTEL_FAM6_ATOM_SILVERMONT_MID, idle_cpu_tangier), @@ -1111,14 +1141,14 @@ static const struct x86_cpu_id intel_idle_ids[] __initconst = { ICPU(INTEL_FAM6_IVYBRIDGE, idle_cpu_ivb), ICPU(INTEL_FAM6_IVYBRIDGE_X, idle_cpu_ivt), ICPU(INTEL_FAM6_HASWELL_CORE, idle_cpu_hsw), - ICPU(INTEL_FAM6_HASWELL_X, idle_cpu_hsw), + ICPU(INTEL_FAM6_HASWELL_X, idle_cpu_hsx), ICPU(INTEL_FAM6_HASWELL_ULT, idle_cpu_hsw), ICPU(INTEL_FAM6_HASWELL_GT3E, idle_cpu_hsw), ICPU(INTEL_FAM6_ATOM_SILVERMONT_X, idle_cpu_avn), ICPU(INTEL_FAM6_BROADWELL_CORE, idle_cpu_bdw), ICPU(INTEL_FAM6_BROADWELL_GT3E, idle_cpu_bdw), - ICPU(INTEL_FAM6_BROADWELL_X, idle_cpu_bdw), - ICPU(INTEL_FAM6_BROADWELL_XEON_D, idle_cpu_bdw), + ICPU(INTEL_FAM6_BROADWELL_X, idle_cpu_bdx), + ICPU(INTEL_FAM6_BROADWELL_XEON_D, idle_cpu_bdx), ICPU(INTEL_FAM6_SKYLAKE_MOBILE, idle_cpu_skl), ICPU(INTEL_FAM6_SKYLAKE_DESKTOP, idle_cpu_skl), ICPU(INTEL_FAM6_KABYLAKE_MOBILE, idle_cpu_skl),

From: "Rafael J. Wysocki" rafael.j.wysocki@intel.com

mainline inclusion from mainline-v5.6-rc1 commit a3299182216397a0b943d2549d1997f4eba2bdd2 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit a3299182216397a0b943d2549d1997f4eba2bdd2 upstream

Add an admin-guide document for the intel_idle driver to describe how it works: how it enumerates idle states, what happens during the initialization of it, how it can be controlled via the kernel command line and so on.

Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Reviewed-by: Randy Dunlap rdunlap@infradead.org Signed-off-by: yjia yingbao.jia@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Hanjun Guo guohanjun@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- Documentation/admin-guide/pm/intel_idle.rst | 246 ++++++++++++++++++ .../admin-guide/pm/working-state.rst | 1 + 2 files changed, 247 insertions(+) create mode 100644 Documentation/admin-guide/pm/intel_idle.rst

diff --git a/Documentation/admin-guide/pm/intel_idle.rst b/Documentation/admin-guide/pm/intel_idle.rst new file mode 100644 index 0000000000000..afbf778035f82 --- /dev/null +++ b/Documentation/admin-guide/pm/intel_idle.rst @@ -0,0 +1,246 @@ +.. SPDX-License-Identifier: GPL-2.0 +.. include:: <isonum.txt> + +============================================== +``intel_idle`` CPU Idle Time Management Driver +============================================== + +:Copyright: |copy| 2020 Intel Corporation + +:Author: Rafael J. Wysocki rafael.j.wysocki@intel.com + + +General Information +=================== + +``intel_idle`` is a part of the +:doc:`CPU idle time management subsystem <cpuidle>` in the Linux kernel +(``CPUIdle``). It is the default CPU idle time management driver for the +Nehalem and later generations of Intel processors, but the level of support for +a particular processor model in it depends on whether or not it recognizes that +processor model and may also depend on information coming from the platform +firmware. [To understand ``intel_idle`` it is necessary to know how ``CPUIdle`` +works in general, so this is the time to get familiar with :doc:`cpuidle` if you +have not done that yet.] + +``intel_idle`` uses the ``MWAIT`` instruction to inform the processor that the +logical CPU executing it is idle and so it may be possible to put some of the +processor's functional blocks into low-power states. That instruction takes two +arguments (passed in the ``EAX`` and ``ECX`` registers of the target CPU), the +first of which, referred to as a *hint*, can be used by the processor to +determine what can be done (for details refer to Intel Software Developer’s +Manual [1]_). Accordingly, ``intel_idle`` refuses to work with processors in +which the support for the ``MWAIT`` instruction has been disabled (for example, +via the platform firmware configuration menu) or which do not support that +instruction at all. + +``intel_idle`` is not modular, so it cannot be unloaded, which means that the +only way to pass early-configuration-time parameters to it is via the kernel +command line. + + +.. _intel-idle-enumeration-of-states: + +Enumeration of Idle States +========================== + +Each ``MWAIT`` hint value is interpreted by the processor as a license to +reconfigure itself in a certain way in order to save energy. The processor +configurations (with reduced power draw) resulting from that are referred to +as C-states (in the ACPI terminology) or idle states. The list of meaningful +``MWAIT`` hint values and idle states (i.e. low-power configurations of the +processor) corresponding to them depends on the processor model and it may also +depend on the configuration of the platform. + +In order to create a list of available idle states required by the ``CPUIdle`` +subsystem (see :ref:`idle-states-representation` in :doc:`cpuidle`), +``intel_idle`` can use two sources of information: static tables of idle states +for different processor models included in the driver itself and the ACPI tables +of the system. The former are always used if the processor model at hand is +recognized by ``intel_idle`` and the latter are used if that is required for +the given processor model (which is the case for all server processor models +recognized by ``intel_idle``) or if the processor model is not recognized. + +If the ACPI tables are going to be used for building the list of available idle +states, ``intel_idle`` first looks for a ``_CST`` object under one of the ACPI +objects corresponding to the CPUs in the system (refer to the ACPI specification +[2]_ for the description of ``_CST`` and its output package). Because the +``CPUIdle`` subsystem expects that the list of idle states supplied by the +driver will be suitable for all of the CPUs handled by it and ``intel_idle`` is +registered as the ``CPUIdle`` driver for all of the CPUs in the system, the +driver looks for the first ``_CST`` object returning at least one valid idle +state description and such that all of the idle states included in its return +package are of the FFH (Functional Fixed Hardware) type, which means that the +``MWAIT`` instruction is expected to be used to tell the processor that it can +enter one of them. The return package of that ``_CST`` is then assumed to be +applicable to all of the other CPUs in the system and the idle state +descriptions extracted from it are stored in a preliminary list of idle states +coming from the ACPI tables. [This step is skipped if ``intel_idle`` is +configured to ignore the ACPI tables; see `below <intel-idle-parameters_>`_.] + +Next, the first (index 0) entry in the list of available idle states is +initialized to represent a "polling idle state" (a pseudo-idle state in which +the target CPU continuously fetches and executes instructions), and the +subsequent (real) idle state entries are populated as follows. + +If the processor model at hand is recognized by ``intel_idle``, there is a +(static) table of idle state descriptions for it in the driver. In that case, +the "internal" table is the primary source of information on idle states and the +information from it is copied to the final list of available idle states. If +using the ACPI tables for the enumeration of idle states is not required +(depending on the processor model), all of the listed idle state are enabled by +default (so all of them will be taken into consideration by ``CPUIdle`` +governors during CPU idle state selection). Otherwise, some of the listed idle +states may not be enabled by default if there are no matching entries in the +preliminary list of idle states coming from the ACPI tables. In that case user +space still can enable them later (on a per-CPU basis) with the help of +the ``disable`` idle state attribute in ``sysfs`` (see +:ref:`idle-states-representation` in :doc:`cpuidle`). This basically means that +the idle states "known" to the driver may not be enabled by default if they have +not been exposed by the platform firmware (through the ACPI tables). + +If the given processor model is not recognized by ``intel_idle``, but it +supports ``MWAIT``, the preliminary list of idle states coming from the ACPI +tables is used for building the final list that will be supplied to the +``CPUIdle`` core during driver registration. For each idle state in that list, +the description, ``MWAIT`` hint and exit latency are copied to the corresponding +entry in the final list of idle states. The name of the idle state represented +by it (to be returned by the ``name`` idle state attribute in ``sysfs``) is +"CX_ACPI", where X is the index of that idle state in the final list (note that +the minimum value of X is 1, because 0 is reserved for the "polling" state), and +its target residency is based on the exit latency value. Specifically, for +C1-type idle states the exit latency value is also used as the target residency +(for compatibility with the majority of the "internal" tables of idle states for +various processor models recognized by ``intel_idle``) and for the other idle +state types (C2 and C3) the target residency value is 3 times the exit latency +(again, that is because it reflects the target residency to exit latency ratio +in the majority of cases for the processor models recognized by ``intel_idle``). +All of the idle states in the final list are enabled by default in this case. + + +.. _intel-idle-initialization: + +Initialization +============== + +The initialization of ``intel_idle`` starts with checking if the kernel command +line options forbid the use of the ``MWAIT`` instruction. If that is the case, +an error code is returned right away. + +The next step is to check whether or not the processor model is known to the +driver, which determines the idle states enumeration method (see +`above <intel-idle-enumeration-of-states_>`_), and whether or not the processor +supports ``MWAIT`` (the initialization fails if that is not the case). Then, +the ``MWAIT`` support in the processor is enumerated through ``CPUID`` and the +driver initialization fails if the level of support is not as expected (for +example, if the total number of ``MWAIT`` substates returned is 0). + +Next, if the driver is not configured to ignore the ACPI tables (see +`below <intel-idle-parameters_>`_), the idle states information provided by the +platform firmware is extracted from them. + +Then, ``CPUIdle`` device objects are allocated for all CPUs and the list of +available idle states is created as explained +`above <intel-idle-enumeration-of-states_>`_. + +Finally, ``intel_idle`` is registered with the help of cpuidle_register_driver() +as the ``CPUIdle`` driver for all CPUs in the system and a CPU online callback +for configuring individual CPUs is registered via cpuhp_setup_state(), which +(among other things) causes the callback routine to be invoked for all of the +CPUs present in the system at that time (each CPU executes its own instance of +the callback routine). That routine registers a ``CPUIdle`` device for the CPU +running it (which enables the ``CPUIdle`` subsystem to operate that CPU) and +optionally performs some CPU-specific initialization actions that may be +required for the given processor model. + + +.. _intel-idle-parameters: + +Kernel Command Line Options and Module Parameters +================================================= + +The *x86* architecture support code recognizes three kernel command line +options related to CPU idle time management: ``idle=poll``, ``idle=halt``, +and ``idle=nomwait``. If any of them is present in the kernel command line, the +``MWAIT`` instruction is not allowed to be used, so the initialization of +``intel_idle`` will fail. + +Apart from that there are two module parameters recognized by ``intel_idle`` +itself that can be set via the kernel command line (they cannot be updated via +sysfs, so that is the only way to change their values). + +The ``max_cstate`` parameter value is the maximum idle state index in the list +of idle states supplied to the ``CPUIdle`` core during the registration of the +driver. It is also the maximum number of regular (non-polling) idle states that +can be used by ``intel_idle``, so the enumeration of idle states is terminated +after finding that number of usable idle states (the other idle states that +potentially might have been used if ``max_cstate`` had been greater are not +taken into consideration at all). Setting ``max_cstate`` can prevent +``intel_idle`` from exposing idle states that are regarded as "too deep" for +some reason to the ``CPUIdle`` core, but it does so by making them effectively +invisible until the system is shut down and started again which may not always +be desirable. In practice, it is only really necessary to do that if the idle +states in question cannot be enabled during system startup, because in the +working state of the system the CPU power management quality of service (PM +QoS) feature can be used to prevent ``CPUIdle`` from touching those idle states +even if they have been enumerated (see :ref:`cpu-pm-qos` in :doc:`cpuidle`). +Setting ``max_cstate`` to 0 causes the ``intel_idle`` initialization to fail. + +The ``noacpi`` module parameter (which is recognized by ``intel_idle`` if the +kernel has been configured with ACPI support), can be set to make the driver +ignore the system's ACPI tables entirely (it is unset by default). + + +.. _intel-idle-core-and-package-idle-states: + +Core and Package Levels of Idle States +====================================== + +Typically, in a processor supporting the ``MWAIT`` instruction there are (at +least) two levels of idle states (or C-states). One level, referred to as +"core C-states", covers individual cores in the processor, whereas the other +level, referred to as "package C-states", covers the entire processor package +and it may also involve other components of the system (GPUs, memory +controllers, I/O hubs etc.). + +Some of the ``MWAIT`` hint values allow the processor to use core C-states only +(most importantly, that is the case for the ``MWAIT`` hint value corresponding +to the ``C1`` idle state), but the majority of them give it a license to put +the target core (i.e. the core containing the logical CPU executing ``MWAIT`` +with the given hint value) into a specific core C-state and then (if possible) +to enter a specific package C-state at the deeper level. For example, the +``MWAIT`` hint value representing the ``C3`` idle state allows the processor to +put the target core into the low-power state referred to as "core ``C3``" (or +``CC3``), which happens if all of the logical CPUs (SMT siblings) in that core +have executed ``MWAIT`` with the ``C3`` hint value (or with a hint value +representing a deeper idle state), and in addition to that (in the majority of +cases) it gives the processor a license to put the entire package (possibly +including some non-CPU components such as a GPU or a memory controller) into the +low-power state referred to as "package ``C3``" (or ``PC3``), which happens if +all of the cores have gone into the ``CC3`` state and (possibly) some additional +conditions are satisfied (for instance, if the GPU is covered by ``PC3``, it may +be required to be in a certain GPU-specific low-power state for ``PC3`` to be +reachable). + +As a rule, there is no simple way to make the processor use core C-states only +if the conditions for entering the corresponding package C-states are met, so +the logical CPU executing ``MWAIT`` with a hint value that is not core-level +only (like for ``C1``) must always assume that this may cause the processor to +enter a package C-state. [That is why the exit latency and target residency +values corresponding to the majority of ``MWAIT`` hint values in the "internal" +tables of idle states in ``intel_idle`` reflect the properties of package +C-states.] If using package C-states is not desirable at all, either +:ref:`PM QoS <cpu-pm-qos>` or the ``max_cstate`` module parameter of +``intel_idle`` described `above <intel-idle-parameters_>`_ must be used to +restrict the range of permissible idle states to the ones with core-level only +``MWAIT`` hint values (like ``C1``). + + +References +========== + +.. [1] *Intel® 64 and IA-32 Architectures Software Developer’s Manual Volume 2B*, + https://www.intel.com/content/www/us/en/architecture-and-technology/64-ia-32... + +.. [2] *Advanced Configuration and Power Interface (ACPI) Specification*, + https://uefi.org/specifications diff --git a/Documentation/admin-guide/pm/working-state.rst b/Documentation/admin-guide/pm/working-state.rst index b6cef9b5e961e..ec695d3c77820 100644 --- a/Documentation/admin-guide/pm/working-state.rst +++ b/Documentation/admin-guide/pm/working-state.rst @@ -6,5 +6,6 @@ Working-State Power Management :maxdepth: 2

cpuidle + intel_idle cpufreq intel_pstate

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.3-rc1 commit 210cc5f9db7a5c66b7ca6290b7d35cc7db7e9dbd category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 210cc5f9db7a5c66b7ca6290b7d35cc7db7e9dbd upstream.

Backport summary: Backport to kernel 4.19.57 for ICX uncore support.

The uncore subsystem on Snow Ridge is similar as previous SKX server. The uncore units on Snow Ridge include Ubox, Chabox, IIO, IRP, M2PCIE, PCU, M2M, PCIE3 and IMC.

- The config register encoding and pci device IDs are changed. - For CHA, the umask_ext and filter_tid fields are changed. - For IIO, the ch_mask and fc_mask fields are changed. - For M2M, the mask_ext field is changed. - Add new PCIe3 unit for PCIe3 root port which provides the interface between PCIe devices, plugged into the PCIe port, and the components (in M2IOSF). - IMC can only be accessed via MMIO on Snow Ridge now. Current common code doesn't support it yet. IMC will be supported in following patches. - There are 9 free running counters for IIO CLOCKS and bandwidth In. - Full uncore event list is not published yet. Event constrain is not included in this patch. It will be added later separately.

Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Thomas Gleixner tglx@linutronix.de Cc: acme@kernel.org Cc: eranian@google.com Link: https://lkml.kernel.org/r/1556672028-119221-3-git-send-email-kan.liang@linux... Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Yunying Sun yunying.sun@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/intel/uncore.c | 6 + arch/x86/events/intel/uncore.h | 2 + arch/x86/events/intel/uncore_snbep.c | 403 +++++++++++++++++++++++++++ 3 files changed, 411 insertions(+)

diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c index 8db21bb828d51..85d4364d2167c 100644 --- a/arch/x86/events/intel/uncore.c +++ b/arch/x86/events/intel/uncore.c @@ -1411,6 +1411,11 @@ static const struct intel_uncore_init_fun icl_uncore_init __initconst = { .pci_init = skl_uncore_pci_init, };

+static const struct intel_uncore_init_fun snr_uncore_init __initconst = { + .cpu_init = snr_uncore_cpu_init, + .pci_init = snr_uncore_pci_init, +}; + static const struct x86_cpu_id intel_uncore_match[] __initconst = { X86_UNCORE_MODEL_MATCH(INTEL_FAM6_NEHALEM_EP, nhm_uncore_init), X86_UNCORE_MODEL_MATCH(INTEL_FAM6_NEHALEM, nhm_uncore_init), @@ -1438,6 +1443,7 @@ static const struct x86_cpu_id intel_uncore_match[] __initconst = { X86_UNCORE_MODEL_MATCH(INTEL_FAM6_KABYLAKE_MOBILE, skl_uncore_init), X86_UNCORE_MODEL_MATCH(INTEL_FAM6_KABYLAKE_DESKTOP, skl_uncore_init), X86_UNCORE_MODEL_MATCH(INTEL_FAM6_ICELAKE_MOBILE, icl_uncore_init), + X86_UNCORE_MODEL_MATCH(INTEL_FAM6_ATOM_TREMONT_X, snr_uncore_init), {}, };

diff --git a/arch/x86/events/intel/uncore.h b/arch/x86/events/intel/uncore.h index 3d331a82b8834..652f7aa992a6d 100644 --- a/arch/x86/events/intel/uncore.h +++ b/arch/x86/events/intel/uncore.h @@ -509,6 +509,8 @@ int knl_uncore_pci_init(void); void knl_uncore_cpu_init(void); int skx_uncore_pci_init(void); void skx_uncore_cpu_init(void); +int snr_uncore_pci_init(void); +void snr_uncore_cpu_init(void);

/* uncore_nhmex.c */ void nhmex_uncore_cpu_init(void); diff --git a/arch/x86/events/intel/uncore_snbep.c b/arch/x86/events/intel/uncore_snbep.c index c06074b847fa6..ff4628e81c4b0 100644 --- a/arch/x86/events/intel/uncore_snbep.c +++ b/arch/x86/events/intel/uncore_snbep.c @@ -324,12 +324,64 @@ #define SKX_M2M_PCI_PMON_CTR0 0x200 #define SKX_M2M_PCI_PMON_BOX_CTL 0x258

+/* SNR Ubox */ +#define SNR_U_MSR_PMON_CTR0 0x1f98 +#define SNR_U_MSR_PMON_CTL0 0x1f91 +#define SNR_U_MSR_PMON_UCLK_FIXED_CTL 0x1f93 +#define SNR_U_MSR_PMON_UCLK_FIXED_CTR 0x1f94 + +/* SNR CHA */ +#define SNR_CHA_RAW_EVENT_MASK_EXT 0x3ffffff +#define SNR_CHA_MSR_PMON_CTL0 0x1c01 +#define SNR_CHA_MSR_PMON_CTR0 0x1c08 +#define SNR_CHA_MSR_PMON_BOX_CTL 0x1c00 +#define SNR_C0_MSR_PMON_BOX_FILTER0 0x1c05 + + +/* SNR IIO */ +#define SNR_IIO_MSR_PMON_CTL0 0x1e08 +#define SNR_IIO_MSR_PMON_CTR0 0x1e01 +#define SNR_IIO_MSR_PMON_BOX_CTL 0x1e00 +#define SNR_IIO_MSR_OFFSET 0x10 +#define SNR_IIO_PMON_RAW_EVENT_MASK_EXT 0x7ffff + +/* SNR IRP */ +#define SNR_IRP0_MSR_PMON_CTL0 0x1ea8 +#define SNR_IRP0_MSR_PMON_CTR0 0x1ea1 +#define SNR_IRP0_MSR_PMON_BOX_CTL 0x1ea0 +#define SNR_IRP_MSR_OFFSET 0x10 + +/* SNR M2PCIE */ +#define SNR_M2PCIE_MSR_PMON_CTL0 0x1e58 +#define SNR_M2PCIE_MSR_PMON_CTR0 0x1e51 +#define SNR_M2PCIE_MSR_PMON_BOX_CTL 0x1e50 +#define SNR_M2PCIE_MSR_OFFSET 0x10 + +/* SNR PCU */ +#define SNR_PCU_MSR_PMON_CTL0 0x1ef1 +#define SNR_PCU_MSR_PMON_CTR0 0x1ef8 +#define SNR_PCU_MSR_PMON_BOX_CTL 0x1ef0 +#define SNR_PCU_MSR_PMON_BOX_FILTER 0x1efc + +/* SNR M2M */ +#define SNR_M2M_PCI_PMON_CTL0 0x468 +#define SNR_M2M_PCI_PMON_CTR0 0x440 +#define SNR_M2M_PCI_PMON_BOX_CTL 0x438 +#define SNR_M2M_PCI_PMON_UMASK_EXT 0xff + +/* SNR PCIE3 */ +#define SNR_PCIE3_PCI_PMON_CTL0 0x508 +#define SNR_PCIE3_PCI_PMON_CTR0 0x4e8 +#define SNR_PCIE3_PCI_PMON_BOX_CTL 0x4e4 + DEFINE_UNCORE_FORMAT_ATTR(event, event, "config:0-7"); DEFINE_UNCORE_FORMAT_ATTR(event2, event, "config:0-6"); DEFINE_UNCORE_FORMAT_ATTR(event_ext, event, "config:0-7,21"); DEFINE_UNCORE_FORMAT_ATTR(use_occ_ctr, use_occ_ctr, "config:7"); DEFINE_UNCORE_FORMAT_ATTR(umask, umask, "config:8-15"); DEFINE_UNCORE_FORMAT_ATTR(umask_ext, umask, "config:8-15,32-43,45-55"); +DEFINE_UNCORE_FORMAT_ATTR(umask_ext2, umask, "config:8-15,32-57"); +DEFINE_UNCORE_FORMAT_ATTR(umask_ext3, umask, "config:8-15,32-39"); DEFINE_UNCORE_FORMAT_ATTR(qor, qor, "config:16"); DEFINE_UNCORE_FORMAT_ATTR(edge, edge, "config:18"); DEFINE_UNCORE_FORMAT_ATTR(tid_en, tid_en, "config:19"); @@ -343,11 +395,14 @@ DEFINE_UNCORE_FORMAT_ATTR(occ_invert, occ_invert, "config:30"); DEFINE_UNCORE_FORMAT_ATTR(occ_edge, occ_edge, "config:14-51"); DEFINE_UNCORE_FORMAT_ATTR(occ_edge_det, occ_edge_det, "config:31"); DEFINE_UNCORE_FORMAT_ATTR(ch_mask, ch_mask, "config:36-43"); +DEFINE_UNCORE_FORMAT_ATTR(ch_mask2, ch_mask, "config:36-47"); DEFINE_UNCORE_FORMAT_ATTR(fc_mask, fc_mask, "config:44-46"); +DEFINE_UNCORE_FORMAT_ATTR(fc_mask2, fc_mask, "config:48-50"); DEFINE_UNCORE_FORMAT_ATTR(filter_tid, filter_tid, "config1:0-4"); DEFINE_UNCORE_FORMAT_ATTR(filter_tid2, filter_tid, "config1:0"); DEFINE_UNCORE_FORMAT_ATTR(filter_tid3, filter_tid, "config1:0-5"); DEFINE_UNCORE_FORMAT_ATTR(filter_tid4, filter_tid, "config1:0-8"); +DEFINE_UNCORE_FORMAT_ATTR(filter_tid5, filter_tid, "config1:0-9"); DEFINE_UNCORE_FORMAT_ATTR(filter_cid, filter_cid, "config1:5"); DEFINE_UNCORE_FORMAT_ATTR(filter_link, filter_link, "config1:5-8"); DEFINE_UNCORE_FORMAT_ATTR(filter_link2, filter_link, "config1:6-8"); @@ -3959,3 +4014,351 @@ int skx_uncore_pci_init(void) }

/* end of SKX uncore support */ + +/* SNR uncore support */ + +static struct intel_uncore_type snr_uncore_ubox = { + .name = "ubox", + .num_counters = 2, + .num_boxes = 1, + .perf_ctr_bits = 48, + .fixed_ctr_bits = 48, + .perf_ctr = SNR_U_MSR_PMON_CTR0, + .event_ctl = SNR_U_MSR_PMON_CTL0, + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, + .fixed_ctr = SNR_U_MSR_PMON_UCLK_FIXED_CTR, + .fixed_ctl = SNR_U_MSR_PMON_UCLK_FIXED_CTL, + .ops = &ivbep_uncore_msr_ops, + .format_group = &ivbep_uncore_format_group, +}; + +static struct attribute *snr_uncore_cha_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask_ext2.attr, + &format_attr_edge.attr, + &format_attr_tid_en.attr, + &format_attr_inv.attr, + &format_attr_thresh8.attr, + &format_attr_filter_tid5.attr, + NULL, +}; +static const struct attribute_group snr_uncore_chabox_format_group = { + .name = "format", + .attrs = snr_uncore_cha_formats_attr, +}; + +static int snr_cha_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + + reg1->reg = SNR_C0_MSR_PMON_BOX_FILTER0 + + box->pmu->type->msr_offset * box->pmu->pmu_idx; + reg1->config = event->attr.config1 & SKX_CHA_MSR_PMON_BOX_FILTER_TID; + reg1->idx = 0; + + return 0; +} + +static void snr_cha_enable_event(struct intel_uncore_box *box, + struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + + if (reg1->idx != EXTRA_REG_NONE) + wrmsrl(reg1->reg, reg1->config); + + wrmsrl(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN); +} + +static struct intel_uncore_ops snr_uncore_chabox_ops = { + .init_box = ivbep_uncore_msr_init_box, + .disable_box = snbep_uncore_msr_disable_box, + .enable_box = snbep_uncore_msr_enable_box, + .disable_event = snbep_uncore_msr_disable_event, + .enable_event = snr_cha_enable_event, + .read_counter = uncore_msr_read_counter, + .hw_config = snr_cha_hw_config, +}; + +static struct intel_uncore_type snr_uncore_chabox = { + .name = "cha", + .num_counters = 4, + .num_boxes = 6, + .perf_ctr_bits = 48, + .event_ctl = SNR_CHA_MSR_PMON_CTL0, + .perf_ctr = SNR_CHA_MSR_PMON_CTR0, + .box_ctl = SNR_CHA_MSR_PMON_BOX_CTL, + .msr_offset = HSWEP_CBO_MSR_OFFSET, + .event_mask = HSWEP_S_MSR_PMON_RAW_EVENT_MASK, + .event_mask_ext = SNR_CHA_RAW_EVENT_MASK_EXT, + .ops = &snr_uncore_chabox_ops, + .format_group = &snr_uncore_chabox_format_group, +}; + +static struct attribute *snr_uncore_iio_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh9.attr, + &format_attr_ch_mask2.attr, + &format_attr_fc_mask2.attr, + NULL, +}; + +static const struct attribute_group snr_uncore_iio_format_group = { + .name = "format", + .attrs = snr_uncore_iio_formats_attr, +}; + +static struct intel_uncore_type snr_uncore_iio = { + .name = "iio", + .num_counters = 4, + .num_boxes = 5, + .perf_ctr_bits = 48, + .event_ctl = SNR_IIO_MSR_PMON_CTL0, + .perf_ctr = SNR_IIO_MSR_PMON_CTR0, + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, + .event_mask_ext = SNR_IIO_PMON_RAW_EVENT_MASK_EXT, + .box_ctl = SNR_IIO_MSR_PMON_BOX_CTL, + .msr_offset = SNR_IIO_MSR_OFFSET, + .ops = &ivbep_uncore_msr_ops, + .format_group = &snr_uncore_iio_format_group, +}; + +static struct intel_uncore_type snr_uncore_irp = { + .name = "irp", + .num_counters = 2, + .num_boxes = 5, + .perf_ctr_bits = 48, + .event_ctl = SNR_IRP0_MSR_PMON_CTL0, + .perf_ctr = SNR_IRP0_MSR_PMON_CTR0, + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, + .box_ctl = SNR_IRP0_MSR_PMON_BOX_CTL, + .msr_offset = SNR_IRP_MSR_OFFSET, + .ops = &ivbep_uncore_msr_ops, + .format_group = &ivbep_uncore_format_group, +}; + +static struct intel_uncore_type snr_uncore_m2pcie = { + .name = "m2pcie", + .num_counters = 4, + .num_boxes = 5, + .perf_ctr_bits = 48, + .event_ctl = SNR_M2PCIE_MSR_PMON_CTL0, + .perf_ctr = SNR_M2PCIE_MSR_PMON_CTR0, + .box_ctl = SNR_M2PCIE_MSR_PMON_BOX_CTL, + .msr_offset = SNR_M2PCIE_MSR_OFFSET, + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, + .ops = &ivbep_uncore_msr_ops, + .format_group = &ivbep_uncore_format_group, +}; + +static int snr_pcu_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + int ev_sel = hwc->config & SNBEP_PMON_CTL_EV_SEL_MASK; + + if (ev_sel >= 0xb && ev_sel <= 0xe) { + reg1->reg = SNR_PCU_MSR_PMON_BOX_FILTER; + reg1->idx = ev_sel - 0xb; + reg1->config = event->attr.config1 & (0xff << reg1->idx); + } + return 0; +} + +static struct intel_uncore_ops snr_uncore_pcu_ops = { + IVBEP_UNCORE_MSR_OPS_COMMON_INIT(), + .hw_config = snr_pcu_hw_config, + .get_constraint = snbep_pcu_get_constraint, + .put_constraint = snbep_pcu_put_constraint, +}; + +static struct intel_uncore_type snr_uncore_pcu = { + .name = "pcu", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + .perf_ctr = SNR_PCU_MSR_PMON_CTR0, + .event_ctl = SNR_PCU_MSR_PMON_CTL0, + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, + .box_ctl = SNR_PCU_MSR_PMON_BOX_CTL, + .num_shared_regs = 1, + .ops = &snr_uncore_pcu_ops, + .format_group = &skx_uncore_pcu_format_group, +}; + +enum perf_uncore_snr_iio_freerunning_type_id { + SNR_IIO_MSR_IOCLK, + SNR_IIO_MSR_BW_IN, + + SNR_IIO_FREERUNNING_TYPE_MAX, +}; + +static struct freerunning_counters snr_iio_freerunning[] = { + [SNR_IIO_MSR_IOCLK] = { 0x1eac, 0x1, 0x10, 1, 48 }, + [SNR_IIO_MSR_BW_IN] = { 0x1f00, 0x1, 0x10, 8, 48 }, +}; + +static struct uncore_event_desc snr_uncore_iio_freerunning_events[] = { + /* Free-Running IIO CLOCKS Counter */ + INTEL_UNCORE_EVENT_DESC(ioclk, "event=0xff,umask=0x10"), + /* Free-Running IIO BANDWIDTH IN Counters */ + INTEL_UNCORE_EVENT_DESC(bw_in_port0, "event=0xff,umask=0x20"), + INTEL_UNCORE_EVENT_DESC(bw_in_port0.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(bw_in_port0.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(bw_in_port1, "event=0xff,umask=0x21"), + INTEL_UNCORE_EVENT_DESC(bw_in_port1.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(bw_in_port1.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(bw_in_port2, "event=0xff,umask=0x22"), + INTEL_UNCORE_EVENT_DESC(bw_in_port2.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(bw_in_port2.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(bw_in_port3, "event=0xff,umask=0x23"), + INTEL_UNCORE_EVENT_DESC(bw_in_port3.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(bw_in_port3.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(bw_in_port4, "event=0xff,umask=0x24"), + INTEL_UNCORE_EVENT_DESC(bw_in_port4.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(bw_in_port4.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(bw_in_port5, "event=0xff,umask=0x25"), + INTEL_UNCORE_EVENT_DESC(bw_in_port5.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(bw_in_port5.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(bw_in_port6, "event=0xff,umask=0x26"), + INTEL_UNCORE_EVENT_DESC(bw_in_port6.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(bw_in_port6.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(bw_in_port7, "event=0xff,umask=0x27"), + INTEL_UNCORE_EVENT_DESC(bw_in_port7.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(bw_in_port7.unit, "MiB"), + { /* end: all zeroes */ }, +}; + +static struct intel_uncore_type snr_uncore_iio_free_running = { + .name = "iio_free_running", + .num_counters = 9, + .num_boxes = 5, + .num_freerunning_types = SNR_IIO_FREERUNNING_TYPE_MAX, + .freerunning = snr_iio_freerunning, + .ops = &skx_uncore_iio_freerunning_ops, + .event_descs = snr_uncore_iio_freerunning_events, + .format_group = &skx_uncore_iio_freerunning_format_group, +}; + +static struct intel_uncore_type *snr_msr_uncores[] = { + &snr_uncore_ubox, + &snr_uncore_chabox, + &snr_uncore_iio, + &snr_uncore_irp, + &snr_uncore_m2pcie, + &snr_uncore_pcu, + &snr_uncore_iio_free_running, + NULL, +}; + +void snr_uncore_cpu_init(void) +{ + uncore_msr_uncores = snr_msr_uncores; +} + +static void snr_m2m_uncore_pci_init_box(struct intel_uncore_box *box) +{ + struct pci_dev *pdev = box->pci_dev; + int box_ctl = uncore_pci_box_ctl(box); + + __set_bit(UNCORE_BOX_FLAG_CTL_OFFS8, &box->flags); + pci_write_config_dword(pdev, box_ctl, IVBEP_PMON_BOX_CTL_INT); +} + +static struct intel_uncore_ops snr_m2m_uncore_pci_ops = { + .init_box = snr_m2m_uncore_pci_init_box, + .disable_box = snbep_uncore_pci_disable_box, + .enable_box = snbep_uncore_pci_enable_box, + .disable_event = snbep_uncore_pci_disable_event, + .enable_event = snbep_uncore_pci_enable_event, + .read_counter = snbep_uncore_pci_read_counter, +}; + +static struct attribute *snr_m2m_uncore_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask_ext3.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh8.attr, + NULL, +}; + +static const struct attribute_group snr_m2m_uncore_format_group = { + .name = "format", + .attrs = snr_m2m_uncore_formats_attr, +}; + +static struct intel_uncore_type snr_uncore_m2m = { + .name = "m2m", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + .perf_ctr = SNR_M2M_PCI_PMON_CTR0, + .event_ctl = SNR_M2M_PCI_PMON_CTL0, + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, + .event_mask_ext = SNR_M2M_PCI_PMON_UMASK_EXT, + .box_ctl = SNR_M2M_PCI_PMON_BOX_CTL, + .ops = &snr_m2m_uncore_pci_ops, + .format_group = &snr_m2m_uncore_format_group, +}; + +static struct intel_uncore_type snr_uncore_pcie3 = { + .name = "pcie3", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + .perf_ctr = SNR_PCIE3_PCI_PMON_CTR0, + .event_ctl = SNR_PCIE3_PCI_PMON_CTL0, + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, + .box_ctl = SNR_PCIE3_PCI_PMON_BOX_CTL, + .ops = &ivbep_uncore_pci_ops, + .format_group = &ivbep_uncore_format_group, +}; + +enum { + SNR_PCI_UNCORE_M2M, + SNR_PCI_UNCORE_PCIE3, +}; + +static struct intel_uncore_type *snr_pci_uncores[] = { + [SNR_PCI_UNCORE_M2M] = &snr_uncore_m2m, + [SNR_PCI_UNCORE_PCIE3] = &snr_uncore_pcie3, + NULL, +}; + +static const struct pci_device_id snr_uncore_pci_ids[] = { + { /* M2M */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x344a), + .driver_data = UNCORE_PCI_DEV_FULL_DATA(12, 0, SNR_PCI_UNCORE_M2M, 0), + }, + { /* PCIe3 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x334a), + .driver_data = UNCORE_PCI_DEV_FULL_DATA(4, 0, SNR_PCI_UNCORE_PCIE3, 0), + }, + { /* end: all zeroes */ } +}; + +static struct pci_driver snr_uncore_pci_driver = { + .name = "snr_uncore", + .id_table = snr_uncore_pci_ids, +}; + +int snr_uncore_pci_init(void) +{ + /* SNR UBOX DID */ + int ret = snbep_pci2phy_map_init(0x3460, SKX_CPUNODEID, + SKX_GIDNIDMAP, true); + + if (ret) + return ret; + + uncore_pci_uncores = snr_pci_uncores; + uncore_pci_driver = &snr_uncore_pci_driver; + return 0; +} + +/* end of SNR uncore support */

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.3-rc1 commit 3da04b8a00dd6d39970b9e764b78c5dfb40ec013 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit 3da04b8a00dd6d39970b9e764b78c5dfb40ec013 upstream.

Backport summary: Backport to kernel 4.19.57 for ICX uncore support.

A new MMIO type uncore box is introduced on Snow Ridge server. The counters of MMIO type uncore box can only be accessed by MMIO.

Add a new uncore type, uncore_mmio_uncores, for MMIO type uncore blocks.

Support MMIO type uncore blocks in CPU hot plug. The MMIO space has to be map/unmap for the first/last CPU. The context also need to be migrated if the bind CPU changes.

Add mmio_init() to init and register PMUs for MMIO type uncore blocks.

Add a helper to calculate the box_ctl address.

The helpers which calculate ctl/ctr can be shared with PCI type uncore blocks.

Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Thomas Gleixner tglx@linutronix.de Cc: acme@kernel.org Cc: eranian@google.com Link: https://lkml.kernel.org/r/1556672028-119221-5-git-send-email-kan.liang@linux... Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Yunying Sun yunying.sun@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/intel/uncore.c | 53 +++++++++++++++++++++++++++++----- arch/x86/events/intel/uncore.h | 21 ++++++++++---- 2 files changed, 61 insertions(+), 13 deletions(-)

diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c index 972fb36d794c6..1cf4985341a13 100644 --- a/arch/x86/events/intel/uncore.c +++ b/arch/x86/events/intel/uncore.c @@ -7,6 +7,7 @@ static struct intel_uncore_type *empty_uncore[] = { NULL, }; struct intel_uncore_type **uncore_msr_uncores = empty_uncore; struct intel_uncore_type **uncore_pci_uncores = empty_uncore; +struct intel_uncore_type **uncore_mmio_uncores = empty_uncore;

static bool pcidrv_registered; struct pci_driver *uncore_pci_driver; @@ -1214,12 +1215,14 @@ static int uncore_event_cpu_offline(unsigned int cpu) target = -1;

uncore_change_context(uncore_msr_uncores, cpu, target); + uncore_change_context(uncore_mmio_uncores, cpu, target); uncore_change_context(uncore_pci_uncores, cpu, target);

unref: /* Clear the references */ die = topology_logical_package_id(cpu); uncore_box_unref(uncore_msr_uncores, die); + uncore_box_unref(uncore_mmio_uncores, die); return 0; }

@@ -1288,12 +1291,13 @@ static int uncore_box_ref(struct intel_uncore_type **types,

static int uncore_event_cpu_online(unsigned int cpu) { - int ret, die, target; + int die, target, msr_ret, mmio_ret;

- die = topology_logical_die_id(cpu); - ret = uncore_box_ref(uncore_msr_uncores, die, cpu); - if (ret) - return ret; + die = topology_logical_die_id(cpu); + msr_ret = uncore_box_ref(uncore_msr_uncores, die, cpu); + mmio_ret = uncore_box_ref(uncore_mmio_uncores, die, cpu); + if (msr_ret && mmio_ret) + return -ENOMEM;

/* * Check if there is an online cpu in the package @@ -1305,7 +1309,10 @@ static int uncore_event_cpu_online(unsigned int cpu)

cpumask_set_cpu(cpu, &uncore_cpu_mask);

- uncore_change_context(uncore_msr_uncores, -1, cpu); + if (!msr_ret) + uncore_change_context(uncore_msr_uncores, -1, cpu); + if (!mmio_ret) + uncore_change_context(uncore_mmio_uncores, -1, cpu); uncore_change_context(uncore_pci_uncores, -1, cpu); return 0; } @@ -1353,12 +1360,35 @@ static int __init uncore_cpu_init(void) return ret; }

+static int __init uncore_mmio_init(void) +{ + struct intel_uncore_type **types = uncore_mmio_uncores; + int ret; + + ret = uncore_types_init(types, true); + if (ret) + goto err; + + for (; *types; types++) { + ret = type_pmu_register(*types); + if (ret) + goto err; + } + return 0; +err: + uncore_types_exit(uncore_mmio_uncores); + uncore_mmio_uncores = empty_uncore; + return ret; +} + + #define X86_UNCORE_MODEL_MATCH(model, init) \ { X86_VENDOR_INTEL, 6, model, X86_FEATURE_ANY, (unsigned long)&init }

struct intel_uncore_init_fun { void (*cpu_init)(void); int (*pci_init)(void); + void (*mmio_init)(void); };

static const struct intel_uncore_init_fun nhm_uncore_init __initconst = { @@ -1471,7 +1501,7 @@ static int __init intel_uncore_init(void) { const struct x86_cpu_id *id; struct intel_uncore_init_fun *uncore_init; - int pret = 0, cret = 0, ret; + int pret = 0, cret = 0, mret = 0, ret;

id = x86_match_cpu(intel_uncore_match); if (!id) @@ -1494,7 +1524,12 @@ static int __init intel_uncore_init(void) cret = uncore_cpu_init(); }

- if (cret && pret) + if (uncore_init->mmio_init) { + uncore_init->mmio_init(); + mret = uncore_mmio_init(); + } + + if (cret && pret && mret) return -ENODEV;

/* Install hotplug callbacks to setup the targets for each package */ @@ -1508,6 +1543,7 @@ static int __init intel_uncore_init(void)

err: uncore_types_exit(uncore_msr_uncores); + uncore_types_exit(uncore_mmio_uncores); uncore_pci_exit(); return ret; } @@ -1517,6 +1553,7 @@ static void __exit intel_uncore_exit(void) { cpuhp_remove_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE); uncore_types_exit(uncore_msr_uncores); + uncore_types_exit(uncore_mmio_uncores); uncore_pci_exit(); } module_exit(intel_uncore_exit); diff --git a/arch/x86/events/intel/uncore.h b/arch/x86/events/intel/uncore.h index 652f7aa992a6d..d1fc89c53c106 100644 --- a/arch/x86/events/intel/uncore.h +++ b/arch/x86/events/intel/uncore.h @@ -56,7 +56,10 @@ struct intel_uncore_type { unsigned fixed_ctr; unsigned fixed_ctl; unsigned box_ctl; - unsigned msr_offset; + union { + unsigned msr_offset; + unsigned mmio_offset; + }; unsigned num_shared_regs:8; unsigned single_fixed:1; unsigned pair_ctr_ctl:1; @@ -183,6 +186,13 @@ static inline bool uncore_pmc_freerunning(int idx) return idx == UNCORE_PMC_IDX_FREERUNNING; }

+static inline +unsigned int uncore_mmio_box_ctl(struct intel_uncore_box *box) +{ + return box->pmu->type->box_ctl + + box->pmu->type->mmio_offset * box->pmu->pmu_idx; +} + static inline unsigned uncore_pci_box_ctl(struct intel_uncore_box *box) { return box->pmu->type->box_ctl; @@ -313,7 +323,7 @@ unsigned uncore_msr_perf_ctr(struct intel_uncore_box *box, int idx) static inline unsigned uncore_fixed_ctl(struct intel_uncore_box *box) { - if (box->pci_dev) + if (box->pci_dev || box->io_addr) return uncore_pci_fixed_ctl(box); else return uncore_msr_fixed_ctl(box); @@ -322,7 +332,7 @@ unsigned uncore_fixed_ctl(struct intel_uncore_box *box) static inline unsigned uncore_fixed_ctr(struct intel_uncore_box *box) { - if (box->pci_dev) + if (box->pci_dev || box->io_addr) return uncore_pci_fixed_ctr(box); else return uncore_msr_fixed_ctr(box); @@ -331,7 +341,7 @@ unsigned uncore_fixed_ctr(struct intel_uncore_box *box) static inline unsigned uncore_event_ctl(struct intel_uncore_box *box, int idx) { - if (box->pci_dev) + if (box->pci_dev || box->io_addr) return uncore_pci_event_ctl(box, idx); else return uncore_msr_event_ctl(box, idx); @@ -340,7 +350,7 @@ unsigned uncore_event_ctl(struct intel_uncore_box *box, int idx) static inline unsigned uncore_perf_ctr(struct intel_uncore_box *box, int idx) { - if (box->pci_dev) + if (box->pci_dev || box->io_addr) return uncore_pci_perf_ctr(box, idx); else return uncore_msr_perf_ctr(box, idx); @@ -478,6 +488,7 @@ u64 uncore_shared_reg_config(struct intel_uncore_box *box, int idx);

extern struct intel_uncore_type **uncore_msr_uncores; extern struct intel_uncore_type **uncore_pci_uncores; +extern struct intel_uncore_type **uncore_mmio_uncores; extern struct pci_driver *uncore_pci_driver; extern raw_spinlock_t pci2phy_map_lock; extern struct list_head pci2phy_map_head;

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.3-rc1 commit ee49532b38dd084650bf715eabe7e3828fb8d275 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit ee49532b38dd084650bf715eabe7e3828fb8d275 upstream.

Backport summary: Backport to kernel 4.19.57 for ICX uncore support.

IMC uncore unit can only be accessed via MMIO on Snow Ridge. The MMIO space of IMC uncore is at the specified offsets from the MEM0_BAR. Add snr_uncore_get_mc_dev() to locate the PCI device with MMIO_BASE and MEM0_BAR register.

Add new ops to access the IMC registers via MMIO.

Add 3 new free running counters for clocks, read and write bandwidth.

Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Cc: Linus Torvalds torvalds@linux-foundation.org Cc: Peter Zijlstra peterz@infradead.org Cc: Thomas Gleixner tglx@linutronix.de Cc: acme@kernel.org Cc: eranian@google.com Link: https://lkml.kernel.org/r/1556672028-119221-7-git-send-email-kan.liang@linux... Signed-off-by: Ingo Molnar mingo@kernel.org Signed-off-by: Yunying Sun yunying.sun@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/intel/uncore.c | 3 +- arch/x86/events/intel/uncore.h | 2 + arch/x86/events/intel/uncore_snbep.c | 197 +++++++++++++++++++++++++++ 3 files changed, 201 insertions(+), 1 deletion(-)

diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c index 6d82a7133daf1..5233bac792bbb 100644 --- a/arch/x86/events/intel/uncore.c +++ b/arch/x86/events/intel/uncore.c @@ -28,7 +28,7 @@ struct event_constraint uncore_constraint_empty =

MODULE_LICENSE("GPL");

-static int uncore_pcibus_to_physid(struct pci_bus *bus) +int uncore_pcibus_to_physid(struct pci_bus *bus) { struct pci2phy_map *map; int phys_id = -1; @@ -1477,6 +1477,7 @@ static const struct intel_uncore_init_fun icl_uncore_init __initconst = { static const struct intel_uncore_init_fun snr_uncore_init __initconst = { .cpu_init = snr_uncore_cpu_init, .pci_init = snr_uncore_pci_init, + .mmio_init = snr_uncore_mmio_init, };

static const struct x86_cpu_id intel_uncore_match[] __initconst = { diff --git a/arch/x86/events/intel/uncore.h b/arch/x86/events/intel/uncore.h index 91bfe52da5b68..35b2d2b7767c8 100644 --- a/arch/x86/events/intel/uncore.h +++ b/arch/x86/events/intel/uncore.h @@ -156,6 +156,7 @@ struct pci2phy_map { };

struct pci2phy_map *__find_pci2phy_map(int segment); +int uncore_pcibus_to_physid(struct pci_bus *bus);

ssize_t uncore_event_show(struct device *dev, struct device_attribute *attr, char *buf); @@ -526,6 +527,7 @@ int skx_uncore_pci_init(void); void skx_uncore_cpu_init(void); int snr_uncore_pci_init(void); void snr_uncore_cpu_init(void); +void snr_uncore_mmio_init(void);

/* uncore_nhmex.c */ void nhmex_uncore_cpu_init(void); diff --git a/arch/x86/events/intel/uncore_snbep.c b/arch/x86/events/intel/uncore_snbep.c index ff4628e81c4b0..cf7321f4b8f96 100644 --- a/arch/x86/events/intel/uncore_snbep.c +++ b/arch/x86/events/intel/uncore_snbep.c @@ -374,6 +374,19 @@ #define SNR_PCIE3_PCI_PMON_CTR0 0x4e8 #define SNR_PCIE3_PCI_PMON_BOX_CTL 0x4e4

+/* SNR IMC */ +#define SNR_IMC_MMIO_PMON_FIXED_CTL 0x54 +#define SNR_IMC_MMIO_PMON_FIXED_CTR 0x38 +#define SNR_IMC_MMIO_PMON_CTL0 0x40 +#define SNR_IMC_MMIO_PMON_CTR0 0x8 +#define SNR_IMC_MMIO_PMON_BOX_CTL 0x22800 +#define SNR_IMC_MMIO_OFFSET 0x4000 +#define SNR_IMC_MMIO_SIZE 0x4000 +#define SNR_IMC_MMIO_BASE_OFFSET 0xd0 +#define SNR_IMC_MMIO_BASE_MASK 0x1FFFFFFF +#define SNR_IMC_MMIO_MEM0_OFFSET 0xd8 +#define SNR_IMC_MMIO_MEM0_MASK 0x7FF + DEFINE_UNCORE_FORMAT_ATTR(event, event, "config:0-7"); DEFINE_UNCORE_FORMAT_ATTR(event2, event, "config:0-6"); DEFINE_UNCORE_FORMAT_ATTR(event_ext, event, "config:0-7,21"); @@ -4361,4 +4374,188 @@ int snr_uncore_pci_init(void) return 0; }

+static struct pci_dev *snr_uncore_get_mc_dev(int id) +{ + struct pci_dev *mc_dev = NULL; + int phys_id, pkg; + + while (1) { + mc_dev = pci_get_device(PCI_VENDOR_ID_INTEL, 0x3451, mc_dev); + if (!mc_dev) + break; + phys_id = uncore_pcibus_to_physid(mc_dev->bus); + if (phys_id < 0) + continue; + pkg = topology_phys_to_logical_pkg(phys_id); + if (pkg < 0) + continue; + else if (pkg == id) + break; + } + return mc_dev; +} + +static void snr_uncore_mmio_init_box(struct intel_uncore_box *box) +{ + struct pci_dev *pdev = snr_uncore_get_mc_dev(box->pkgid); + unsigned int box_ctl = uncore_mmio_box_ctl(box); + resource_size_t addr; + u32 pci_dword; + + if (!pdev) + return; + + pci_read_config_dword(pdev, SNR_IMC_MMIO_BASE_OFFSET, &pci_dword); + addr = (pci_dword & SNR_IMC_MMIO_BASE_MASK) << 23; + + pci_read_config_dword(pdev, SNR_IMC_MMIO_MEM0_OFFSET, &pci_dword); + addr |= (pci_dword & SNR_IMC_MMIO_MEM0_MASK) << 12; + + addr += box_ctl; + + box->io_addr = ioremap(addr, SNR_IMC_MMIO_SIZE); + if (!box->io_addr) + return; + + writel(IVBEP_PMON_BOX_CTL_INT, box->io_addr); +} + +static void snr_uncore_mmio_disable_box(struct intel_uncore_box *box) +{ + u32 config; + + if (!box->io_addr) + return; + + config = readl(box->io_addr); + config |= SNBEP_PMON_BOX_CTL_FRZ; + writel(config, box->io_addr); +} + +static void snr_uncore_mmio_enable_box(struct intel_uncore_box *box) +{ + u32 config; + + if (!box->io_addr) + return; + + config = readl(box->io_addr); + config &= ~SNBEP_PMON_BOX_CTL_FRZ; + writel(config, box->io_addr); +} + +static void snr_uncore_mmio_enable_event(struct intel_uncore_box *box, + struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + if (!box->io_addr) + return; + + writel(hwc->config | SNBEP_PMON_CTL_EN, + box->io_addr + hwc->config_base); +} + +static void snr_uncore_mmio_disable_event(struct intel_uncore_box *box, + struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + if (!box->io_addr) + return; + + writel(hwc->config, box->io_addr + hwc->config_base); +} + +static struct intel_uncore_ops snr_uncore_mmio_ops = { + .init_box = snr_uncore_mmio_init_box, + .exit_box = uncore_mmio_exit_box, + .disable_box = snr_uncore_mmio_disable_box, + .enable_box = snr_uncore_mmio_enable_box, + .disable_event = snr_uncore_mmio_disable_event, + .enable_event = snr_uncore_mmio_enable_event, + .read_counter = uncore_mmio_read_counter, +}; + +static struct uncore_event_desc snr_uncore_imc_events[] = { + INTEL_UNCORE_EVENT_DESC(clockticks, "event=0x00,umask=0x00"), + INTEL_UNCORE_EVENT_DESC(cas_count_read, "event=0x04,umask=0x0f"), + INTEL_UNCORE_EVENT_DESC(cas_count_read.scale, "6.103515625e-5"), + INTEL_UNCORE_EVENT_DESC(cas_count_read.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(cas_count_write, "event=0x04,umask=0x30"), + INTEL_UNCORE_EVENT_DESC(cas_count_write.scale, "6.103515625e-5"), + INTEL_UNCORE_EVENT_DESC(cas_count_write.unit, "MiB"), + { /* end: all zeroes */ }, +}; + +static struct intel_uncore_type snr_uncore_imc = { + .name = "imc", + .num_counters = 4, + .num_boxes = 2, + .perf_ctr_bits = 48, + .fixed_ctr_bits = 48, + .fixed_ctr = SNR_IMC_MMIO_PMON_FIXED_CTR, + .fixed_ctl = SNR_IMC_MMIO_PMON_FIXED_CTL, + .event_descs = snr_uncore_imc_events, + .perf_ctr = SNR_IMC_MMIO_PMON_CTR0, + .event_ctl = SNR_IMC_MMIO_PMON_CTL0, + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, + .box_ctl = SNR_IMC_MMIO_PMON_BOX_CTL, + .mmio_offset = SNR_IMC_MMIO_OFFSET, + .ops = &snr_uncore_mmio_ops, + .format_group = &skx_uncore_format_group, +}; + +enum perf_uncore_snr_imc_freerunning_type_id { + SNR_IMC_DCLK, + SNR_IMC_DDR, + + SNR_IMC_FREERUNNING_TYPE_MAX, +}; + +static struct freerunning_counters snr_imc_freerunning[] = { + [SNR_IMC_DCLK] = { 0x22b0, 0x0, 0, 1, 48 }, + [SNR_IMC_DDR] = { 0x2290, 0x8, 0, 2, 48 }, +}; + +static struct uncore_event_desc snr_uncore_imc_freerunning_events[] = { + INTEL_UNCORE_EVENT_DESC(dclk, "event=0xff,umask=0x10"), + + INTEL_UNCORE_EVENT_DESC(read, "event=0xff,umask=0x20"), + INTEL_UNCORE_EVENT_DESC(read.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(read.unit, "MiB"), + INTEL_UNCORE_EVENT_DESC(write, "event=0xff,umask=0x21"), + INTEL_UNCORE_EVENT_DESC(write.scale, "3.814697266e-6"), + INTEL_UNCORE_EVENT_DESC(write.unit, "MiB"), +}; + +static struct intel_uncore_ops snr_uncore_imc_freerunning_ops = { + .init_box = snr_uncore_mmio_init_box, + .exit_box = uncore_mmio_exit_box, + .read_counter = uncore_mmio_read_counter, + .hw_config = uncore_freerunning_hw_config, +}; + +static struct intel_uncore_type snr_uncore_imc_free_running = { + .name = "imc_free_running", + .num_counters = 3, + .num_boxes = 1, + .num_freerunning_types = SNR_IMC_FREERUNNING_TYPE_MAX, + .freerunning = snr_imc_freerunning, + .ops = &snr_uncore_imc_freerunning_ops, + .event_descs = snr_uncore_imc_freerunning_events, + .format_group = &skx_uncore_iio_freerunning_format_group, +}; + +static struct intel_uncore_type *snr_mmio_uncores[] = { + &snr_uncore_imc, + &snr_uncore_imc_free_running, + NULL, +}; + +void snr_uncore_mmio_init(void) +{ + uncore_mmio_uncores = snr_mmio_uncores; +} + /* end of SNR uncore support */

From: Kan Liang kan.liang@linux.intel.com

mainline inclusion from mainline-v5.7-rc1 commit bc88a2fe216a51e8ab46d61f89d0c1b5a400470e category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I47H3V CVE: NA

--------------------------------

commit bc88a2fe216a51e8ab46d61f89d0c1b5a400470e upstream.

Backport summary: Backport to kernel 4.19.57 for ICX uncore support.

The offset between uncore boxes of free-running counters varies, e.g. IIO free-running counters on Ice Lake server.

Add box_offsets, an array of offsets between adjacent uncore boxes.

Signed-off-by: Kan Liang kan.liang@linux.intel.com Signed-off-by: Peter Zijlstra (Intel) peterz@infradead.org Link: https://lkml.kernel.org/r/1584470314-46657-1-git-send-email-kan.liang@linux.... Signed-off-by: Yunying Sun yunying.sun@intel.com Signed-off-by: Jackie Liu liuyun01@kylinos.cn Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com Reviewed-by: Yang Jihong yangjihong1@huawei.com Reviewed-by: Xie XiuQi xiexiuqi@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- arch/x86/events/intel/uncore.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/arch/x86/events/intel/uncore.h b/arch/x86/events/intel/uncore.h index 35b2d2b7767c8..e456dc6359abf 100644 --- a/arch/x86/events/intel/uncore.h +++ b/arch/x86/events/intel/uncore.h @@ -147,6 +147,7 @@ struct freerunning_counters { unsigned int box_offset; unsigned int num_counters; unsigned int bits; + unsigned *box_offsets; };

struct pci2phy_map { @@ -303,7 +304,9 @@ unsigned int uncore_freerunning_counter(struct intel_uncore_box *box,

return pmu->type->freerunning[type].counter_base + pmu->type->freerunning[type].counter_offset * idx + - pmu->type->freerunning[type].box_offset * pmu->pmu_idx; + (pmu->type->freerunning[type].box_offsets ? + pmu->type->freerunning[type].box_offsets[pmu->pmu_idx] : + pmu->type->freerunning[type].box_offset * pmu->pmu_idx); }

static inline