- Kernel - mailweb.openeuler.org

[PATCH OLK-6.6 v3 00/13] Cgroup-Based Interference Statistics
by Pu Lehui 27 May '25

27 May '25

Pu Lehui (5): interference: Add CGROUP_IFS config interference: Add CGROUP_IFS_DEFAULT_ENABLED config cgroup: Export cgroup_init_cftypes as internal interface cgroup: Export cgroup_addrm_files as internal interface openeuler_defconfig: Disable CGROUP_IFS config by default Tengda Wu (5): interference: Add sched rundelay interference track support interference: Add sched sleep interference track support interference: Add sched throttle interference track support interference: Exclude throttle time from rundelay interference: Add sched wakelat interference track support Xu Kuohai (3): interference: Add cgroup-based inteference statistics framework interference: Add smt interference track support interference: Add irq interference track support arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + arch/x86/kernel/smpboot.c | 5 +- drivers/base/arch_topology.c | 4 + include/linux/cgroup-defs.h | 4 + include/linux/cgroup.h | 189 ++++++++++++++ include/linux/sched.h | 9 + init/Kconfig | 23 ++ kernel/cgroup/Makefile | 1 + kernel/cgroup/cgroup-internal.h | 20 ++ kernel/cgroup/cgroup.c | 27 +- kernel/cgroup/ifs.c | 329 +++++++++++++++++++++++++ kernel/sched/core.c | 23 +- kernel/sched/cputime.c | 8 +- kernel/sched/deadline.c | 4 +- kernel/sched/fair.c | 14 +- kernel/sched/rt.c | 4 +- kernel/sched/sched.h | 11 +- kernel/sched/stats.c | 2 + kernel/sched/stats.h | 54 +++- 20 files changed, 704 insertions(+), 29 deletions(-) create mode 100644 kernel/cgroup/ifs.c -- 2.34.1

2 14

[PATCH OLK-6.6 00/11] RDMA/hns: Trace support and bugfixes
by Junxian Huang 27 May '25

27 May '25

From: huangdonghua <huangdonghua3(a)h-partners.com> Trace support and bugfixes Guofeng Yue (1): RDMA/hns: Inappropriate format characters cleanup Junxian Huang (7): RDMA/hns: Add trace for flush CQE RDMA/hns: Add trace for WQE dumping RDMA/hns: Add trace for AEQE dumping RDMA/hns: Add trace for MR/MTR attribute dumping RDMA/hns: Include hnae3.h in hns_roce_hw_v2.h RDMA/hns: Add trace for CMDQ dumping RDMA/hns: Optimize HW performance by limiting ACK request frequency Yuyu Li (2): RDMA/hns: Fix congestion control algorithm parameter range RDMA/hns: Fix scc_param failed logic judgments wenglianfa (1): RDMA/hns: Fix address information leakage of DCA memory drivers/infiniband/hw/hns/Makefile | 1 + drivers/infiniband/hw/hns/hns_roce_ah.c | 1 - drivers/infiniband/hw/hns/hns_roce_dca.c | 3 +- drivers/infiniband/hw/hns/hns_roce_device.h | 22 ++ drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 42 +++- drivers/infiniband/hw/hns/hns_roce_hw_v2.h | 23 +- drivers/infiniband/hw/hns/hns_roce_main.c | 3 +- drivers/infiniband/hw/hns/hns_roce_mr.c | 6 +- drivers/infiniband/hw/hns/hns_roce_qp.c | 4 +- drivers/infiniband/hw/hns/hns_roce_restrack.c | 1 - drivers/infiniband/hw/hns/hns_roce_srq.c | 2 +- drivers/infiniband/hw/hns/hns_roce_sysfs.c | 3 +- drivers/infiniband/hw/hns/hns_roce_trace.h | 213 ++++++++++++++++++ 13 files changed, 306 insertions(+), 18 deletions(-) create mode 100644 drivers/infiniband/hw/hns/hns_roce_trace.h -- 2.33.0

2 12

[PATCH OLK-6.6] arm64/watchdog: fix watchdog failure in low power scenarios
by Bowen You 27 May '25

27 May '25

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/ICAO4O -------------------------------- During testing, it was found that in low-power scenarios, the CPU core frequently powers on and off, which causes the SDEI to fail. This is because the secure timer in the BIOS does not save and restore the state during the CPU core power-on and power-off process. The OS needs to add enable/disable operations in the suspend/resume process. Fixes: 06f3c8d59324 ("watchdog: add nmi_watchdog support for arm64 based on SDEI") Signed-off-by: Bowen You <youbowen2(a)huawei.com> --- arch/arm64/kernel/watchdog_sdei.c | 40 +++++++++++++++++++++++++++++++ 1 file changed, 40 insertions(+) diff --git a/arch/arm64/kernel/watchdog_sdei.c b/arch/arm64/kernel/watchdog_sdei.c index b0302b22dc72..c891ea766684 100644 --- a/arch/arm64/kernel/watchdog_sdei.c +++ b/arch/arm64/kernel/watchdog_sdei.c @@ -16,6 +16,7 @@ #include <linux/arm_sdei.h> #include <linux/kprobes.h> #include <linux/nmi.h> +#include <linux/cpu_pm.h> /* We use the secure physical timer as SDEI NMI watchdog timer */ #define SDEI_NMI_WATCHDOG_HWIRQ 29 @@ -106,6 +107,40 @@ void sdei_watchdog_clear_eoi(void) sdei_api_clear_eoi(SDEI_NMI_WATCHDOG_HWIRQ); } +static int sdei_watchdog_pm_notifier(struct notifier_block *nb, + unsigned long action, void *data) +{ + int rv; + + WARN_ON_ONCE(preemptible()); + + /* + * sdei watchdog not initialized yet, do nothing. + */ + if (sdei_watchdog_event_num == 0) + return NOTIFY_DONE; + + switch (action) { + case CPU_PM_ENTER: + rv = sdei_api_event_disable(sdei_watchdog_event_num); + break; + case CPU_PM_EXIT: + rv = sdei_api_event_enable(sdei_watchdog_event_num); + break; + default: + return NOTIFY_DONE; + } + + if (rv) + return notifier_from_errno(rv); + + return NOTIFY_OK; +} + +static struct notifier_block sdei_watchdog_pm_nb = { + .notifier_call = sdei_watchdog_pm_notifier, +}; + int __init sdei_watchdog_hardlockup_probe(void) { int ret; @@ -144,6 +179,11 @@ int __init sdei_watchdog_hardlockup_probe(void) return ret; } + ret = cpu_pm_register_notifier(&sdei_watchdog_pm_nb); + if (ret) { + pr_warn("Failed to register sdei PM notifier...\n"); + return ret; + } sdei_watchdog_registered = true; pr_info("SDEI Watchdog registered successfully\n"); -- 2.34.1

2 1

[PATCH OLK-6.6 0/9] v4: ucounts: rlimit scalability issues
by Chen Ridong 27 May '25

27 May '25

v4: ucounts: rlimit scalability issues Chen Ridong (4): ucounts: add CONFIG_UCOUNTS_PERCPU_COUNTER ucounts: add input for inc_rlimit_[get_]ucounts_limit ucounts: reinplement rlimit with percpu_counter ucount: fix kabi for ucounts MengEn Sun (1): ucounts: move kfree() out of critical zone protected by ucounts_lock Sebastian Andrzej Siewior (4): rcu: provide a static initializer for hlist_nulls_head ucount: replace get_ucounts_or_wrap() with atomic_inc_not_zero() ucount: use RCU for ucounts lookups ucount: use rcuref_t for reference counting arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/list_nulls.h | 1 + include/linux/user_namespace.h | 44 +++- init/Kconfig | 9 + init/main.c | 1 + ipc/mqueue.c | 3 +- kernel/signal.c | 2 +- kernel/ucount.c | 268 +++++++++++++++++++------ mm/mlock.c | 2 +- 10 files changed, 268 insertions(+), 64 deletions(-) -- 2.34.1

2 10

[PATCH OLK-6.6] tpm: Add a driver for Loongson TPM device
by Hongchen Zhang 27 May '25

27 May '25

From: Qunqin Zhao <zhaoqunqin(a)loongson.cn> LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/ICATLF CVE: NA -------------------------------- Loongson security engine supports random number generation, hash, symmetric encryption and asymmetric encryption. Based on these encryption functions, TPM2 have been implemented in the Loongson security engine firmware. This driver is responsible for copying data into the memory visible to the firmware and receiving data from the firmware. Co-developed-by: Yinggang Gu <guyinggang(a)loongson.cn> Signed-off-by: Yinggang Gu <guyinggang(a)loongson.cn> Signed-off-by: Qunqin Zhao <zhaoqunqin(a)loongson.cn> Change-Id: I029740a822afd797d211a868b36e5ba8dd3b6b6a --- MAINTAINERS | 5 ++ drivers/char/Kconfig | 1 + drivers/char/loongson_se.c | 33 ++++++---- drivers/char/tpm/Kconfig | 9 +++ drivers/char/tpm/Makefile | 1 + drivers/char/tpm/tpm_loongson.c | 103 ++++++++++++++++++++++++++++++++ include/soc/loongson/se.h | 1 + 7 files changed, 143 insertions(+), 10 deletions(-) create mode 100644 drivers/char/tpm/tpm_loongson.c diff --git a/MAINTAINERS b/MAINTAINERS index c6a3ac61989c..546459e25d4a 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -7337,6 +7337,11 @@ F: Documentation/networking/net_dim.rst F: include/linux/dim.h F: lib/dim/ +LOONGSON TPM DRIVER +M: Qunqin Zhao <zhaoqunqin(a)loongson.cn> +S: Maintained +F: drivers/char/tpm/tpm_loongson.c + DYNAMIC THERMAL POWER MANAGEMENT (DTPM) M: Daniel Lezcano <daniel.lezcano(a)kernel.org> L: linux-pm(a)vger.kernel.org diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index ea7ace87a9df..1d2957bc100d 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -394,6 +394,7 @@ config UV_MMTIMER config LOONGSON_SE tristate "LOONGSON SECURITY MODULE Interface" depends on LOONGARCH + select MFD_CORE default m help If you have LOONGSON security module (SE) support say Yes and it diff --git a/drivers/char/loongson_se.c b/drivers/char/loongson_se.c index 853f7e10cd41..c9bd51adc792 100644 --- a/drivers/char/loongson_se.c +++ b/drivers/char/loongson_se.c @@ -14,6 +14,7 @@ #include <linux/iopoll.h> #include <linux/kernel.h> #include <linux/module.h> +#include <linux/mfd/core.h> #include <linux/platform_device.h> #include <soc/loongson/se.h> @@ -309,19 +310,25 @@ int se_send_ch_requeset(struct lsse_ch *ch) { struct loongson_se *se; u32 status, int_bit; + int err; se = ch->se; int_bit = ch->int_bit; + mutex_lock(&se->ch_init_lock); if ((se_readl(se, SE_L2SINT_STAT) & int_bit) || - !(se_readl(se, SE_L2SINT_EN) & int_bit)) + !(se_readl(se, SE_L2SINT_EN) & int_bit)) { + mutex_unlock(&se->ch_init_lock); return -EBUSY; + } se_enable_int(se, int_bit); se_writel(se, int_bit, SE_L2SINT_SET); - return readl_relaxed_poll_timeout_atomic(se->base + SE_L2SINT_STAT, status, + err = readl_relaxed_poll_timeout_atomic(se->base + SE_L2SINT_STAT, status, !(status & int_bit), 10, 10000); + mutex_unlock(&se->ch_init_lock); + return err; } EXPORT_SYMBOL_GPL(se_send_ch_requeset); @@ -337,10 +344,10 @@ struct lsse_ch *se_init_ch(struct device *dev, int id, int data_size, int msg_si { struct loongson_se *se = dev_get_drvdata(dev); struct lsse_ch *ch; - unsigned long flag; int data_first, data_nr; int msg_first, msg_nr; + mutex_lock(&se->ch_init_lock); if (!se) { pr_err("SE has bot been initialized\n"); return NULL; @@ -356,8 +363,6 @@ struct lsse_ch *se_init_ch(struct device *dev, int id, int data_size, int msg_si return NULL; } - spin_lock_irqsave(&se->dev_lock, flag); - ch = &se->chs[id]; ch->se = se; ch->id = id; @@ -369,7 +374,7 @@ struct lsse_ch *se_init_ch(struct device *dev, int id, int data_size, int msg_si 0, data_nr, 0); if (data_first >= se->mem_map_pages) { dev_err(se->dev, "Insufficient memory space\n"); - spin_unlock_irqrestore(&se->dev_lock, flag); + mutex_unlock(&se->ch_init_lock); return NULL; } @@ -384,7 +389,7 @@ struct lsse_ch *se_init_ch(struct device *dev, int id, int data_size, int msg_si if (msg_first >= se->mem_map_pages) { dev_err(se->dev, "Insufficient memory space\n"); bitmap_clear(se->mem_map, data_first, data_nr); - spin_unlock_irqrestore(&se->dev_lock, flag); + mutex_unlock(&se->ch_init_lock); return NULL; } @@ -396,14 +401,15 @@ struct lsse_ch *se_init_ch(struct device *dev, int id, int data_size, int msg_si ch->priv = priv; spin_lock_init(&ch->ch_lock); - spin_unlock_irqrestore(&se->dev_lock, flag); if (loongson_se_set_msg(ch)) { dev_err(se->dev, "Channel %d setup message address failed\n", id); + mutex_unlock(&se->ch_init_lock); return NULL; } se_enable_int(se, ch->int_bit); + mutex_unlock(&se->ch_init_lock); return ch; } @@ -448,6 +454,10 @@ void se_deinit_ch(struct lsse_ch *ch) } EXPORT_SYMBOL_GPL(se_deinit_ch); +static const struct mfd_cell engines[] = { + { .name = "loongson-tpm" }, +}; + static int loongson_se_probe(struct platform_device *pdev) { struct loongson_se *se; @@ -460,6 +470,7 @@ static int loongson_se_probe(struct platform_device *pdev) se->dev = dev; dev_set_drvdata(dev, se); init_completion(&se->cmd_completion); + mutex_init(&se->ch_init_lock); spin_lock_init(&se->cmd_lock); spin_lock_init(&se->dev_lock); /* Setup DMA buffer */ @@ -497,10 +508,12 @@ static int loongson_se_probe(struct platform_device *pdev) } err = se_init_hw(se, se->mem_addr, size); - if (err) + if (err) { se_disable_hw(se); + return err; + } - return err; + return devm_mfd_add_devices(dev, 0, engines, ARRAY_SIZE(engines), NULL, 0, NULL); } static const struct acpi_device_id loongson_se_acpi_match[] = { diff --git a/drivers/char/tpm/Kconfig b/drivers/char/tpm/Kconfig index 301284e07603..3524b2e6086e 100644 --- a/drivers/char/tpm/Kconfig +++ b/drivers/char/tpm/Kconfig @@ -174,6 +174,15 @@ config TCG_IBMVTPM will be accessible from within Linux. To compile this driver as a module, choose M here; the module will be called tpm_ibmvtpm. +config TCG_LOONGSON + tristate "Loongson TPM Interface" + depends on LOONGSON_SE + help + If you want to make Loongson TPM support available, say Yes and + it will be accessible from within Linux. To compile this + driver as a module, choose M here; the module will be called + tpm_loongson. + config TCG_XEN tristate "XEN TPM Interface" depends on TCG_TPM && XEN diff --git a/drivers/char/tpm/Makefile b/drivers/char/tpm/Makefile index 8f868c9b9ce7..ea2919ddbd6e 100644 --- a/drivers/char/tpm/Makefile +++ b/drivers/char/tpm/Makefile @@ -44,3 +44,4 @@ obj-$(CONFIG_TCG_VTPM_PROXY) += tpm_vtpm_proxy.o obj-$(CONFIG_TCG_FTPM_TEE) += tpm_ftpm_tee.o obj-$(CONFIG_TCG_HYGON) += tpm_hygon.o obj-$(CONFIG_TCM_HYGON) += tcm_hygon.o +obj-$(CONFIG_TCG_LOONGSON) += tpm_loongson.o diff --git a/drivers/char/tpm/tpm_loongson.c b/drivers/char/tpm/tpm_loongson.c new file mode 100644 index 000000000000..3c27d86b77ba --- /dev/null +++ b/drivers/char/tpm/tpm_loongson.c @@ -0,0 +1,103 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2025 Loongson Technology Corporation Limited. */ + +#include <linux/device.h> +#include <soc/loongson/se.h> +#include <linux/platform_device.h> +#include <linux/wait.h> + +#include "tpm.h" + +struct tpm_loongson_msg { + u32 cmd; + u32 data_off; + u32 data_len; + u32 info[5]; +}; + +struct tpm_loongson_dev { + struct lsse_ch *se_ch; + struct completion tpm_loongson_completion; +}; + +static void tpm_loongson_complete(struct lsse_ch *ch) +{ + struct tpm_loongson_dev *td = ch->priv; + + complete(&td->tpm_loongson_completion); +} + +static int tpm_loongson_recv(struct tpm_chip *chip, u8 *buf, size_t count) +{ + struct tpm_loongson_dev *td = dev_get_drvdata(&chip->dev); + struct tpm_loongson_msg *rmsg; + int sig; + + sig = wait_for_completion_interruptible(&td->tpm_loongson_completion); + if (sig) + return sig; + + rmsg = td->se_ch->rmsg; + memcpy(buf, td->se_ch->data_buffer, rmsg->data_len); + + return rmsg->data_len; +} + +static int tpm_loongson_send(struct tpm_chip *chip, u8 *buf, size_t count) +{ + struct tpm_loongson_dev *td = dev_get_drvdata(&chip->dev); + struct tpm_loongson_msg *smsg = td->se_ch->smsg; + + memcpy(td->se_ch->data_buffer, buf, count); + smsg->data_len = count; + + return se_send_ch_requeset(td->se_ch); +} + +static const struct tpm_class_ops tpm_loongson_ops = { + .flags = TPM_OPS_AUTO_STARTUP, + .recv = tpm_loongson_recv, + .send = tpm_loongson_send, +}; + +static int tpm_loongson_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct tpm_loongson_msg *smsg; + struct tpm_loongson_dev *td; + struct tpm_chip *chip; + + td = devm_kzalloc(dev, sizeof(struct tpm_loongson_dev), GFP_KERNEL); + if (!td) + return -ENOMEM; + + init_completion(&td->tpm_loongson_completion); + td->se_ch = se_init_ch(dev->parent, SE_CH_TPM, PAGE_SIZE, + 2 * sizeof(struct tpm_loongson_msg), td, + tpm_loongson_complete); + if (!td->se_ch) + return -ENODEV; + smsg = td->se_ch->smsg; + smsg->cmd = SE_CMD_TPM; + smsg->data_off = td->se_ch->data_buffer - td->se_ch->se->mem_base; + + chip = tpmm_chip_alloc(dev, &tpm_loongson_ops); + if (IS_ERR(chip)) + return PTR_ERR(chip); + chip->flags = TPM_CHIP_FLAG_TPM2 | TPM_CHIP_FLAG_IRQ; + dev_set_drvdata(&chip->dev, td); + + return tpm_chip_register(chip); +} + +static struct platform_driver tpm_loongson_driver = { + .probe = tpm_loongson_probe, + .driver = { + .name = "loongson-tpm", + }, +}; +module_platform_driver(tpm_loongson_driver); + +MODULE_ALIAS("platform:loongson-tpm"); +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("Loongson TPM driver"); diff --git a/include/soc/loongson/se.h b/include/soc/loongson/se.h index d9864e3f3b85..22872f7d1a19 100644 --- a/include/soc/loongson/se.h +++ b/include/soc/loongson/se.h @@ -109,6 +109,7 @@ struct loongson_se { void __iomem *base; u32 version; u32 ch_status; + struct mutex ch_init_lock; spinlock_t cmd_lock; spinlock_t dev_lock; -- 2.43.0

2 1

[PATCH OLK-6.6 0/9] v4: ucounts: rlimit scalability issues
by Chen Ridong 27 May '25

27 May '25

v4: ucounts: rlimit scalability issues Chen Ridong (4): ucounts: add CONFIG_UCOUNTS_PERCPU_COUNTER ucounts: add input for inc_rlimit_[get_]ucounts_limit ucounts: reinplement rlimit with percpu_counter ucount: fix kabi for ucounts MengEn Sun (1): ucounts: move kfree() out of critical zone protected by ucounts_lock Sebastian Andrzej Siewior (4): rcu: provide a static initializer for hlist_nulls_head ucount: replace get_ucounts_or_wrap() with atomic_inc_not_zero() ucount: use RCU for ucounts lookups ucount: use rcuref_t for reference counting include/linux/list_nulls.h | 1 + include/linux/user_namespace.h | 44 +++++- init/Kconfig | 9 ++ init/main.c | 1 + ipc/mqueue.c | 3 +- kernel/signal.c | 2 +- kernel/ucount.c | 268 ++++++++++++++++++++++++++------- mm/mlock.c | 2 +- 8 files changed, 266 insertions(+), 64 deletions(-) -- 2.34.1

2 10

[PATCH OLK-6.6 v2 00/13] Cgroup-Based Interference Statistics
by Pu Lehui 27 May '25

27 May '25

Pu Lehui (5): interference: Add CGROUP_IFS config interference: Add CGROUP_IFS_DEFAULT_ENABLED config cgroup: Export cgroup_init_cftypes as internal interface cgroup: Export cgroup_addrm_files as internal interface openeuler_defconfig: Disable IFS configs by default Tengda Wu (5): interference: Add sched rundelay interference track support interference: Add sched sleep interference track support interference: Add sched throttle interference track support interference: Exclude throttle time from rundelay interference: Add sched wakelat interference track support Xu Kuohai (3): interference: Add cgroup-based inteference statistics framework interference: Add smt interference track support interference: Add irq interference track support arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + arch/x86/kernel/smpboot.c | 5 +- drivers/base/arch_topology.c | 4 + include/linux/cgroup-defs.h | 4 + include/linux/cgroup.h | 189 ++++++++++++++ include/linux/sched.h | 9 + init/Kconfig | 23 ++ kernel/cgroup/Makefile | 1 + kernel/cgroup/cgroup-internal.h | 20 ++ kernel/cgroup/cgroup.c | 27 +- kernel/cgroup/ifs.c | 329 +++++++++++++++++++++++++ kernel/sched/core.c | 23 +- kernel/sched/cputime.c | 8 +- kernel/sched/deadline.c | 4 +- kernel/sched/fair.c | 14 +- kernel/sched/rt.c | 4 +- kernel/sched/sched.h | 11 +- kernel/sched/stats.c | 2 + kernel/sched/stats.h | 54 +++- 20 files changed, 706 insertions(+), 29 deletions(-) create mode 100644 kernel/cgroup/ifs.c -- 2.34.1

2 14

[PATCH OLK-6.6 00/12] Cgroup-Based Interference Statistics
by Pu Lehui 27 May '25

27 May '25

Pu Lehui (4): interference: Add CGROUP_IFS config interference: Add CGROUP_IFS_DEFAULT_ENABLED config cgroup: Export cgroup_init_cftypes as internal interface cgroup: Export cgroup_addrm_files as internal interface Tengda Wu (5): interference: Add sched rundelay interference track support interference: Add sched sleep interference track support interference: Add sched throttle interference track support interference: Exclude throttle time from rundelay interference: Add sched wakelat interference track support Xu Kuohai (3): interference: Add cgroup-based inteference statistics framework interference: Add smt interference track support interference: Add irq interference track support arch/x86/kernel/smpboot.c | 5 +- drivers/base/arch_topology.c | 4 + include/linux/cgroup-defs.h | 4 + include/linux/cgroup.h | 189 ++++++++++++++++++ include/linux/sched.h | 9 + init/Kconfig | 23 +++ kernel/cgroup/Makefile | 1 + kernel/cgroup/cgroup-internal.h | 20 ++ kernel/cgroup/cgroup.c | 27 ++- kernel/cgroup/ifs.c | 329 ++++++++++++++++++++++++++++++++ kernel/sched/core.c | 23 ++- kernel/sched/cputime.c | 8 +- kernel/sched/deadline.c | 4 +- kernel/sched/fair.c | 14 +- kernel/sched/rt.c | 4 +- kernel/sched/sched.h | 11 +- kernel/sched/stats.c | 2 + kernel/sched/stats.h | 54 +++++- 18 files changed, 702 insertions(+), 29 deletions(-) create mode 100644 kernel/cgroup/ifs.c -- 2.34.1

2 13

[PATCH OLK-6.6] fs/jfs: Prevent integer overflow in AG size calculation
by Zhang Kunbo 27 May '25

27 May '25

From: Rand Deeb <rand.sec96(a)gmail.com> stable inclusion from stable-v6.6.88 commit 7ccf3b35274512b60ecb614e0637e76bd6f2d829 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IC6PV5 CVE: CVE-2025-37858 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit 7fcbf789629cdb9fbf4e2172ce31136cfed11e5e ] The JFS filesystem calculates allocation group (AG) size using 1 << l2agsize in dbExtendFS(). When l2agsize exceeds 31 (possible with >2TB aggregates on 32-bit systems), this 32-bit shift operation causes undefined behavior and improper AG sizing. On 32-bit architectures: - Left-shifting 1 by 32+ bits results in 0 due to integer overflow - This creates invalid AG sizes (0 or garbage values) in sbi->bmap->db_agsize - Subsequent block allocations would reference invalid AG structures - Could lead to: - Filesystem corruption during extend operations - Kernel crashes due to invalid memory accesses - Security vulnerabilities via malformed on-disk structures Fix by casting to s64 before shifting: bmp->db_agsize = (s64)1 << l2agsize; This ensures 64-bit arithmetic even on 32-bit architectures. The cast matches the data type of db_agsize (s64) and follows similar patterns in JFS block calculation code. Found by Linux Verification Center (linuxtesting.org) with SVACE. Signed-off-by: Rand Deeb <rand.sec96(a)gmail.com> Signed-off-by: Dave Kleikamp <dave.kleikamp(a)oracle.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Zhang Kunbo <zhangkunbo(a)huawei.com> --- fs/jfs/jfs_dmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/jfs/jfs_dmap.c b/fs/jfs/jfs_dmap.c index f9009e4f9ffd..edb22cf9521a 100644 --- a/fs/jfs/jfs_dmap.c +++ b/fs/jfs/jfs_dmap.c @@ -3403,7 +3403,7 @@ int dbExtendFS(struct inode *ipbmap, s64 blkno, s64 nblocks) oldl2agsize = bmp->db_agl2size; bmp->db_agl2size = l2agsize; - bmp->db_agsize = 1 << l2agsize; + bmp->db_agsize = (s64)1 << l2agsize; /* compute new number of AG */ agno = bmp->db_numag; -- 2.34.1

2 1

[OLK-6.6 0/8] bugfix for hisilicon accelerator driver
by Weili Qian 27 May '25

27 May '25

Chenghai Huang (1): uacce: fix unsigned long print format Weili Qian (7): uacce: fix isolate sysfs check condition uacce: delete some redundant code crypto: hisilicon/qm - fix vfs_num error crypto: hisilicon/qm - alloc power of 2 interrupts crypto: hisilicon - enable error reporting again crypto: hisilicon/trng - use DEFINE_MUTEX() for mutex lock crypto: hisilicon/zip - do not expose hashagg algorithm when uacce mode is 2 drivers/crypto/hisilicon/hpre/hpre_main.c | 11 +++++ drivers/crypto/hisilicon/qm.c | 60 ++++++++++++++++------- drivers/crypto/hisilicon/sec2/sec_main.c | 11 +++++ drivers/crypto/hisilicon/trng/trng.c | 38 +++++++------- drivers/crypto/hisilicon/zip/dae_main.c | 10 +++- drivers/crypto/hisilicon/zip/zip_main.c | 11 +++++ drivers/misc/uacce/uacce.c | 10 ++-- 7 files changed, 107 insertions(+), 44 deletions(-) -- 2.33.0

1 8