- Linuxarm - mailweb.openeuler.org

Re: [PATCH v11 12/13] vfio/pci: Register a DMA fault response region
by Shameerali Kolothum Thodi 18 Feb '21

18 Feb '21

Hi Eric, > > -----Original Message----- > > From: Eric Auger [mailto:eric.auger@redhat.com] > > Sent: 16 November 2020 11:00 > > To: eric.auger.pro(a)gmail.com; eric.auger(a)redhat.com; > > iommu(a)lists.linux-foundation.org; linux-kernel(a)vger.kernel.org; > > kvm(a)vger.kernel.org; kvmarm(a)lists.cs.columbia.edu; will(a)kernel.org; > > joro(a)8bytes.org; maz(a)kernel.org; robin.murphy(a)arm.com; > > alex.williamson(a)redhat.com > > Cc: jean-philippe(a)linaro.org; zhangfei.gao(a)linaro.org; > > zhangfei.gao(a)gmail.com; vivek.gautam(a)arm.com; Shameerali Kolothum > > Thodi <shameerali.kolothum.thodi(a)huawei.com>; > > jacob.jun.pan(a)linux.intel.com; yi.l.liu(a)intel.com; tn(a)semihalf.com; > > nicoleotsuka(a)gmail.com; yuzenghui <yuzenghui(a)huawei.com> > > Subject: [PATCH v11 12/13] vfio/pci: Register a DMA fault response > > region > > > > In preparation for vSVA, let's register a DMA fault response region, > > where the userspace will push the page responses and increment the > > head of the buffer. The kernel will pop those responses and inject > > them on iommu side. > > > > Signed-off-by: Eric Auger <eric.auger(a)redhat.com> > > --- > > drivers/vfio/pci/vfio_pci.c | 114 +++++++++++++++++++++++++--- > > drivers/vfio/pci/vfio_pci_private.h | 5 ++ > > drivers/vfio/pci/vfio_pci_rdwr.c | 39 ++++++++++ > > include/uapi/linux/vfio.h | 32 ++++++++ > > 4 files changed, 181 insertions(+), 9 deletions(-) > > > > diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c > > index 65a83fd0e8c0..e9a904ce3f0d 100644 > > --- a/drivers/vfio/pci/vfio_pci.c > > +++ b/drivers/vfio/pci/vfio_pci.c > > @@ -318,9 +318,20 @@ static void vfio_pci_dma_fault_release(struct > > vfio_pci_device *vdev, > > kfree(vdev->fault_pages); > > } > > > > -static int vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev, > > - struct vfio_pci_region *region, > > - struct vm_area_struct *vma) > > +static void > > +vfio_pci_dma_fault_response_release(struct vfio_pci_device *vdev, > > + struct vfio_pci_region *region) { > > + if (vdev->dma_fault_response_wq) > > + destroy_workqueue(vdev->dma_fault_response_wq); > > + kfree(vdev->fault_response_pages); > > + vdev->fault_response_pages = NULL; > > +} > > + > > +static int __vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev, > > + struct vfio_pci_region *region, > > + struct vm_area_struct *vma, > > + u8 *pages) > > { > > u64 phys_len, req_len, pgoff, req_start; > > unsigned long long addr; > > @@ -333,14 +344,14 @@ static int vfio_pci_dma_fault_mmap(struct > > vfio_pci_device *vdev, > > ((1U << (VFIO_PCI_OFFSET_SHIFT - PAGE_SHIFT)) - 1); > > req_start = pgoff << PAGE_SHIFT; > > > > - /* only the second page of the producer fault region is mmappable */ > > + /* only the second page of the fault region is mmappable */ > > if (req_start < PAGE_SIZE) > > return -EINVAL; > > > > if (req_start + req_len > phys_len) > > return -EINVAL; > > > > - addr = virt_to_phys(vdev->fault_pages); > > + addr = virt_to_phys(pages); > > vma->vm_private_data = vdev; > > vma->vm_pgoff = (addr >> PAGE_SHIFT) + pgoff; > > > > @@ -349,13 +360,29 @@ static int vfio_pci_dma_fault_mmap(struct > > vfio_pci_device *vdev, > > return ret; > > } > > > > -static int vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev, > > - struct vfio_pci_region *region, > > - struct vfio_info_cap *caps) > > +static int vfio_pci_dma_fault_mmap(struct vfio_pci_device *vdev, > > + struct vfio_pci_region *region, > > + struct vm_area_struct *vma) > > +{ > > + return __vfio_pci_dma_fault_mmap(vdev, region, vma, > > vdev->fault_pages); > > +} > > + > > +static int > > +vfio_pci_dma_fault_response_mmap(struct vfio_pci_device *vdev, > > + struct vfio_pci_region *region, > > + struct vm_area_struct *vma) > > +{ > > + return __vfio_pci_dma_fault_mmap(vdev, region, vma, > > vdev->fault_response_pages); > > +} > > + > > +static int __vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev, > > + struct vfio_pci_region *region, > > + struct vfio_info_cap *caps, > > + u32 cap_id) > > { > > struct vfio_region_info_cap_sparse_mmap *sparse = NULL; > > struct vfio_region_info_cap_fault cap = { > > - .header.id = VFIO_REGION_INFO_CAP_DMA_FAULT, > > + .header.id = cap_id, > > .header.version = 1, > > .version = 1, > > }; > > @@ -383,6 +410,14 @@ static int > > vfio_pci_dma_fault_add_capability(struct > > vfio_pci_device *vdev, > > return ret; > > } > > > > +static int vfio_pci_dma_fault_add_capability(struct vfio_pci_device *vdev, > > + struct vfio_pci_region *region, > > + struct vfio_info_cap *caps) { > > + return __vfio_pci_dma_fault_add_capability(vdev, region, caps, > > + VFIO_REGION_INFO_CAP_DMA_FAULT); } > > + > > static const struct vfio_pci_regops vfio_pci_dma_fault_regops = { > > .rw = vfio_pci_dma_fault_rw, > > .release = vfio_pci_dma_fault_release, > > @@ -390,6 +425,13 @@ static const struct vfio_pci_regops > > vfio_pci_dma_fault_regops = { > > .add_capability = vfio_pci_dma_fault_add_capability, > > }; > > > > +static const struct vfio_pci_regops vfio_pci_dma_fault_response_regops = { > > + .rw = vfio_pci_dma_fault_response_rw, > > + .release = vfio_pci_dma_fault_response_release, > > + .mmap = vfio_pci_dma_fault_response_mmap, > > + .add_capability = vfio_pci_dma_fault_add_capability, As I mentioned in the Qemu patch ([RFC v7 26/26] vfio/pci: Implement return_page_response page response callback), it looks like we are using the VFIO_REGION_INFO_CAP_DMA_FAULT cap id for the dma_fault_response here as well. Is that intentional? (Was wondering how it worked in the first place and noted this). Please check. Thanks, Shameer

2 1

Re: [PATCH v2] iommu: Check dev->iommu in iommu_dev_xxx functions
by Shameerali Kolothum Thodi 16 Feb '21

16 Feb '21

Hi Robin/Joerg, > -----Original Message----- > From: Shameer Kolothum [mailto:shameerali.kolothum.thodi@huawei.com] > Sent: 01 February 2021 12:41 > To: linux-kernel(a)vger.kernel.org; iommu(a)lists.linux-foundation.org > Cc: joro(a)8bytes.org; robin.murphy(a)arm.com; jean-philippe(a)linaro.org; > will(a)kernel.org; Zengtao (B) <prime.zeng(a)hisilicon.com>; > linuxarm(a)openeuler.org > Subject: [Linuxarm] [PATCH v2] iommu: Check dev->iommu in iommu_dev_xxx > functions > > The device iommu probe/attach might have failed leaving dev->iommu > to NULL and device drivers may still invoke these functions resulting > in a crash in iommu vendor driver code. Hence make sure we check that. > > Also added iommu_ops to the "struct dev_iommu" and set it if the dev > is successfully associated with an iommu. > > Fixes: a3a195929d40 ("iommu: Add APIs for multiple domains per device") > Signed-off-by: Shameer Kolothum <shameerali.kolothum.thodi(a)huawei.com> > --- > v1 --> v2: > -Added iommu_ops to struct dev_iommu based on the discussion with Robin. > -Rebased against iommu-tree core branch. A gentle ping on this... Thanks, Shameer > --- > drivers/iommu/iommu.c | 19 +++++++------------ > include/linux/iommu.h | 2 ++ > 2 files changed, 9 insertions(+), 12 deletions(-) > > diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c > index fd76e2f579fe..6023d0b7c542 100644 > --- a/drivers/iommu/iommu.c > +++ b/drivers/iommu/iommu.c > @@ -217,6 +217,7 @@ static int __iommu_probe_device(struct device *dev, > struct list_head *group_list > } > > dev->iommu->iommu_dev = iommu_dev; > + dev->iommu->ops = iommu_dev->ops; > > group = iommu_group_get_for_dev(dev); > if (IS_ERR(group)) { > @@ -2865,10 +2866,8 @@ EXPORT_SYMBOL_GPL(iommu_fwspec_add_ids); > */ > int iommu_dev_enable_feature(struct device *dev, enum > iommu_dev_features feat) > { > - const struct iommu_ops *ops = dev->bus->iommu_ops; > - > - if (ops && ops->dev_enable_feat) > - return ops->dev_enable_feat(dev, feat); > + if (dev->iommu && dev->iommu->ops->dev_enable_feat) > + return dev->iommu->ops->dev_enable_feat(dev, feat); > > return -ENODEV; > } > @@ -2881,10 +2880,8 @@ > EXPORT_SYMBOL_GPL(iommu_dev_enable_feature); > */ > int iommu_dev_disable_feature(struct device *dev, enum > iommu_dev_features feat) > { > - const struct iommu_ops *ops = dev->bus->iommu_ops; > - > - if (ops && ops->dev_disable_feat) > - return ops->dev_disable_feat(dev, feat); > + if (dev->iommu && dev->iommu->ops->dev_disable_feat) > + return dev->iommu->ops->dev_disable_feat(dev, feat); > > return -EBUSY; > } > @@ -2892,10 +2889,8 @@ > EXPORT_SYMBOL_GPL(iommu_dev_disable_feature); > > bool iommu_dev_feature_enabled(struct device *dev, enum > iommu_dev_features feat) > { > - const struct iommu_ops *ops = dev->bus->iommu_ops; > - > - if (ops && ops->dev_feat_enabled) > - return ops->dev_feat_enabled(dev, feat); > + if (dev->iommu && dev->iommu->ops->dev_feat_enabled) > + return dev->iommu->ops->dev_feat_enabled(dev, feat); > > return false; > } > diff --git a/include/linux/iommu.h b/include/linux/iommu.h > index 524ffc2ff64f..ff0c76bdfb67 100644 > --- a/include/linux/iommu.h > +++ b/include/linux/iommu.h > @@ -354,6 +354,7 @@ struct iommu_fault_param { > * @fault_param: IOMMU detected device fault reporting data > * @fwspec: IOMMU fwspec data > * @iommu_dev: IOMMU device this device is linked to > + * @ops: iommu-ops for talking to the iommu_dev > * @priv: IOMMU Driver private data > * > * TODO: migrate other per device data pointers under iommu_dev_data, > e.g. > @@ -364,6 +365,7 @@ struct dev_iommu { > struct iommu_fault_param *fault_param; > struct iommu_fwspec *fwspec; > struct iommu_device *iommu_dev; > + const struct iommu_ops *ops; > void *priv; > }; > > -- > 2.17.1 > _______________________________________________ > Linuxarm mailing list -- linuxarm(a)openeuler.org > To unsubscribe send an email to linuxarm-leave(a)openeuler.org

2 4

[PATCH V2 net-next 00/13] net: hns3: some cleanups for -next
by Huazhong Tan 13 Feb '21

13 Feb '21

To improve code readability and maintainability, the series refactor out some bloated functions in the HNS3 ethernet driver. change log: V2: remove an unused variable in #5 previous version: V1: https://patchwork.kernel.org/project/netdevbpf/cover/1612943005-59416-1-git… Hao Chen (1): net: hns3: refactor out hclge_rm_vport_all_mac_table() Huazhong Tan (2): net: hns3: refactor out hclge_set_rss_tuple() net: hns3: refactor out hclgevf_set_rss_tuple() Jian Shen (3): net: hns3: refactor out hclge_get_rss_tuple() net: hns3: refactor out hclgevf_get_rss_tuple() net: hns3: split out hclge_dbg_dump_qos_buf_cfg() Jiaran Zhang (1): net: hns3: use ipv6_addr_any() helper Peng Li (4): net: hns3: refactor out hclge_cmd_convert_err_code() net: hns3: refactor out hclgevf_cmd_convert_err_code() net: hns3: clean up hns3_dbg_cmd_write() net: hns3: refactor out hclge_set_vf_vlan_common() Yufeng Mo (2): net: hns3: split out hclge_cmd_send() net: hns3: split out hclgevf_cmd_send() .../ethernet/hisilicon/hns3/hns3_debugfs.c | 44 +-- .../hisilicon/hns3/hns3pf/hclge_cmd.c | 155 ++++++----- .../hisilicon/hns3/hns3pf/hclge_debugfs.c | 158 ++++++++--- .../hisilicon/hns3/hns3pf/hclge_main.c | 254 +++++++++++------- .../hisilicon/hns3/hns3vf/hclgevf_cmd.c | 194 +++++++------ .../hisilicon/hns3/hns3vf/hclgevf_main.c | 114 +++++--- 6 files changed, 571 insertions(+), 348 deletions(-) -- 2.25.1

2 9

[PATCH for next v1 0/2] gpio: few clean up patches to replace spin_lock_irqsave with spin_lock
by Luo Jiaxing 13 Feb '21

13 Feb '21

There is no need to use API with _irqsave in hard IRQ handler, So replace those with spin_lock. Luo Jiaxing (2): gpio: omap: Replace raw_spin_lock_irqsave with raw_spin_lock in omap_gpio_irq_handler() gpio: grgpio: Replace spin_lock_irqsave with spin_lock in grgpio_irq_handler() drivers/gpio/gpio-grgpio.c | 5 ++--- drivers/gpio/gpio-omap.c | 15 ++++++--------- 2 files changed, 8 insertions(+), 12 deletions(-) -- 2.7.4

6 29

[PATCH] i2c: busses: Replace spin_lock_irqsave with spin_lock in hard IRQ
by Weihang Li 12 Feb '21

12 Feb '21

There is no need to do irqsave and irqrestore in context of hard IRQ. Cc: Baruch Siach <baruch(a)tkos.co.il> Cc: Paul Cercueil <paul(a)crapouillou.net> Signed-off-by: Weihang Li <liweihang(a)huawei.com> --- drivers/i2c/busses/i2c-digicolor.c | 5 ++--- drivers/i2c/busses/i2c-jz4780.c | 5 ++--- 2 files changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/i2c/busses/i2c-digicolor.c b/drivers/i2c/busses/i2c-digicolor.c index f67639d..60c838c 100644 --- a/drivers/i2c/busses/i2c-digicolor.c +++ b/drivers/i2c/busses/i2c-digicolor.c @@ -160,12 +160,11 @@ static irqreturn_t dc_i2c_irq(int irq, void *dev_id) { struct dc_i2c *i2c = dev_id; int cmd_status = dc_i2c_cmd_status(i2c); - unsigned long flags; u8 addr_cmd; writeb_relaxed(1, i2c->regs + II_INTFLAG_CLEAR); - spin_lock_irqsave(&i2c->lock, flags); + spin_lock(&i2c->lock); if (cmd_status == II_CMD_STATUS_ACK_BAD || cmd_status == II_CMD_STATUS_ABORT) { @@ -207,7 +206,7 @@ static irqreturn_t dc_i2c_irq(int irq, void *dev_id) } out: - spin_unlock_irqrestore(&i2c->lock, flags); + spin_unlock(&i2c->lock); return IRQ_HANDLED; } diff --git a/drivers/i2c/busses/i2c-jz4780.c b/drivers/i2c/busses/i2c-jz4780.c index cb4a25e..8509c5f 100644 --- a/drivers/i2c/busses/i2c-jz4780.c +++ b/drivers/i2c/busses/i2c-jz4780.c @@ -437,9 +437,8 @@ static irqreturn_t jz4780_i2c_irq(int irqno, void *dev_id) unsigned short intst; unsigned short intmsk; struct jz4780_i2c *i2c = dev_id; - unsigned long flags; - spin_lock_irqsave(&i2c->lock, flags); + spin_lock(&i2c->lock); intmsk = jz4780_i2c_readw(i2c, JZ4780_I2C_INTM); intst = jz4780_i2c_readw(i2c, JZ4780_I2C_INTST); @@ -551,7 +550,7 @@ static irqreturn_t jz4780_i2c_irq(int irqno, void *dev_id) } done: - spin_unlock_irqrestore(&i2c->lock, flags); + spin_unlock(&i2c->lock); return IRQ_HANDLED; } -- 2.8.1

2 1

[PATCH V2 net-next 09/13] net: hns3: split out hclge_cmd_send()
by Huazhong Tan 12 Feb '21

12 Feb '21

From: Yufeng Mo <moyufeng(a)huawei.com> hclge_cmd_send() is bloated, so split it into separate functions for readability and maintainability. Signed-off-by: Yufeng Mo <moyufeng(a)huawei.com> Signed-off-by: Huazhong Tan <tanhuazhong(a)huawei.com> --- .../hisilicon/hns3/hns3pf/hclge_cmd.c | 100 +++++++++++------- 1 file changed, 59 insertions(+), 41 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c index cb2c955ce52c..1bd0ddfaec4d 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c @@ -194,6 +194,22 @@ struct errcode { int common_errno; }; +static void hclge_cmd_copy_desc(struct hclge_hw *hw, struct hclge_desc *desc, + int num) +{ + struct hclge_desc *desc_to_use; + int handle = 0; + + while (handle < num) { + desc_to_use = &hw->cmq.csq.desc[hw->cmq.csq.next_to_use]; + *desc_to_use = desc[handle]; + (hw->cmq.csq.next_to_use)++; + if (hw->cmq.csq.next_to_use >= hw->cmq.csq.desc_num) + hw->cmq.csq.next_to_use = 0; + handle++; + } +} + static int hclge_cmd_convert_err_code(u16 desc_ret) { struct errcode hclge_cmd_errcode[] = { @@ -243,6 +259,44 @@ static int hclge_cmd_check_retval(struct hclge_hw *hw, struct hclge_desc *desc, return hclge_cmd_convert_err_code(desc_ret); } +static int hclge_cmd_check_result(struct hclge_hw *hw, struct hclge_desc *desc, + int num, int ntc) +{ + struct hclge_dev *hdev = container_of(hw, struct hclge_dev, hw); + bool is_completed = false; + u32 timeout = 0; + int handle, ret; + + /** + * If the command is sync, wait for the firmware to write back, + * if multi descriptors to be sent, use the first one to check + */ + if (HCLGE_SEND_SYNC(le16_to_cpu(desc->flag))) { + do { + if (hclge_cmd_csq_done(hw)) { + is_completed = true; + break; + } + udelay(1); + timeout++; + } while (timeout < hw->cmq.tx_timeout); + } + + if (!is_completed) + ret = -EBADE; + else + ret = hclge_cmd_check_retval(hw, desc, num, ntc); + + /* Clean the command send queue */ + handle = hclge_cmd_csq_clean(hw); + if (handle < 0) + ret = handle; + else if (handle != num) + dev_warn(&hdev->pdev->dev, + "cleaned %d, need to clean %d\n", handle, num); + return ret; +} + /** * hclge_cmd_send - send command to command queue * @hw: pointer to the hw struct @@ -256,11 +310,7 @@ int hclge_cmd_send(struct hclge_hw *hw, struct hclge_desc *desc, int num) { struct hclge_dev *hdev = container_of(hw, struct hclge_dev, hw); struct hclge_cmq_ring *csq = &hw->cmq.csq; - struct hclge_desc *desc_to_use; - bool complete = false; - u32 timeout = 0; - int handle = 0; - int retval; + int ret; int ntc; spin_lock_bh(&hw->cmq.csq.lock); @@ -284,49 +334,17 @@ int hclge_cmd_send(struct hclge_hw *hw, struct hclge_desc *desc, int num) * which will be use for hardware to write back */ ntc = hw->cmq.csq.next_to_use; - while (handle < num) { - desc_to_use = &hw->cmq.csq.desc[hw->cmq.csq.next_to_use]; - *desc_to_use = desc[handle]; - (hw->cmq.csq.next_to_use)++; - if (hw->cmq.csq.next_to_use >= hw->cmq.csq.desc_num) - hw->cmq.csq.next_to_use = 0; - handle++; - } + + hclge_cmd_copy_desc(hw, desc, num); /* Write to hardware */ hclge_write_dev(hw, HCLGE_NIC_CSQ_TAIL_REG, hw->cmq.csq.next_to_use); - /** - * If the command is sync, wait for the firmware to write back, - * if multi descriptors to be sent, use the first one to check - */ - if (HCLGE_SEND_SYNC(le16_to_cpu(desc->flag))) { - do { - if (hclge_cmd_csq_done(hw)) { - complete = true; - break; - } - udelay(1); - timeout++; - } while (timeout < hw->cmq.tx_timeout); - } - - if (!complete) - retval = -EBADE; - else - retval = hclge_cmd_check_retval(hw, desc, num, ntc); - - /* Clean the command send queue */ - handle = hclge_cmd_csq_clean(hw); - if (handle < 0) - retval = handle; - else if (handle != num) - dev_warn(&hdev->pdev->dev, - "cleaned %d, need to clean %d\n", handle, num); + ret = hclge_cmd_check_result(hw, desc, num, ntc); spin_unlock_bh(&hw->cmq.csq.lock); - return retval; + return ret; } static void hclge_set_default_capability(struct hclge_dev *hdev) -- 2.25.1

1 4

Re: [PATCH for-next 00/32] spin lock usage optimization for SCSI drivers
by Song Bao Hua (Barry Song) 12 Feb '21

12 Feb '21

> > > > On Wed, 10 Feb 2021, Song Bao Hua (Barry Song) wrote: > > > > > > On Wed, 10 Feb 2021, Song Bao Hua (Barry Song) wrote: > > > > > > > > > > On Wed, 10 Feb 2021, Song Bao Hua (Barry Song) wrote: > > > > > > > > > > > > > > There is no warning from m68k builds. That's because > > > > > > > > arch_irqs_disabled() returns true when the IPL is non-zero. > > > > > > > > > > > > > > So for m68k, the case is arch_irqs_disabled() is true, but > > > > > > > interrupts can still come? > > > > > > > > > > > > > > Then it seems it is very confusing. If prioritized interrupts > > > > > > > can still come while arch_irqs_disabled() is true, > > > > > > > > > > > > Yes, on m68k CPUs, an IRQ having a priority level higher than the > > > > > > present priority mask will get serviced. > > > > > > > > > > > > Non-Maskable Interrupt (NMI) is not subject to this rule and gets > > > > > > serviced regardless. > > > > > > > > > > > > > how could spin_lock_irqsave() block the prioritized interrupts? > > > > > > > > > > > > It raises the the mask level to 7. Again, please see > > > > > > arch/m68k/include/asm/irqflags.h > > > > > > > > > > Hi Finn, > > > > > Thanks for your explanation again. > > > > > > > > > > TBH, that is why m68k is so confusing. irqs_disabled() on m68k > > > > > should just reflect the status of all interrupts have been disabled > > > > > except NMI. > > > > > > > > > > irqs_disabled() should be consistent with the calling of APIs such > > > > > as local_irq_disable, local_irq_save, spin_lock_irqsave etc. > > > > > > > > > > > > > When irqs_disabled() returns true, we cannot infer that > > > > arch_local_irq_disable() was called. But I have not yet found driver > > > > code or core kernel code attempting that inference. > > > > > > > > > > > > > > > > > Isn't arch_irqs_disabled() a status reflection of irq disable > > > > > > > API? > > > > > > > > > > > > > > > > > > > Why not? > > > > > > > > > > If so, arch_irqs_disabled() should mean all interrupts have been > > > > > masked except NMI as NMI is unmaskable. > > > > > > > > > > > > > Can you support that claim with a reference to core kernel code or > > > > documentation? (If some arch code agrees with you, that's neither here > > > > nor there.) > > > > > > I think those links I share you have supported this. Just you don't > > > believe :-) > > > > > > > Your links show that the distinction between fast and slow handlers was > > removed. Your links don't support your claim that "arch_irqs_disabled() > > should mean all interrupts have been masked". Where is the code that makes > > that inference? Where is the documentation that supports your claim? > > (1) > https://lwn.net/Articles/380931/ > Looking at all these worries, one might well wonder if a system which *disabled > interrupts for all handlers* would function well at all. So it is interesting > to note one thing: any system which has the lockdep locking checker enabled > has been running all handlers that way for some years now. Many developers > and testers run lockdep-enabled kernels, and they are available for some of > the more adventurous distributions (Rawhide, for example) as well. So we > have quite a bit of test coverage for this mode of operation already. > > (2) > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/ > ?id=b738a50a > > "We run all handlers *with interrupts disabled* and expect them not to > enable them. Warn when we catch one who does." > > (3) > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/ > ?id=e58aa3d2d0cc > genirq: Run irq handlers *with interrupts disabled* > > Running interrupt handlers with interrupts enabled can cause stack > overflows. That has been observed with multiqueue NICs delivering all > their interrupts to a single core. We might band aid that somehow by > checking the interrupt stacks, but the real safe fix is to *run the irq > handlers with interrupts disabled*. > > > All these documents say we are running irq handler with interrupts > disabled. but it seems you think high-prio interrupts don't belong > to "interrupts" in those documents :-) > > that is why we can't get agreement. I think "interrupts" mean > all except NMI in these documents, but you insist high-prio IRQ > is an exception. > > > > > > > > > > > > > > > > > > > Are all interrupts (including NMI) masked whenever > > > > > > arch_irqs_disabled() returns true on your platforms? > > > > > > > > > > On my platform, once irqs_disabled() is true, all interrupts are > > > > > masked except NMI. NMI just ignore spin_lock_irqsave or > > > > > local_irq_disable. > > > > > > > > > > On ARM64, we also have high-priority interrupts, but they are > > > > > running as PESUDO_NMI: > > > > > https://lwn.net/Articles/755906/ > > > > > > > > > > > > > A glance at the ARM GIC specification suggests that your hardware > > > > works much like 68000 hardware. > > > > > > > > When enabled, a CPU interface takes the highest priority pending > > > > interrupt for its connected processor and determines whether the > > > > interrupt has sufficient priority for it to signal the interrupt > > > > request to the processor. [...] > > > > > > > > When the processor acknowledges the interrupt at the CPU interface, > > > > the Distributor changes the status of the interrupt from pending to > > > > either active, or active and pending. At this point the CPU > > > > interface can signal another interrupt to the processor, to preempt > > > > interrupts that are active on the processor. If there is no pending > > > > interrupt with sufficient priority for signaling to the processor, > > > > the interface deasserts the interrupt request signal to the > > > > processor. > > > > > > > > https://developer.arm.com/documentation/ihi0048/b/ > > > > > > > > Have you considered that Linux/arm might benefit if it could fully > > > > exploit hardware features already available, such as the interrupt > > > > priority masking feature in the GIC in existing arm systems? > > > > > > I guess no:-) there are only two levels: IRQ and NMI. Injecting a > > > high-prio IRQ level between them makes no sense. > > > > > > To me, arm64's design is quite clear and has no any confusion. > > > > > > > Are you saying that the ARM64 hardware design is confusing because it > > implements a priority mask, and that's why you had to simplify it with a > > pseudo-nmi scheme in software? > > No, I was not saying this. I think both m68k and arm64 have good hardware > design. Just Linux's implementation is running irq-handlers with interrupts > disabled. So ARM64's pseudo-nmi is adapted to Linux better. > > > > > > > > > > > > On m68k, it seems you mean： > > > > > irq_disabled() is true, but high-priority interrupts can still come; > > > > > local_irq_disable() can disable high-priority interrupts, and at that > > > > > time, irq_disabled() is also true. > > > > > > > > > > TBH, this is wrong and confusing on m68k. > > > > > > > > > > > > > Like you, I was surprised when I learned about it. But that doesn't mean > > > > it's wrong. The fact that it works should tell you something. > > > > > > > > > > The fact is that m68k lets arch_irq_disabled() return true to pretend > > > all IRQs are disabled while high-priority IRQ is still open, thus "pass" > > > all sanitizing check in genirq and kernel core. > > > > > > > The fact is that m68k has arch_irq_disabled() return false when all IRQs > > are enabled. So there is no bug. > > But it has arch_irq_disabled() return true while some interrupts(not NMI) > are still open. > > > > > > > Things could always be made simpler. But discarding features isn't > > > > necessarily an improvement. > > > > > > This feature could be used by calling local_irq_enable_in_hardirq() in > > > those IRQ handlers who hope high-priority interrupts to preempt it for a > > > while. > > > > > > > So, if one handler is sensitive to interrupt latency, all other handlers > > should be modified? I don't think that's workable. > > I think we just enable preempt_rt or force threaded_irq, and then improve > the priority of the irq thread who is sensitive to latency. No need to > touch all threads. > > I also understand your point, we let one high-prio interrupt preempt > low priority interrupt, then we don't need to change the whole system. > But I think Linux prefers the method of threaded_irq or preempt_rt > for this kind of problems. > > > > > In anycase, what you're describing is a completely different nested > > interrupt scheme that would defeat the priority level mechanism that the > > hardware provides us with. > > Yes. Indeed. > > > > > > It shouldn't hide somewhere and make confusion. > > > > > > > The problem is hiding so well that no-one has found it! I say it doesn't > > exist. > > Long long ago(before 2.6.38), we had a kernel supporting IRQF_DISABLED and > nested interrupts were widely supported, but system also ran well in most > cases. That means nested interrupts don't really matter in most cases. > That is why m68k is also running well even though it is still nesting. > > > > > > On the other hand, those who care about realtime should use threaded IRQ > > > and let IRQ threads preempt each other. > > > > > > > Yes. And those threads also have priority levels. > > Finn, I am not a m68k guy, would you help check if this could activate a > warning on m68k. maybe we can discuss this question in genirq maillist from > this warning if you are happy. Thanks very much. > > diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h > index 7c9d6a2d7e90..b8ca27555c76 100644 > --- a/include/linux/hardirq.h > +++ b/include/linux/hardirq.h > @@ -32,6 +32,7 @@ static __always_inline void rcu_irq_enter_check_tick(void) > */ > #define __irq_enter() \ > do { \ > + WARN_ONCE(in_hardirq() && irqs_disabled(), "nested > interrupts\n"); \ > preempt_count_add(HARDIRQ_OFFSET); \ > lockdep_hardirq_enter(); \ > account_hardirq_enter(current); \ > @@ -44,6 +45,7 @@ static __always_inline void rcu_irq_enter_check_tick(void) > */ > #define __irq_enter_raw() \ > do { \ > + WARN_ONCE(in_hardirq() && irqs_disabled(), "nested > interrupts\n"); \ > preempt_count_add(HARDIRQ_OFFSET); \ > lockdep_hardirq_enter(); \ > } while (0) > I am not a m68k guy, here I can show you some code in cortex-m, which supports full interrupt priority preemption in hardware, but its arch code just disables it to satisfy Linux: arch/arm/kernel/entry-header.S .macro v7m_exception_entry ... @ Linux expects to have irqs off. Do it here before taking stack space cpsid i arch/arm/kernel/entry-v7m.S __irq_entry: v7m_exception_entry @ @ Invoke the IRQ handler @ ... @ routine called with r0 = irq number, r1 = struct pt_regs * bl nvic_handle_irq pop {lr} Another example in arch/arc/kernel/entry-arcv2.S: ENTRY(handle_interrupt) INTERRUPT_PROLOGUE # irq control APIs local_irq_save/restore/disable/enable fiddle with # global interrupt enable bits in STATUS32 (.IE for 1 prio, .E[] for 2 prio) # However a taken interrupt doesn't clear these bits. Thus irqs_disabled() # query in hard ISR path would return false (since .IE is set) which would # trips genirq interrupt handling asserts. # # So do a "soft" disable of interrutps here. # # Note this disable is only for consistent book-keeping as further interrupts # will be disabled anyways even w/o this. Hardware tracks active interrupts # seperately in AUX_IRQ_ACT.active and will not take new interrupts # unless this one returns (or higher prio becomes pending in 2-prio scheme) IRQ_DISABLE ; icause is banked: one per priority level ; so a higher prio interrupt taken here won't clobber prev prio icause lr r0, [ICAUSE] mov blink, ret_from_exception b.d arch_do_IRQ mov r1, sp Actually in m68k, I also saw its IRQ entry disabled interrupts by ' move #0x2700,%sr /* disable intrs */' arch/m68k/include/asm/entry.h: .macro SAVE_ALL_SYS move #0x2700,%sr /* disable intrs */ btst #5,%sp@(2) /* from user? */ bnes 6f /* no, skip */ movel %sp,sw_usp /* save user sp */ ... .macro SAVE_ALL_INT SAVE_ALL_SYS moveq #-1,%d0 /* not system call entry */ movel %d0,%sp@(PT_OFF_ORIG_D0) .endm arch/m68k/kernel/entry.S: /* This is the main interrupt handler for autovector interrupts */ ENTRY(auto_inthandler) SAVE_ALL_INT GET_CURRENT(%d0) | put exception # in d0 bfextu %sp@(PT_OFF_FORMATVEC){#4,#10},%d0 subw #VEC_SPUR,%d0 movel %sp,%sp@- movel %d0,%sp@- | put vector # on stack auto_irqhandler_fixup = . + 2 jsr do_IRQ | process the IRQ addql #8,%sp | pop parameters off stack jra ret_from_exception So my question is that " move #0x2700,%sr" is actually disabling all interrupts? And is m68k actually running irq handlers with interrupts disabled? Best Regards Barry

2 4

[PATCH v3 for-next] RDMA/hns: Adjust definition of FRMR fields
by Weihang Li 11 Feb '21

11 Feb '21

From: Yixing Liu <liuyixing1(a)huawei.com> FRMR is not well-supported on HIP08, it is re-designed for HIP09 and the position of related fields is changed. Then the ULPs should be forbidden to use FRMR on older hardwares. Signed-off-by: Yixing Liu <liuyixing1(a)huawei.com> Signed-off-by: Weihang Li <liweihang(a)huawei.com> --- This patch is one of series "RDMA/hns: Updates for 5.12", the others have been applied. Changes since v2: * Aovid setting IB_DEVICE_MEM_MGT_EXTENSIONS for HIP08 or older hardwares. * Link: https://patchwork.kernel.org/project/linux-rdma/patch/1612859923-44041-1-gi… Changes since v1: * Add check for HW version and rewrite the description. * Link: https://patchwork.kernel.org/project/linux-rdma/patch/1612517974-31867-6-gi… drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 45 +++++++++++++++++------------- drivers/infiniband/hw/hns/hns_roce_hw_v2.h | 12 ++++---- drivers/infiniband/hw/hns/hns_roce_main.c | 3 +- 3 files changed, 34 insertions(+), 26 deletions(-) diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c index 012076a..c3934ab 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c @@ -99,16 +99,16 @@ static void set_frmr_seg(struct hns_roce_v2_rc_send_wqe *rc_sq_wqe, u64 pbl_ba; /* use ib_access_flags */ - roce_set_bit(rc_sq_wqe->byte_4, V2_RC_FRMR_WQE_BYTE_4_BIND_EN_S, - wr->access & IB_ACCESS_MW_BIND ? 1 : 0); - roce_set_bit(rc_sq_wqe->byte_4, V2_RC_FRMR_WQE_BYTE_4_ATOMIC_S, - wr->access & IB_ACCESS_REMOTE_ATOMIC ? 1 : 0); - roce_set_bit(rc_sq_wqe->byte_4, V2_RC_FRMR_WQE_BYTE_4_RR_S, - wr->access & IB_ACCESS_REMOTE_READ ? 1 : 0); - roce_set_bit(rc_sq_wqe->byte_4, V2_RC_FRMR_WQE_BYTE_4_RW_S, - wr->access & IB_ACCESS_REMOTE_WRITE ? 1 : 0); - roce_set_bit(rc_sq_wqe->byte_4, V2_RC_FRMR_WQE_BYTE_4_LW_S, - wr->access & IB_ACCESS_LOCAL_WRITE ? 1 : 0); + roce_set_bit(fseg->byte_40, V2_RC_FRMR_WQE_BYTE_40_BIND_EN_S, + !!(wr->access & IB_ACCESS_MW_BIND)); + roce_set_bit(fseg->byte_40, V2_RC_FRMR_WQE_BYTE_40_ATOMIC_S, + !!(wr->access & IB_ACCESS_REMOTE_ATOMIC)); + roce_set_bit(fseg->byte_40, V2_RC_FRMR_WQE_BYTE_40_RR_S, + !!(wr->access & IB_ACCESS_REMOTE_READ)); + roce_set_bit(fseg->byte_40, V2_RC_FRMR_WQE_BYTE_40_RW_S, + !!(wr->access & IB_ACCESS_REMOTE_WRITE)); + roce_set_bit(fseg->byte_40, V2_RC_FRMR_WQE_BYTE_40_LW_S, + !!(wr->access & IB_ACCESS_LOCAL_WRITE)); /* Data structure reuse may lead to confusion */ pbl_ba = mr->pbl_mtr.hem_cfg.root_ba; @@ -121,12 +121,10 @@ static void set_frmr_seg(struct hns_roce_v2_rc_send_wqe *rc_sq_wqe, rc_sq_wqe->va = cpu_to_le64(wr->mr->iova); fseg->pbl_size = cpu_to_le32(mr->npages); - roce_set_field(fseg->mode_buf_pg_sz, - V2_RC_FRMR_WQE_BYTE_40_PBL_BUF_PG_SZ_M, + roce_set_field(fseg->byte_40, V2_RC_FRMR_WQE_BYTE_40_PBL_BUF_PG_SZ_M, V2_RC_FRMR_WQE_BYTE_40_PBL_BUF_PG_SZ_S, to_hr_hw_page_shift(mr->pbl_mtr.hem_cfg.buf_pg_shift)); - roce_set_bit(fseg->mode_buf_pg_sz, - V2_RC_FRMR_WQE_BYTE_40_BLK_MODE_S, 0); + roce_set_bit(fseg->byte_40, V2_RC_FRMR_WQE_BYTE_40_BLK_MODE_S, 0); } static void set_atomic_seg(const struct ib_send_wr *wr, @@ -522,10 +520,12 @@ static inline int set_ud_wqe(struct hns_roce_qp *qp, return 0; } -static int set_rc_opcode(struct hns_roce_v2_rc_send_wqe *rc_sq_wqe, +static int set_rc_opcode(struct hns_roce_dev *hr_dev, + struct hns_roce_v2_rc_send_wqe *rc_sq_wqe, const struct ib_send_wr *wr) { u32 ib_op = wr->opcode; + int ret = 0; rc_sq_wqe->immtdata = get_immtdata(wr); @@ -545,7 +545,10 @@ static int set_rc_opcode(struct hns_roce_v2_rc_send_wqe *rc_sq_wqe, rc_sq_wqe->va = cpu_to_le64(atomic_wr(wr)->remote_addr); break; case IB_WR_REG_MR: - set_frmr_seg(rc_sq_wqe, reg_wr(wr)); + if (hr_dev->pci_dev->revision >= PCI_REVISION_ID_HIP09) + set_frmr_seg(rc_sq_wqe, reg_wr(wr)); + else + ret = -EOPNOTSUPP; break; case IB_WR_LOCAL_INV: roce_set_bit(rc_sq_wqe->byte_4, V2_RC_SEND_WQE_BYTE_4_SO_S, 1); @@ -554,19 +557,23 @@ static int set_rc_opcode(struct hns_roce_v2_rc_send_wqe *rc_sq_wqe, rc_sq_wqe->inv_key = cpu_to_le32(wr->ex.invalidate_rkey); break; default: - return -EINVAL; + ret = -EINVAL; } + if (unlikely(ret)) + return ret; + roce_set_field(rc_sq_wqe->byte_4, V2_RC_SEND_WQE_BYTE_4_OPCODE_M, V2_RC_SEND_WQE_BYTE_4_OPCODE_S, to_hr_opcode(ib_op)); - return 0; + return ret; } static inline int set_rc_wqe(struct hns_roce_qp *qp, const struct ib_send_wr *wr, void *wqe, unsigned int *sge_idx, unsigned int owner_bit) { + struct hns_roce_dev *hr_dev = to_hr_dev(qp->ibqp.device); struct hns_roce_v2_rc_send_wqe *rc_sq_wqe = wqe; unsigned int curr_idx = *sge_idx; unsigned int valid_num_sge; @@ -577,7 +584,7 @@ static inline int set_rc_wqe(struct hns_roce_qp *qp, rc_sq_wqe->msg_len = cpu_to_le32(msg_len); - ret = set_rc_opcode(rc_sq_wqe, wr); + ret = set_rc_opcode(hr_dev, rc_sq_wqe, wr); if (WARN_ON(ret)) return ret; diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h index 7a0f8ac..39621fb 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h @@ -1255,15 +1255,15 @@ struct hns_roce_v2_rc_send_wqe { #define V2_RC_SEND_WQE_BYTE_4_INLINE_S 12 -#define V2_RC_FRMR_WQE_BYTE_4_BIND_EN_S 19 +#define V2_RC_FRMR_WQE_BYTE_40_BIND_EN_S 10 -#define V2_RC_FRMR_WQE_BYTE_4_ATOMIC_S 20 +#define V2_RC_FRMR_WQE_BYTE_40_ATOMIC_S 11 -#define V2_RC_FRMR_WQE_BYTE_4_RR_S 21 +#define V2_RC_FRMR_WQE_BYTE_40_RR_S 12 -#define V2_RC_FRMR_WQE_BYTE_4_RW_S 22 +#define V2_RC_FRMR_WQE_BYTE_40_RW_S 13 -#define V2_RC_FRMR_WQE_BYTE_4_LW_S 23 +#define V2_RC_FRMR_WQE_BYTE_40_LW_S 14 #define V2_RC_SEND_WQE_BYTE_4_FLAG_S 31 @@ -1280,7 +1280,7 @@ struct hns_roce_v2_rc_send_wqe { struct hns_roce_wqe_frmr_seg { __le32 pbl_size; - __le32 mode_buf_pg_sz; + __le32 byte_40; }; #define V2_RC_FRMR_WQE_BYTE_40_PBL_BUF_PG_SZ_S 4 diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c index c29215a..c9c0836 100644 --- a/drivers/infiniband/hw/hns/hns_roce_main.c +++ b/drivers/infiniband/hw/hns/hns_roce_main.c @@ -201,7 +201,8 @@ static int hns_roce_query_device(struct ib_device *ib_dev, props->max_srq_sge = hr_dev->caps.max_srq_sges; } - if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_FRMR) { + if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_FRMR && + hr_dev->pci_dev->revision >= PCI_REVISION_ID_HIP09) { props->device_cap_flags |= IB_DEVICE_MEM_MGT_EXTENSIONS; props->max_fast_reg_page_list_len = HNS_ROCE_FRMR_MAX_PA; } -- 2.8.1

2 1

[PATCH for next v1 0/2] gpio: few clean up patches to replace spin_lock_irqsave with spin_lock
by Luo Jiaxing 10 Feb '21

10 Feb '21

There is no need to use API with _irqsave in hard IRQ handler, So replace those with spin_lock. Luo Jiaxing (2): gpio: omap: Replace raw_spin_lock_irqsave with raw_spin_lock in omap_gpio_irq_handler() gpio: grgpio: Replace spin_lock_irqsave with spin_lock in grgpio_irq_handler() drivers/gpio/gpio-grgpio.c | 5 ++--- drivers/gpio/gpio-omap.c | 15 ++++++--------- 2 files changed, 8 insertions(+), 12 deletions(-) -- 2.7.4

2 3

[PATCH v2 0/4] Use subdir-ccflags-* to inherit debug flag
by Yicong Yang 10 Feb '21

10 Feb '21

Few drivers use ccflags-* in their top directory to enable -DDEBUG, but don't have config options to enable debug in the sub-directories, or they use per subdirectory ccflags-* to have DEBUG with the same kconfig option. Considering they intends to enable debug for all the files under the directory with the same kconfig option, it will be clearer to use subdir-ccflags-* instead of ccflags-* to inherit the debug settings from Kconfig when traversing subdirectories. We primarily find this issue when debugging PCIe and thought other drivers may also have this issues. Previous discussion can be find at https://lore.kernel.org/linux-pci/1612438215-33105-1-git-send-email-yangyic… Change since v1: - reword the commits to illustrate the reasons of the change and the benefits. v1: https://lore.kernel.org/lkml/1612518255-23052-1-git-send-email-yangyicong@h… Junhao He (4): driver core: Use subdir-ccflags-* to inherit debug flag hwmon: Use subdir-ccflags-* to inherit debug flag pps: Use subdir-ccflags-* to inherit debug flag staging: comedi: Use subdir-ccflags-* to inherit debug flag drivers/base/Makefile | 2 +- drivers/base/power/Makefile | 2 -- drivers/hwmon/Makefile | 2 +- drivers/pps/Makefile | 2 +- drivers/staging/comedi/Makefile | 2 +- drivers/staging/comedi/drivers/Makefile | 1 - drivers/staging/comedi/drivers/tests/Makefile | 2 -- drivers/staging/comedi/kcomedilib/Makefile | 2 -- 8 files changed, 4 insertions(+), 11 deletions(-) -- 2.8.1

3 7