Skip to contentSkip to conversation

6.3.7-rc1 review

PATCH In review Greg Kroah-Hartman 12 participants 299 messages Jun 7, 2023 – Jun 13, 2023 Page 1 of 2

Subsystems (7)
lore
Share & tools
lore
author maintainer
#1 lore raw

PATCH 1/286 RDMA/bnxt_re: Fix the page_size used during the MR creation

In reply to Greg Kroah-Hartman

From: Selvin Xavier <[email protected]>

[ Upstream commit 08c7f09356e45d093d1867c7a3c6ac6526e2f98b ]

Driver populates the list of pages used for Memory region wrongly when page size is more than system page size. This is causing a failure when some of the applications that creates MR with page size as 2M. Since HW can support multiple page sizes, pass the correct page size while creating the MR.

Also, driver need not adjust the number of pages when HW Queues are created with user memory. It should work with the number of dma blocks returned by ib_umem_num_dma_blocks. Fix this calculation also.

Fixes 0c4dcd602817 ("RDMA/bnxt_re: Refactor hardware queue memory allocation")
Fixes f6919d56388c ("RDMA/bnxt_re: Code refactor while populating user MRs")
---
 drivers/infiniband/hw/bnxt_re/qplib_res.c | 12 ++----------
 drivers/infiniband/hw/bnxt_re/qplib_sp.c  |  7 +++----
 2 files changed, 5 insertions(+), 14 deletions(-)
drivers/infiniband/hw/bnxt_re/qplib_res.c +2 -10
index 126d4f26f75ad..81b0c5e879f9e 100644
@@ -215,17 +215,9 @@ int bnxt_qplib_alloc_init_hwq(struct bnxt_qplib_hwq *hwq,
 			return -EINVAL;
 		hwq_attr->sginfo->npages = npages;
 	} else {
-		unsigned long sginfo_num_pages = ib_umem_num_dma_blocks(
-			hwq_attr->sginfo->umem, hwq_attr->sginfo->pgsize);
-
+		npages = ib_umem_num_dma_blocks(hwq_attr->sginfo->umem,
+						hwq_attr->sginfo->pgsize);
 		hwq->is_user = true;
-		npages = sginfo_num_pages;
-		npages = (npages * PAGE_SIZE) /
-			  BIT_ULL(hwq_attr->sginfo->pgshft);
-		if ((sginfo_num_pages * PAGE_SIZE) %
-		     BIT_ULL(hwq_attr->sginfo->pgshft))
-			if (!npages)
-				npages++;
 	}
 
 	if (npages == MAX_PBL_LVL_0_PGS && !hwq_attr->sginfo->nopte) {
drivers/infiniband/hw/bnxt_re/qplib_sp.c +3 -4
index b802981b71716..bae7d89261439 100644
@@ -584,16 +584,15 @@ int bnxt_qplib_reg_mr(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mr,
 		/* Free the hwq if it already exist, must be a rereg */
 		if (mr->hwq.max_elements)
 			bnxt_qplib_free_hwq(res, &mr->hwq);
-		/* Use system PAGE_SIZE */
 		hwq_attr.res = res;
 		hwq_attr.depth = pages;
-		hwq_attr.stride = buf_pg_size;
+		hwq_attr.stride = sizeof(dma_addr_t);
 		hwq_attr.type = HWQ_TYPE_MR;
 		hwq_attr.sginfo = &sginfo;
 		hwq_attr.sginfo->umem = umem;
 		hwq_attr.sginfo->npages = pages;
-		hwq_attr.sginfo->pgsize = PAGE_SIZE;
-		hwq_attr.sginfo->pgshft = PAGE_SHIFT;
+		hwq_attr.sginfo->pgsize = buf_pg_size;
+		hwq_attr.sginfo->pgshft = ilog2(buf_pg_size);
 		rc = bnxt_qplib_alloc_init_hwq(&mr->hwq, &hwq_attr);
 		if (rc) {
 			dev_err(&res->pdev->dev,
Signature · 1 line
2.39.2
author maintainer
#10 lore raw

PATCH 10/286 iommu: Make IPMMU_VMSA dependencies more strict

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit e332003bb216a9f91e08004b9e2de0745f321290 ]

On riscv64, linux-next-20233030 (and for several days earlier), there is a kconfig warning:

WARNING: unmet direct dependencies detected for IOMMU_IO_PGTABLE_LPAE
Depends on [n]: IOMMU_SUPPORT [=y] && (ARM || ARM64 || COMPILE_TEST [=n]) && !GENERIC_ATOMIC64 [=n] Selected by [y]:
- IPMMU_VMSA [=y] && IOMMU_SUPPORT [=y] && (ARCH_RENESAS [=y] || COMPILE_TEST [=n]) && !GENERIC_ATOMIC64 [=n]

and build errors:

riscv64-linux-ld: drivers/iommu/io-pgtable-arm.o: in function `.L140':
io-pgtable-arm.c:(.init.text+0x1e8): undefined reference to `alloc_io_pgtable_ops'
riscv64-linux-ld: drivers/iommu/io-pgtable-arm.o: in function `.L168':
io-pgtable-arm.c:(.init.text+0xab0): undefined reference to `free_io_pgtable_ops'
riscv64-linux-ld: drivers/iommu/ipmmu-vmsa.o: in function `.L140':
ipmmu-vmsa.c:(.text+0xbc4): undefined reference to `free_io_pgtable_ops'
riscv64-linux-ld: drivers/iommu/ipmmu-vmsa.o: in function `.L0 ':
ipmmu-vmsa.c:(.text+0x145e): undefined reference to `alloc_io_pgtable_ops'

Add ARM || ARM64 || COMPILE_TEST dependencies to IPMMU_VMSA to prevent these issues, i.e., so that ARCH_RENESAS on RISC-V is not allowed.

This makes the ARCH dependencies become:

	depends on (ARCH_RENESAS && (ARM || ARM64)) || COMPILE_TEST

but that can be a bit hard to read.

Fixes 8292493c22c8 ("riscv: Kconfig.socs: Add ARCH_RENESAS kconfig option")
Cc Joerg Roedel <[email protected]>
Cc Will Deacon <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Conor Dooley <[email protected]>
Cc Lad Prabhakar <[email protected]>
---
 drivers/iommu/Kconfig | 1 +
 1 file changed, 1 insertion(+)
drivers/iommu/Kconfig +1 -0
index 889c7efd050bc..18e68fbaec884 100644
@@ -287,6 +287,7 @@ config EXYNOS_IOMMU_DEBUG
 config IPMMU_VMSA
 	bool "Renesas VMSA-compatible IPMMU"
 	depends on ARCH_RENESAS || COMPILE_TEST
+	depends on ARM || ARM64 || COMPILE_TEST
 	depends on !GENERIC_ATOMIC64	# for IOMMU_IO_PGTABLE_LPAE
 	select IOMMU_API
 	select IOMMU_IO_PGTABLE_LPAE
Signature · 1 line
2.39.2
author maintainer
#11 lore raw

PATCH 11/286 iommu/rockchip: Fix unwind goto issue

In reply to Greg Kroah-Hartman

From: Chao Wang <[email protected]>

[ Upstream commit ec014683c564fb74fc68e8f5e84691d3b3839d24 ]

Smatch complains that
drivers/iommu/rockchip-iommu.c:1306 rk_iommu_probe() warn: missing unwind goto?

The rk_iommu_probe function, after obtaining the irq value through platform_get_irq, directly returns an error if the returned value is negative, without releasing any resources.

Fix this by adding a new error handling label "err_pm_disable" and use a goto statement to redirect to the error handling process. In order to preserve the original semantics, set err to the value of irq.

Fixes 1aa55ca9b14a ("iommu/rockchip: Move irq request past pm_runtime_enable")
---
 drivers/iommu/rockchip-iommu.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
drivers/iommu/rockchip-iommu.c +8 -6
index f30db22ea5d7a..31cd1e2929e9f 100644
@@ -1302,20 +1302,22 @@ static int rk_iommu_probe(struct platform_device *pdev)
 	for (i = 0; i < iommu->num_irq; i++) {
 		int irq = platform_get_irq(pdev, i);
 
-		if (irq < 0)
-			return irq;
+		if (irq < 0) {
+			err = irq;
+			goto err_pm_disable;
+		}
 
 		err = devm_request_irq(iommu->dev, irq, rk_iommu_irq,
 				       IRQF_SHARED, dev_name(dev), iommu);
-		if (err) {
-			pm_runtime_disable(dev);
-			goto err_remove_sysfs;
-		}
+		if (err)
+			goto err_pm_disable;
 	}
 
 	dma_set_mask_and_coherent(dev, rk_ops->dma_bit_mask);
 
 	return 0;
+err_pm_disable:
+	pm_runtime_disable(dev);
 err_remove_sysfs:
 	iommu_device_sysfs_remove(&iommu->iommu);
 err_put_group:
Signature · 1 line
2.39.2
author maintainer
#12 lore raw

PATCH 12/286 iommu/amd: Dont block updates to GATag if guest mode is on

In reply to Greg Kroah-Hartman

From: Joao Martins <[email protected]>

[ Upstream commit ed8a2f4ddef2eaaf864ab1efbbca9788187036ab ]

On KVM GSI routing table updates, specially those where they have vIOMMUs with interrupt remapping enabled (to boot >255vcpus setups without relying on KVM_FEATURE_MSI_EXT_DEST_ID), a VMM may update the backing VF MSIs with a new VCPU affinity.

On AMD with AVIC enabled, the new vcpu affinity info is updated via:
avic_pi_update_irte()
irq_set_vcpu_affinity()
amd_ir_set_vcpu_affinity()
amd_iommu_{de}activate_guest_mode()

Where the IRTE[GATag] is updated with the new vcpu affinity. The GATag contains VM ID and VCPU ID, and is used by IOMMU hardware to signal KVM (via GALog) when interrupt cannot be delivered due to vCPU is in blocking state.

The issue is that amd_iommu_activate_guest_mode() will essentially only change IRTE fields on transitions from non-guest-mode to guest-mode and otherwise returns *with no changes to IRTE* on already configured guest-mode interrupts. To the guest this means that the VF interrupts remain affined to the first vCPU they were first configured, and guest will be unable to issue VF interrupts and receive messages like this from spurious interrupts (e.g. from waking the wrong vCPU in GALog):

[  167.759472] __common_interrupt: 3.34 No irq handler for vector
[  230.680927] mlx5_core 0000:00:02.0: mlx5_cmd_eq_recover:247:(pid
3122): Recovered 1 EQEs on cmd_eq
[  230.681799] mlx5_core 0000:00:02.0:
wait_func_handle_exec_timeout:1113:(pid 3122): cmd[0]: CREATE_CQ(0x400)
recovered after timeout
[  230.683266] __common_interrupt: 3.34 No irq handler for vector

Given the fact that amd_ir_set_vcpu_affinity() uses amd_iommu_activate_guest_mode() underneath it essentially means that VCPU affinity changes of IRTEs are nops. Fix it by dropping the check for guest-mode at amd_iommu_activate_guest_mode(). Same thing is applicable to amd_iommu_deactivate_guest_mode() although, even if the IRTE doesn't change underlying DestID on the host, the VFIO IRQ handler will still be able to poke at the right guest-vCPU.

Fixes b9c6ff94e43a ("iommu/amd: Re-factor guest virtual APIC (de-)activation code")
---
 drivers/iommu/amd/iommu.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/iommu/amd/iommu.c +1 -2
index 167da5b1a5e31..99ec06b0e09f6 100644
@@ -3482,8 +3482,7 @@ int amd_iommu_activate_guest_mode(void *data)
 	struct irte_ga *entry = (struct irte_ga *) ir_data->entry;
 	u64 valid;
 
-	if (!AMD_IOMMU_GUEST_IR_VAPIC(amd_iommu_guest_ir) ||
-	    !entry || entry->lo.fields_vapic.guest_mode)
+	if (!AMD_IOMMU_GUEST_IR_VAPIC(amd_iommu_guest_ir) || !entry)
 		return 0;
 
 	valid = entry->lo.fields_vapic.valid;
Signature · 1 line
2.39.2
author maintainer
#13 lore raw

PATCH 13/286 iommu/amd: Handle GALog overflows

In reply to Greg Kroah-Hartman

From: Joao Martins <[email protected]>

[ Upstream commit af47b0a24058e56e983881993752f88288ca6511 ]

GALog exists to propagate interrupts into all vCPUs in the system when interrupts are marked as non running (e.g. when vCPUs aren't running). A GALog overflow happens when there's in no space in the log to record the GATag of the interrupt. So when the GALOverflow condition happens, the GALog queue is processed and the GALog is restarted, as the IOMMU manual indicates in section "2.7.4 Guest Virtual APIC Log Restart Procedure":

| * Wait until MMIO Offset 2020h[GALogRun]=0b so that all request
|   entries are completed as circumstances allow. GALogRun must be 0b to
|   modify the guest virtual APIC log registers safely.
| * Write MMIO Offset 0018h[GALogEn]=0b.
| * As necessary, change the following values (e.g., to relocate or
| resize the guest virtual APIC event log):
|   - the Guest Virtual APIC Log Base Address Register
|      [MMIO Offset 00E0h],
|   - the Guest Virtual APIC Log Head Pointer Register
|      [MMIO Offset 2040h][GALogHead], and
|   - the Guest Virtual APIC Log Tail Pointer Register
|      [MMIO Offset 2048h][GALogTail].
| * Write MMIO Offset 2020h[GALOverflow] = 1b to clear the bit (W1C).
| * Write MMIO Offset 0018h[GALogEn] = 1b, and either set
|   MMIO Offset 0018h[GAIntEn] to enable the GA log interrupt or clear
|   the bit to disable it.

Failing to handle the GALog overflow means that none of the VFs (in any guest) will work with IOMMU AVIC forcing the user to power cycle the host. When handling the event it resumes the GALog without resizing much like how it is done in the event handler overflow. The [MMIO Offset 2020h][GALOverflow] bit might be set in status register without the [MMIO Offset 2020h][GAInt] bit, so when deciding to poll for GA events (to clear space in the galog), also check the overflow bit.

[suravee: Check for GAOverflow without GAInt, toggle CONTROL_GAINT_EN]

Stable-dep-of: 8ec4e2befef1 ("iommu/amd: Fix up merge conflict resolution")

---
 drivers/iommu/amd/amd_iommu.h |  1 +
 drivers/iommu/amd/init.c      | 24 ++++++++++++++++++++++++
 drivers/iommu/amd/iommu.c     |  9 ++++++++-
 3 files changed, 33 insertions(+), 1 deletion(-)
drivers/iommu/amd/amd_iommu.h +1 -0
index c160a332ce339..24c7e6c6c0de9 100644
@@ -15,6 +15,7 @@ extern irqreturn_t amd_iommu_int_thread(int irq, void *data);
 extern irqreturn_t amd_iommu_int_handler(int irq, void *data);
 extern void amd_iommu_apply_erratum_63(struct amd_iommu *iommu, u16 devid);
 extern void amd_iommu_restart_event_logging(struct amd_iommu *iommu);
+extern void amd_iommu_restart_ga_log(struct amd_iommu *iommu);
 extern int amd_iommu_init_devices(void);
 extern void amd_iommu_uninit_devices(void);
 extern void amd_iommu_init_notifier(void);
drivers/iommu/amd/init.c +24 -0
index 19a46b9f73574..fd487c33b28aa 100644
@@ -751,6 +751,30 @@ void amd_iommu_restart_event_logging(struct amd_iommu *iommu)
 	iommu_feature_enable(iommu, CONTROL_EVT_LOG_EN);
 }
 
+/*
+ * This function restarts event logging in case the IOMMU experienced
+ * an GA log overflow.
+ */
+void amd_iommu_restart_ga_log(struct amd_iommu *iommu)
+{
+	u32 status;
+
+	status = readl(iommu->mmio_base + MMIO_STATUS_OFFSET);
+	if (status & MMIO_STATUS_GALOG_RUN_MASK)
+		return;
+
+	pr_info_ratelimited("IOMMU GA Log restarting\n");
+
+	iommu_feature_disable(iommu, CONTROL_GALOG_EN);
+	iommu_feature_disable(iommu, CONTROL_GAINT_EN);
+
+	writel(MMIO_STATUS_GALOG_OVERFLOW_MASK,
+	       iommu->mmio_base + MMIO_STATUS_OFFSET);
+
+	iommu_feature_enable(iommu, CONTROL_GAINT_EN);
+	iommu_feature_enable(iommu, CONTROL_GALOG_EN);
+}
+
 /*
  * This function resets the command buffer if the IOMMU stopped fetching
  * commands from it.
drivers/iommu/amd/iommu.c +8 -1
index 99ec06b0e09f6..478da9b4a1b14 100644
@@ -845,6 +845,7 @@ amd_iommu_set_pci_msi_domain(struct device *dev, struct amd_iommu *iommu) { }
 	(MMIO_STATUS_EVT_OVERFLOW_INT_MASK | \
 	 MMIO_STATUS_EVT_INT_MASK | \
 	 MMIO_STATUS_PPR_INT_MASK | \
+	 MMIO_STATUS_GALOG_OVERFLOW_MASK | \
 	 MMIO_STATUS_GALOG_INT_MASK)
 
 irqreturn_t amd_iommu_int_thread(int irq, void *data)
@@ -868,10 +869,16 @@ irqreturn_t amd_iommu_int_thread(int irq, void *data)
 		}
 
 #ifdef CONFIG_IRQ_REMAP
-		if (status & MMIO_STATUS_GALOG_INT_MASK) {
+		if (status & (MMIO_STATUS_GALOG_INT_MASK |
+			      MMIO_STATUS_GALOG_OVERFLOW_MASK)) {
 			pr_devel("Processing IOMMU GA Log\n");
 			iommu_poll_ga_log(iommu);
 		}
+
+		if (status & MMIO_STATUS_GALOG_OVERFLOW_MASK) {
+			pr_info_ratelimited("IOMMU GA Log overflow\n");
+			amd_iommu_restart_ga_log(iommu);
+		}
 #endif
 
 		if (status & MMIO_STATUS_EVT_OVERFLOW_INT_MASK) {
Signature · 1 line
2.39.2
author maintainer
#14 lore raw

PATCH 14/286 iommu/amd: Fix up merge conflict resolution

In reply to Greg Kroah-Hartman

From: Jerry Snitselaar <[email protected]>

[ Upstream commit 8ec4e2befef10c7679cd59251956a428e783c0b5 ]

Merge commit e17c6debd4b2 ("Merge branches 'arm/mediatek', 'arm/msm', 'arm/renesas', 'arm/rockchip', 'arm/smmu', 'x86/vt-d' and 'x86/amd' into next") added amd_iommu_init_devices, amd_iommu_uninit_devices,
and amd_iommu_init_notifier back to drivers/iommu/amd/amd_iommu.h.
The only references to them are here, so clean them up.

Fixes e17c6debd4b2 ("Merge branches 'arm/mediatek', 'arm/msm', 'arm/renesas', 'arm/rockchip', 'arm/smmu', 'x86/vt-d' and 'x86/amd' into next")
Cc Joerg Roedel <[email protected]>
Cc Suravee Suthikulpanit <[email protected]>
Cc Will Deacon <[email protected]>
Cc Robin Murphy <[email protected]>
---
 drivers/iommu/amd/amd_iommu.h | 3 ---
 1 file changed, 3 deletions(-)
drivers/iommu/amd/amd_iommu.h +0 -3
index 24c7e6c6c0de9..471f40351f4c8 100644
@@ -16,9 +16,6 @@ extern irqreturn_t amd_iommu_int_handler(int irq, void *data);
 extern void amd_iommu_apply_erratum_63(struct amd_iommu *iommu, u16 devid);
 extern void amd_iommu_restart_event_logging(struct amd_iommu *iommu);
 extern void amd_iommu_restart_ga_log(struct amd_iommu *iommu);
-extern int amd_iommu_init_devices(void);
-extern void amd_iommu_uninit_devices(void);
-extern void amd_iommu_init_notifier(void);
 extern void amd_iommu_set_rlookup_table(struct amd_iommu *iommu, u16 devid);
 
 #ifdef CONFIG_AMD_IOMMU_DEBUGFS
Signature · 1 line
2.39.2
author maintainer
#15 lore raw

PATCH 15/286 iommu/amd: Add missing domain type checks

In reply to Greg Kroah-Hartman

From: Jason Gunthorpe <[email protected]>

[ Upstream commit 29f54745f24547a84b18582e054df9bea1a7bf3e ]

Drivers are supposed to list the domain types they support in their domain_alloc() ops so when we add new domain types, like BLOCKING or SVA, they don't start breaking.

This ended up providing an empty UNMANAGED domain when the core code asked for a BLOCKING domain, which happens to be the fallback for drivers that don't support it, but this is completely wrong for SVA.

Check for the DMA types AMD supports and reject every other kind.

---
 drivers/iommu/amd/iommu.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/iommu/amd/iommu.c +5 -1
index 478da9b4a1b14..8bd5390808784 100644
@@ -2065,7 +2065,7 @@ static struct protection_domain *protection_domain_alloc(unsigned int type)
 {
 	struct io_pgtable_ops *pgtbl_ops;
 	struct protection_domain *domain;
-	int pgtable = amd_iommu_pgtable;
+	int pgtable;
 	int mode = DEFAULT_PGTABLE_LEVEL;
 	int ret;
 
@@ -2082,6 +2082,10 @@ static struct protection_domain *protection_domain_alloc(unsigned int type)
 		mode = PAGE_MODE_NONE;
 	} else if (type == IOMMU_DOMAIN_UNMANAGED) {
 		pgtable = AMD_IOMMU_V1;
+	} else if (type == IOMMU_DOMAIN_DMA || type == IOMMU_DOMAIN_DMA_FQ) {
+		pgtable = amd_iommu_pgtable;
+	} else {
+		return NULL;
 	}
 
 	switch (pgtable) {
Signature · 1 line
2.39.2
author maintainer
#16 lore raw

PATCH 16/286 nfsd: make a copy of struct iattr before calling notify_change

In reply to Greg Kroah-Hartman

From: Jeff Layton <[email protected]>

[ Upstream commit d53d70084d27f56bcdf5074328f2c9ec861be596 ]

notify_change can modify the iattr structure. In particular it can end up setting ATTR_MODE when ATTR_KILL_SUID is already set, causing a BUG() if the same iattr is passed to notify_change more than once.

Make a copy of the struct iattr before calling notify_change.

Fixes 34b91dda7124 ("NFSD: Make nfsd4_setattr() wait before returning NFS4ERR_DELAY")
---
 fs/nfsd/vfs.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)
fs/nfsd/vfs.c +9 -1
index 5783209f17fc5..e4884dde048ce 100644
@@ -536,7 +536,15 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 	inode_lock(inode);
 	for (retries = 1;;) {
-		host_err = __nfsd_setattr(dentry, iap);
+		struct iattr attrs;
+
+		/*
+		 * notify_change() can alter its iattr argument, making
+		 * @iap unsuitable for submission multiple times. Make a
+		 * copy for every loop iteration.
+		 */
+		attrs = *iap;
+		host_err = __nfsd_setattr(dentry, &attrs);
 		if (host_err != -EAGAIN || !retries--)
 			break;
 		if (!nfsd_wait_for_delegreturn(rqstp, inode))
Signature · 1 line
2.39.2
author maintainer
#17 lore raw

PATCH 17/286 dmaengine: pl330: rename _start to prevent build error

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit a1a5f2c887252dec161c1e12e04303ca9ba56fa9 ]

"_start" is used in several arches and proably should be reserved for ARCH usage. Using it in a driver for a private symbol can cause a build error when it conflicts with ARCH usage of the same symbol.

Therefore rename pl330's "_start" to "pl330_start_thread" so that there is no conflict and no build error.

drivers/dma/pl330.c:1053:13: error: '_start' redeclared as different kind of symbol
1053 | static bool _start(struct pl330_thread *thrd)

      |             ^~~~~~
In file included from ../include/linux/interrupt.h:21,
                 from ../drivers/dma/pl330.c:18:
arch/riscv/include/asm/sections.h:11:13: note: previous declaration of '_start' with type 'char[]'
   11 | extern char _start[];
      |             ^~~~~~
Fixes b7d861d93945 ("DMA: PL330: Merge PL330 driver into drivers/dma/")
Fixes ae43b3289186 ("ARM: 8202/1: dmaengine: pl330: Add runtime Power Management support v12")
Cc Jaswinder Singh <[email protected]>
Cc Boojin Kim <[email protected]>
Cc Krzysztof Kozlowski <[email protected]>
Cc Russell King <[email protected]>
Cc Vinod Koul <[email protected]>
---
 drivers/dma/pl330.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/dma/pl330.c +4 -4
index 0d9257fbdfb0d..b4731fe6bbc14 100644
@@ -1050,7 +1050,7 @@ static bool _trigger(struct pl330_thread *thrd)
 	return true;
 }
 
-static bool _start(struct pl330_thread *thrd)
+static bool pl330_start_thread(struct pl330_thread *thrd)
 {
 	switch (_state(thrd)) {
 	case PL330_STATE_FAULT_COMPLETING:
@@ -1702,7 +1702,7 @@ static int pl330_update(struct pl330_dmac *pl330)
 			thrd->req_running = -1;
 
 			/* Get going again ASAP */
-			_start(thrd);
+			pl330_start_thread(thrd);
 
 			/* For now, just make a list of callbacks to be done */
 			list_add_tail(&descdone->rqd, &pl330->req_done);
@@ -2089,7 +2089,7 @@ static void pl330_tasklet(struct tasklet_struct *t)
 	} else {
 		/* Make sure the PL330 Channel thread is active */
 		spin_lock(&pch->thread->dmac->lock);
-		_start(pch->thread);
+		pl330_start_thread(pch->thread);
 		spin_unlock(&pch->thread->dmac->lock);
 	}
 
@@ -2107,7 +2107,7 @@ static void pl330_tasklet(struct tasklet_struct *t)
 			if (power_down) {
 				pch->active = true;
 				spin_lock(&pch->thread->dmac->lock);
-				_start(pch->thread);
+				pl330_start_thread(pch->thread);
 				spin_unlock(&pch->thread->dmac->lock);
 				power_down = false;
 			}
Signature · 1 line
2.39.2
author maintainer
#18 lore raw

PATCH 18/286 crypto: x86/aria - Use 16 byte alignment for GFNI constant vectors

In reply to Greg Kroah-Hartman

From: Ard Biesheuvel <[email protected]>

[ Upstream commit 6ab39f99927eed605728b02d512438d828183c97 ]

The GFNI routines in the AVX version of the ARIA implementation now use explicit VMOVDQA instructions to load the constant input vectors, which means they must be 16 byte aligned. So ensure that this is the case, by dropping the section split and the incorrect .align 8 directive, and emitting the constants into the 16-byte aligned section instead.

Note that the AVX2 version of this code deviates from this pattern, and does not require a similar fix, given that it loads these contants as 8-byte memory operands, for which AVX2 permits any alignment.

Cc Taehee Yoo <[email protected]>
Fixes 8b84475318641c2b ("crypto: x86/aria-avx - Do not use avx2 instructions")
Reported-by [email protected]
---
 arch/x86/crypto/aria-aesni-avx-asm_64.S | 2 --
 1 file changed, 2 deletions(-)
arch/x86/crypto/aria-aesni-avx-asm_64.S +0 -2
index 9243f6289d34b..ed6c22fb16720 100644
@@ -773,8 +773,6 @@
 	.octa 0x3F893781E95FE1576CDA64D2BA0CB204
 
 #ifdef CONFIG_AS_GFNI
-.section	.rodata.cst8, "aM", @progbits, 8
-.align 8
 /* AES affine: */
 #define tf_aff_const BV8(1, 1, 0, 0, 0, 1, 1, 0)
 .Ltf_aff_bitmatrix:
Signature · 1 line
2.39.2
author maintainer
#19 lore raw

PATCH 19/286 riscv: Fix unused variable warning when BUILTIN_DTB is set

In reply to Greg Kroah-Hartman

From: Alexandre Ghiti <[email protected]>

[ Upstream commit 33d418da6f476b15e4510e0a590062583f63cd36 ]

commit ef69d2559fe9 ("riscv: Move early dtb mapping into the fixmap region") wrongly moved the #ifndef CONFIG_BUILTIN_DTB surrounding the pa variable definition in create_fdt_early_page_table(), so move it back to its right place to quiet the following warning:

../arch/riscv/mm/init.c: In function ‘create_fdt_early_page_table’:
../arch/riscv/mm/init.c:925:12: warning: unused variable ‘pa’ [-Wunused-variable]
925 | uintptr_t pa = dtb_pa & ~(PMD_SIZE - 1);

Fixes ef69d2559fe9 ("riscv: Move early dtb mapping into the fixmap region")
---
 arch/riscv/mm/init.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/riscv/mm/init.c +1 -1
index 6ebb75a9a6b9f..dc1793bf01796 100644
@@ -846,9 +846,9 @@ static void __init create_kernel_page_table(pgd_t *pgdir, bool early)
 static void __init create_fdt_early_page_table(uintptr_t fix_fdt_va,
 					       uintptr_t dtb_pa)
 {
+#ifndef CONFIG_BUILTIN_DTB
 	uintptr_t pa = dtb_pa & ~(PMD_SIZE - 1);
 
-#ifndef CONFIG_BUILTIN_DTB
 	/* Make sure the fdt fixmap address is always aligned on PMD size */
 	BUILD_BUG_ON(FIX_FDT % (PMD_SIZE / PAGE_SIZE));
 
Signature · 1 line
2.39.2
author maintainer
#2 lore raw

PATCH 2/286 phy: amlogic: phy-meson-g12a-mipi-dphy-analog: fix CNTL2_DIF_TX_CTL0 value

In reply to Greg Kroah-Hartman

From: Neil Armstrong <[email protected]>

[ Upstream commit b949193011540bb17cf1da7795ec42af1b875203 ]

Use the same CNTL2_DIF_TX_CTL0 value used by the vendor, it was reported fixing timings issues.

---
 drivers/phy/amlogic/phy-meson-g12a-mipi-dphy-analog.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/phy/amlogic/phy-meson-g12a-mipi-dphy-analog.c +1 -1
index c14089fa7db49..cabdddbbabfd7 100644
@@ -70,7 +70,7 @@ static int phy_g12a_mipi_dphy_analog_power_on(struct phy *phy)
 		     HHI_MIPI_CNTL1_BANDGAP);
 
 	regmap_write(priv->regmap, HHI_MIPI_CNTL2,
-		     FIELD_PREP(HHI_MIPI_CNTL2_DIF_TX_CTL0, 0x459) |
+		     FIELD_PREP(HHI_MIPI_CNTL2_DIF_TX_CTL0, 0x45a) |
 		     FIELD_PREP(HHI_MIPI_CNTL2_DIF_TX_CTL1, 0x2680));
 
 	reg = DSI_LANE_CLK;
Signature · 1 line
2.39.2
author maintainer
#20 lore raw

PATCH 20/286 net/mlx5e: TC, Remove unused vf_tun variable

In reply to Greg Kroah-Hartman

From: Paul Blakey <[email protected]>

[ Upstream commit 7195d9a0c8df0ab78c9d7a587809d16b00432426 ]

vf_tun is being assigned but never being used so remove it.

Stable-dep-of: 37c3b9fa7ccf ("net/mlx5e: Prevent encap offload when neigh update is running")

---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 15 +++------------
 1 file changed, 3 insertions(+), 12 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +3 -12
index 53acd9a8a4c35..d9e651d33f8b8 100644
@@ -1792,8 +1792,7 @@ set_encap_dests(struct mlx5e_priv *priv,
 static void
 clean_encap_dests(struct mlx5e_priv *priv,
 		  struct mlx5e_tc_flow *flow,
-		  struct mlx5_flow_attr *attr,
-		  bool *vf_tun)
+		  struct mlx5_flow_attr *attr)
 {
 	struct mlx5_esw_flow_attr *esw_attr;
 	int out_index;
@@ -1802,17 +1801,11 @@ clean_encap_dests(struct mlx5e_priv *priv,
 		return;
 
 	esw_attr = attr->esw_attr;
-	*vf_tun = false;
 
 	for (out_index = 0; out_index < MLX5_MAX_FLOW_FWD_VPORTS; out_index++) {
 		if (!(esw_attr->dests[out_index].flags & MLX5_ESW_DEST_ENCAP))
 			continue;
 
-		if (esw_attr->dests[out_index].flags &
-		    MLX5_ESW_DEST_CHAIN_WITH_SRC_PORT_CHANGE &&
-		    !esw_attr->dest_int_port)
-			*vf_tun = true;
-
 		mlx5e_detach_encap(priv, flow, attr, out_index);
 		kfree(attr->parse_attr->tun_info[out_index]);
 	}
@@ -2046,7 +2039,6 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
 	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
 	struct mlx5_flow_attr *attr = flow->attr;
 	struct mlx5_esw_flow_attr *esw_attr;
-	bool vf_tun;
 
 	esw_attr = attr->esw_attr;
 	mlx5e_put_flow_tunnel_id(flow);
@@ -2068,7 +2060,7 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
 	if (flow->decap_route)
 		mlx5e_detach_decap_route(priv, flow);
 
-	clean_encap_dests(priv, flow, attr, &vf_tun);
+	clean_encap_dests(priv, flow, attr);
 
 	mlx5_tc_ct_match_del(get_ct_priv(priv), &flow->attr->ct_attr);
 
@@ -4452,7 +4444,6 @@ static void
 mlx5_free_flow_attr(struct mlx5e_tc_flow *flow, struct mlx5_flow_attr *attr)
 {
 	struct mlx5_core_dev *counter_dev = get_flow_counter_dev(flow);
-	bool vf_tun;
 
 	if (!attr)
 		return;
@@ -4460,7 +4451,7 @@ mlx5_free_flow_attr(struct mlx5e_tc_flow *flow, struct mlx5_flow_attr *attr)
 	if (attr->post_act_handle)
 		mlx5e_tc_post_act_del(get_post_action(flow->priv), attr->post_act_handle);
 
-	clean_encap_dests(flow->priv, flow, attr, &vf_tun);
+	clean_encap_dests(flow->priv, flow, attr);
 
 	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT)
 		mlx5_fc_destroy(counter_dev, attr->counter);
Signature · 1 line
2.39.2
author maintainer
#21 lore raw

PATCH 21/286 net/mlx5e: TC, Move main flow attribute cleanup to helper func

In reply to Greg Kroah-Hartman

From: Paul Blakey <[email protected]>

[ Upstream commit a830ec485e8368a29e328d08e2eb28750bbc483f ]

Actions that can be setup per flow attribute (so per split rule) are cleaned up from mlx5_free_flow_attr(), mlx5e_tc_del_fdb_flow(), and free_flow_post_acts().

Remove the duplication by re-using the helper function for the main flow attribute and split rules attributes.

Stable-dep-of: 37c3b9fa7ccf ("net/mlx5e: Prevent encap offload when neigh update is running")

---
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   | 27 +++++++------------
 1 file changed, 9 insertions(+), 18 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +9 -18
index d9e651d33f8b8..3827dbc2b39c5 100644
@@ -183,7 +183,8 @@ static struct lock_class_key tc_ht_wq_key;
 
 static void mlx5e_put_flow_tunnel_id(struct mlx5e_tc_flow *flow);
 static void free_flow_post_acts(struct mlx5e_tc_flow *flow);
-static void mlx5_free_flow_attr(struct mlx5e_tc_flow *flow, struct mlx5_flow_attr *attr);
+static void mlx5_free_flow_attr_actions(struct mlx5e_tc_flow *flow,
+					struct mlx5_flow_attr *attr);
 
 void
 mlx5e_tc_match_to_reg_match(struct mlx5_flow_spec *spec,
@@ -2028,7 +2029,7 @@ static void free_branch_attr(struct mlx5e_tc_flow *flow, struct mlx5_flow_attr *
 	if (!attr)
 		return;
 
-	mlx5_free_flow_attr(flow, attr);
+	mlx5_free_flow_attr_actions(flow, attr);
 	kvfree(attr->parse_attr);
 	kfree(attr);
 }
@@ -2060,18 +2061,8 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
 	if (flow->decap_route)
 		mlx5e_detach_decap_route(priv, flow);
 
-	clean_encap_dests(priv, flow, attr);
-
 	mlx5_tc_ct_match_del(get_ct_priv(priv), &flow->attr->ct_attr);
 
-	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) {
-		mlx5e_mod_hdr_dealloc(&attr->parse_attr->mod_hdr_acts);
-		mlx5e_tc_detach_mod_hdr(priv, flow, attr);
-	}
-
-	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT)
-		mlx5_fc_destroy(esw_attr->counter_dev, attr->counter);
-
 	if (esw_attr->int_port)
 		mlx5e_tc_int_port_put(mlx5e_get_int_port_priv(priv), esw_attr->int_port);
 
@@ -2084,8 +2075,7 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
 	mlx5e_tc_act_stats_del_flow(get_act_stats_handle(priv), flow);
 
 	free_flow_post_acts(flow);
-	free_branch_attr(flow, attr->branch_true);
-	free_branch_attr(flow, attr->branch_false);
+	mlx5_free_flow_attr_actions(flow, attr);
 
 	kvfree(attr->esw_attr->rx_tun_attr);
 	kvfree(attr->parse_attr);
@@ -3804,9 +3794,7 @@ free_flow_post_acts(struct mlx5e_tc_flow *flow)
 		if (list_is_last(&attr->list, &flow->attrs))
 			break;
 
-		mlx5_free_flow_attr(flow, attr);
-		free_branch_attr(flow, attr->branch_true);
-		free_branch_attr(flow, attr->branch_false);
+		mlx5_free_flow_attr_actions(flow, attr);
 
 		list_del(&attr->list);
 		kvfree(attr->parse_attr);
@@ -4441,7 +4429,7 @@ mlx5_alloc_flow_attr(enum mlx5_flow_namespace_type type)
 }
 
 static void
-mlx5_free_flow_attr(struct mlx5e_tc_flow *flow, struct mlx5_flow_attr *attr)
+mlx5_free_flow_attr_actions(struct mlx5e_tc_flow *flow, struct mlx5_flow_attr *attr)
 {
 	struct mlx5_core_dev *counter_dev = get_flow_counter_dev(flow);
 
@@ -4460,6 +4448,9 @@ mlx5_free_flow_attr(struct mlx5e_tc_flow *flow, struct mlx5_flow_attr *attr)
 		mlx5e_mod_hdr_dealloc(&attr->parse_attr->mod_hdr_acts);
 		mlx5e_tc_detach_mod_hdr(flow->priv, flow, attr);
 	}
+
+	free_branch_attr(flow, attr->branch_true);
+	free_branch_attr(flow, attr->branch_false);
 }
 
 static int
Signature · 1 line
2.39.2
author maintainer
#22 lore raw

PATCH 22/286 net/mlx5e: Extract remaining tunnel encap code to dedicated file

In reply to Greg Kroah-Hartman

From: Chris Mi <[email protected]>

[ Upstream commit e2ab5aa11f191b54514f063a5b5c29f3559f4ab7 ]

Move set_encap_dests() and clean_encap_dests() to the tunnel encap dedicated file. And rename them to mlx5e_tc_tun_encap_dests_set() and mlx5e_tc_tun_encap_dests_unset().

No functional change in this patch. It is needed in the next patch.

Stable-dep-of: 37c3b9fa7ccf ("net/mlx5e: Prevent encap offload when neigh update is running")

---
 .../mellanox/mlx5/core/en/tc_tun_encap.c      | 83 +++++++++++++++++
 .../mellanox/mlx5/core/en/tc_tun_encap.h      |  9 ++
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   | 89 +------------------
 3 files changed, 94 insertions(+), 87 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c +83 -0
index fbb392d54fa51..7655526222570 100644
@@ -985,6 +985,89 @@ int mlx5e_attach_decap(struct mlx5e_priv *priv,
 	return err;
 }
 
+int mlx5e_tc_tun_encap_dests_set(struct mlx5e_priv *priv,
+				 struct mlx5e_tc_flow *flow,
+				 struct mlx5_flow_attr *attr,
+				 struct netlink_ext_ack *extack,
+				 bool *vf_tun)
+{
+	struct mlx5e_tc_flow_parse_attr *parse_attr;
+	struct mlx5_esw_flow_attr *esw_attr;
+	struct net_device *encap_dev = NULL;
+	struct mlx5e_rep_priv *rpriv;
+	struct mlx5e_priv *out_priv;
+	int out_index;
+	int err = 0;
+
+	if (!mlx5e_is_eswitch_flow(flow))
+		return 0;
+
+	parse_attr = attr->parse_attr;
+	esw_attr = attr->esw_attr;
+	*vf_tun = false;
+
+	for (out_index = 0; out_index < MLX5_MAX_FLOW_FWD_VPORTS; out_index++) {
+		struct net_device *out_dev;
+		int mirred_ifindex;
+
+		if (!(esw_attr->dests[out_index].flags & MLX5_ESW_DEST_ENCAP))
+			continue;
+
+		mirred_ifindex = parse_attr->mirred_ifindex[out_index];
+		out_dev = dev_get_by_index(dev_net(priv->netdev), mirred_ifindex);
+		if (!out_dev) {
+			NL_SET_ERR_MSG_MOD(extack, "Requested mirred device not found");
+			err = -ENODEV;
+			goto out;
+		}
+		err = mlx5e_attach_encap(priv, flow, attr, out_dev, out_index,
+					 extack, &encap_dev);
+		dev_put(out_dev);
+		if (err)
+			goto out;
+
+		if (esw_attr->dests[out_index].flags &
+		    MLX5_ESW_DEST_CHAIN_WITH_SRC_PORT_CHANGE &&
+		    !esw_attr->dest_int_port)
+			*vf_tun = true;
+
+		out_priv = netdev_priv(encap_dev);
+		rpriv = out_priv->ppriv;
+		esw_attr->dests[out_index].rep = rpriv->rep;
+		esw_attr->dests[out_index].mdev = out_priv->mdev;
+	}
+
+	if (*vf_tun && esw_attr->out_count > 1) {
+		NL_SET_ERR_MSG_MOD(extack, "VF tunnel encap with mirroring is not supported");
+		err = -EOPNOTSUPP;
+		goto out;
+	}
+
+out:
+	return err;
+}
+
+void mlx5e_tc_tun_encap_dests_unset(struct mlx5e_priv *priv,
+				    struct mlx5e_tc_flow *flow,
+				    struct mlx5_flow_attr *attr)
+{
+	struct mlx5_esw_flow_attr *esw_attr;
+	int out_index;
+
+	if (!mlx5e_is_eswitch_flow(flow))
+		return;
+
+	esw_attr = attr->esw_attr;
+
+	for (out_index = 0; out_index < MLX5_MAX_FLOW_FWD_VPORTS; out_index++) {
+		if (!(esw_attr->dests[out_index].flags & MLX5_ESW_DEST_ENCAP))
+			continue;
+
+		mlx5e_detach_encap(flow->priv, flow, attr, out_index);
+		kfree(attr->parse_attr->tun_info[out_index]);
+	}
+}
+
 static int cmp_route_info(struct mlx5e_route_key *a,
 			  struct mlx5e_route_key *b)
 {
drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.h +9 -0
index 8ad273dde40ee..5d7d67687cbcd 100644
@@ -30,6 +30,15 @@ int mlx5e_attach_decap_route(struct mlx5e_priv *priv,
 void mlx5e_detach_decap_route(struct mlx5e_priv *priv,
 			      struct mlx5e_tc_flow *flow);
 
+int mlx5e_tc_tun_encap_dests_set(struct mlx5e_priv *priv,
+				 struct mlx5e_tc_flow *flow,
+				 struct mlx5_flow_attr *attr,
+				 struct netlink_ext_ack *extack,
+				 bool *vf_tun);
+void mlx5e_tc_tun_encap_dests_unset(struct mlx5e_priv *priv,
+				    struct mlx5e_tc_flow *flow,
+				    struct mlx5_flow_attr *attr);
+
 struct ip_tunnel_info *mlx5e_dup_tun_info(const struct ip_tunnel_info *tun_info);
 
 int mlx5e_tc_set_attr_rx_tun(struct mlx5e_tc_flow *flow,
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +2 -87

107 lines not on this page · raw message

Signature · 1 line
2.39.2
author maintainer
#23 lore raw

PATCH 23/286 net/mlx5e: Prevent encap offload when neigh update is running

In reply to Greg Kroah-Hartman

From: Chris Mi <[email protected]>

[ Upstream commit 37c3b9fa7ccf5caad6d87ba4d42bf00be46be1cf ]

The cited commit adds a compeletion to remove dependency on rtnl lock. But it causes a deadlock for multiple encapsulations:

crash> bt ffff8aece8a64000
PID: 1514557 TASK: ffff8aece8a64000 CPU: 3 COMMAND: "tc"
#0 [ffffa6d14183f368] __schedule at ffffffffb8ba7f45
#1 [ffffa6d14183f3f8] schedule at ffffffffb8ba8418
#2 [ffffa6d14183f418] schedule_preempt_disabled at ffffffffb8ba8898 #3 [ffffa6d14183f428] __mutex_lock at ffffffffb8baa7f8
#4 [ffffa6d14183f4d0] mutex_lock_nested at ffffffffb8baabeb
#5 [ffffa6d14183f4e0] mlx5e_attach_encap at ffffffffc0f48c17 [mlx5_core] #6 [ffffa6d14183f628] mlx5e_tc_add_fdb_flow at ffffffffc0f39680 [mlx5_core] #7 [ffffa6d14183f688] __mlx5e_add_fdb_flow at ffffffffc0f3b636 [mlx5_core] #8 [ffffa6d14183f6f0] mlx5e_tc_add_flow at ffffffffc0f3bcdf [mlx5_core] #9 [ffffa6d14183f728] mlx5e_configure_flower at ffffffffc0f3c1d1 [mlx5_core]
#10 [ffffa6d14183f790] mlx5e_rep_setup_tc_cls_flower at ffffffffc0f3d529 [mlx5_core] #11 [ffffa6d14183f7a0] mlx5e_rep_setup_tc_cb at ffffffffc0f3d714 [mlx5_core] #12 [ffffa6d14183f7b0] tc_setup_cb_add at ffffffffb8931bb8
#13 [ffffa6d14183f810] fl_hw_replace_filter at ffffffffc0dae901 [cls_flower] #14 [ffffa6d14183f8d8] fl_change at ffffffffc0db5c57 [cls_flower]
#15 [ffffa6d14183f970] tc_new_tfilter at ffffffffb8936047
#16 [ffffa6d14183fac8] rtnetlink_rcv_msg at ffffffffb88c7c31
#17 [ffffa6d14183fb50] netlink_rcv_skb at ffffffffb8942853
#18 [ffffa6d14183fbc0] rtnetlink_rcv at ffffffffb88c1835
#19 [ffffa6d14183fbd0] netlink_unicast at ffffffffb8941f27
#20 [ffffa6d14183fc18] netlink_sendmsg at ffffffffb8942245
#21 [ffffa6d14183fc98] sock_sendmsg at ffffffffb887d482
#22 [ffffa6d14183fcb8] ____sys_sendmsg at ffffffffb887d81a
#23 [ffffa6d14183fd38] ___sys_sendmsg at ffffffffb88806e2
#24 [ffffa6d14183fe90] __sys_sendmsg at ffffffffb88807a2
#25 [ffffa6d14183ff28] __x64_sys_sendmsg at ffffffffb888080f
#26 [ffffa6d14183ff38] do_syscall_64 at ffffffffb8b9b6a8
#27 [ffffa6d14183ff50] entry_SYSCALL_64_after_hwframe at ffffffffb8c0007c crash> bt 0xffff8aeb07544000
PID: 1110766 TASK: ffff8aeb07544000 CPU: 0 COMMAND: "kworker/u20:9"
#0 [ffffa6d14e6b7bd8] __schedule at ffffffffb8ba7f45
#1 [ffffa6d14e6b7c68] schedule at ffffffffb8ba8418
#2 [ffffa6d14e6b7c88] schedule_timeout at ffffffffb8baef88
#3 [ffffa6d14e6b7d10] wait_for_completion at ffffffffb8ba968b
#4 [ffffa6d14e6b7d60] mlx5e_take_all_encap_flows at ffffffffc0f47ec4 [mlx5_core] #5 [ffffa6d14e6b7da0] mlx5e_rep_update_flows at ffffffffc0f3e734 [mlx5_core] #6 [ffffa6d14e6b7df8] mlx5e_rep_neigh_update at ffffffffc0f400bb [mlx5_core] #7 [ffffa6d14e6b7e50] process_one_work at ffffffffb80acc9c
#8 [ffffa6d14e6b7ed0] worker_thread at ffffffffb80ad012
#9 [ffffa6d14e6b7f10] kthread at ffffffffb80b615d
#10 [ffffa6d14e6b7f50] ret_from_fork at ffffffffb8001b2f

After the first encap is attached, flow will be added to encap entry's flows list. If neigh update is running at this time, the following encaps of the flow can't hold the encap_tbl_lock and sleep. If neigh update thread is waiting for that flow's init_done, deadlock happens.

Fix it by holding lock outside of the for loop. If neigh update is running, prevent encap flows from offloading. Since the lock is held outside of the for loop, concurrent creation of encap entries is not allowed. So remove unnecessary wait_for_completion call for res_ready.

Fixes 95435ad7999b ("net/mlx5e: Only access fully initialized flows in neigh update")
---
 .../mellanox/mlx5/core/en/tc_tun_encap.c      | 37 ++++++++++---------
 1 file changed, 20 insertions(+), 17 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c +20 -17
index 7655526222570..bbab164eab546 100644
@@ -492,6 +492,19 @@ void mlx5e_encap_put(struct mlx5e_priv *priv, struct mlx5e_encap_entry *e)
 	mlx5e_encap_dealloc(priv, e);
 }
 
+static void mlx5e_encap_put_locked(struct mlx5e_priv *priv, struct mlx5e_encap_entry *e)
+{
+	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
+
+	lockdep_assert_held(&esw->offloads.encap_tbl_lock);
+
+	if (!refcount_dec_and_test(&e->refcnt))
+		return;
+	list_del(&e->route_list);
+	hash_del_rcu(&e->encap_hlist);
+	mlx5e_encap_dealloc(priv, e);
+}
+
 static void mlx5e_decap_put(struct mlx5e_priv *priv, struct mlx5e_decap_entry *d)
 {
 	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
@@ -785,6 +798,8 @@ int mlx5e_attach_encap(struct mlx5e_priv *priv,
 	uintptr_t hash_key;
 	int err = 0;
 
+	lockdep_assert_held(&esw->offloads.encap_tbl_lock);
+
 	parse_attr = attr->parse_attr;
 	tun_info = parse_attr->tun_info[out_index];
 	mpls_info = &parse_attr->mpls_info[out_index];
@@ -798,7 +813,6 @@ int mlx5e_attach_encap(struct mlx5e_priv *priv,
 
 	hash_key = hash_encap_info(&key);
 
-	mutex_lock(&esw->offloads.encap_tbl_lock);
 	e = mlx5e_encap_get(priv, &key, hash_key);
 
 	/* must verify if encap is valid or not */
@@ -809,15 +823,6 @@ int mlx5e_attach_encap(struct mlx5e_priv *priv,
 			goto out_err;
 		}
 
-		mutex_unlock(&esw->offloads.encap_tbl_lock);
-		wait_for_completion(&e->res_ready);
-
-		/* Protect against concurrent neigh update. */
-		mutex_lock(&esw->offloads.encap_tbl_lock);
-		if (e->compl_result < 0) {
-			err = -EREMOTEIO;
-			goto out_err;
-		}
 		goto attach_flow;
 	}
 
@@ -846,15 +851,12 @@ int mlx5e_attach_encap(struct mlx5e_priv *priv,
 	INIT_LIST_HEAD(&e->flows);
 	hash_add_rcu(esw->offloads.encap_tbl, &e->encap_hlist, hash_key);
 	tbl_time_before = mlx5e_route_tbl_get_last_update(priv);
-	mutex_unlock(&esw->offloads.encap_tbl_lock);
 
 	if (family == AF_INET)
 		err = mlx5e_tc_tun_create_header_ipv4(priv, mirred_dev, e);
 	else if (family == AF_INET6)
 		err = mlx5e_tc_tun_create_header_ipv6(priv, mirred_dev, e);
 
-	/* Protect against concurrent neigh update. */
-	mutex_lock(&esw->offloads.encap_tbl_lock);
 	complete_all(&e->res_ready);
 	if (err) {
 		e->compl_result = err;
@@ -889,18 +891,15 @@ int mlx5e_attach_encap(struct mlx5e_priv *priv,
 	} else {
 		flow_flag_set(flow, SLOW);
 	}
-	mutex_unlock(&esw->offloads.encap_tbl_lock);
 
 	return err;
 
 out_err:
-	mutex_unlock(&esw->offloads.encap_tbl_lock);
 	if (e)
-		mlx5e_encap_put(priv, e);
+		mlx5e_encap_put_locked(priv, e);
 	return err;
 
 out_err_init:
-	mutex_unlock(&esw->offloads.encap_tbl_lock);
 	kfree(tun_info);
 	kfree(e);
 	return err;
@@ -996,6 +995,7 @@ int mlx5e_tc_tun_encap_dests_set(struct mlx5e_priv *priv,
 	struct net_device *encap_dev = NULL;
 	struct mlx5e_rep_priv *rpriv;
 	struct mlx5e_priv *out_priv;
+	struct mlx5_eswitch *esw;
 	int out_index;
 	int err = 0;
 
@@ -1006,6 +1006,8 @@ int mlx5e_tc_tun_encap_dests_set(struct mlx5e_priv *priv,
 	esw_attr = attr->esw_attr;
 	*vf_tun = false;
 
+	esw = priv->mdev->priv.eswitch;
+	mutex_lock(&esw->offloads.encap_tbl_lock);
 	for (out_index = 0; out_index < MLX5_MAX_FLOW_FWD_VPORTS; out_index++) {
 		struct net_device *out_dev;
 		int mirred_ifindex;
@@ -1044,6 +1046,7 @@ int mlx5e_tc_tun_encap_dests_set(struct mlx5e_priv *priv,
 	}
 
 out:
+	mutex_unlock(&esw->offloads.encap_tbl_lock);
 	return err;
 }
 
Signature · 1 line
2.39.2
author maintainer
#24 lore raw

PATCH 24/286 net/mlx5e: Consider internal buffers size in port buffer calculations

In reply to Greg Kroah-Hartman

From: Maher Sanalla <[email protected]>

[ Upstream commit 81fe2be062915e2a2fdc494c3cd90e946e946c25 ]

Currently, when a user triggers a change in port buffer headroom (buffers 0-7), the driver checks that the requested headroom does not exceed the total port buffer size. However, this check does not take into account the internal buffers (buffers 8-9), which are also part of the total port buffer. This can result in treating invalid port buffer change requests as valid, causing unintended changes to the shared buffer.

To address this, include the internal buffers size in the calculation of available port buffer space which ensures that port buffer requests do not exceed the correct limit.

Furthermore, remove internal buffers (8-9) size from the total_size calculation as these buffers are reserved for internal use and are not exposed to the user.

While at it, add verbosity to the debug prints in mlx5e_port_query_buffer() function to ease future debugging.

Fixes ecdf2dadee8e ("net/mlx5e: Receive buffer support for DCBX")
---
 .../mellanox/mlx5/core/en/port_buffer.c       | 42 ++++++++++++-------
 .../mellanox/mlx5/core/en/port_buffer.h       |  8 ++--
 .../ethernet/mellanox/mlx5/core/en_dcbnl.c    |  7 ++--
 3 files changed, 36 insertions(+), 21 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/port_buffer.c +27 -15
index 7ac1ad9c46de0..0d78527451bca 100644
@@ -51,7 +51,7 @@ int mlx5e_port_query_buffer(struct mlx5e_priv *priv,
 	if (err)
 		goto out;
 
-	for (i = 0; i < MLX5E_MAX_BUFFER; i++) {
+	for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++) {
 		buffer = MLX5_ADDR_OF(pbmc_reg, out, buffer[i]);
 		port_buffer->buffer[i].lossy =
 			MLX5_GET(bufferx_reg, buffer, lossy);
@@ -73,14 +73,24 @@ int mlx5e_port_query_buffer(struct mlx5e_priv *priv,
 			  port_buffer->buffer[i].lossy);
 	}
 
-	port_buffer->headroom_size = total_used;
+	port_buffer->internal_buffers_size = 0;
+	for (i = MLX5E_MAX_NETWORK_BUFFER; i < MLX5E_TOTAL_BUFFERS; i++) {
+		buffer = MLX5_ADDR_OF(pbmc_reg, out, buffer[i]);
+		port_buffer->internal_buffers_size +=
+			MLX5_GET(bufferx_reg, buffer, size) * port_buff_cell_sz;
+	}
+
 	port_buffer->port_buffer_size =
 		MLX5_GET(pbmc_reg, out, port_buffer_size) * port_buff_cell_sz;
-	port_buffer->spare_buffer_size =
-		port_buffer->port_buffer_size - total_used;
-
-	mlx5e_dbg(HW, priv, "total buffer size=%d, spare buffer size=%d\n",
-		  port_buffer->port_buffer_size,
+	port_buffer->headroom_size = total_used;
+	port_buffer->spare_buffer_size = port_buffer->port_buffer_size -
+					 port_buffer->internal_buffers_size -
+					 port_buffer->headroom_size;
+
+	mlx5e_dbg(HW, priv,
+		  "total buffer size=%u, headroom buffer size=%u, internal buffers size=%u, spare buffer size=%u\n",
+		  port_buffer->port_buffer_size, port_buffer->headroom_size,
+		  port_buffer->internal_buffers_size,
 		  port_buffer->spare_buffer_size);
 out:
 	kfree(out);
@@ -206,11 +216,11 @@ static int port_update_pool_cfg(struct mlx5_core_dev *mdev,
 	if (!MLX5_CAP_GEN(mdev, sbcam_reg))
 		return 0;
 
-	for (i = 0; i < MLX5E_MAX_BUFFER; i++)
+	for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++)
 		lossless_buff_count += ((port_buffer->buffer[i].size) &&
 				       (!(port_buffer->buffer[i].lossy)));
 
-	for (i = 0; i < MLX5E_MAX_BUFFER; i++) {
+	for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++) {
 		p = select_sbcm_params(&port_buffer->buffer[i], lossless_buff_count);
 		err = mlx5e_port_set_sbcm(mdev, 0, i,
 					  MLX5_INGRESS_DIR,
@@ -293,7 +303,7 @@ static int port_set_buffer(struct mlx5e_priv *priv,
 	if (err)
 		goto out;
 
-	for (i = 0; i < MLX5E_MAX_BUFFER; i++) {
+	for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++) {
 		void *buffer = MLX5_ADDR_OF(pbmc_reg, in, buffer[i]);
 		u64 size = port_buffer->buffer[i].size;
 		u64 xoff = port_buffer->buffer[i].xoff;
@@ -351,7 +361,7 @@ static int update_xoff_threshold(struct mlx5e_port_buffer *port_buffer,
 {
 	int i;
 
-	for (i = 0; i < MLX5E_MAX_BUFFER; i++) {
+	for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++) {
 		if (port_buffer->buffer[i].lossy) {
 			port_buffer->buffer[i].xoff = 0;
 			port_buffer->buffer[i].xon  = 0;
@@ -408,7 +418,7 @@ static int update_buffer_lossy(struct mlx5_core_dev *mdev,
 	int err;
 	int i;
 
-	for (i = 0; i < MLX5E_MAX_BUFFER; i++) {
+	for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++) {
 		prio_count = 0;
 		lossy_count = 0;
 
@@ -515,7 +525,7 @@ int mlx5e_port_manual_buffer_config(struct mlx5e_priv *priv,
 
 	if (change & MLX5E_PORT_BUFFER_PRIO2BUFFER) {
 		update_prio2buffer = true;
-		for (i = 0; i < MLX5E_MAX_BUFFER; i++)
+		for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++)
 			mlx5e_dbg(HW, priv, "%s: requested to map prio[%d] to buffer %d\n",
 				  __func__, i, prio2buffer[i]);
 
@@ -530,7 +540,7 @@ int mlx5e_port_manual_buffer_config(struct mlx5e_priv *priv,
 	}
 
 	if (change & MLX5E_PORT_BUFFER_SIZE) {
-		for (i = 0; i < MLX5E_MAX_BUFFER; i++) {
+		for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++) {
 			mlx5e_dbg(HW, priv, "%s: buffer[%d]=%d\n", __func__, i, buffer_size[i]);
 			if (!port_buffer.buffer[i].lossy && !buffer_size[i]) {
 				mlx5e_dbg(HW, priv, "%s: lossless buffer[%d] size cannot be zero\n",
@@ -544,7 +554,9 @@ int mlx5e_port_manual_buffer_config(struct mlx5e_priv *priv,
 
 		mlx5e_dbg(HW, priv, "%s: total buffer requested=%d\n", __func__, total_used);
 
-		if (total_used > port_buffer.port_buffer_size)
+		if (total_used > port_buffer.headroom_size &&
+		    (total_used - port_buffer.headroom_size) >
+			    port_buffer.spare_buffer_size)
 			return -EINVAL;
 
 		update_buffer = true;
drivers/net/ethernet/mellanox/mlx5/core/en/port_buffer.h +5 -3
index a6ef118de758f..f4a19ffbb641c 100644
@@ -35,7 +35,8 @@
 #include "en.h"
 #include "port.h"
 
-#define MLX5E_MAX_BUFFER 8
+#define MLX5E_MAX_NETWORK_BUFFER 8
+#define MLX5E_TOTAL_BUFFERS 10
 #define MLX5E_DEFAULT_CABLE_LEN 7 /* 7 meters */
 
 #define MLX5_BUFFER_SUPPORTED(mdev) (MLX5_CAP_GEN(mdev, pcam_reg) && \
@@ -60,8 +61,9 @@ struct mlx5e_bufferx_reg {
 struct mlx5e_port_buffer {
 	u32                       port_buffer_size;
 	u32                       spare_buffer_size;
-	u32                       headroom_size;
-	struct mlx5e_bufferx_reg  buffer[MLX5E_MAX_BUFFER];
+	u32                       headroom_size;	  /* Buffers 0-7 */
+	u32                       internal_buffers_size;  /* Buffers 8-9 */
+	struct mlx5e_bufferx_reg  buffer[MLX5E_MAX_NETWORK_BUFFER];
 };
 
 int mlx5e_port_manual_buffer_config(struct mlx5e_priv *priv,
drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c +4 -3
index 89de92d064836..ebee52a8361aa 100644
@@ -926,9 +926,10 @@ static int mlx5e_dcbnl_getbuffer(struct net_device *dev,
 	if (err)
 		return err;
 
-	for (i = 0; i < MLX5E_MAX_BUFFER; i++)
+	for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++)
 		dcb_buffer->buffer_size[i] = port_buffer.buffer[i].size;
-	dcb_buffer->total_size = port_buffer.port_buffer_size;
+	dcb_buffer->total_size = port_buffer.port_buffer_size -
+				 port_buffer.internal_buffers_size;
 
 	return 0;
 }
@@ -970,7 +971,7 @@ static int mlx5e_dcbnl_setbuffer(struct net_device *dev,
 	if (err)
 		return err;
 
-	for (i = 0; i < MLX5E_MAX_BUFFER; i++) {
+	for (i = 0; i < MLX5E_MAX_NETWORK_BUFFER; i++) {
 		if (port_buffer.buffer[i].size != dcb_buffer->buffer_size[i]) {
 			changed |= MLX5E_PORT_BUFFER_SIZE;
 			buffer_size = dcb_buffer->buffer_size;
Signature · 1 line
2.39.2
author maintainer
#25 lore raw

PATCH 25/286 net/mlx5e: Do not update SBCM when prio2buffer command is invalid

In reply to Greg Kroah-Hartman

From: Maher Sanalla <[email protected]>

[ Upstream commit 623efc4cbd6115db36716e31037cb6d1f3ce6754 ]

The shared buffer pools configuration which are stored in the SBCM register are updated when the user changes the prio2buffer mapping.

However, in case the user desired prio2buffer change is invalid, which can occur due to mapping a lossless priority to a not large enough buffer, the SBCM update should not be performed, as the user command is failed.

Thus, Perform the SBCM update only after xoff threshold calculation is performed and the user prio2buffer mapping is validated.

Fixes a440030d8946 ("net/mlx5e: Update shared buffer along with device buffer changes")
---
 drivers/net/ethernet/mellanox/mlx5/core/en/port_buffer.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/port_buffer.c +2 -2
index 0d78527451bca..7e8e96cc5cd08 100644
@@ -442,11 +442,11 @@ static int update_buffer_lossy(struct mlx5_core_dev *mdev,
 	}
 
 	if (changed) {
-		err = port_update_pool_cfg(mdev, port_buffer);
+		err = update_xoff_threshold(port_buffer, xoff, max_mtu, port_buff_cell_sz);
 		if (err)
 			return err;
 
-		err = update_xoff_threshold(port_buffer, xoff, max_mtu, port_buff_cell_sz);
+		err = port_update_pool_cfg(mdev, port_buffer);
 		if (err)
 			return err;
 
Signature · 1 line
2.39.2
author maintainer
#26 lore raw

PATCH 26/286 net/mlx5: Drain health before unregistering devlink

In reply to Greg Kroah-Hartman

From: Shay Drory <[email protected]>

[ Upstream commit 824c8dc4a470040bf0e56ba716543839c2498d49 ]

mlx5 health mechanism is using devlink APIs, which are using devlink notify APIs. After the cited patch, using devlink notify APIs after devlink is unregistered triggers a WARN_ON().
Hence, drain health WQ before devlink is unregistered.

Fixes cf530217408e ("devlink: Notify users when objects are accessible")
---
 drivers/net/ethernet/mellanox/mlx5/core/main.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/main.c +4 -3
index 62327b52f1acf..504179dbff9dd 100644
@@ -1790,14 +1790,15 @@ static void remove_one(struct pci_dev *pdev)
 	struct devlink *devlink = priv_to_devlink(dev);
 
 	set_bit(MLX5_BREAK_FW_WAIT, &dev->intf_state);
-	/* mlx5_drain_fw_reset() is using devlink APIs. Hence, we must drain
-	 * fw_reset before unregistering the devlink.
+	/* mlx5_drain_fw_reset() and mlx5_drain_health_wq() are using
+	 * devlink notify APIs.
+	 * Hence, we must drain them before unregistering the devlink.
 	 */
 	mlx5_drain_fw_reset(dev);
+	mlx5_drain_health_wq(dev);
 	devlink_unregister(devlink);
 	mlx5_sriov_disable(pdev);
 	mlx5_crdump_disable(dev);
-	mlx5_drain_health_wq(dev);
 	mlx5_uninit_one(dev);
 	mlx5_pci_close(dev);
 	mlx5_mdev_uninit(dev);
Signature · 1 line
2.39.2
author maintainer
#27 lore raw

PATCH 27/286 net/mlx5: SF, Drain health before removing device

In reply to Greg Kroah-Hartman

From: Shay Drory <[email protected]>

[ Upstream commit b4646da0573fae9dfa2b8f1f10936cb6eedd7230 ]

There is no point in recovery during device removal. Also, if health work started need to wait for it to avoid races and NULL pointer access.

Hence, drain health WQ before removing device.

Fixes 1958fc2f0712 ("net/mlx5: SF, Add auxiliary device driver")
---
 drivers/net/ethernet/mellanox/mlx5/core/sf/dev/driver.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/mellanox/mlx5/core/sf/dev/driver.c +1 -0
index a7377619ba6f2..2424cdf9cca99 100644
@@ -63,6 +63,7 @@ static void mlx5_sf_dev_remove(struct auxiliary_device *adev)
 	struct mlx5_sf_dev *sf_dev = container_of(adev, struct mlx5_sf_dev, adev);
 	struct devlink *devlink = priv_to_devlink(sf_dev->mdev);
 
+	mlx5_drain_health_wq(sf_dev->mdev);
 	devlink_unregister(devlink);
 	mlx5_uninit_one(sf_dev->mdev);
 	iounmap(sf_dev->mdev->iseg);
Signature · 1 line
2.39.2
author maintainer
#28 lore raw

PATCH 28/286 net/mlx5: fw_tracer, Fix event handling

In reply to Greg Kroah-Hartman

From: Shay Drory <[email protected]>

[ Upstream commit 341a80de2468f481b1f771683709b5649cbfe513 ]

mlx5 driver needs to parse traces with event_id inside the range of first_string_trace and num_string_trace. However, mlx5 is parsing all events with event_id >= first_string_trace.

Fix it by checking for the correct range.

Fixes c71ad41ccb0c ("net/mlx5: FW tracer, events handling")
---
 drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.c +1 -1
index f40497823e65f..7c0f2adbea000 100644
@@ -490,7 +490,7 @@ static void poll_trace(struct mlx5_fw_tracer *tracer,
 				(u64)timestamp_low;
 		break;
 	default:
-		if (tracer_event->event_id >= tracer->str_db.first_string_trace ||
+		if (tracer_event->event_id >= tracer->str_db.first_string_trace &&
 		    tracer_event->event_id <= tracer->str_db.first_string_trace +
 					      tracer->str_db.num_string_trace) {
 			tracer_event->type = TRACER_EVENT_TYPE_STRING;
Signature · 1 line
2.39.2
author maintainer
#29 lore raw

PATCH 29/286 net/mlx5e: Use query_special_contexts cmd only once per mdev

In reply to Greg Kroah-Hartman

From: Dragos Tatulea <[email protected]>

[ Upstream commit 1db1f21caebbb1b6e9b1e7657df613616be3fb49 ]

Don't query the firmware so many times (num rqs * num wqes * wqe frags) because it slows down linearly the interface creation time when the product is larger. Do it only once per mdev and store the result in mlx5e_param.

Due to helper function being called from different files, move it to an appropriate location. Rename the function with a proper prefix and add a small cleanup.

This fix applies only for legacy rq.

Fixes 1b1e4868836a ("net/mlx5e: Use query_special_contexts for mkeys")
---
 drivers/net/ethernet/mellanox/mlx5/core/en.h  |  1 +
 .../net/ethernet/mellanox/mlx5/core/en_main.c | 24 +++----------------
 drivers/net/ethernet/mellanox/mlx5/core/mr.c  | 21 ++++++++++++++++
 include/linux/mlx5/driver.h                   |  1 +
 4 files changed, 26 insertions(+), 21 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en.h +1 -0
index 4a19ef4a98110..5ee90a394fff9 100644
@@ -327,6 +327,7 @@ struct mlx5e_params {
 	unsigned int sw_mtu;
 	int hard_mtu;
 	bool ptp_rx;
+	__be32 terminate_lkey_be;
 };
 
 static inline u8 mlx5e_get_dcb_num_tc(struct mlx5e_params *params)
drivers/net/ethernet/mellanox/mlx5/core/en_main.c +3 -21
index 579c2d217fdc6..ff579d7d8432c 100644
@@ -668,26 +668,6 @@ static void mlx5e_rq_free_shampo(struct mlx5e_rq *rq)
 	mlx5e_rq_shampo_hd_free(rq);
 }
 
-static __be32 mlx5e_get_terminate_scatter_list_mkey(struct mlx5_core_dev *dev)
-{
-	u32 out[MLX5_ST_SZ_DW(query_special_contexts_out)] = {};
-	u32 in[MLX5_ST_SZ_DW(query_special_contexts_in)] = {};
-	int res;
-
-	if (!MLX5_CAP_GEN(dev, terminate_scatter_list_mkey))
-		return MLX5_TERMINATE_SCATTER_LIST_LKEY;
-
-	MLX5_SET(query_special_contexts_in, in, opcode,
-		 MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS);
-	res = mlx5_cmd_exec_inout(dev, query_special_contexts, in, out);
-	if (res)
-		return MLX5_TERMINATE_SCATTER_LIST_LKEY;
-
-	res = MLX5_GET(query_special_contexts_out, out,
-		       terminate_scatter_list_mkey);
-	return cpu_to_be32(res);
-}
-
 static int mlx5e_alloc_rq(struct mlx5e_params *params,
 			  struct mlx5e_xsk_param *xsk,
 			  struct mlx5e_rq_param *rqp,
@@ -852,7 +832,7 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params,
 			/* check if num_frags is not a pow of two */
 			if (rq->wqe.info.num_frags < (1 << rq->wqe.info.log_num_frags)) {
 				wqe->data[f].byte_count = 0;
-				wqe->data[f].lkey = mlx5e_get_terminate_scatter_list_mkey(mdev);
+				wqe->data[f].lkey = params->terminate_lkey_be;
 				wqe->data[f].addr = 0;
 			}
 		}
@@ -4973,6 +4953,8 @@ void mlx5e_build_nic_params(struct mlx5e_priv *priv, struct mlx5e_xsk *xsk, u16
 	/* RQ */
 	mlx5e_build_rq_params(mdev, params);
 
+	params->terminate_lkey_be = mlx5_core_get_terminate_scatter_list_mkey(mdev);
+
 	params->packet_merge.timeout = mlx5e_choose_lro_timeout(mdev, MLX5E_DEFAULT_LRO_TIMEOUT);
 
 	/* CQ moderation params */
drivers/net/ethernet/mellanox/mlx5/core/mr.c +21 -0
index 9d735c343a3b8..678f0be813752 100644
@@ -32,6 +32,7 @@
 
 #include <linux/kernel.h>
 #include <linux/mlx5/driver.h>
+#include <linux/mlx5/qp.h>
 #include "mlx5_core.h"
 
 int mlx5_core_create_mkey(struct mlx5_core_dev *dev, u32 *mkey, u32 *in,
@@ -122,3 +123,23 @@ int mlx5_core_destroy_psv(struct mlx5_core_dev *dev, int psv_num)
 	return mlx5_cmd_exec_in(dev, destroy_psv, in);
 }
 EXPORT_SYMBOL(mlx5_core_destroy_psv);
+
+__be32 mlx5_core_get_terminate_scatter_list_mkey(struct mlx5_core_dev *dev)
+{
+	u32 out[MLX5_ST_SZ_DW(query_special_contexts_out)] = {};
+	u32 in[MLX5_ST_SZ_DW(query_special_contexts_in)] = {};
+	u32 mkey;
+
+	if (!MLX5_CAP_GEN(dev, terminate_scatter_list_mkey))
+		return MLX5_TERMINATE_SCATTER_LIST_LKEY;
+
+	MLX5_SET(query_special_contexts_in, in, opcode,
+		 MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS);
+	if (mlx5_cmd_exec_inout(dev, query_special_contexts, in, out))
+		return MLX5_TERMINATE_SCATTER_LIST_LKEY;
+
+	mkey = MLX5_GET(query_special_contexts_out, out,
+			terminate_scatter_list_mkey);
+	return cpu_to_be32(mkey);
+}
+EXPORT_SYMBOL(mlx5_core_get_terminate_scatter_list_mkey);
include/linux/mlx5/driver.h +1 -0
index 7e225e41d55b8..68a3183d5d589 100644
@@ -1088,6 +1088,7 @@ void mlx5_cmdif_debugfs_cleanup(struct mlx5_core_dev *dev);
 int mlx5_core_create_psv(struct mlx5_core_dev *dev, u32 pdn,
 			 int npsvs, u32 *sig_index);
 int mlx5_core_destroy_psv(struct mlx5_core_dev *dev, int psv_num);
+__be32 mlx5_core_get_terminate_scatter_list_mkey(struct mlx5_core_dev *dev);
 void mlx5_core_put_rsc(struct mlx5_core_rsc_common *common);
 int mlx5_query_odp_caps(struct mlx5_core_dev *dev,
 			struct mlx5_odp_caps *odp_caps);
Signature · 1 line
2.39.2
author maintainer
#3 lore raw

PATCH 3/286 RDMA/efa: Fix unsupported page sizes in device

In reply to Greg Kroah-Hartman

From: Yonatan Nachum <[email protected]>

[ Upstream commit 866422cdddcdf59d8c68e9472d49ba1be29b5fcf ]

Device uses 4KB size blocks for user pages indirect list while the driver creates those blocks with the size of PAGE_SIZE of the kernel. On kernels with PAGE_SIZE different than 4KB (ARM RHEL), this leads to a failure on register MR with indirect list because of the miss communication between driver and device.

---
 drivers/infiniband/hw/efa/efa_verbs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/efa/efa_verbs.c +1 -1
index 31454643f8c54..f9526a4c75b26 100644
@@ -1397,7 +1397,7 @@ static int pbl_continuous_initialize(struct efa_dev *dev,
  */
 static int pbl_indirect_initialize(struct efa_dev *dev, struct pbl_context *pbl)
 {
-	u32 size_in_pages = DIV_ROUND_UP(pbl->pbl_buf_size_in_bytes, PAGE_SIZE);
+	u32 size_in_pages = DIV_ROUND_UP(pbl->pbl_buf_size_in_bytes, EFA_CHUNK_PAYLOAD_SIZE);
 	struct scatterlist *sgl;
 	int sg_dma_cnt, err;
 
Signature · 1 line
2.39.2
author maintainer
#30 lore raw

PATCH 30/286 net/mlx5e: CT: Use per action stats

In reply to Greg Kroah-Hartman

From: Paul Blakey <[email protected]>

[ Upstream commit 13aca17b450e87a8de4e4a3b3ad454efbc576740 ]

CT action can miss in a middle of an action list, use per action stats to correctly report stats for missed packets.

Stable-dep-of: 5d862ec631f3 ("net/mlx5: Fix post parse infra to only parse every action once")

---
 .../mellanox/mlx5/core/en/tc/act/act.h        |  2 ++
 .../mellanox/mlx5/core/en/tc/act/ct.c         |  9 ++++++++-
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   | 19 +++++++++++++++++--
 3 files changed, 27 insertions(+), 3 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h +2 -0
index 8346557eeaf63..cdcddf6e1b08b 100644
@@ -56,6 +56,8 @@ struct mlx5e_tc_act {
 				   const struct flow_action_entry *act,
 				   struct mlx5_flow_attr *attr);
 
+	bool (*is_missable)(const struct flow_action_entry *act);
+
 	int (*offload_action)(struct mlx5e_priv *priv,
 			      struct flow_offload_action *fl_act,
 			      struct flow_action_entry *act);
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c +8 -1
index a829c94289c10..fce1c0fd24535 100644
@@ -95,10 +95,17 @@ tc_act_is_multi_table_act_ct(struct mlx5e_priv *priv,
 	return true;
 }
 
+static bool
+tc_act_is_missable_ct(const struct flow_action_entry *act)
+{
+	return !(act->ct.action & TCA_CT_ACT_CLEAR);
+}
+
 struct mlx5e_tc_act mlx5e_tc_act_ct = {
 	.can_offload = tc_act_can_offload_ct,
 	.parse_action = tc_act_parse_ct,
-	.is_multi_table_act = tc_act_is_multi_table_act_ct,
 	.post_parse = tc_act_post_parse_ct,
+	.is_multi_table_act = tc_act_is_multi_table_act_ct,
+	.is_missable = tc_act_is_missable_ct,
 };
 
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +17 -2
index b5d23b159f345..82b76dcc05ac1 100644
@@ -3974,7 +3974,9 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 	enum mlx5_flow_namespace_type ns_type;
 	struct mlx5e_priv *priv = flow->priv;
 	struct flow_action_entry *act, **_act;
+	struct mlx5_flow_attr *prev_attr;
 	struct mlx5e_tc_act *tc_act;
+	bool is_missable;
 	int err, i;
 
 	flow_action_reorder.num_entries = flow_action->num_entries;
@@ -3991,6 +3993,9 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 	flow_action_for_each(i, _act, &flow_action_reorder) {
 		jump_state.jump_target = false;
 		act = *_act;
+		is_missable = false;
+		prev_attr = attr;
+
 		tc_act = mlx5e_tc_act_get(act->id, ns_type);
 		if (!tc_act) {
 			NL_SET_ERR_MSG_MOD(extack, "Not implemented offload action");
@@ -4014,14 +4019,14 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 			goto out_free;
 
 		parse_state->actions |= attr->action;
-		if (!tc_act->stats_action)
-			attr->tc_act_cookies[attr->tc_act_cookies_count++] = act->cookie;
 
 		/* Split attr for multi table act if not the last act. */
 		if (jump_state.jump_target ||
 		    (tc_act->is_multi_table_act &&
 		    tc_act->is_multi_table_act(priv, act, attr) &&
 		    i < flow_action_reorder.num_entries - 1)) {
+			is_missable = tc_act->is_missable ? tc_act->is_missable(act) : false;
+
 			err = mlx5e_tc_act_post_parse(parse_state, flow_action, attr, ns_type);
 			if (err)
 				goto out_free;
@@ -4034,6 +4039,16 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 
 			list_add(&attr->list, &flow->attrs);
 		}
+
+		if (is_missable) {
+			/* Add counter to prev, and assign act to new (next) attr */
+			prev_attr->action |= MLX5_FLOW_CONTEXT_ACTION_COUNT;
+			flow_flag_set(flow, USE_ACT_STATS);
+
+			attr->tc_act_cookies[attr->tc_act_cookies_count++] = act->cookie;
+		} else if (!tc_act->stats_action) {
+			prev_attr->tc_act_cookies[prev_attr->tc_act_cookies_count++] = act->cookie;
+		}
 	}
 
 	kfree(flow_action_reorder.entries);
Signature · 1 line
2.39.2
author maintainer
#4 lore raw

PATCH 4/286 RDMA/hns: Fix timeout attr in query qp for HIP08

In reply to Greg Kroah-Hartman

From: Chengchang Tang <[email protected]>

[ Upstream commit 58caa2a51ad4fd21763696cc6c4defc9fc1b4b4f ]

On HIP08, the queried timeout attr is different from the timeout attr configured by the user.

It is found by rdma-core testcase test_rdmacm_async_traffic:

======================================================================
FAIL: test_rdmacm_async_traffic (tests.test_rdmacm.CMTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
File "./tests/test_rdmacm.py", line 33, in test_rdmacm_async_traffic
self.two_nodes_rdmacm_traffic(CMAsyncConnection, self.rdmacm_traffic,
File "./tests/base.py", line 382, in two_nodes_rdmacm_traffic
raise(res)
AssertionError

Fixes 926a01dc000d ("RDMA/hns: Add QP operations support for hip08 SoC")
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 17 ++++++++++++++---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h |  2 ++
 2 files changed, 16 insertions(+), 3 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +14 -3
index dbf97fe5948ff..79954d99dc7a3 100644
@@ -5093,7 +5093,6 @@ static int hns_roce_v2_set_abs_fields(struct ib_qp *ibqp,
 static bool check_qp_timeout_cfg_range(struct hns_roce_dev *hr_dev, u8 *timeout)
 {
 #define QP_ACK_TIMEOUT_MAX_HIP08 20
-#define QP_ACK_TIMEOUT_OFFSET 10
 #define QP_ACK_TIMEOUT_MAX 31
 
 	if (hr_dev->pci_dev->revision == PCI_REVISION_ID_HIP08) {
@@ -5102,7 +5101,7 @@ static bool check_qp_timeout_cfg_range(struct hns_roce_dev *hr_dev, u8 *timeout)
 				   "local ACK timeout shall be 0 to 20.\n");
 			return false;
 		}
-		*timeout += QP_ACK_TIMEOUT_OFFSET;
+		*timeout += HNS_ROCE_V2_QP_ACK_TIMEOUT_OFS_HIP08;
 	} else if (hr_dev->pci_dev->revision > PCI_REVISION_ID_HIP08) {
 		if (*timeout > QP_ACK_TIMEOUT_MAX) {
 			ibdev_warn(&hr_dev->ib_dev,
@@ -5388,6 +5387,18 @@ static int hns_roce_v2_query_qpc(struct hns_roce_dev *hr_dev, u32 qpn,
 	return ret;
 }
 
+static u8 get_qp_timeout_attr(struct hns_roce_dev *hr_dev,
+			      struct hns_roce_v2_qp_context *context)
+{
+	u8 timeout;
+
+	timeout = (u8)hr_reg_read(context, QPC_AT);
+	if (hr_dev->pci_dev->revision == PCI_REVISION_ID_HIP08)
+		timeout -= HNS_ROCE_V2_QP_ACK_TIMEOUT_OFS_HIP08;
+
+	return timeout;
+}
+
 static int hns_roce_v2_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
 				int qp_attr_mask,
 				struct ib_qp_init_attr *qp_init_attr)
@@ -5465,7 +5476,7 @@ static int hns_roce_v2_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
 	qp_attr->max_dest_rd_atomic = 1 << hr_reg_read(&context, QPC_RR_MAX);
 
 	qp_attr->min_rnr_timer = (u8)hr_reg_read(&context, QPC_MIN_RNR_TIME);
-	qp_attr->timeout = (u8)hr_reg_read(&context, QPC_AT);
+	qp_attr->timeout = get_qp_timeout_attr(hr_dev, &context);
 	qp_attr->retry_cnt = hr_reg_read(&context, QPC_RETRY_NUM_INIT);
 	qp_attr->rnr_retry = hr_reg_read(&context, QPC_RNR_NUM_INIT);
 
drivers/infiniband/hw/hns/hns_roce_hw_v2.h +2 -0
index af9d00225cdf5..b5a336e182f83 100644
@@ -72,6 +72,8 @@
 #define HNS_ROCE_V2_IDX_ENTRY_SZ		4
 
 #define HNS_ROCE_V2_SCCC_SZ			32
+#define HNS_ROCE_V2_QP_ACK_TIMEOUT_OFS_HIP08    10
+
 #define HNS_ROCE_V3_SCCC_SZ			64
 #define HNS_ROCE_V3_GMV_ENTRY_SZ		32
 
Signature · 1 line
2.39.2
author maintainer
#5 lore raw

PATCH 5/286 RDMA/hns: Fix base address table allocation

In reply to Greg Kroah-Hartman

From: Chengchang Tang <[email protected]>

[ Upstream commit 7f3969b14f356dd65fa95b3528eb05c32e68bc06 ]

For hns, the specification of an entry like resource (E.g. WQE/CQE/EQE) depends on BT page size, buf page size and hopnum. For user mode, the buf page size depends on UMEM. Therefore, the actual specification is controlled by BT page size and hopnum.

The current BT page size and hopnum are obtained from firmware. This makes the driver inflexible and introduces unnecessary constraints. Resource allocation failures occur in many scenarios.

This patch will calculate whether the BT page size set by firmware is sufficient before allocating BT, and increase the BT page size if it is insufficient.

Fixes 1133401412a9 ("RDMA/hns: Optimize base address table config flow for qp buffer")
---
 drivers/infiniband/hw/hns/hns_roce_mr.c | 43 +++++++++++++++++++++++++
 1 file changed, 43 insertions(+)
drivers/infiniband/hw/hns/hns_roce_mr.c +43 -0
index 37a5cf62f88b4..14376490ac226 100644
@@ -33,6 +33,7 @@
 
 #include <linux/vmalloc.h>
 #include <rdma/ib_umem.h>
+#include <linux/math.h>
 #include "hns_roce_device.h"
 #include "hns_roce_cmd.h"
 #include "hns_roce_hem.h"
@@ -909,6 +910,44 @@ static int mtr_init_buf_cfg(struct hns_roce_dev *hr_dev,
 	return page_cnt;
 }
 
+static u64 cal_pages_per_l1ba(unsigned int ba_per_bt, unsigned int hopnum)
+{
+	return int_pow(ba_per_bt, hopnum - 1);
+}
+
+static unsigned int cal_best_bt_pg_sz(struct hns_roce_dev *hr_dev,
+				      struct hns_roce_mtr *mtr,
+				      unsigned int pg_shift)
+{
+	unsigned long cap = hr_dev->caps.page_size_cap;
+	struct hns_roce_buf_region *re;
+	unsigned int pgs_per_l1ba;
+	unsigned int ba_per_bt;
+	unsigned int ba_num;
+	int i;
+
+	for_each_set_bit_from(pg_shift, &cap, sizeof(cap) * BITS_PER_BYTE) {
+		if (!(BIT(pg_shift) & cap))
+			continue;
+
+		ba_per_bt = BIT(pg_shift) / BA_BYTE_LEN;
+		ba_num = 0;
+		for (i = 0; i < mtr->hem_cfg.region_count; i++) {
+			re = &mtr->hem_cfg.region[i];
+			if (re->hopnum == 0)
+				continue;
+
+			pgs_per_l1ba = cal_pages_per_l1ba(ba_per_bt, re->hopnum);
+			ba_num += DIV_ROUND_UP(re->count, pgs_per_l1ba);
+		}
+
+		if (ba_num <= ba_per_bt)
+			return pg_shift;
+	}
+
+	return 0;
+}
+
 static int mtr_alloc_mtt(struct hns_roce_dev *hr_dev, struct hns_roce_mtr *mtr,
 			 unsigned int ba_page_shift)
 {
@@ -917,6 +956,10 @@ static int mtr_alloc_mtt(struct hns_roce_dev *hr_dev, struct hns_roce_mtr *mtr,
 
 	hns_roce_hem_list_init(&mtr->hem_list);
 	if (!cfg->is_direct) {
+		ba_page_shift = cal_best_bt_pg_sz(hr_dev, mtr, ba_page_shift);
+		if (!ba_page_shift)
+			return -ERANGE;
+
 		ret = hns_roce_hem_list_request(hr_dev, &mtr->hem_list,
 						cfg->region, cfg->region_count,
 						ba_page_shift);
Signature · 1 line
2.39.2
author maintainer
#6 lore raw

PATCH 6/286 RDMA/hns: Modify the value of long message loopback slice

In reply to Greg Kroah-Hartman

From: Yangyang Li <[email protected]>

[ Upstream commit 56518a603fd2bf74762d176ac980572db84a3e14 ]

Long message loopback slice is used for achieving traffic balance between QPs. It prevents the problem that QPs with large traffic occupying the hardware pipeline for a long time and QPs with small traffic cannot be scheduled.

Currently, its maximum value is set to 16K, which means only after a QP sends 16K will the second QP be scheduled. This value is too large, which will lead to unbalanced traffic scheduling, and thus it needs to be modified.

The setting range of the long message loopback slice is modified to be from 1024 (the lower limit supported by hardware) to mtu. Actual testing shows that this value can significantly reduce error in hardware traffic scheduling.

This solution is compatible with both HIP08 and HIP09. The modified lp_pktn_ini has a maximum value of 2 (when mtu is 256), so the range checking code for lp_pktn_ini is no longer necessary and needs to be deleted.

Fixes 0e60778efb07 ("RDMA/hns: Modify the value of MAX_LP_MSG_LEN to meet hardware compatibility")
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +3 -5
index 79954d99dc7a3..9369f93afaedd 100644
@@ -4664,11 +4664,9 @@ static int modify_qp_init_to_rtr(struct ib_qp *ibqp,
 	mtu = ib_mtu_enum_to_int(ib_mtu);
 	if (WARN_ON(mtu <= 0))
 		return -EINVAL;
-#define MAX_LP_MSG_LEN 16384
-	/* MTU * (2 ^ LP_PKTN_INI) shouldn't be bigger than 16KB */
-	lp_pktn_ini = ilog2(MAX_LP_MSG_LEN / mtu);
-	if (WARN_ON(lp_pktn_ini >= 0xF))
-		return -EINVAL;
+#define MIN_LP_MSG_LEN 1024
+	/* mtu * (2 ^ lp_pktn_ini) should be in the range of 1024 to mtu */
+	lp_pktn_ini = ilog2(max(mtu, MIN_LP_MSG_LEN) / mtu);
 
 	if (attr_mask & IB_QP_PATH_MTU) {
 		hr_reg_write(context, QPC_MTU, ib_mtu);
Signature · 1 line
2.39.2
author maintainer
#7 lore raw

PATCH 7/286 dmaengine: at_xdmac: fix potential Oops in at_xdmac_prep_interleaved()

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

[ Upstream commit 4d43acb145c363626d76f49febb4240c488cd1cf ]

There are two place if the at_xdmac_interleaved_queue_desc() fails which could lead to a NULL dereference where "first" is NULL and we call list_add_tail(&first->desc_node, ...). In the first caller, the return is not checked so add a check for that. In the next caller, the return is checked but if it fails on the first iteration through the loop then it will lead to a NULL pointer dereference.

Fixes 4e5385784e69 ("dmaengine: at_xdmac: handle numf > 1")
Fixes 62b5cb757f1d ("dmaengine: at_xdmac: fix memory leak in interleaved mode")
---
 drivers/dma/at_xdmac.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/dma/at_xdmac.c +5 -2
index 96f1b69f8a75e..ab13704f27f11 100644
@@ -1102,6 +1102,8 @@ at_xdmac_prep_interleaved(struct dma_chan *chan,
 							NULL,
 							src_addr, dst_addr,
 							xt, xt->sgl);
+		if (!first)
+			return NULL;
 
 		/* Length of the block is (BLEN+1) microblocks. */
 		for (i = 0; i < xt->numf - 1; i++)
@@ -1132,8 +1134,9 @@ at_xdmac_prep_interleaved(struct dma_chan *chan,
 							       src_addr, dst_addr,
 							       xt, chunk);
 			if (!desc) {
-				list_splice_tail_init(&first->descs_list,
-						      &atchan->free_descs_list);
+				if (first)
+					list_splice_tail_init(&first->descs_list,
+							      &atchan->free_descs_list);
 				return NULL;
 			}
 
Signature · 1 line
2.39.2
author maintainer
#8 lore raw

PATCH 8/286 RDMA/bnxt_re: Fix a possible memory leak

In reply to Greg Kroah-Hartman

From: Kalesh AP <[email protected]>

[ Upstream commit 349e3c0cf239cc01d58a1e6c749e171de014cd6a ]

Inside bnxt_qplib_create_cq(), when the check for NULL DPI fails, driver returns directly without freeing the memory allocated inside bnxt_qplib_alloc_init_hwq() routine.

Fixed this by moving the check for NULL DPI before invoking bnxt_qplib_alloc_init_hwq().

---
 drivers/infiniband/hw/bnxt_re/qplib_fp.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/infiniband/hw/bnxt_re/qplib_fp.c +6 -5
index 96e581ced50e2..ab2cc1c67f70b 100644
@@ -2043,6 +2043,12 @@ int bnxt_qplib_create_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq)
 	u32 pg_sz_lvl;
 	int rc;
 
+	if (!cq->dpi) {
+		dev_err(&rcfw->pdev->dev,
+			"FP: CREATE_CQ failed due to NULL DPI\n");
+		return -EINVAL;
+	}
+
 	hwq_attr.res = res;
 	hwq_attr.depth = cq->max_wqe;
 	hwq_attr.stride = sizeof(struct cq_base);
@@ -2054,11 +2060,6 @@ int bnxt_qplib_create_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq)
 
 	RCFW_CMD_PREP(req, CREATE_CQ, cmd_flags);
 
-	if (!cq->dpi) {
-		dev_err(&rcfw->pdev->dev,
-			"FP: CREATE_CQ failed due to NULL DPI\n");
-		return -EINVAL;
-	}
 	req.dpi = cpu_to_le32(cq->dpi->dpi);
 	req.cq_handle = cpu_to_le64(cq->cq_handle);
 	req.cq_size = cpu_to_le32(cq->hwq.max_elements);
Signature · 1 line
2.39.2
author maintainer
#9 lore raw

PATCH 9/286 RDMA/bnxt_re: Fix return value of bnxt_re_process_raw_qp_pkt_rx

In reply to Greg Kroah-Hartman

From: Kalesh AP <[email protected]>

[ Upstream commit 0fa0d520e2a878cb4c94c4dc84395905d3f14f54 ]

bnxt_re_process_raw_qp_pkt_rx() always return 0 and ignores the return value of bnxt_re_post_send_shadow_qp().

---
 drivers/infiniband/hw/bnxt_re/ib_verbs.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/infiniband/hw/bnxt_re/ib_verbs.c +1 -3
index 989edc7896338..94222de1d3719 100644
@@ -3241,9 +3241,7 @@ static int bnxt_re_process_raw_qp_pkt_rx(struct bnxt_re_qp *gsi_qp,
 	udwr.remote_qkey = gsi_sqp->qplib_qp.qkey;
 
 	/* post data received  in the send queue */
-	rc = bnxt_re_post_send_shadow_qp(rdev, gsi_sqp, swr);
-
-	return 0;
+	return bnxt_re_post_send_shadow_qp(rdev, gsi_sqp, swr);
 }
 
 static void bnxt_re_process_res_rawqp1_wc(struct ib_wc *wc,
Signature · 1 line
2.39.2

Greg Kroah-Hartman

author maintainer
#0 lore raw

cover 6.3.7-rc1 review

This is the start of the stable review cycle for the 6.3.7 release. There are 286 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri, 09 Jun 2023 20:07:31 +0000. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

	https://www.kernel.org/pub/linux/kernel/v6.x/stable-review/patch-6.3.7-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-6.3.y
and the diffstat can be found below.

thanks,

greg k-h

-------------
Pseudo-Shortlog of commits:

Greg Kroah-Hartman <[email protected]>

    Linux 6.3.7-rc1

Theodore Ts'o <[email protected]>
ext4: enable the lazy init thread when remounting read/write

Nathan Chancellor <[email protected]>
riscv: vmlinux.lds.S: Explicitly handle '.got' section

Darrick J. Wong <[email protected]>
xfs: verify buffer contents when we skip log replay

Vasant Hegde <[email protected]>
iommu/amd/pgtbl_v2: Fix domain max address

Lino Sanfilippo <[email protected]>
tpm, tpm_tis: Request threaded interrupt handler

Jim Wylder <[email protected]>
regmap: Account for register length when chunking

Roberto Sassu <[email protected]>
KEYS: asymmetric: Copy sig and digest in public_key_verify_signature()

Kuan-Ting Chen <[email protected]>
ksmbd: fix multiple out-of-bounds read during context decoding

Kuan-Ting Chen <[email protected]>
ksmbd: fix slab-out-of-bounds read in smb2_handle_negotiate

Namjae Jeon <[email protected]>
ksmbd: fix incorrect AllocationSize set in smb2_get_info

Namjae Jeon <[email protected]>

    ksmbd: fix UAF issue from opinfo->conn

Namjae Jeon <[email protected]>
ksmbd: fix credit count leakage

Sean Christopherson <[email protected]>
KVM: x86: Bail from kvm_recalculate_phys_map() if x2APIC ID is out-of-bounds

Sean Christopherson <[email protected]>
KVM: x86: Account fastpath-only VM-Exits in vCPU stats

Sean Christopherson <[email protected]>
KVM: x86/mmu: Grab memslot for correct address space in NX recovery worker

Oliver Upton <[email protected]>
KVM: arm64: Drop last page ref in kvm_pgtable_stage2_free_removed()

Akihiko Odaki <[email protected]>
KVM: arm64: Populate fault info for watchpoint

Mirsad Goran Todorovac <[email protected]>
test_firmware: fix the memory leak of the allocated firmware buffer

Mirsad Goran Todorovac <[email protected]>
test_firmware: fix a memory leak with reqs buffer

Mirsad Goran Todorovac <[email protected]>
test_firmware: prevent race conditions by a correct implementation of locking

Maninder Singh <[email protected]>
powerpc/xmon: Use KSYM_NAME_LEN in array size

Herve Codina <[email protected]>
serial: cpm_uart: Fix a COMPILE_TEST dependency

Christophe JAILLET <[email protected]>
serial: 8250_tegra: Fix an error handling path in tegra_uart_probe()

Helge Deller <[email protected]>
fbcon: Fix null-ptr-deref in soft_cursor

Theodore Ts'o <[email protected]>
ext4: add lockdep annotations for i_data_sem for ea_inode's

Theodore Ts'o <[email protected]>
ext4: disallow ea_inodes with extended attributes

Theodore Ts'o <[email protected]>
ext4: set lockdep subclass for the ea_inode in ext4_xattr_inode_cache_find()

Theodore Ts'o <[email protected]>
ext4: add EA_INODE checking to ext4_iget()

Paolo Abeni <[email protected]>
mptcp: fix active subflow finalization

Paolo Abeni <[email protected]>
mptcp: fix connect timeout handling

Matthieu Baerts <[email protected]>
selftests: mptcp: userspace pm: skip if MPTCP is not supported

Matthieu Baerts <[email protected]>
selftests: mptcp: sockopt: skip if MPTCP is not supported

Matthieu Baerts <[email protected]>
selftests: mptcp: simult flows: skip if MPTCP is not supported

Matthieu Baerts <[email protected]>
selftests: mptcp: diag: skip if MPTCP is not supported

Matthieu Baerts <[email protected]>
selftests: mptcp: join: avoid using 'cmp --bytes'

Matthieu Baerts <[email protected]>
selftests: mptcp: join: skip if MPTCP is not supported

Matthieu Baerts <[email protected]>
selftests: mptcp: pm nl: skip if MPTCP is not supported

Matthieu Baerts <[email protected]>
selftests: mptcp: connect: skip if MPTCP is not supported

Pietro Borrello <[email protected]>
tracing/probe: trace_probe_primary_from_call(): checked list_first_entry

Steven Rostedt (Google) <[email protected]>
tracing/histograms: Allow variables to have some modifiers

Daniel Bristot de Oliveira <[email protected]>
tracing/timerlat: Always wakeup the timerlat thread

Arnd Bergmann <[email protected]>
mtdchar: mark bits of ioctl handler noinline

Paul Moore <[email protected]>
selinux: don't use make's grouped targets feature yet

Ben Noordhuis <[email protected]>
io_uring: undeprecate epoll_ctl support

Ism Hong <[email protected]>
riscv: perf: Fix callchain parse error with kernel tracepoint events

Peter Rosin <[email protected]>
dmaengine: at_hdmac: Extend the Flow Controller bitfield to three bits

Peter Rosin <[email protected]>
dmaengine: at_hdmac: Repair bitfield macros for peripheral ID handling

Benjamin Tissoires <[email protected]>
HID: hidpp: terminate retry loop on success

Sasha Levin <[email protected]>
btrfs: call btrfs_orig_bbio_end_io in btrfs_end_bio_work

Ekansh Gupta <[email protected]>
misc: fastrpc: Pass proper scm arguments for secure map request

Elliot Berman <[email protected]>
firmware: qcom_scm: Use fixed width src vm bitmap

Lucas De Marchi <[email protected]>
module/decompress: Fix error checking on zstd decompression

Lino Sanfilippo <[email protected]>
tpm, tpm_tis: correct tpm_tis_flags enumeration values

Jon Pan-Doh <[email protected]>
iommu/amd: Fix domain flush size when syncing iotlb

Gaurav Batra <[email protected]>
powerpc/iommu: Limit number of TCEs to 512 for H_STUFF_TCE hcall

Damien Le Moal <[email protected]>
block: fix revalidate performance regression

Johan Hovold <[email protected]>
phy: qcom-qmp-pcie-msm8996: fix init-count imbalance

Johan Hovold <[email protected]>
phy: qcom-qmp-combo: fix init-count imbalance

pengfuyuan <[email protected]>
btrfs: fix csum_tree_block page iteration to avoid tripping on -Werror=array-bounds

Sherry Sun <[email protected]>
tty: serial: fsl_lpuart: use UARTCTRL_TXINV to send break instead of UARTCTRL_SBK

Marek Vasut <[email protected]>
mmc: pwrseq: sd8787: Fix WILC CHIP_EN and RESETN toggling order

Deren Wu <[email protected]>
mmc: vub300: fix invalid response handling

Tim Huang <[email protected]>
drm/amd/pm: reverse mclk and fclk clocks levels for renoir

Tim Huang <[email protected]>
drm/amd/pm: reverse mclk and fclk clocks levels for yellow carp

Tim Huang <[email protected]>
drm/amd/pm: reverse mclk clocks levels for SMU v13.0.5

Guchun Chen <[email protected]>
drm/amd/pm: resolve reboot exception for si oland

Tim Huang <[email protected]>
drm/amd/pm: reverse mclk and fclk clocks levels for vangogh

Tim Huang <[email protected]>
drm/amd/pm: reverse mclk and fclk clocks levels for SMU v13.0.4

Ikshwaku Chauhan <[email protected]>
drm/amdgpu: enable tmz by default for GC 11.0.1

Damien Le Moal <[email protected]>
ata: libata-scsi: Use correct device no in ata_find_dev()

Bart Van Assche <[email protected]>
scsi: stex: Fix gcc 13 warnings

Richard Acayan <[email protected]>
misc: fastrpc: reject new invocations during device removal

Richard Acayan <[email protected]>
misc: fastrpc: return -EPIPE to invocations on device removal

Ekansh Gupta <[email protected]>
misc: fastrpc: Reassign memory ownership only for remote heap

Yu Kuai <[email protected]>
md/raid5: fix miscalculation of 'end_sector' in raid5_read_one_chunk()

Uttkarsh Aggarwal <[email protected]>
usb: gadget: f_fs: Add unbind event before functionfs_unbind

Frank Li <[email protected]>

    usb: cdns3: fix NCM gadget RX speed 20x slow than expection at iMX8QM

Marek Vasut <[email protected]>
dt-bindings: usb: snps,dwc3: Fix "snps,hsphy_interface" type

Sebastian Krzyszkowiak <[email protected]>
net: usb: qmi_wwan: Set DTR quirk for BroadMobi BM818

Lukas Bulwahn <[email protected]>
iio: dac: build ad5758 driver when AD5758 is selected

Sean Nyekjaer <[email protected]>
iio: adc: stm32-adc: skip adc-diff-channels setup if none is present

Paul Cercueil <[email protected]>
iio: adc: ad7192: Change "shorted" channels to differential

Rasmus Villemoes <[email protected]>
iio: addac: ad74413: fix resistance input processing

Marek Vasut <[email protected]>
iio: dac: mcp4725: Fix i2c_master_send() return value handling

Masahiro Honda <[email protected]>
iio: adc: ad_sigma_delta: Fix IRQ issue by setting IRQ_DISABLE_UNLAZY flag

Sean Nyekjaer <[email protected]>
iio: adc: stm32-adc: skip adc-channels setup if none is present

Matti Vaittinen <[email protected]>
iio: accel: kx022a fix irq getting

Frank Li <[email protected]>
iio: light: vcnl4035: fixed chip ID check

Geert Uytterhoeven <[email protected]>
dt-bindings: iio: adc: renesas,rcar-gyroadc: Fix adi,ad7476 compatible value

Jean-Baptiste Maneyrol <[email protected]>
iio: imu: inv_icm42600: fix timestamp reset

Nikita Zhandarovich <[email protected]>
HID: wacom: avoid integer overflow in wacom_intuos_inout()

Sung-Chi Li <[email protected]>
HID: google: add jewel USB id

ChiaEn Wu <[email protected]>
iio: adc: mt6370: Fix ibus and ibat scaling value of some specific vendor ID chips

Lars-Peter Clausen <[email protected]>
iio: ad4130: Make sure clock provider gets removed

Lars-Peter Clausen <[email protected]>
iio: tmag5273: Fix runtime PM leak on measurement error

Jiakai Luo <[email protected]>
iio: adc: mxs-lradc: fix the order of two cleanup operations

Hans de Goede <[email protected]>
iio: accel: st_accel: Fix invalid mount_matrix on devices without ACPI _ONT method

Aric Cyr <[email protected]>
drm/amd/display: Only wait for blank completion if OTG active

Masami Hiramatsu (Google) <[email protected]>
selftests/ftrace: Choose target function for filter test from samples

Laurent Pinchart <[email protected]>
media: uvcvideo: Don't expose unsupported formats to userspace

Francesco Dolcini <[email protected]>

    dt-bindings: serial: 8250_omap: add rs485-rts-active-high

K Prateek Nayak <[email protected]>
drivers: base: cacheinfo: Update cpu_map_populated during CPU Hotplug

K Prateek Nayak <[email protected]>
drivers: base: cacheinfo: Fix shared_cpu_map changes in event of CPU hotplug

Dan Carpenter <[email protected]>
mailbox: mailbox-test: fix a locking issue in mbox_test_message_write()

Pin-yen Lin <[email protected]>
media: mediatek: vcodec: Only apply 4K frame sizes on decoder formats

Fuad Tabba <[email protected]>
KVM: arm64: Reload PTE after invoking walker callback on preorder traversal

Like Xu <[email protected]>

    perf/x86/intel: Save/restore cpuc->active_pebs_data_cfg when using guest PEBS

Gleb Chesnokov <[email protected]>
scsi: qla2xxx: Fix NULL pointer dereference in target mode

Will Deacon <[email protected]>
KVM: arm64: Prevent unconditional donation of unmapped regions from the host

Jean-Philippe Brucker <[email protected]>
KVM: arm64: vgic: Fix locking comment

Jean-Philippe Brucker <[email protected]>
KVM: arm64: vgic: Wrap vgic_its_create() with config_lock

Jean-Philippe Brucker <[email protected]>
KVM: arm64: vgic: Fix a circular locking issue

Dan Carpenter <[email protected]>
iio: adc: imx93: fix a signedness bug in imx93_adc_read_raw()

Loic Poulain <[email protected]>
block: Deny writable memory mapping if block is read-only

Daniel Smith <[email protected]>
nvme-pci: Add quirk for Teamgroup MP33 SSD

Ming Lei <[email protected]>
ublk: fix AB-BA lockdep warning

Guchun Chen <[email protected]>
drm/amdgpu: skip disabling fence driver src_irqs when device is unplugged

Xiubo Li <[email protected]>
ceph: silence smatch warning in reconnect_caps_cb()

Arnd Bergmann <[email protected]>
atm: hide unused procfs functions

Rob Clark <[email protected]>
drm/msm: Be more shouty if per-process pgtables aren't working

Arnd Bergmann <[email protected]>
ALSA: oss: avoid missing-prototype warnings

Maurizio Lombardi <[email protected]>
nvme: do not let the user delete a ctrl before a complete initialization

Christoph Hellwig <[email protected]>
nvme-multipath: don't call blk_mark_disk_dead in nvme_mpath_remove_disk

Tom Rix <[email protected]>
netfilter: conntrack: define variables exp_nat_nla_policy and any_addr with CONFIG_NF_NAT

Kai-Heng Feng <[email protected]>
net: wwan: t7xx: Ensure init is completed before system sleep

Arnd Bergmann <[email protected]>
wifi: b43: fix incorrect __packed annotation

Wenchao Hao <[email protected]>
scsi: core: Decrease scsi_device's iorequest_cnt if dispatch failed

Po-Wen Kao <[email protected]>
scsi: ufs: core: Fix MCQ nr_hw_queues

Po-Wen Kao <[email protected]>
scsi: ufs: core: Rename symbol sizeof_utp_transfer_cmd_desc()

Po-Wen Kao <[email protected]>
scsi: ufs: core: Fix MCQ tag calculation

Ariel Malamud <[email protected]>
wifi: iwlwifi: mvm: Add locking to the rate read flow

Johannes Berg <[email protected]>
wifi: mac80211: recalc chanctx mindef before assigning

Johannes Berg <[email protected]>
wifi: mac80211: consider reserved chanctx for mindef

Johannes Berg <[email protected]>
wifi: mac80211: simplify chanctx allocation

Linus Walleij <[email protected]>
arm64: vdso: Pass (void *) to virt_to_page()

Min-Hua Chen <[email protected]>
arm64/mm: mark private VM_FAULT_X defines as vm_fault_t

Dario Binacchi <[email protected]>
ARM: dts: stm32: add pin map for CAN controller on stm32f7

Yun Lu <[email protected]>
wifi: rtl8xxxu: fix authentication timeout due to incorrect RCR value

Rubén Gómez <[email protected]>
ACPI: resource: Add IRQ override quirk for LG UltraPC 17U70P

Alexander Gordeev <[email protected]>
s390/ipl: fix IPIB virtual vs physical address confusion

Alexander Gordeev <[email protected]>
s390/topology: honour nr_cpu_ids when adding CPUs

Holger Dengler <[email protected]>
s390/pkey: zeroize key blobs

Kai Vehmanen <[email protected]>
ASoC: SOF: pm: save io region state in case of errors in resume

Pierre-Louis Bossart <[email protected]>
ASoC: SOF: sof-client-probes: fix pm_runtime imbalance in error handling

Pierre-Louis Bossart <[email protected]>
ASoC: SOF: pcm: fix pm_runtime imbalance in error handling

Pierre-Louis Bossart <[email protected]>
ASoC: SOF: debug: conditionally bump runtime_pm counter on exceptions

Hyunwoo Kim <[email protected]>
media: dvb-core: Fix use-after-free due to race condition at dvb_ca_en50221

Takashi Iwai <[email protected]>
media: dvb-core: Fix kernel WARNING for blocking operation in wait_event*()

Hyunwoo Kim <[email protected]>
media: dvb-core: Fix use-after-free due to race at dvb_register_device()

Hyunwoo Kim <[email protected]>
media: dvb-core: Fix use-after-free due on race condition at dvb_net

Hyunwoo Kim <[email protected]>
media: dvb-core: Fix use-after-free on race condition at dvb_frontend

Krzysztof Kozlowski <[email protected]>
media: mn88443x: fix !CONFIG_OF error by drop of_match_ptr from ID table

Hyunwoo Kim <[email protected]>
media: ttusb-dec: fix memory leak in ttusb_dec_exit_dvb()

YongSu Yoo <[email protected]>
media: dvb_ca_en50221: fix a size write bug

Wei Chen <[email protected]>
media: netup_unidvb: fix irq init by register it at the end of probe

Wei Chen <[email protected]>
media: dvb-usb: dw2102: fix uninit-value in su3000_read_mac_address

Wei Chen <[email protected]>
media: dvb-usb: digitv: fix null-ptr-deref in digitv_i2c_xfer()

Zhang Shurong <[email protected]>
media: dvb-usb-v2: rtl28xxu: fix null-ptr-deref in rtl28xxu_i2c_xfer

Wei Chen <[email protected]>

    media: dvb-usb-v2: ce6230: fix null-ptr-deref in ce6230_i2c_master_xfer()

Wei Chen <[email protected]>

    media: dvb-usb-v2: ec168: fix null-ptr-deref in ec168_i2c_xfer()

Wei Chen <[email protected]>

    media: dvb-usb: az6027: fix three null-ptr-deref in az6027_i2c_xfer()

YongSu Yoo <[email protected]>
media: dvb_demux: fix a bug for the continuity counter

Paweł Anikiel <[email protected]>
ASoC: ssm2602: Add workaround for playback distortions

Alexandru Sorodoc <[email protected]>
ALSA: hda/realtek: Add quirks for ASUS GU604V and GU603V

Martin Povišer <[email protected]>
ASoC: dt-bindings: Adjust #sound-dai-cells on TI's single-DAI codecs

Aidan MacDonald <[email protected]>
ASoC: jz4740-i2s: Make I2S divider calculations more robust

Benedict Wong <[email protected]>
xfrm: Check if_id in inbound policy/secpath match

Johannes Berg <[email protected]>
um: harddog: fix modular build

V sujith kumar Reddy <[email protected]>
ASoC: SOF: amd: Fix NULL pointer crash in acp_sof_ipc_msg_data function

Hao Zeng <[email protected]>
cpupower:Fix resource leaks in sysfs_get_enabled()

Maxim Kochetkov <[email protected]>
ASoC: dwc: limit the number of overrun messages

Jeremy Soller <[email protected]>
ASoC: amd: yc: Add DMI entry to support System76 Pangolin 12

Adrian Huang <[email protected]>
nvme-pci: clamp max_hw_sectors based on DMA optimized limitation

Hristo Venev <[email protected]>
nvme-pci: add quirk for missing secondary temperature thresholds

Sagi Grimberg <[email protected]>
nvme-pci: add NVME_QUIRK_BOGUS_NID for HS-SSD-FUTURE 2048G

Guoqing Jiang <[email protected]>
block/rnbd: replace REQ_OP_FLUSH with REQ_OP_WRITE

Ivan Orlov <[email protected]>
nbd: Fix debugfs_create_dir error checking

Helge Deller <[email protected]>
fbdev: stifb: Fix info entry in sti_struct on error path

Helge Deller <[email protected]>

    fbdev: modedb: Add 1920x1080 at 60 Hz video mode

Zheng Wang <[email protected]>
fbdev: imsttfb: Fix use after free bug in imsttfb_probe

Yifan Zhang <[email protected]>
drm/amdgpu: set gfx9 onwards APU atomics support to be true

Thong Thai <[email protected]>
drm/amdgpu/nv: update VCN 3 max HEVC encoding resolution

Bob Peterson <[email protected]>
gfs2: Don't deref jdesc in evict

Liming Sun <[email protected]>
platform/mellanox: fix potential race in mlxbf-tmfifo driver

Julian Winkler <[email protected]>
platform/x86: intel_scu_pcidrv: Add back PCI ID for Medfield

Niklas Söderlund <[email protected]>
media: rcar-vin: Select correct interrupt mode for V4L2_FIELD_ALTERNATE

Niklas Söderlund <[email protected]>
media: rcar-vin: Fix NV12 size alignment

Niklas Söderlund <[email protected]>
media: rcar-vin: Gen3 can not scale NV12

Mario Limonciello <[email protected]>
hwmon: (k10temp) Add PCI ID for family 19, model 78h

Kan Liang <[email protected]>
perf/x86/intel/ds: Flush PEBS DS when changing PEBS_DATA_CFG

Haibo Li <[email protected]>
ARM: 9295/1: unwind:fix unwind abort for uleb128 case

Filipe Manana <[email protected]>
btrfs: abort transaction when sibling keys check fails for leaves

Jammy Huang <[email protected]>
drm/ast: Fix ARM compatibility

Lee Jones <[email protected]>
mailbox: mailbox-test: Fix potential double-free in mbox_test_message_write()

lyndonli <[email protected]>
drm/amdgpu: Use the default reset when loading or reloading the driver

Hans de Goede <[email protected]>
ASoC: Intel: soc-acpi-cht: Add quirk for Nextbook Ares 8A tablet

Qing Zhang <[email protected]>
LoongArch: Add ARCH_HAS_FORTIFY_SOURCE selection

WANG Xuerui <[email protected]>
LoongArch: Relay BCE exceptions to userland as SIGSEGV with si_code=SEGV_BNDERR

Hersen Wu <[email protected]>

    drm/amd/display: fix memleak in aconnector->timing_requested

jasontao <[email protected]>
ALSA: hda: Glenfly: add HD Audio PCI IDs and HDMI Codec Vendor IDs.

Johannes Thumshirn <[email protected]>
watchdog: menz069_wdt: fix watchdog initialisation

Chong Li <[email protected]>
drm/amdgpu: release gpu full access after "amdgpu_device_ip_late_init"

Paolo Abeni <[email protected]>

    mptcp: add annotations around sk->sk_shutdown accesses

Paolo Abeni <[email protected]>

    mptcp: fix data race around msk->first access

Paolo Abeni <[email protected]>
mptcp: consolidate passive msk socket initialization

Paolo Abeni <[email protected]>
mptcp: simplify subflow_syn_recv_sock()

Paolo Abeni <[email protected]>
mptcp: avoid unneeded address copy

Paolo Abeni <[email protected]>

    mptcp: add annotations around msk->subflow accesses

Paolo Abeni <[email protected]>
mptcp: avoid unneeded __mptcp_nmpc_socket() usage

Xin Long <[email protected]>
rtnetlink: add the missing IFLA_GRO_ tb check in validate_linkmsg

Xin Long <[email protected]>
rtnetlink: move IFLA_GSO_ tb check to validate_linkmsg

Xin Long <[email protected]>
rtnetlink: call validate_linkmsg in rtnl_create_link

Maciej Fijalkowski <[email protected]>
ice: recycle/free all of the fragments from multi-buffer frame

Xu Liang <[email protected]>
net: phy: mxl-gpy: extend interrupt fix to all impacted variants

Yoshihiro Shimoda <[email protected]>
net: renesas: rswitch: Fix return value in error path of xmit

Chris Packham <[email protected]>
mtd: rawnand: marvell: don't set the NAND frequency select

Chris Packham <[email protected]>
mtd: rawnand: marvell: ensure timing values are written

Andreas Svensson <[email protected]>
net: dsa: mv88e6xxx: Increase wait after reset deactivation

Bert Karwatzki <[email protected]>
net: ipa: Use correct value for IPA_STATUS_SIZE

fuyuanli <[email protected]>
tcp: fix mishandling when the sack compression is deferred.

Hangyu Hua <[email protected]>
net/sched: flower: fix possible OOB write in fl_set_geneve_opt()

Chen-Yu Tsai <[email protected]>
iommu/mediatek: Flush IOTLB completely only if domain has been attached

Edward Cree <[email protected]>
sfc: fix error unwinds in TC offload

Moshe Shemesh <[email protected]>
net/mlx5: Read embedded cpu after init bit cleared

Saeed Mahameed <[email protected]>
net/mlx5e: Fix error handling in mlx5e_refresh_tirs

Christoph Hellwig <[email protected]>
nvme: fix the name of Zone Append for verbose logging

Bastien Nocera <[email protected]>
HID: logitech-hidpp: Handle timeout differently from busy

Dan Carpenter <[email protected]>
nfsd: fix double fget() bug in __write_ports_addfd()

Vladislav Efanov <[email protected]>
udp6: Fix race condition in udp6_sendmsg & connect

Pedro Tammela <[email protected]>
net/netlink: fix NETLINK_LIST_MEMBERSHIPS length report

Zhengchao Shao <[email protected]>
net: sched: fix NULL pointer dereference in mq_attach

Peilin Ye <[email protected]>
net/sched: Prohibit regrafting ingress or clsact Qdiscs

Peilin Ye <[email protected]>
net/sched: Reserve TC_H_INGRESS (TC_H_CLSACT) for ingress (clsact) Qdiscs

Peilin Ye <[email protected]>
net/sched: sch_clsact: Only create under TC_H_CLSACT

Peilin Ye <[email protected]>
net/sched: sch_ingress: Only create under TC_H_INGRESS

Wen Gu <[email protected]>
net/smc: Don't use RMBs not mapped to new link in SMCRv2 ADD LINK

Wen Gu <[email protected]>
net/smc: Scan from current RMB list when no position specified

David Howells <[email protected]>
rxrpc: Truncate UTS_RELEASE for rxrpc version

Cambda Zhu <[email protected]>
tcp: Return user_mss for TCP_MAXSEG in CLOSE/LISTEN state if user_mss set

Eric Dumazet <[email protected]>
tcp: deny tcp_disconnect() when threads are waiting

Eric Dumazet <[email protected]>
af_packet: do not use READ_ONCE() in packet_bind()

Mustafa Ismail <[email protected]>
RDMA/irdma: Fix Local Invalidate fencing

Mustafa Ismail <[email protected]>
RDMA/irdma: Prevent QP use after free

Akihiro Suda <[email protected]>
efi: Bump stub image version for macOS HVF compatibility

Arnd Bergmann <[email protected]>
mtd: rawnand: ingenic: fix empty stub helper definitions

Namhyung Kim <[email protected]>
perf ftrace latency: Remove unnecessary "--" from --use-nsec option

Raju Rangoju <[email protected]>
amd-xgbe: fix the false linkup in xgbe_phy_status

Jakub Kicinski <[email protected]>
tls: improve lockless access safety of tls_err_abort()

Kuniyuki Iwashima <[email protected]>

    af_packet: Fix data-races of pkt_sk(sk)->num.

Eric Dumazet <[email protected]>
netrom: fix info-leak in nr_write_internal()

Wei Fang <[email protected]>
net: stmmac: fix call trace when stmmac_xdp_xmit() is invoked

Thomas Bogendoerfer <[email protected]>
net: mellanox: mlxbf_gige: Fix skb_panic splat under memory pressure

Jianbo Liu <[email protected]>
net/mlx5e: Move Ethernet driver debugfs to profile init callback

Dmytro Linkin <[email protected]>
net/mlx5e: Don't attach netdev profile while handling internal error

Vlad Buslov <[email protected]>
net/mlx5: Fix post parse infra to only parse every action once

Paul Blakey <[email protected]>
net/mlx5e: TC, Remove CT action reordering

Paul Blakey <[email protected]>
net/mlx5e: CT: Use per action stats

Dragos Tatulea <[email protected]>
net/mlx5e: Use query_special_contexts cmd only once per mdev

Shay Drory <[email protected]>
net/mlx5: fw_tracer, Fix event handling

Shay Drory <[email protected]>
net/mlx5: SF, Drain health before removing device

Shay Drory <[email protected]>
net/mlx5: Drain health before unregistering devlink

Maher Sanalla <[email protected]>
net/mlx5e: Do not update SBCM when prio2buffer command is invalid

Maher Sanalla <[email protected]>
net/mlx5e: Consider internal buffers size in port buffer calculations

Chris Mi <[email protected]>
net/mlx5e: Prevent encap offload when neigh update is running

Chris Mi <[email protected]>
net/mlx5e: Extract remaining tunnel encap code to dedicated file

Paul Blakey <[email protected]>
net/mlx5e: TC, Move main flow attribute cleanup to helper func

Paul Blakey <[email protected]>
net/mlx5e: TC, Remove unused vf_tun variable

Alexandre Ghiti <[email protected]>
riscv: Fix unused variable warning when BUILTIN_DTB is set

Ard Biesheuvel <[email protected]>
crypto: x86/aria - Use 16 byte alignment for GFNI constant vectors

Randy Dunlap <[email protected]>
dmaengine: pl330: rename _start to prevent build error

Jeff Layton <[email protected]>
nfsd: make a copy of struct iattr before calling notify_change

Jason Gunthorpe <[email protected]>
iommu/amd: Add missing domain type checks

Jerry Snitselaar <[email protected]>
iommu/amd: Fix up merge conflict resolution

Joao Martins <[email protected]>
iommu/amd: Handle GALog overflows

Joao Martins <[email protected]>
iommu/amd: Don't block updates to GATag if guest mode is on

Chao Wang <[email protected]>
iommu/rockchip: Fix unwind goto issue

Randy Dunlap <[email protected]>
iommu: Make IPMMU_VMSA dependencies more strict

Kalesh AP <[email protected]>
RDMA/bnxt_re: Fix return value of bnxt_re_process_raw_qp_pkt_rx

Kalesh AP <[email protected]>
RDMA/bnxt_re: Fix a possible memory leak

Dan Carpenter <[email protected]>
dmaengine: at_xdmac: fix potential Oops in at_xdmac_prep_interleaved()

Yangyang Li <[email protected]>
RDMA/hns: Modify the value of long message loopback slice

Chengchang Tang <[email protected]>
RDMA/hns: Fix base address table allocation

Chengchang Tang <[email protected]>
RDMA/hns: Fix timeout attr in query qp for HIP08

Yonatan Nachum <[email protected]>
RDMA/efa: Fix unsupported page sizes in device

Neil Armstrong <[email protected]>
phy: amlogic: phy-meson-g12a-mipi-dphy-analog: fix CNTL2_DIF_TX_CTL0 value

Selvin Xavier <[email protected]>
RDMA/bnxt_re: Fix the page_size used during the MR creation

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

Diffstat:

 .../bindings/iio/adc/renesas,rcar-gyroadc.yaml     |   2 +-
 .../devicetree/bindings/serial/8250_omap.yaml      |   1 +
 .../devicetree/bindings/sound/tas2562.yaml         |   6 +-
 .../devicetree/bindings/sound/tas2770.yaml         |   6 +-
 .../devicetree/bindings/sound/tas27xx.yaml         |   6 +-
 .../devicetree/bindings/usb/snps,dwc3.yaml         |   2 +-
 Makefile                                           |   4 +-
 arch/arm/boot/dts/stm32f7-pinctrl.dtsi             |  82 +++++++++
 arch/arm/kernel/unwind.c                           |  25 ++-
 arch/arm64/include/asm/kvm_pgtable.h               |   6 +-
 arch/arm64/kernel/vdso.c                           |   2 +-
 arch/arm64/kvm/hyp/include/hyp/switch.h            |   8 +-
 arch/arm64/kvm/hyp/nvhe/mem_protect.c              |  14 +-
 arch/arm64/kvm/hyp/nvhe/switch.c                   |   2 +
 arch/arm64/kvm/hyp/pgtable.c                       |  17 +-
 arch/arm64/kvm/hyp/vhe/switch.c                    |   1 +
 arch/arm64/kvm/vgic/vgic-init.c                    |  25 ++-
 arch/arm64/kvm/vgic/vgic-its.c                     |  14 +-
 arch/arm64/kvm/vgic/vgic-kvm-device.c              |  10 +-
 arch/arm64/kvm/vgic/vgic-mmio-v3.c                 |  31 ++--
 arch/arm64/kvm/vgic/vgic-mmio.c                    |   9 +-
 arch/arm64/kvm/vgic/vgic-v2.c                      |   6 -
 arch/arm64/kvm/vgic/vgic-v3.c                      |   7 -
 arch/arm64/kvm/vgic/vgic-v4.c                      |   3 +-
 arch/arm64/mm/fault.c                              |   4 +-
 arch/loongarch/Kconfig                             |   1 +
 arch/loongarch/include/asm/inst.h                  |  26 +++
 arch/loongarch/kernel/genex.S                      |   1 +
 arch/loongarch/kernel/traps.c                      |  92 +++++++++++
 arch/powerpc/platforms/pseries/iommu.c             |  13 +-
 arch/powerpc/xmon/xmon.c                           |   2 +-
 arch/riscv/include/asm/perf_event.h                |   7 +
 arch/riscv/kernel/vmlinux.lds.S                    |   2 +
 arch/riscv/mm/init.c                               |   2 +-
 arch/s390/kernel/ipl.c                             |   3 +-
 arch/s390/kernel/topology.c                        |  32 ++--
 arch/um/drivers/Makefile                           |   4 +-
 arch/um/drivers/harddog.h                          |   9 +
 arch/um/drivers/harddog_kern.c                     |   7 +-
 arch/um/drivers/harddog_user.c                     |   1 +
 arch/um/drivers/harddog_user_exp.c                 |   9 +
 arch/x86/crypto/aria-aesni-avx-asm_64.S            |   2 -
 arch/x86/events/intel/core.c                       |   2 +-
 arch/x86/events/intel/ds.c                         |  56 ++++---
 arch/x86/include/asm/perf_event.h                  |   3 +
 arch/x86/kvm/lapic.c                               |  20 ++-
 arch/x86/kvm/mmu/mmu.c                             |   5 +-
 arch/x86/kvm/x86.c                                 |   3 +
 block/blk-settings.c                               |   3 +-
 block/fops.c                                       |  12 +-
 crypto/asymmetric_keys/public_key.c                |  38 +++--
 drivers/acpi/resource.c                            |  12 ++
 drivers/ata/libata-scsi.c                          |  34 +++-
 drivers/base/cacheinfo.c                           |  26 +++
 drivers/base/regmap/regmap.c                       |   6 +-
 drivers/block/nbd.c                                |   4 +-
 drivers/block/rnbd/rnbd-proto.h                    |   2 +-
 drivers/block/ublk_drv.c                           |   9 +-
 drivers/char/tpm/tpm_tis_core.c                    |   7 +-
 drivers/char/tpm/tpm_tis_core.h                    |   8 +-
 drivers/dma/at_hdmac.c                             |  17 +-
 drivers/dma/at_xdmac.c                             |   7 +-
 drivers/dma/pl330.c                                |   8 +-
 drivers/firmware/qcom_scm.c                        |  12 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c         |  45 +++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c          |   3 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c            |   3 +-
 drivers/gpu/drm/amd/amdgpu/nv.c                    |  22 ++-
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c  |   9 +-
 drivers/gpu/drm/amd/display/dc/core/dc.c           |   4 +-
 drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c         |  29 ----
 drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c   |  10 +-
 drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c    |   5 +-
 .../gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c   |   5 +-
 .../gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c   |   5 +-
 .../gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c   |   5 +-
 drivers/gpu/drm/ast/ast_main.c                     |   9 +-
 drivers/gpu/drm/msm/msm_iommu.c                    |   7 +-
 drivers/hid/hid-google-hammer.c                    |   2 +
 drivers/hid/hid-ids.h                              |   1 +
 drivers/hid/hid-logitech-hidpp.c                   |  12 +-
 drivers/hid/wacom_wac.c                            |   2 +-
 drivers/hwmon/k10temp.c                            |   1 +
 drivers/iio/accel/kionix-kx022a.c                  |   2 +-
 drivers/iio/accel/st_accel_core.c                  |   4 +-
 drivers/iio/adc/ad4130.c                           |  12 +-
 drivers/iio/adc/ad7192.c                           |   8 +-
 drivers/iio/adc/ad_sigma_delta.c                   |   4 +
 drivers/iio/adc/imx93_adc.c                        |   7 +-
 drivers/iio/adc/mt6370-adc.c                       |  53 +++++-
 drivers/iio/adc/mxs-lradc-adc.c                    |  10 +-
 drivers/iio/adc/stm32-adc.c                        |  61 +++----
 drivers/iio/addac/ad74413r.c                       |   2 +-
 drivers/iio/dac/Makefile                           |   2 +-
 drivers/iio/dac/mcp4725.c                          |  16 +-
 drivers/iio/imu/inv_icm42600/inv_icm42600_buffer.c |  10 +-
 drivers/iio/light/vcnl4035.c                       |   3 +
 drivers/iio/magnetometer/tmag5273.c                |   5 +-
 drivers/infiniband/hw/bnxt_re/ib_verbs.c           |   4 +-
 drivers/infiniband/hw/bnxt_re/qplib_fp.c           |  11 +-
 drivers/infiniband/hw/bnxt_re/qplib_res.c          |  12 +-
 drivers/infiniband/hw/bnxt_re/qplib_sp.c           |   7 +-
 drivers/infiniband/hw/efa/efa_verbs.c              |   2 +-
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c         |  25 ++-
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h         |   2 +
 drivers/infiniband/hw/hns/hns_roce_mr.c            |  43 +++++
 drivers/infiniband/hw/irdma/verbs.c                |  12 +-
 drivers/iommu/Kconfig                              |   1 +
 drivers/iommu/amd/amd_iommu.h                      |   4 +-
 drivers/iommu/amd/init.c                           |  24 +++
 drivers/iommu/amd/iommu.c                          |  31 +++-
 drivers/iommu/mtk_iommu.c                          |   3 +-
 drivers/iommu/rockchip-iommu.c                     |  14 +-
 drivers/mailbox/mailbox-test.c                     |  13 +-
 drivers/md/raid5.c                                 |   2 +-
 drivers/media/dvb-core/dvb_ca_en50221.c            |  49 +++++-
 drivers/media/dvb-core/dvb_demux.c                 |   4 +-
 drivers/media/dvb-core/dvb_frontend.c              |  69 ++++++--
 drivers/media/dvb-core/dvb_net.c                   |  38 ++++-
 drivers/media/dvb-core/dvbdev.c                    |  84 +++++++---
 drivers/media/dvb-frontends/mn88443x.c             |   2 +-
 drivers/media/pci/netup_unidvb/netup_unidvb_core.c |  17 +-
 .../mediatek/vcodec/mtk_vcodec_dec_stateful.c      |   3 +
 drivers/media/platform/renesas/rcar-vin/rcar-dma.c |  21 ++-
 drivers/media/usb/dvb-usb-v2/ce6230.c              |   8 +
 drivers/media/usb/dvb-usb-v2/ec168.c               |  12 ++
 drivers/media/usb/dvb-usb-v2/rtl28xxu.c            |  20 +++
 drivers/media/usb/dvb-usb/az6027.c                 |  12 ++
 drivers/media/usb/dvb-usb/digitv.c                 |   4 +
 drivers/media/usb/dvb-usb/dw2102.c                 |   2 +-
 drivers/media/usb/ttusb-dec/ttusb_dec.c            |   3 +-
 drivers/media/usb/uvc/uvc_driver.c                 |  16 +-
 drivers/misc/fastrpc.c                             |  33 +++-
 drivers/mmc/core/pwrseq_sd8787.c                   |  34 +++-
 drivers/mmc/host/vub300.c                          |   3 +
 drivers/mtd/mtdchar.c                              |   8 +-
 drivers/mtd/nand/raw/ingenic/ingenic_ecc.h         |   8 +-
 drivers/mtd/nand/raw/marvell_nand.c                |  10 +-
 drivers/net/dsa/mv88e6xxx/chip.c                   |   2 +-
 drivers/net/ethernet/amd/xgbe/xgbe-mdio.c          |  12 +-
 drivers/net/ethernet/intel/ice/ice_txrx.c          |   2 +-
 .../ethernet/mellanox/mlx5/core/diag/fw_tracer.c   |   2 +-
 drivers/net/ethernet/mellanox/mlx5/core/en.h       |   1 +
 .../ethernet/mellanox/mlx5/core/en/port_buffer.c   |  44 +++--
 .../ethernet/mellanox/mlx5/core/en/port_buffer.h   |   8 +-
 .../ethernet/mellanox/mlx5/core/en/tc/act/act.c    |  27 +--
 .../ethernet/mellanox/mlx5/core/en/tc/act/act.h    |   8 +-
 .../net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c |   9 +-
 .../ethernet/mellanox/mlx5/core/en/tc_tun_encap.c  | 120 ++++++++++++--
 .../ethernet/mellanox/mlx5/core/en/tc_tun_encap.h  |   9 +
 .../net/ethernet/mellanox/mlx5/core/en_common.c    |  11 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c |   7 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c  |  69 ++++----
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c   |   6 +
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c    | 184 +++++----------------
 drivers/net/ethernet/mellanox/mlx5/core/main.c     |   9 +-
 drivers/net/ethernet/mellanox/mlx5/core/mr.c       |  21 +++
 .../ethernet/mellanox/mlx5/core/sf/dev/driver.c    |   1 +
 .../ethernet/mellanox/mlxbf_gige/mlxbf_gige_rx.c   |  13 +-
 drivers/net/ethernet/renesas/rswitch.c             |   2 +-
 drivers/net/ethernet/sfc/tc.c                      |   4 +-
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c  |   3 +-
 drivers/net/ethernet/stmicro/stmmac/stmmac_xdp.c   |   6 +
 drivers/net/ipa/ipa_endpoint.c                     |   2 +-
 drivers/net/phy/mxl-gpy.c                          |  16 +-
 drivers/net/usb/qmi_wwan.c                         |   2 +-
 drivers/net/wireless/ath/ath10k/qmi.c              |   4 +-
 drivers/net/wireless/broadcom/b43/b43.h            |   2 +-
 .../net/wireless/broadcom/b43legacy/b43legacy.h    |   2 +-
 drivers/net/wireless/intel/iwlwifi/mvm/rs.c        |   3 +
 drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu.h   |   1 +
 .../net/wireless/realtek/rtl8xxxu/rtl8xxxu_core.c  |   4 +-
 drivers/net/wwan/t7xx/t7xx_pci.c                   |  18 ++
 drivers/net/wwan/t7xx/t7xx_pci.h                   |   1 +
 drivers/nvme/host/constants.c                      |   2 +-
 drivers/nvme/host/core.c                           |   6 +-
 drivers/nvme/host/hwmon.c                          |   4 +-
 drivers/nvme/host/multipath.c                      |   1 -
 drivers/nvme/host/nvme.h                           |   5 +
 drivers/nvme/host/pci.c                            |   8 +-
 .../phy/amlogic/phy-meson-g12a-mipi-dphy-analog.c  |   2 +-
 drivers/phy/qualcomm/phy-qcom-qmp-combo.c          |   5 +-
 drivers/phy/qualcomm/phy-qcom-qmp-pcie-msm8996.c   |   5 +-
 drivers/platform/mellanox/mlxbf-tmfifo.c           |  11 +-
 drivers/platform/x86/intel_scu_pcidrv.c            |   1 +
 drivers/remoteproc/qcom_q6v5_mss.c                 |   8 +-
 drivers/remoteproc/qcom_q6v5_pas.c                 |   2 +-
 drivers/s390/crypto/pkey_api.c                     |   3 +
 drivers/scsi/qla2xxx/qla_def.h                     |   1 +
 drivers/scsi/qla2xxx/qla_init.c                    |   3 +
 drivers/scsi/qla2xxx/qla_inline.h                  |   3 +
 drivers/scsi/qla2xxx/qla_isr.c                     |   3 +
 drivers/scsi/scsi_lib.c                            |   2 +
 drivers/scsi/stex.c                                |   4 +
 drivers/soc/qcom/rmtfs_mem.c                       |   2 +-
 drivers/tty/serial/8250/8250_tegra.c               |   4 +-
 drivers/tty/serial/Kconfig                         |   2 +-
 drivers/tty/serial/cpm_uart/cpm_uart.h             |   2 -
 drivers/tty/serial/fsl_lpuart.c                    |  44 ++---
 drivers/ufs/core/ufs-mcq.c                         |   5 +-
 drivers/ufs/core/ufshcd.c                          |  10 +-
 drivers/usb/cdns3/cdns3-gadget.c                   |  13 ++
 drivers/usb/gadget/function/f_fs.c                 |   2 +-
 drivers/video/fbdev/core/bitblit.c                 |   3 +
 drivers/video/fbdev/core/modedb.c                  |   5 +
 drivers/video/fbdev/imsttfb.c                      |  15 +-
 drivers/video/fbdev/stifb.c                        |   1 +
 drivers/watchdog/menz69_wdt.c                      |  16 +-
 fs/btrfs/bio.c                                     |   2 +-
 fs/btrfs/ctree.c                                   |   2 +
 fs/btrfs/disk-io.c                                 |   2 +-
 fs/ceph/mds_client.c                               |   3 +-
 fs/ext4/ext4.h                                     |   5 +-
 fs/ext4/inode.c                                    |  34 +++-
 fs/ext4/super.c                                    |  24 +--
 fs/ext4/xattr.c                                    |  41 ++---
 fs/gfs2/super.c                                    |   8 +
 fs/ksmbd/oplock.c                                  |  72 +++++---
 fs/ksmbd/smb2pdu.c                                 |  90 +++++-----
 fs/nfsd/nfsctl.c                                   |   7 +-
 fs/nfsd/vfs.c                                      |  10 +-
 fs/xfs/xfs_buf_item_recover.c                      |  10 ++
 include/linux/firmware/qcom/qcom_scm.h             |   2 +-
 include/linux/mlx5/driver.h                        |   1 +
 include/linux/pe.h                                 |  25 +--
 include/linux/sunrpc/svcsock.h                     |   7 +-
 include/media/dvb_frontend.h                       |   6 +-
 include/media/dvb_net.h                            |   4 +
 include/media/dvbdev.h                             |  15 ++
 include/net/sock.h                                 |   4 +
 include/net/tcp.h                                  |   1 +
 include/ufs/ufshcd.h                               |   2 +-
 io_uring/epoll.c                                   |   4 -
 kernel/module/decompress.c                         |   2 +-
 kernel/trace/trace_events_hist.c                   |  23 ++-
 kernel/trace/trace_osnoise.c                       |   2 +
 kernel/trace/trace_probe.h                         |   2 +-
 lib/test_firmware.c                                |  85 +++++++---
 net/atm/resources.c                                |   2 +
 net/core/rtnetlink.c                               |  54 ++++--
 net/core/sock.c                                    |   2 +-
 net/ipv4/af_inet.c                                 |   2 +
 net/ipv4/inet_connection_sock.c                    |   1 +
 net/ipv4/tcp.c                                     |   9 +-
 net/ipv4/tcp_input.c                               |   2 +-
 net/ipv4/tcp_timer.c                               |  16 +-
 net/mac80211/chan.c                                |  75 +++++----
 net/mac80211/ieee80211_i.h                         |   3 +-
 net/mac80211/util.c                                |   2 +-
 net/mptcp/protocol.c                               | 143 +++++++++-------
 net/mptcp/protocol.h                               |  15 +-
 net/mptcp/subflow.c                                |  63 ++-----
 net/netfilter/nf_conntrack_netlink.c               |   4 +
 net/netlink/af_netlink.c                           |   2 +-
 net/netrom/nr_subr.c                               |   7 +-
 net/packet/af_packet.c                             |   8 +-
 net/packet/diag.c                                  |   2 +-
 net/rxrpc/af_rxrpc.c                               |   1 +
 net/rxrpc/ar-internal.h                            |   1 +
 net/rxrpc/local_event.c                            |  11 +-
 net/sched/cls_flower.c                             |   3 +
 net/sched/sch_api.c                                |  16 +-
 net/sched/sch_ingress.c                            |  16 +-
 net/smc/smc_llc.c                                  |   9 +-
 net/sunrpc/svcsock.c                               |  24 +--
 net/tls/tls_strp.c                                 |   4 +-
 net/tls/tls_sw.c                                   |   4 +-
 net/xfrm/xfrm_policy.c                             |  11 +-
 security/selinux/Makefile                          |   6 +-
 sound/core/oss/pcm_plugin.h                        |  16 +-
 sound/pci/hda/hda_intel.c                          |  21 +++
 sound/pci/hda/patch_hdmi.c                         |  22 +++
 sound/pci/hda/patch_realtek.c                      |  20 +++
 sound/soc/amd/yc/acp6x-mach.c                      |   7 +
 sound/soc/codecs/ssm2602.c                         |  15 ++
 sound/soc/dwc/dwc-i2s.c                            |   4 +-
 sound/soc/intel/common/soc-acpi-intel-cht-match.c  |  26 +++
 sound/soc/jz4740/jz4740-i2s.c                      |  54 +++++-
 sound/soc/sof/amd/acp-ipc.c                        |   7 +-
 sound/soc/sof/debug.c                              |   4 +-
 sound/soc/sof/pcm.c                                |  17 +-
 sound/soc/sof/pm.c                                 |  14 +-
 sound/soc/sof/sof-client-probes.c                  |  14 +-
 tools/perf/builtin-ftrace.c                        |   2 +-
 tools/power/cpupower/lib/powercap.c                |  23 ++-
 .../ftrace/test.d/filter/event-filter-function.tc  |  47 +++---
 tools/testing/selftests/net/mptcp/Makefile         |   2 +-
 tools/testing/selftests/net/mptcp/diag.sh          |   4 +
 tools/testing/selftests/net/mptcp/mptcp_connect.sh |   4 +
 tools/testing/selftests/net/mptcp/mptcp_join.sh    |  17 +-
 tools/testing/selftests/net/mptcp/mptcp_lib.sh     |  40 +++++
 tools/testing/selftests/net/mptcp/mptcp_sockopt.sh |   4 +
 tools/testing/selftests/net/mptcp/pm_netlink.sh    |   4 +
 tools/testing/selftests/net/mptcp/simult_flows.sh  |   4 +
 tools/testing/selftests/net/mptcp/userspace_pm.sh  |   4 +
 295 files changed, 2806 insertions(+), 1289 deletions(-)
author maintainer
#32 lore raw

PATCH 32/286 net/mlx5: Fix post parse infra to only parse every action once

From: Vlad Buslov <[email protected]>

[ Upstream commit 5d862ec631f3d3cc3b4f8cdb5b9fc5879663f1d3 ]

Caller of mlx5e_tc_act_post_parse() needs it to parse only the subset of actions starting after previous split and ending at the current action. However, that range is not provided as arguments and mlx5e_tc_act_post_parse() uses generic flow_action_for_each() that iterates over all flow actions. Not only this is redundant, it also causes a bug when mlx5e_tc_act->post_parse() callback is not idempotent since it will be called for every split. For example, ct action tc_act_post_parse_ct() callback obtains a reference to mlx5_ct_ft instance and calling it several times during parsing stage will cause reference counter imbalance.

Fix the issue by providing a proper action range of the current split subset to mlx5e_tc_act_post_parse() and only calling mlx5e_tc_act->post_parse() for actions inside the subset range.

Fixes 8300f225268b ("net/mlx5e: Create new flow attr for multi table actions")
---
 drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.c | 7 ++++++-
 drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h | 2 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c         | 8 +++++---
 3 files changed, 12 insertions(+), 5 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.c +6 -1
index fc923a99b6a48..0380a04c3691c 100644
@@ -84,7 +84,7 @@ mlx5e_tc_act_init_parse_state(struct mlx5e_tc_act_parse_state *parse_state,
 
 int
 mlx5e_tc_act_post_parse(struct mlx5e_tc_act_parse_state *parse_state,
-			struct flow_action *flow_action,
+			struct flow_action *flow_action, int from, int to,
 			struct mlx5_flow_attr *attr,
 			enum mlx5_flow_namespace_type ns_type)
 {
@@ -96,6 +96,11 @@ mlx5e_tc_act_post_parse(struct mlx5e_tc_act_parse_state *parse_state,
 	priv = parse_state->flow->priv;
 
 	flow_action_for_each(i, act, flow_action) {
+		if (i < from)
+			continue;
+		else if (i > to)
+			break;
+
 		tc_act = mlx5e_tc_act_get(act->id, ns_type);
 		if (!tc_act || !tc_act->post_parse)
 			continue;
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h +1 -1
index d7615e329e6d9..84c78d5f5bed8 100644
@@ -114,7 +114,7 @@ mlx5e_tc_act_init_parse_state(struct mlx5e_tc_act_parse_state *parse_state,
 
 int
 mlx5e_tc_act_post_parse(struct mlx5e_tc_act_parse_state *parse_state,
-			struct flow_action *flow_action,
+			struct flow_action *flow_action, int from, int to,
 			struct mlx5_flow_attr *attr,
 			enum mlx5_flow_namespace_type ns_type);
 
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +5 -3
index a78e201fd883b..82b96196e97b7 100644
@@ -3975,8 +3975,8 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 	struct mlx5_flow_attr *prev_attr;
 	struct flow_action_entry *act;
 	struct mlx5e_tc_act *tc_act;
+	int err, i, i_split = 0;
 	bool is_missable;
-	int err, i;
 
 	ns_type = mlx5e_get_flow_namespace(flow);
 	list_add(&attr->list, &flow->attrs);
@@ -4017,7 +4017,8 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 		    i < flow_action->num_entries - 1)) {
 			is_missable = tc_act->is_missable ? tc_act->is_missable(act) : false;
 
-			err = mlx5e_tc_act_post_parse(parse_state, flow_action, attr, ns_type);
+			err = mlx5e_tc_act_post_parse(parse_state, flow_action, i_split, i, attr,
+						      ns_type);
 			if (err)
 				goto out_free_post_acts;
 
@@ -4027,6 +4028,7 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 				goto out_free_post_acts;
 			}
 
+			i_split = i + 1;
 			list_add(&attr->list, &flow->attrs);
 		}
 
@@ -4041,7 +4043,7 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 		}
 	}
 
-	err = mlx5e_tc_act_post_parse(parse_state, flow_action, attr, ns_type);
+	err = mlx5e_tc_act_post_parse(parse_state, flow_action, i_split, i, attr, ns_type);
 	if (err)
 		goto out_free_post_acts;
 
Signature · 1 line
2.39.2
author maintainer
#41 lore raw

PATCH 41/286 perf ftrace latency: Remove unnecessary "--" from --use-nsec option

In reply to Greg Kroah-Hartman

From: Namhyung Kim <[email protected]>

[ Upstream commit 8d73259ef23f449329294dc187932f7470268126 ]

The option name should not have the dashes. Current version shows four dashes for the option.

  $ perf ftrace latency -h

Usage: perf ftrace [<options>] [<command>]
or: perf ftrace [<options>] -- [<command>] [<options>]
or: perf ftrace {trace|latency} [<options>] [<command>]
or: perf ftrace {trace|latency} [<options>] -- [<command>] [<options>]

-b, --use-bpf Use BPF to measure function latency
-n, ----use-nsec Use nano-second histogram
-T, --trace-funcs <func>
Show latency of given function

Fixes 84005bb6148618cc ("perf ftrace latency: Add -n/--use-nsec option")
Cc Adrian Hunter <[email protected]>
Cc Changbin Du <[email protected]>
Cc Ian Rogers <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Jiri Olsa <[email protected]>
Cc Peter Zijlstra <[email protected]>
---
 tools/perf/builtin-ftrace.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/builtin-ftrace.c +1 -1
index fb1b66ef2e167..ce482ef58e6f2 100644
@@ -1175,7 +1175,7 @@ int cmd_ftrace(int argc, const char **argv)
 	OPT_BOOLEAN('b', "use-bpf", &ftrace.target.use_bpf,
 		    "Use BPF to measure function latency"),
 #endif
-	OPT_BOOLEAN('n', "--use-nsec", &ftrace.use_nsec,
+	OPT_BOOLEAN('n', "use-nsec", &ftrace.use_nsec,
 		    "Use nano-second histogram"),
 	OPT_PARENT(common_options),
 	};
Signature · 1 line
2.39.2
author maintainer
#42 lore raw

PATCH 42/286 mtd: rawnand: ingenic: fix empty stub helper definitions

In reply to Greg Kroah-Hartman

From: Arnd Bergmann <[email protected]>

[ Upstream commit 650a8884a364ff2568b51cde9009cfd43cdae6ad ]

A few functions provide an empty interface definition when CONFIG_MTD_NAND_INGENIC_ECC is disabled, but they are accidentally defined as global functions in the header:

drivers/mtd/nand/raw/ingenic/ingenic_ecc.h:39:5: error: no previous prototype for 'ingenic_ecc_calculate' drivers/mtd/nand/raw/ingenic/ingenic_ecc.h:46:5: error: no previous prototype for 'ingenic_ecc_correct' drivers/mtd/nand/raw/ingenic/ingenic_ecc.h:53:6: error: no previous prototype for 'ingenic_ecc_release' drivers/mtd/nand/raw/ingenic/ingenic_ecc.h:57:21: error: no previous prototype for 'of_ingenic_ecc_get'

Turn them into 'static inline' definitions instead.

Fixes 15de8c6efd0e ("mtd: rawnand: ingenic: Separate top-level and SoC specific code")
---
 drivers/mtd/nand/raw/ingenic/ingenic_ecc.h | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/mtd/nand/raw/ingenic/ingenic_ecc.h +4 -4
index 2cda439b5e11b..017868f59f222 100644
@@ -36,25 +36,25 @@ int ingenic_ecc_correct(struct ingenic_ecc *ecc,
 void ingenic_ecc_release(struct ingenic_ecc *ecc);
 struct ingenic_ecc *of_ingenic_ecc_get(struct device_node *np);
 #else /* CONFIG_MTD_NAND_INGENIC_ECC */
-int ingenic_ecc_calculate(struct ingenic_ecc *ecc,
+static inline int ingenic_ecc_calculate(struct ingenic_ecc *ecc,
 			  struct ingenic_ecc_params *params,
 			  const u8 *buf, u8 *ecc_code)
 {
 	return -ENODEV;
 }
 
-int ingenic_ecc_correct(struct ingenic_ecc *ecc,
+static inline int ingenic_ecc_correct(struct ingenic_ecc *ecc,
 			struct ingenic_ecc_params *params, u8 *buf,
 			u8 *ecc_code)
 {
 	return -ENODEV;
 }
 
-void ingenic_ecc_release(struct ingenic_ecc *ecc)
+static inline void ingenic_ecc_release(struct ingenic_ecc *ecc)
 {
 }
 
-struct ingenic_ecc *of_ingenic_ecc_get(struct device_node *np)
+static inline struct ingenic_ecc *of_ingenic_ecc_get(struct device_node *np)
 {
 	return ERR_PTR(-ENODEV);
 }
Signature · 1 line
2.39.2
author maintainer
#43 lore raw

PATCH 43/286 efi: Bump stub image version for macOS HVF compatibility

In reply to Greg Kroah-Hartman

From: Akihiro Suda <[email protected]>

[ Upstream commit 36e4fc57fc1619f462e669e939209c45763bc8f5 ]

The macOS hypervisor framework includes a host-side VMM called VZLinuxBootLoader [1] which implements native support for booting the Linux kernel inside a guest directly (instead of, e.g., via GRUB installed inside the guest). On x86, it incorporates a BIOS style loader that does not implement or expose EFI to the loaded kernel. However, this loader appears to fail when the 'image minor version' field in the kernel image's PE/COFF header (which is generally only used by EFI based bootloaders) is set to any value other than 0x0. [2]

Commit e346bebbd36b1576 ("efi: libstub: Always enable initrd command line loader and bump version") incremented the EFI stub image minor version to convey that all EFI stub kernels now implement support for the initrd= command line option, and do so in a way where it can load initrd images from any filesystem known to the EFI firmware (as opposed to prior implementations that could only load initrds from the same volume that the kernel image was loaded from).

Unfortunately, bumping the version to v1.1 triggers this issue in VZLinuxBootLoader, breaking the boot on x86. So let's keep the image minor version at 0x0, and bump the image major version instead.

While at it, convert this field to a bit field, so that individual features are discoverable from it, as suggested by Linus. So let's bump the major version to v3, and document the initrd= command line loading feature as being represented by bit 1 in the mask.

Note that, due to the prior interpretation as a monotonically increasing version field, loaders are still permitted to assume that the LoadFile2 initrd loading feature is supported for any major version value >= 1, even if bit 0 is not set.

[1] https://developer.apple.com/documentation/virtualization/vzlinuxbootloader
[2] https://lore.kernel.org/linux-efi/CAG8fp8Teu4G9JuenQrqGndFt2Gy+V4YgJ=hN1xX7AD940YKf3A@mail.gmail.com/

Fixes e346bebbd36b1576 ("efi: libstub: Always enable initrd command ...")
[ardb: rewrite comment and commit log]
---
 include/linux/pe.h | 25 +++++++++++++------------
 1 file changed, 13 insertions(+), 12 deletions(-)
include/linux/pe.h +13 -12
index 6ffabf1e6d039..16754fb2f954a 100644
@@ -11,25 +11,26 @@
 #include <linux/types.h>
 
 /*
- * Linux EFI stub v1.0 adds the following functionality:
- * - Loading initrd from the LINUX_EFI_INITRD_MEDIA_GUID device path,
- * - Loading/starting the kernel from firmware that targets a different
- *   machine type, via the entrypoint exposed in the .compat PE/COFF section.
+ * Starting from version v3.0, the major version field should be interpreted as
+ * a bit mask of features supported by the kernel's EFI stub:
+ * - 0x1: initrd loading from the LINUX_EFI_INITRD_MEDIA_GUID device path,
+ * - 0x2: initrd loading using the initrd= command line option, where the file
+ *        may be specified using device path notation, and is not required to
+ *        reside on the same volume as the loaded kernel image.
  *
  * The recommended way of loading and starting v1.0 or later kernels is to use
  * the LoadImage() and StartImage() EFI boot services, and expose the initrd
  * via the LINUX_EFI_INITRD_MEDIA_GUID device path.
  *
- * Versions older than v1.0 support initrd loading via the image load options
- * (using initrd=, limited to the volume from which the kernel itself was
- * loaded), or via arch specific means (bootparams, DT, etc).
+ * Versions older than v1.0 may support initrd loading via the image load
+ * options (using initrd=, limited to the volume from which the kernel itself
+ * was loaded), or only via arch specific means (bootparams, DT, etc).
  *
- * On x86, LoadImage() and StartImage() can be omitted if the EFI handover
- * protocol is implemented, which can be inferred from the version,
- * handover_offset and xloadflags fields in the bootparams structure.
+ * The minor version field must remain 0x0.
+ * (https://lore.kernel.org/all/[email protected]/)
  */
-#define LINUX_EFISTUB_MAJOR_VERSION		0x1
-#define LINUX_EFISTUB_MINOR_VERSION		0x1
+#define LINUX_EFISTUB_MAJOR_VERSION		0x3
+#define LINUX_EFISTUB_MINOR_VERSION		0x0
 
 /*
  * LINUX_PE_MAGIC appears at offset 0x38 into the MS-DOS header of EFI bootable
Signature · 1 line
2.39.2
author maintainer
#44 lore raw

PATCH 44/286 RDMA/irdma: Prevent QP use after free

In reply to Greg Kroah-Hartman

From: Mustafa Ismail <[email protected]>

[ Upstream commit c8f304d75f6c6cc679a73f89591f9a915da38f09 ]

There is a window where the poll cq may use a QP that has been freed. This can happen if a CQE is polled before irdma_clean_cqes() can clear the CQE's related to the QP and the destroy QP races to free the QP memory. then the QP structures are used in irdma_poll_cq. Fix this by moving the clearing of CQE's before the reference is removed and the QP is destroyed.

Fixes b48c24c2d710 ("RDMA/irdma: Implement device supported verb APIs")
---
 drivers/infiniband/hw/irdma/verbs.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/infiniband/hw/irdma/verbs.c +6 -5
index 1b2e3e800c9a6..446a0ab3faaa5 100644
@@ -522,11 +522,6 @@ static int irdma_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata)
 	if (!iwqp->user_mode)
 		cancel_delayed_work_sync(&iwqp->dwork_flush);
 
-	irdma_qp_rem_ref(&iwqp->ibqp);
-	wait_for_completion(&iwqp->free_qp);
-	irdma_free_lsmm_rsrc(iwqp);
-	irdma_cqp_qp_destroy_cmd(&iwdev->rf->sc_dev, &iwqp->sc_qp);
-
 	if (!iwqp->user_mode) {
 		if (iwqp->iwscq) {
 			irdma_clean_cqes(iwqp, iwqp->iwscq);
@@ -534,6 +529,12 @@ static int irdma_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata)
 				irdma_clean_cqes(iwqp, iwqp->iwrcq);
 		}
 	}
+
+	irdma_qp_rem_ref(&iwqp->ibqp);
+	wait_for_completion(&iwqp->free_qp);
+	irdma_free_lsmm_rsrc(iwqp);
+	irdma_cqp_qp_destroy_cmd(&iwdev->rf->sc_dev, &iwqp->sc_qp);
+
 	irdma_remove_push_mmap_entries(iwqp);
 	irdma_free_qp_rsrc(iwqp);
 
Signature · 1 line
2.39.2
author maintainer
#45 lore raw

PATCH 45/286 RDMA/irdma: Fix Local Invalidate fencing

In reply to Greg Kroah-Hartman

From: Mustafa Ismail <[email protected]>

[ Upstream commit 5842d1d9c1b0d17e0c29eae65ae1f245f83682dd ]

If the local invalidate fence is indicated in the WR, only the read fence is currently being set in WQE. Fix this to set both the read and local fence in the WQE.

Fixes b48c24c2d710 ("RDMA/irdma: Implement device supported verb APIs")
---
 drivers/infiniband/hw/irdma/verbs.c | 1 +
 1 file changed, 1 insertion(+)
drivers/infiniband/hw/irdma/verbs.c +1 -0
index 446a0ab3faaa5..d0bb21d3007c2 100644
@@ -3297,6 +3297,7 @@ static int irdma_post_send(struct ib_qp *ibqp,
 			break;
 		case IB_WR_LOCAL_INV:
 			info.op_type = IRDMA_OP_TYPE_INV_STAG;
+			info.local_fence = info.read_fence;
 			info.op.inv_local_stag.target_stag = ib_wr->ex.invalidate_rkey;
 			err = irdma_uk_stag_local_invalidate(ukqp, &info, true);
 			break;
Signature · 1 line
2.39.2
author maintainer
#46 lore raw

PATCH 46/286 af_packet: do not use READ_ONCE() in packet_bind()

In reply to Greg Kroah-Hartman

From: Eric Dumazet <[email protected]>

[ Upstream commit 6ffc57ea004234d9373c57b204fd10370a69f392 ]

A recent patch added READ_ONCE() in packet_bind() and packet_bind_spkt()

This is better handled by reading pkt_sk(sk)->num later in packet_do_bind() while appropriate lock is held.

READ_ONCE() in writers are often an evidence of something being wrong.

---
 net/packet/af_packet.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
net/packet/af_packet.c +5 -3
index f3513316743ad..b79d2fa788061 100644
@@ -3193,6 +3193,9 @@ static int packet_do_bind(struct sock *sk, const char *name, int ifindex,
 
 	lock_sock(sk);
 	spin_lock(&po->bind_lock);
+	if (!proto)
+		proto = po->num;
+
 	rcu_read_lock();
 
 	if (po->fanout) {
@@ -3291,7 +3294,7 @@ static int packet_bind_spkt(struct socket *sock, struct sockaddr *uaddr,
 	memcpy(name, uaddr->sa_data, sizeof(uaddr->sa_data_min));
 	name[sizeof(uaddr->sa_data_min)] = 0;
 
-	return packet_do_bind(sk, name, 0, READ_ONCE(pkt_sk(sk)->num));
+	return packet_do_bind(sk, name, 0, 0);
 }
 
 static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
@@ -3308,8 +3311,7 @@ static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len
 	if (sll->sll_family != AF_PACKET)
 		return -EINVAL;
 
-	return packet_do_bind(sk, NULL, sll->sll_ifindex,
-			      sll->sll_protocol ? : READ_ONCE(pkt_sk(sk)->num));
+	return packet_do_bind(sk, NULL, sll->sll_ifindex, sll->sll_protocol);
 }
 
 static struct proto packet_proto = {
Signature · 1 line
2.39.2
author maintainer
#47 lore raw

PATCH 47/286 tcp: deny tcp_disconnect() when threads are waiting

In reply to Greg Kroah-Hartman

From: Eric Dumazet <[email protected]>

[ Upstream commit 4faeee0cf8a5d88d63cdbc3bab124fb0e6aed08c ]

Historically connect(AF_UNSPEC) has been abused by syzkaller and other fuzzers to trigger various bugs.

A recent one triggers a divide-by-zero [1], and Paolo Abeni was able to diagnose the issue.

tcp_recvmsg_locked() has tests about sk_state being not TCP_LISTEN and TCP REPAIR mode being not used.

Then later if socket lock is released in sk_wait_data(), another thread can call connect(AF_UNSPEC), then make this socket a TCP listener.

When recvmsg() is resumed, it can eventually call tcp_cleanup_rbuf() and attempt a divide by 0 in tcp_rcv_space_adjust() [1]

This patch adds a new socket field, counting number of threads blocked in sk_wait_event() and inet_wait_for_connect().

If this counter is not zero, tcp_disconnect() returns an error.

This patch adds code in blocking socket system calls, thus should not hurt performance of non blocking ones.

Note that we probably could revert commit 499350a5a6e7 ("tcp:
initialize rcv_mss to TCP_MIN_MSS instead of 0") to restore original tcpi_rcv_mss meaning (was 0 if no payload was ever received on a socket)

[1]
divide error: 0000 [#1] PREEMPT SMP KASAN

CPU: 0 PID: 13832 Comm: syz-executor.5 Not tainted 6.3.0-rc4-syzkaller-00224-g00c7b5f4ddc5 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/02/2023
RIP: 0010:tcp_rcv_space_adjust+0x36e/0x9d0 net/ipv4/tcp_input.c:740
Code: 00 00 00 00 fc ff df 4c 89 64 24 48 8b 44 24 04 44 89 f9 41 81 c7 80 03 00 00 c1 e1 04 44 29 f0 48 63 c9 48 01 e9 48 0f af c1 <49> f7 f6 48 8d 04 41 48 89 44 24 40 48 8b 44 24 30 48 c1 e8 03 48
RSP: 0018:ffffc900033af660 EFLAGS: 00010206
RAX: 4a66b76cbade2c48 RBX: ffff888076640cc0 RCX: 00000000c334e4ac
RDX: 0000000000000000 RSI: dffffc0000000000 RDI: 0000000000000001
RBP: 00000000c324e86c R08: 0000000000000001 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000000 R12: ffff8880766417f8
R13: ffff888028fbb980 R14: 0000000000000000 R15: 0000000000010344
FS: 00007f5bffbfe700(0000) GS:ffff8880b9800000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000001b32f25000 CR3: 000000007ced0000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<TASK>
tcp_recvmsg_locked+0x100e/0x22e0 net/ipv4/tcp.c:2616
tcp_recvmsg+0x117/0x620 net/ipv4/tcp.c:2681
inet6_recvmsg+0x114/0x640 net/ipv6/af_inet6.c:670
sock_recvmsg_nosec net/socket.c:1017 [inline]
sock_recvmsg+0xe2/0x160 net/socket.c:1038
____sys_recvmsg+0x210/0x5a0 net/socket.c:2720
___sys_recvmsg+0xf2/0x180 net/socket.c:2762
do_recvmmsg+0x25e/0x6e0 net/socket.c:2856
__sys_recvmmsg net/socket.c:2935 [inline]
__do_sys_recvmmsg net/socket.c:2958 [inline]
__se_sys_recvmmsg net/socket.c:2951 [inline]
__x64_sys_recvmmsg+0x20f/0x260 net/socket.c:2951
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x39/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x7f5c0108c0f9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 f1 19 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f5bffbfe168 EFLAGS: 00000246 ORIG_RAX: 000000000000012b
RAX: ffffffffffffffda RBX: 00007f5c011ac050 RCX: 00007f5c0108c0f9
RDX: 0000000000000001 RSI: 0000000020000bc0 RDI: 0000000000000003
RBP: 00007f5c010e7b39 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000122 R11: 0000000000000246 R12: 0000000000000000
R13: 00007f5c012cfb1f R14: 00007f5bffbfe300 R15: 0000000000022000
</TASK>
---
 include/net/sock.h              | 4 ++++
 net/ipv4/af_inet.c              | 2 ++
 net/ipv4/inet_connection_sock.c | 1 +
 net/ipv4/tcp.c                  | 6 ++++++
 4 files changed, 13 insertions(+)
include/net/sock.h +4 -0
index 9cd0354221507..45e46a1c4afc6 100644
@@ -336,6 +336,7 @@ struct sk_filter;
   *	@sk_cgrp_data: cgroup data for this cgroup
   *	@sk_memcg: this socket's memory cgroup association
   *	@sk_write_pending: a write to stream socket waits to start
+  *	@sk_wait_pending: number of threads blocked on this socket
   *	@sk_state_change: callback to indicate change in the state of the sock
   *	@sk_data_ready: callback to indicate there is data to be processed
   *	@sk_write_space: callback to indicate there is bf sending space available
@@ -428,6 +429,7 @@ struct sock {
 	unsigned int		sk_napi_id;
 #endif
 	int			sk_rcvbuf;
+	int			sk_wait_pending;
 
 	struct sk_filter __rcu	*sk_filter;
 	union {
@@ -1174,6 +1176,7 @@ static inline void sock_rps_reset_rxhash(struct sock *sk)
 
 #define sk_wait_event(__sk, __timeo, __condition, __wait)		\
 	({	int __rc;						\
+		__sk->sk_wait_pending++;				\
 		release_sock(__sk);					\
 		__rc = __condition;					\
 		if (!__rc) {						\
@@ -1183,6 +1186,7 @@ static inline void sock_rps_reset_rxhash(struct sock *sk)
 		}							\
 		sched_annotate_sleep();					\
 		lock_sock(__sk);					\
+		__sk->sk_wait_pending--;				\
 		__rc = __condition;					\
 		__rc;							\
 	})
net/ipv4/af_inet.c +2 -0
index 70fd769f1174b..daeec363b0976 100644
@@ -586,6 +586,7 @@ static long inet_wait_for_connect(struct sock *sk, long timeo, int writebias)
 
 	add_wait_queue(sk_sleep(sk), &wait);
 	sk->sk_write_pending += writebias;
+	sk->sk_wait_pending++;
 
 	/* Basic assumption: if someone sets sk->sk_err, he _must_
 	 * change state of the socket from TCP_SYN_*.
@@ -601,6 +602,7 @@ static long inet_wait_for_connect(struct sock *sk, long timeo, int writebias)
 	}
 	remove_wait_queue(sk_sleep(sk), &wait);
 	sk->sk_write_pending -= writebias;
+	sk->sk_wait_pending--;
 	return timeo;
 }
 
net/ipv4/inet_connection_sock.c +1 -0
index 65ad4251f6fd8..1386787eaf1a5 100644
@@ -1142,6 +1142,7 @@ struct sock *inet_csk_clone_lock(const struct sock *sk,
 	if (newsk) {
 		struct inet_connection_sock *newicsk = inet_csk(newsk);
 
+		newsk->sk_wait_pending = 0;
 		inet_sk_set_state(newsk, TCP_SYN_RECV);
 		newicsk->icsk_bind_hash = NULL;
 		newicsk->icsk_bind2_hash = NULL;
net/ipv4/tcp.c +6 -0
index ed63ee8f0d7e3..f4b243cc7f4a5 100644
@@ -3080,6 +3080,12 @@ int tcp_disconnect(struct sock *sk, int flags)
 	int old_state = sk->sk_state;
 	u32 seq;
 
+	/* Deny disconnect if other threads are blocked in sk_wait_event()
+	 * or inet_wait_for_connect().
+	 */
+	if (sk->sk_wait_pending)
+		return -EBUSY;
+
 	if (old_state != TCP_CLOSE)
 		tcp_set_state(sk, TCP_CLOSE);
 
Signature · 1 line
2.39.2
author maintainer
#48 lore raw

PATCH 48/286 tcp: Return user_mss for TCP_MAXSEG in CLOSE/LISTEN state if user_mss set

In reply to Greg Kroah-Hartman

From: Cambda Zhu <[email protected]>

[ Upstream commit 34dfde4ad87b84d21278a7e19d92b5b2c68e6c4d ]

This patch replaces the tp->mss_cache check in getting TCP_MAXSEG with tp->rx_opt.user_mss check for CLOSE/LISTEN sock. Since tp->mss_cache is initialized with TCP_MSS_DEFAULT, checking if it's zero is probably a bug.

With this change, getting TCP_MAXSEG before connecting will return default MSS normally, and return user_mss if user_mss is set.

---
 net/ipv4/tcp.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
net/ipv4/tcp.c +2 -1
index f4b243cc7f4a5..6bb8eb8031051 100644
@@ -4077,7 +4077,8 @@ int do_tcp_getsockopt(struct sock *sk, int level,
 	switch (optname) {
 	case TCP_MAXSEG:
 		val = tp->mss_cache;
-		if (!val && ((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN)))
+		if (tp->rx_opt.user_mss &&
+		    ((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN)))
 			val = tp->rx_opt.user_mss;
 		if (tp->repair)
 			val = tp->rx_opt.mss_clamp;
Signature · 1 line
2.39.2
author maintainer
#49 lore raw

PATCH 49/286 rxrpc: Truncate UTS_RELEASE for rxrpc version

In reply to Greg Kroah-Hartman

From: David Howells <[email protected]>

[ Upstream commit 020c69c1a793ed29d28793808eddd75210c858dd ]

UTS_RELEASE has a maximum length of 64 which can cause rxrpc_version to exceed the 65 byte message limit.

Per the rx spec[1]: "If a server receives a packet with a type value of 13, and the client-initiated flag set, it should respond with a 65-byte payload containing a string that identifies the version of AFS software it is running."

The current implementation causes a compile error when WERROR is turned on and/or UTS_RELEASE exceeds the length of 49 (making the version string more than 64 characters).

Fix this by generating the string during module initialisation and limiting the UTS_RELEASE segment of the string does not exceed 49 chars. We need to make sure that the 64 bytes includes "linux-" at the front and " AF_RXRPC" at the back as this may be used in pattern matching.

---
 net/rxrpc/af_rxrpc.c    |  1 +
 net/rxrpc/ar-internal.h |  1 +
 net/rxrpc/local_event.c | 11 ++++++++++-
 3 files changed, 12 insertions(+), 1 deletion(-)
net/rxrpc/af_rxrpc.c +1 -0
index a6f0d29f35ef9..f5d1fc1266a5a 100644
@@ -967,6 +967,7 @@ static int __init af_rxrpc_init(void)
 	BUILD_BUG_ON(sizeof(struct rxrpc_skb_priv) > sizeof_field(struct sk_buff, cb));
 
 	ret = -ENOMEM;
+	rxrpc_gen_version_string();
 	rxrpc_call_jar = kmem_cache_create(
 		"rxrpc_call_jar", sizeof(struct rxrpc_call), 0,
 		SLAB_HWCACHE_ALIGN, NULL);
net/rxrpc/ar-internal.h +1 -0
index 5d44dc08f66d0..e8e14c6f904d9 100644
@@ -1068,6 +1068,7 @@ int rxrpc_get_server_data_key(struct rxrpc_connection *, const void *, time64_t,
 /*
  * local_event.c
  */
+void rxrpc_gen_version_string(void);
 void rxrpc_send_version_request(struct rxrpc_local *local,
 				struct rxrpc_host_header *hdr,
 				struct sk_buff *skb);
net/rxrpc/local_event.c +10 -1
index 5e69ea6b233da..993c69f97488c 100644
@@ -16,7 +16,16 @@
 #include <generated/utsrelease.h>
 #include "ar-internal.h"
 
-static const char rxrpc_version_string[65] = "linux-" UTS_RELEASE " AF_RXRPC";
+static char rxrpc_version_string[65]; // "linux-" UTS_RELEASE " AF_RXRPC";
+
+/*
+ * Generate the VERSION packet string.
+ */
+void rxrpc_gen_version_string(void)
+{
+	snprintf(rxrpc_version_string, sizeof(rxrpc_version_string),
+		 "linux-%.49s AF_RXRPC", UTS_RELEASE);
+}
 
 /*
  * Reply to a version request
Signature · 1 line
2.39.2
author maintainer
#50 lore raw

PATCH 50/286 net/smc: Scan from current RMB list when no position specified

In reply to Greg Kroah-Hartman

From: Wen Gu <[email protected]>

[ Upstream commit b24aa141c2ff26c919237aee61ea1818fc6780d9 ]

When finding the first RMB of link group, it should start from the current RMB list whose index is 0. So fix it.

Fixes b4ba4652b3f8 ("net/smc: extend LLC layer for SMC-Rv2")
---
 net/smc/smc_llc.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
net/smc/smc_llc.c +4 -1
index a0840b8c935b8..8423e8e0063f4 100644
@@ -578,7 +578,10 @@ static struct smc_buf_desc *smc_llc_get_next_rmb(struct smc_link_group *lgr,
 {
 	struct smc_buf_desc *buf_next;
 
-	if (!buf_pos || list_is_last(&buf_pos->list, &lgr->rmbs[*buf_lst])) {
+	if (!buf_pos)
+		return _smc_llc_get_next_rmb(lgr, buf_lst);
+
+	if (list_is_last(&buf_pos->list, &lgr->rmbs[*buf_lst])) {
 		(*buf_lst)++;
 		return _smc_llc_get_next_rmb(lgr, buf_lst);
 	}
Signature · 1 line
2.39.2
author maintainer
#33 lore raw

PATCH 33/286 net/mlx5e: Dont attach netdev profile while handling internal error

In reply to Greg Kroah-Hartman

From: Dmytro Linkin <[email protected]>

[ Upstream commit bdf274750fca17b289404ef03453c4070725302c ]

As part of switchdev mode disablement, driver changes port netdevice profile from uplink to nic. If this process is triggered by health recovery flow (PCI reset, for ex.) profile attach would fail because all fw commands aborted when internal error flag is set. As a result, nic netdevice profile is not attached and driver fails to rollback to uplink profile, which leave driver in broken state and cause crash later.

To handle broken state do netdevice profile initialization only instead of full attachment and release mdev resources on driver suspend as expected. Actual netdevice attachment is done during driver load.

Fixes c4d7eb57687f ("net/mxl5e: Add change profile method")
---
 .../net/ethernet/mellanox/mlx5/core/en_main.c | 35 ++++++++++++++++---
 1 file changed, 31 insertions(+), 4 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en_main.c +31 -4
index ff579d7d8432c..a60610c7a7bb7 100644
@@ -5798,8 +5798,8 @@ void mlx5e_detach_netdev(struct mlx5e_priv *priv)
 }
 
 static int
-mlx5e_netdev_attach_profile(struct net_device *netdev, struct mlx5_core_dev *mdev,
-			    const struct mlx5e_profile *new_profile, void *new_ppriv)
+mlx5e_netdev_init_profile(struct net_device *netdev, struct mlx5_core_dev *mdev,
+			  const struct mlx5e_profile *new_profile, void *new_ppriv)
 {
 	struct mlx5e_priv *priv = netdev_priv(netdev);
 	int err;
@@ -5815,6 +5815,25 @@ mlx5e_netdev_attach_profile(struct net_device *netdev, struct mlx5_core_dev *mde
 	err = new_profile->init(priv->mdev, priv->netdev);
 	if (err)
 		goto priv_cleanup;
+
+	return 0;
+
+priv_cleanup:
+	mlx5e_priv_cleanup(priv);
+	return err;
+}
+
+static int
+mlx5e_netdev_attach_profile(struct net_device *netdev, struct mlx5_core_dev *mdev,
+			    const struct mlx5e_profile *new_profile, void *new_ppriv)
+{
+	struct mlx5e_priv *priv = netdev_priv(netdev);
+	int err;
+
+	err = mlx5e_netdev_init_profile(netdev, mdev, new_profile, new_ppriv);
+	if (err)
+		return err;
+
 	err = mlx5e_attach_netdev(priv);
 	if (err)
 		goto profile_cleanup;
@@ -5822,7 +5841,6 @@ mlx5e_netdev_attach_profile(struct net_device *netdev, struct mlx5_core_dev *mde
 
 profile_cleanup:
 	new_profile->cleanup(priv);
-priv_cleanup:
 	mlx5e_priv_cleanup(priv);
 	return err;
 }
@@ -5841,6 +5859,12 @@ int mlx5e_netdev_change_profile(struct mlx5e_priv *priv,
 	priv->profile->cleanup(priv);
 	mlx5e_priv_cleanup(priv);
 
+	if (mdev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) {
+		mlx5e_netdev_init_profile(netdev, mdev, new_profile, new_ppriv);
+		set_bit(MLX5E_STATE_DESTROYING, &priv->state);
+		return -EIO;
+	}
+
 	err = mlx5e_netdev_attach_profile(netdev, mdev, new_profile, new_ppriv);
 	if (err) { /* roll back to original profile */
 		netdev_warn(netdev, "%s: new profile init failed, %d\n", __func__, err);
@@ -5902,8 +5926,11 @@ static int mlx5e_suspend(struct auxiliary_device *adev, pm_message_t state)
 	struct net_device *netdev = priv->netdev;
 	struct mlx5_core_dev *mdev = priv->mdev;
 
-	if (!netif_device_present(netdev))
+	if (!netif_device_present(netdev)) {
+		if (test_bit(MLX5E_STATE_DESTROYING, &priv->state))
+			mlx5e_destroy_mdev_resources(mdev);
 		return -ENODEV;
+	}
 
 	mlx5e_detach_netdev(priv);
 	mlx5e_destroy_mdev_resources(mdev);
Signature · 1 line
2.39.2
author maintainer
#51 lore raw

PATCH 51/286 net/smc: Dont use RMBs not mapped to new link in SMCRv2 ADD LINK

In reply to Greg Kroah-Hartman

From: Wen Gu <[email protected]>

[ Upstream commit 71c6aa0305e3d2365d3bfd0134b4025d9e7ba388 ]

We encountered a crash when using SMCRv2. It is caused by a logical error in smc_llc_fill_ext_v2().

 BUG: kernel NULL pointer dereference, address: 0000000000000014
 #PF: supervisor read access in kernel mode
 #PF: error_code(0x0000) - not-present page
 PGD 0 P4D 0
 Oops: 0000 [#1] PREEMPT SMP PTI
 CPU: 7 PID: 453 Comm: kworker/7:4 Kdump: loaded Tainted: G        W   E      6.4.0-rc3+ #44
 Workqueue: events smc_llc_add_link_work [smc]
 RIP: 0010:smc_llc_fill_ext_v2+0x117/0x280 [smc]
 RSP: 0018:ffffacb5c064bd88 EFLAGS: 00010282
 RAX: ffff9a6bc1c3c02c RBX: ffff9a6be3558000 RCX: 0000000000000000
 RDX: 0000000000000002 RSI: 0000000000000002 RDI: 000000000000000a
 RBP: ffffacb5c064bdb8 R08: 0000000000000040 R09: 000000000000000c
 R10: ffff9a6bc0910300 R11: 0000000000000002 R12: 0000000000000000
 R13: 0000000000000002 R14: ffff9a6bc1c3c02c R15: ffff9a6be3558250
 FS:  0000000000000000(0000) GS:ffff9a6eefdc0000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 CR2: 0000000000000014 CR3: 000000010b078003 CR4: 00000000003706e0
 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
 Call Trace:
  <TASK>
  smc_llc_send_add_link+0x1ae/0x2f0 [smc]
  smc_llc_srv_add_link+0x2c9/0x5a0 [smc]
  ? cc_mkenc+0x40/0x60
  smc_llc_add_link_work+0xb8/0x140 [smc]
  process_one_work+0x1e5/0x3f0
  worker_thread+0x4d/0x2f0
  ? __pfx_worker_thread+0x10/0x10
  kthread+0xe5/0x120
  ? __pfx_kthread+0x10/0x10
  ret_from_fork+0x2c/0x50
  </TASK>

When an alernate RNIC is available in system, SMC will try to add a new link based on the RNIC for resilience. All the RMBs in use will be mapped to the new link. Then the RMBs' MRs corresponding to the new link will be filled into SMCRv2 LLC ADD LINK messages.

However, smc_llc_fill_ext_v2() mistakenly accesses to unused RMBs which haven't been mapped to the new link and have no valid MRs, thus causing a crash. So this patch fixes the logic.

Fixes b4ba4652b3f8 ("net/smc: extend LLC layer for SMC-Rv2")
---
 net/smc/smc_llc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/smc/smc_llc.c +2 -2
index 8423e8e0063f4..7a8d9163d186e 100644
@@ -617,6 +617,8 @@ static int smc_llc_fill_ext_v2(struct smc_llc_msg_add_link_v2_ext *ext,
 		goto out;
 	buf_pos = smc_llc_get_first_rmb(lgr, &buf_lst);
 	for (i = 0; i < ext->num_rkeys; i++) {
+		while (buf_pos && !(buf_pos)->used)
+			buf_pos = smc_llc_get_next_rmb(lgr, &buf_lst, buf_pos);
 		if (!buf_pos)
 			break;
 		rmb = buf_pos;
@@ -626,8 +628,6 @@ static int smc_llc_fill_ext_v2(struct smc_llc_msg_add_link_v2_ext *ext,
 			cpu_to_be64((uintptr_t)rmb->cpu_addr) :
 			cpu_to_be64((u64)sg_dma_address(rmb->sgt[lnk_idx].sgl));
 		buf_pos = smc_llc_get_next_rmb(lgr, &buf_lst, buf_pos);
-		while (buf_pos && !(buf_pos)->used)
-			buf_pos = smc_llc_get_next_rmb(lgr, &buf_lst, buf_pos);
 	}
 	len += i * sizeof(ext->rt[0]);
 out:
Signature · 1 line
2.39.2
author maintainer
#52 lore raw

PATCH 52/286 net/sched: sch_ingress: Only create under TC_H_INGRESS

In reply to Greg Kroah-Hartman

From: Peilin Ye <[email protected]>

[ Upstream commit c7cfbd115001f94de9e4053657946a383147e803 ]

ingress Qdiscs are only supposed to be created under TC_H_INGRESS. Return -EOPNOTSUPP if 'parent' is not TC_H_INGRESS, similar to mq_init().

---
 net/sched/sch_ingress.c | 6 ++++++
 1 file changed, 6 insertions(+)
net/sched/sch_ingress.c +6 -0
index 84838128b9c5b..f9ef6deb27709 100644
@@ -80,6 +80,9 @@ static int ingress_init(struct Qdisc *sch, struct nlattr *opt,
 	struct net_device *dev = qdisc_dev(sch);
 	int err;
 
+	if (sch->parent != TC_H_INGRESS)
+		return -EOPNOTSUPP;
+
 	net_inc_ingress_queue();
 
 	mini_qdisc_pair_init(&q->miniqp, sch, &dev->miniq_ingress);
@@ -101,6 +104,9 @@ static void ingress_destroy(struct Qdisc *sch)
 {
 	struct ingress_sched_data *q = qdisc_priv(sch);
 
+	if (sch->parent != TC_H_INGRESS)
+		return;
+
 	tcf_block_put_ext(q->block, sch, &q->block_info);
 	net_dec_ingress_queue();
 }
Signature · 1 line
2.39.2
author maintainer
#53 lore raw

PATCH 53/286 net/sched: sch_clsact: Only create under TC_H_CLSACT

In reply to Greg Kroah-Hartman

From: Peilin Ye <[email protected]>

[ Upstream commit 5eeebfe6c493192b10d516abfd72742900f2a162 ]

clsact Qdiscs are only supposed to be created under TC_H_CLSACT (which equals TC_H_INGRESS). Return -EOPNOTSUPP if 'parent' is not TC_H_CLSACT.

---
 net/sched/sch_ingress.c | 6 ++++++
 1 file changed, 6 insertions(+)
net/sched/sch_ingress.c +6 -0
index f9ef6deb27709..35963929e1178 100644
@@ -225,6 +225,9 @@ static int clsact_init(struct Qdisc *sch, struct nlattr *opt,
 	struct net_device *dev = qdisc_dev(sch);
 	int err;
 
+	if (sch->parent != TC_H_CLSACT)
+		return -EOPNOTSUPP;
+
 	net_inc_ingress_queue();
 	net_inc_egress_queue();
 
@@ -254,6 +257,9 @@ static void clsact_destroy(struct Qdisc *sch)
 {
 	struct clsact_sched_data *q = qdisc_priv(sch);
 
+	if (sch->parent != TC_H_CLSACT)
+		return;
+
 	tcf_block_put_ext(q->egress_block, sch, &q->egress_block_info);
 	tcf_block_put_ext(q->ingress_block, sch, &q->ingress_block_info);
 
Signature · 1 line
2.39.2
author maintainer
#54 lore raw

PATCH 54/286 net/sched: Reserve TC_H_INGRESS (TC_H_CLSACT) for ingress (clsact) Qdiscs

In reply to Greg Kroah-Hartman

From: Peilin Ye <[email protected]>

[ Upstream commit f85fa45d4a9408d98c46c8fa45ba2e3b2f4bf219 ]

Currently it is possible to add e.g. an HTB Qdisc under ffff:fff1 (TC_H_INGRESS, TC_H_CLSACT):

  $ ip link add name ifb0 type ifb
  $ tc qdisc add dev ifb0 parent ffff:fff1 htb
  $ tc qdisc add dev ifb0 clsact
  Error: Exclusivity flag on, cannot modify.
  $ drgn
  ...
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

ifb0 = netdev_get_by_name(prog, "ifb0")
qdisc = ifb0.ingress_queue.qdisc_sleeping
print(qdisc.ops.id.string_().decode())

htb

Quoted 1 line
Quoted 1 line
Quoted 1 line

qdisc.flags.value_() # TCQ_F_INGRESS

2

Only allow ingress and clsact Qdiscs under ffff:fff1. Return -EINVAL for everything else. Make TCQ_F_INGRESS a static flag of ingress and clsact Qdiscs.

Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
Fixes 1f211a1b929c ("net, sched: add clsact qdisc")
---
 net/sched/sch_api.c     | 7 ++++++-
 net/sched/sch_ingress.c | 4 ++--
 2 files changed, 8 insertions(+), 3 deletions(-)
net/sched/sch_api.c +6 -1
index aba789c30a2eb..b7c230f9c5b19 100644
@@ -1250,7 +1250,12 @@ static struct Qdisc *qdisc_create(struct net_device *dev,
 	sch->parent = parent;
 
 	if (handle == TC_H_INGRESS) {
-		sch->flags |= TCQ_F_INGRESS;
+		if (!(sch->flags & TCQ_F_INGRESS)) {
+			NL_SET_ERR_MSG(extack,
+				       "Specified parent ID is reserved for ingress and clsact Qdiscs");
+			err = -EINVAL;
+			goto err_out3;
+		}
 		handle = TC_H_MAKE(TC_H_INGRESS, 0);
 	} else {
 		if (handle == 0) {
net/sched/sch_ingress.c +2 -2
index 35963929e1178..e43a454993723 100644
@@ -140,7 +140,7 @@ static struct Qdisc_ops ingress_qdisc_ops __read_mostly = {
 	.cl_ops			=	&ingress_class_ops,
 	.id			=	"ingress",
 	.priv_size		=	sizeof(struct ingress_sched_data),
-	.static_flags		=	TCQ_F_CPUSTATS,
+	.static_flags		=	TCQ_F_INGRESS | TCQ_F_CPUSTATS,
 	.init			=	ingress_init,
 	.destroy		=	ingress_destroy,
 	.dump			=	ingress_dump,
@@ -281,7 +281,7 @@ static struct Qdisc_ops clsact_qdisc_ops __read_mostly = {
 	.cl_ops			=	&clsact_class_ops,
 	.id			=	"clsact",
 	.priv_size		=	sizeof(struct clsact_sched_data),
-	.static_flags		=	TCQ_F_CPUSTATS,
+	.static_flags		=	TCQ_F_INGRESS | TCQ_F_CPUSTATS,
 	.init			=	clsact_init,
 	.destroy		=	clsact_destroy,
 	.dump			=	ingress_dump,
Signature · 1 line
2.39.2
author maintainer
#55 lore raw

PATCH 55/286 net/sched: Prohibit regrafting ingress or clsact Qdiscs

In reply to Greg Kroah-Hartman

From: Peilin Ye <[email protected]>

[ Upstream commit 9de95df5d15baa956c2b70b9e794842e790a8a13 ]

Currently, after creating an ingress (or clsact) Qdisc and grafting it under TC_H_INGRESS (TC_H_CLSACT), it is possible to graft it again under e.g. a TBF Qdisc:

  $ ip link add ifb0 type ifb
  $ tc qdisc add dev ifb0 handle 1: root tbf rate 20kbit buffer 1600 limit 3000
  $ tc qdisc add dev ifb0 clsact
  $ tc qdisc link dev ifb0 handle ffff: parent 1:1
  $ tc qdisc show dev ifb0
  qdisc tbf 1: root refcnt 2 rate 20Kbit burst 1600b lat 560.0ms
  qdisc clsact ffff: parent ffff:fff1 refcnt 2
                                      ^^^^^^^^

clsact's refcount has increased: it is now grafted under both TC_H_CLSACT and 1:1.

ingress and clsact Qdiscs should only be used under TC_H_INGRESS (TC_H_CLSACT). Prohibit regrafting them.

Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
Fixes 1f211a1b929c ("net, sched: add clsact qdisc")
---
 net/sched/sch_api.c | 5 +++++
 1 file changed, 5 insertions(+)
net/sched/sch_api.c +5 -0
index b7c230f9c5b19..a686d7cae5130 100644
@@ -1594,6 +1594,11 @@ static int tc_modify_qdisc(struct sk_buff *skb, struct nlmsghdr *n,
 					NL_SET_ERR_MSG(extack, "Invalid qdisc name");
 					return -EINVAL;
 				}
+				if (q->flags & TCQ_F_INGRESS) {
+					NL_SET_ERR_MSG(extack,
+						       "Cannot regraft ingress or clsact Qdiscs");
+					return -EINVAL;
+				}
 				if (q == p ||
 				    (p && check_loop(q, p, 0))) {
 					NL_SET_ERR_MSG(extack, "Qdisc parent/child loop detected");
Signature · 1 line
2.39.2
author maintainer
#56 lore raw

PATCH 56/286 net: sched: fix NULL pointer dereference in mq_attach

In reply to Greg Kroah-Hartman

From: Zhengchao Shao <[email protected]>

[ Upstream commit 36eec020fab668719b541f34d97f44e232ffa165 ]

When use the following command to test:
1)ip link add bond0 type bond
2)ip link set bond0 up
3)tc qdisc add dev bond0 root handle ffff: mq
4)tc qdisc replace dev bond0 parent ffff:fff1 handle ffff: mq

The kernel reports NULL pointer dereference issue. The stack information is as follows:

Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000
Internal error: Oops: 0000000096000006 [#1] SMP
Modules linked in:
pstate: 20000005 (nzCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : mq_attach+0x44/0xa0
lr : qdisc_graft+0x20c/0x5cc

sp : ffff80000e2236a0
x29: ffff80000e2236a0 x28: ffff0000c0e59d80 x27: ffff0000c0be19c0
x26: ffff0000cae3e800 x25: 0000000000000010 x24: 00000000fffffff1
x23: 0000000000000000 x22: ffff0000cae3e800 x21: ffff0000c9df4000
x20: ffff0000c9df4000 x19: 0000000000000000 x18: ffff80000a934000
x17: ffff8000f5b56000 x16: ffff80000bb08000 x15: 0000000000000000
x14: 0000000000000000 x13: 6b6b6b6b6b6b6b6b x12: 6b6b6b6b00000001
x11: 0000000000000000 x10: 0000000000000000 x9 : 0000000000000000 x8 : ffff0000c0be0730 x7 : bbbbbbbbbbbbbbbb x6 : 0000000000000008 x5 : ffff0000cae3e864 x4 : 0000000000000000 x3 : 0000000000000001 x2 : 0000000000000001 x1 : ffff8000090bc23c x0 : 0000000000000000

Call trace:
mq_attach+0x44/0xa0
qdisc_graft+0x20c/0x5cc
tc_modify_qdisc+0x1c4/0x664
rtnetlink_rcv_msg+0x354/0x440
netlink_rcv_skb+0x64/0x144
rtnetlink_rcv+0x28/0x34
netlink_unicast+0x1e8/0x2a4
netlink_sendmsg+0x308/0x4a0
sock_sendmsg+0x64/0xac
____sys_sendmsg+0x29c/0x358
___sys_sendmsg+0x90/0xd0
__sys_sendmsg+0x7c/0xd0
__arm64_sys_sendmsg+0x2c/0x38
invoke_syscall+0x54/0x114
el0_svc_common.constprop.1+0x90/0x174
do_el0_svc+0x3c/0xb0
el0_svc+0x24/0xec
el0t_64_sync_handler+0x90/0xb4
el0t_64_sync+0x174/0x178

This is because when mq is added for the first time, qdiscs in mq is set to NULL in mq_attach(). Therefore, when replacing mq after adding mq, we need to initialize qdiscs in the mq before continuing to graft. Otherwise, it will couse NULL pointer dereference issue in mq_attach(). And the same issue will occur in the attach functions of mqprio, taprio and htb. ffff:fff1 means that the repalce qdisc is ingress. Ingress does not allow any qdisc to be attached. Therefore, ffff:fff1 is incorrectly used, and the command should be dropped.

---
 net/sched/sch_api.c | 4 ++++
 1 file changed, 4 insertions(+)
net/sched/sch_api.c +4 -0
index a686d7cae5130..7045b67b5533e 100644
@@ -1604,6 +1604,10 @@ static int tc_modify_qdisc(struct sk_buff *skb, struct nlmsghdr *n,
 					NL_SET_ERR_MSG(extack, "Qdisc parent/child loop detected");
 					return -ELOOP;
 				}
+				if (clid == TC_H_INGRESS) {
+					NL_SET_ERR_MSG(extack, "Ingress cannot graft directly");
+					return -EINVAL;
+				}
 				qdisc_refcount_inc(q);
 				goto graft;
 			} else {
Signature · 1 line
2.39.2
author maintainer
#57 lore raw

PATCH 57/286 net/netlink: fix NETLINK_LIST_MEMBERSHIPS length report

In reply to Greg Kroah-Hartman

From: Pedro Tammela <[email protected]>

[ Upstream commit f4e4534850a9d18c250a93f8d7fbb51310828110 ]

The current code for the length calculation wrongly truncates the reported length of the groups array, causing an under report of the subscribed groups. To fix this, use 'BITS_TO_BYTES()' which rounds up the division by 8.

Fixes b42be38b2778 ("netlink: add API to retrieve all group memberships")
---
 net/netlink/af_netlink.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/netlink/af_netlink.c +1 -1
index 45d47b39de225..717e27a4b66a0 100644
@@ -1779,7 +1779,7 @@ static int netlink_getsockopt(struct socket *sock, int level, int optname,
 				break;
 			}
 		}
-		if (put_user(ALIGN(nlk->ngroups / 8, sizeof(u32)), optlen))
+		if (put_user(ALIGN(BITS_TO_BYTES(nlk->ngroups), sizeof(u32)), optlen))
 			err = -EFAULT;
 		netlink_unlock_table();
 		return err;
Signature · 1 line
2.39.2
author maintainer
#58 lore raw

PATCH 58/286 udp6: Fix race condition in udp6_sendmsg & connect

In reply to Greg Kroah-Hartman

From: Vladislav Efanov <[email protected]>

[ Upstream commit 448a5ce1120c5bdbce1f1ccdabcd31c7d029f328 ]

Syzkaller got the following report:

BUG: KASAN: use-after-free in sk_setup_caps+0x621/0x690 net/core/sock.c:2018
Read of size 8 at addr ffff888027f82780 by task syz-executor276/3255

The function sk_setup_caps (called by ip6_sk_dst_store_flow-> ip6_dst_store) referenced already freed memory as this memory was freed by parallel task in udpv6_sendmsg->ip6_sk_dst_lookup_flow-> sk_dst_check.

task1 (connect) task2 (udp6_sendmsg)
sk_setup_caps->sk_dst_set |
| sk_dst_check->
| sk_dst_set
| dst_release
sk_setup_caps references |
to already freed dst_entry|

The reason for this race condition is: sk_setup_caps() keeps using the dst after transferring the ownership to the dst cache.

Found by Linux Verification Center (linuxtesting.org) with syzkaller.

Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
---
 net/core/sock.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/core/sock.c +1 -1
index c258887953905..3fd71f343c9f2 100644
@@ -2386,7 +2386,6 @@ void sk_setup_caps(struct sock *sk, struct dst_entry *dst)
 {
 	u32 max_segs = 1;
 
-	sk_dst_set(sk, dst);
 	sk->sk_route_caps = dst->dev->features;
 	if (sk_is_tcp(sk))
 		sk->sk_route_caps |= NETIF_F_GSO;
@@ -2405,6 +2404,7 @@ void sk_setup_caps(struct sock *sk, struct dst_entry *dst)
 		}
 	}
 	sk->sk_gso_max_segs = max_segs;
+	sk_dst_set(sk, dst);
 }
 EXPORT_SYMBOL_GPL(sk_setup_caps);
 
Signature · 1 line
2.39.2
author maintainer
#59 lore raw

PATCH 59/286 nfsd: fix double fget() bug in __write_ports_addfd()

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

[ Upstream commit c034203b6a9dae6751ef4371c18cb77983e30c28 ]

The bug here is that you cannot rely on getting the same socket from multiple calls to fget() because userspace can influence that. This is a kind of double fetch bug.

The fix is to delete the svc_alien_sock() function and instead do the checking inside the svc_addsock() function.

Fixes 3064639423c4 ("nfsd: check passed socket's net matches NFSd superblock's one")
---
 fs/nfsd/nfsctl.c               |  7 +------
 include/linux/sunrpc/svcsock.h |  7 +++----
 net/sunrpc/svcsock.c           | 24 ++++++------------------
 3 files changed, 10 insertions(+), 28 deletions(-)
fs/nfsd/nfsctl.c +1 -6
index 7b8f17ee52243..ba07757f3cd0a 100644
@@ -702,16 +702,11 @@ static ssize_t __write_ports_addfd(char *buf, struct net *net, const struct cred
 	if (err != 0 || fd < 0)
 		return -EINVAL;
 
-	if (svc_alien_sock(net, fd)) {
-		printk(KERN_ERR "%s: socket net is different to NFSd's one\n", __func__);
-		return -EINVAL;
-	}
-
 	err = nfsd_create_serv(net);
 	if (err != 0)
 		return err;
 
-	err = svc_addsock(nn->nfsd_serv, fd, buf, SIMPLE_TRANSACTION_LIMIT, cred);
+	err = svc_addsock(nn->nfsd_serv, net, fd, buf, SIMPLE_TRANSACTION_LIMIT, cred);
 
 	if (err >= 0 &&
 	    !nn->nfsd_serv->sv_nrthreads && !xchg(&nn->keep_active, 1))
include/linux/sunrpc/svcsock.h +3 -4
index bcc555c7ae9c6..13aff355d5a13 100644
@@ -59,10 +59,9 @@ int		svc_recv(struct svc_rqst *, long);
 int		svc_send(struct svc_rqst *);
 void		svc_drop(struct svc_rqst *);
 void		svc_sock_update_bufs(struct svc_serv *serv);
-bool		svc_alien_sock(struct net *net, int fd);
-int		svc_addsock(struct svc_serv *serv, const int fd,
-					char *name_return, const size_t len,
-					const struct cred *cred);
+int		svc_addsock(struct svc_serv *serv, struct net *net,
+			    const int fd, char *name_return, const size_t len,
+			    const struct cred *cred);
 void		svc_init_xprt_sock(void);
 void		svc_cleanup_xprt_sock(void);
 struct svc_xprt *svc_sock_create(struct svc_serv *serv, int prot);
net/sunrpc/svcsock.c +6 -18
index bf2d2cdca1185..e2a94589dd5df 100644
@@ -1340,25 +1340,10 @@ static struct svc_sock *svc_setup_socket(struct svc_serv *serv,
 	return svsk;
 }
 
-bool svc_alien_sock(struct net *net, int fd)
-{
-	int err;
-	struct socket *sock = sockfd_lookup(fd, &err);
-	bool ret = false;
-
-	if (!sock)
-		goto out;
-	if (sock_net(sock->sk) != net)
-		ret = true;
-	sockfd_put(sock);
-out:
-	return ret;
-}
-EXPORT_SYMBOL_GPL(svc_alien_sock);
-
 /**
  * svc_addsock - add a listener socket to an RPC service
  * @serv: pointer to RPC service to which to add a new listener
+ * @net: caller's network namespace
  * @fd: file descriptor of the new listener
  * @name_return: pointer to buffer to fill in with name of listener
  * @len: size of the buffer
@@ -1368,8 +1353,8 @@ EXPORT_SYMBOL_GPL(svc_alien_sock);
  * Name is terminated with '\n'.  On error, returns a negative errno
  * value.
  */
-int svc_addsock(struct svc_serv *serv, const int fd, char *name_return,
-		const size_t len, const struct cred *cred)
+int svc_addsock(struct svc_serv *serv, struct net *net, const int fd,
+		char *name_return, const size_t len, const struct cred *cred)
 {
 	int err = 0;
 	struct socket *so = sockfd_lookup(fd, &err);
@@ -1380,6 +1365,9 @@ int svc_addsock(struct svc_serv *serv, const int fd, char *name_return,
 
 	if (!so)
 		return err;
+	err = -EINVAL;
+	if (sock_net(so->sk) != net)
+		goto out;
 	err = -EAFNOSUPPORT;
 	if ((so->sk->sk_family != PF_INET) && (so->sk->sk_family != PF_INET6))
 		goto out;
Signature · 1 line
2.39.2
author maintainer
#60 lore raw

PATCH 60/286 HID: logitech-hidpp: Handle timeout differently from busy

In reply to Greg Kroah-Hartman

From: Bastien Nocera <[email protected]>

[ Upstream commit 6199d23c91ce53bfed455f09a8c5ed170d516824 ]

If an attempt at contacting a receiver or a device fails because the receiver or device never responds, don't restart the communication, only restart it if the receiver or device answers that it's busy, as originally intended.

This was the behaviour on communication timeout before commit 586e8fede795 ("HID: logitech-hidpp: Retry commands when device is busy").

This fixes some overly long waits in a critical path on boot, when checking whether the device is connected by getting its HID++ version.

Fixes 586e8fede795 ("HID: logitech-hidpp: Retry commands when device is busy")
---
 drivers/hid/hid-logitech-hidpp.c | 1 +
 1 file changed, 1 insertion(+)
drivers/hid/hid-logitech-hidpp.c +1 -0
index da89e84c9cbeb..0853114e16a08 100644
@@ -311,6 +311,7 @@ static int hidpp_send_message_sync(struct hidpp_device *hidpp,
 			dbg_hid("%s:timeout waiting for response\n", __func__);
 			memset(response, 0, sizeof(struct hidpp_report));
 			ret = -ETIMEDOUT;
+			goto exit;
 		}
 
 		if (response->report_id == REPORT_ID_HIDPP_SHORT &&
Signature · 1 line
2.39.2
author maintainer
#34 lore raw

PATCH 34/286 net/mlx5e: Move Ethernet driver debugfs to profile init callback

In reply to Greg Kroah-Hartman

From: Jianbo Liu <[email protected]>

[ Upstream commit c4c24fc30cc417ace332ceceaba4f70f81dcd521 ]

As priv->dfs_root is cleared, and therefore missed, when change eswitch mode, move the creation of the root debugfs to the init callback of mlx5e_nic_profile and mlx5e_uplink_rep_profile, and the destruction to the cleanup callback for symmeter.

Fixes 288eca60cc31 ("net/mlx5e: Add Ethernet driver debugfs")
---
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 10 +++++-----
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c  |  6 ++++++
 2 files changed, 11 insertions(+), 5 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en_main.c +5 -5
index a60610c7a7bb7..7c72bed7f81aa 100644
@@ -5226,12 +5226,16 @@ static int mlx5e_nic_init(struct mlx5_core_dev *mdev,
 
 	mlx5e_timestamp_init(priv);
 
+	priv->dfs_root = debugfs_create_dir("nic",
+					    mlx5_debugfs_get_dev_root(mdev));
+
 	fs = mlx5e_fs_init(priv->profile, mdev,
 			   !test_bit(MLX5E_STATE_DESTROYING, &priv->state),
 			   priv->dfs_root);
 	if (!fs) {
 		err = -ENOMEM;
 		mlx5_core_err(mdev, "FS initialization failed, %d\n", err);
+		debugfs_remove_recursive(priv->dfs_root);
 		return err;
 	}
 	priv->fs = fs;
@@ -5252,6 +5256,7 @@ static void mlx5e_nic_cleanup(struct mlx5e_priv *priv)
 	mlx5e_health_destroy_reporters(priv);
 	mlx5e_ktls_cleanup(priv);
 	mlx5e_fs_cleanup(priv->fs);
+	debugfs_remove_recursive(priv->dfs_root);
 	priv->fs = NULL;
 }
 
@@ -5976,9 +5981,6 @@ static int mlx5e_probe(struct auxiliary_device *adev,
 	priv->profile = profile;
 	priv->ppriv = NULL;
 
-	priv->dfs_root = debugfs_create_dir("nic",
-					    mlx5_debugfs_get_dev_root(priv->mdev));
-
 	err = profile->init(mdev, netdev);
 	if (err) {
 		mlx5_core_err(mdev, "mlx5e_nic_profile init failed, %d\n", err);
@@ -6007,7 +6009,6 @@ static int mlx5e_probe(struct auxiliary_device *adev,
 err_profile_cleanup:
 	profile->cleanup(priv);
 err_destroy_netdev:
-	debugfs_remove_recursive(priv->dfs_root);
 	mlx5e_destroy_netdev(priv);
 err_devlink_port_unregister:
 	mlx5e_devlink_port_unregister(mlx5e_dev);
@@ -6027,7 +6028,6 @@ static void mlx5e_remove(struct auxiliary_device *adev)
 	unregister_netdev(priv->netdev);
 	mlx5e_suspend(adev, state);
 	priv->profile->cleanup(priv);
-	debugfs_remove_recursive(priv->dfs_root);
 	mlx5e_destroy_netdev(priv);
 	mlx5e_devlink_port_unregister(mlx5e_dev);
 	mlx5e_destroy_devlink(mlx5e_dev);
drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +6 -0
index 6e18d91c3d766..992f3f9c11925 100644
@@ -30,6 +30,7 @@
  * SOFTWARE.
  */
 
+#include <linux/debugfs.h>
 #include <linux/mlx5/fs.h>
 #include <net/switchdev.h>
 #include <net/pkt_cls.h>
@@ -811,11 +812,15 @@ static int mlx5e_init_ul_rep(struct mlx5_core_dev *mdev,
 {
 	struct mlx5e_priv *priv = netdev_priv(netdev);
 
+	priv->dfs_root = debugfs_create_dir("nic",
+					    mlx5_debugfs_get_dev_root(mdev));
+
 	priv->fs = mlx5e_fs_init(priv->profile, mdev,
 				 !test_bit(MLX5E_STATE_DESTROYING, &priv->state),
 				 priv->dfs_root);
 	if (!priv->fs) {
 		netdev_err(priv->netdev, "FS allocation failed\n");
+		debugfs_remove_recursive(priv->dfs_root);
 		return -ENOMEM;
 	}
 
@@ -828,6 +833,7 @@ static int mlx5e_init_ul_rep(struct mlx5_core_dev *mdev,
 static void mlx5e_cleanup_rep(struct mlx5e_priv *priv)
 {
 	mlx5e_fs_cleanup(priv->fs);
+	debugfs_remove_recursive(priv->dfs_root);
 	priv->fs = NULL;
 }
 
Signature · 1 line
2.39.2
author maintainer
#61 lore raw

PATCH 61/286 nvme: fix the name of Zone Append for verbose logging

In reply to Greg Kroah-Hartman

From: Christoph Hellwig <[email protected]>

[ Upstream commit 856303797724d28f1d65b702f0eadcee1ea7abf5 ]

No Management involved in Zone Appened.

Fixes bd83fe6f2cd2 ("nvme: add verbose error logging")
---
 drivers/nvme/host/constants.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/nvme/host/constants.c +1 -1
index bc523ca022548..5e4f8848dce08 100644
@@ -21,7 +21,7 @@ static const char * const nvme_ops[] = {
 	[nvme_cmd_resv_release] = "Reservation Release",
 	[nvme_cmd_zone_mgmt_send] = "Zone Management Send",
 	[nvme_cmd_zone_mgmt_recv] = "Zone Management Receive",
-	[nvme_cmd_zone_append] = "Zone Management Append",
+	[nvme_cmd_zone_append] = "Zone Append",
 };
 
 static const char * const nvme_admin_ops[] = {
Signature · 1 line
2.39.2
author maintainer
#62 lore raw

PATCH 62/286 net/mlx5e: Fix error handling in mlx5e_refresh_tirs

In reply to Greg Kroah-Hartman

From: Saeed Mahameed <[email protected]>

[ Upstream commit b6193d7030e3c59f1d4c75648c9c8fa40cad2bcd ]

Allocation failure is outside the critical lock section and should return immediately rather than jumping to the unlock section.

Also unlock as soon as required and remove the now redundant jump label.

Fixes 80a2a9026b24 ("net/mlx5e: Add a lock on tir list")
---
 drivers/net/ethernet/mellanox/mlx5/core/en_common.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en_common.c +4 -7
index 993af4c12d909..21cd232c2c20c 100644
@@ -149,10 +149,8 @@ int mlx5e_refresh_tirs(struct mlx5e_priv *priv, bool enable_uc_lb,
 
 	inlen = MLX5_ST_SZ_BYTES(modify_tir_in);
 	in = kvzalloc(inlen, GFP_KERNEL);
-	if (!in) {
-		err = -ENOMEM;
-		goto out;
-	}
+	if (!in)
+		return -ENOMEM;
 
 	if (enable_uc_lb)
 		lb_flags = MLX5_TIRC_SELF_LB_BLOCK_BLOCK_UNICAST;
@@ -170,14 +168,13 @@ int mlx5e_refresh_tirs(struct mlx5e_priv *priv, bool enable_uc_lb,
 		tirn = tir->tirn;
 		err = mlx5_core_modify_tir(mdev, tirn, in);
 		if (err)
-			goto out;
+			break;
 	}
+	mutex_unlock(&mdev->mlx5e_res.hw_objs.td.list_lock);
 
-out:
 	kvfree(in);
 	if (err)
 		netdev_err(priv->netdev, "refresh tir(0x%x) failed, %d\n", tirn, err);
-	mutex_unlock(&mdev->mlx5e_res.hw_objs.td.list_lock);
 
 	return err;
 }
Signature · 1 line
2.39.2
author maintainer
#63 lore raw

PATCH 63/286 net/mlx5: Read embedded cpu after init bit cleared

In reply to Greg Kroah-Hartman

From: Moshe Shemesh <[email protected]>

[ Upstream commit bbfa4b58997e3d38ba629c9f6fc0bd1c163aaf43 ]

During driver load it reads embedded_cpu bit from initialization segment, but the initialization segment is readable only after initialization bit is cleared.

Move the call to mlx5_read_embedded_cpu() right after initialization bit cleared.

Fixes 591905ba9679 ("net/mlx5: Introduce Mellanox SmartNIC and modify page management logic")
---
 drivers/net/ethernet/mellanox/mlx5/core/main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/mellanox/mlx5/core/main.c +1 -1
index 504179dbff9dd..9058fa8c5b657 100644
@@ -913,7 +913,6 @@ static int mlx5_pci_init(struct mlx5_core_dev *dev, struct pci_dev *pdev,
 	}
 
 	mlx5_pci_vsc_init(dev);
-	dev->caps.embedded_cpu = mlx5_read_embedded_cpu(dev);
 	return 0;
 
 err_clr_master:
@@ -1147,6 +1146,7 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot, u64 timeout
 		goto err_cmd_cleanup;
 	}
 
+	dev->caps.embedded_cpu = mlx5_read_embedded_cpu(dev);
 	mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_UP);
 
 	mlx5_start_health_poll(dev);
Signature · 1 line
2.39.2
author maintainer
#64 lore raw

PATCH 64/286 sfc: fix error unwinds in TC offload

In reply to Greg Kroah-Hartman

From: Edward Cree <[email protected]>

[ Upstream commit 622ab656344a288acf4fb03d628c3bb5dd241f34 ]

Failure ladders weren't exactly unwinding what the function had done up
to that point; most seriously, when we encountered an already offloaded rule, the failure path tried to remove the new rule from the hashtable, which would in fact remove the already-present 'old' rule (since it has the same key) from the table, and leak its resources.

Fixes d902e1a737d4 ("sfc: bare bones TC offload on EF100")
Fixes 17654d84b47c ("sfc: add offloading of 'foreign' TC (decap) rules")
---
 drivers/net/ethernet/sfc/tc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/sfc/tc.c +2 -2
index deeaab9ee761d..217f3876af722 100644
@@ -379,9 +379,9 @@ static int efx_tc_flower_replace(struct efx_nic *efx,
 	if (old) {
 		netif_dbg(efx, drv, efx->net_dev,
 			  "Already offloaded rule (cookie %lx)\n", tc->cookie);
-		rc = -EEXIST;
 		NL_SET_ERR_MSG_MOD(extack, "Rule already offloaded");
-		goto release;
+		kfree(rule);
+		return -EEXIST;
 	}
 
 	/* Parse actions */
Signature · 1 line
2.39.2
author maintainer
#35 lore raw

PATCH 35/286 net: mellanox: mlxbf_gige: Fix skb_panic splat under memory pressure

In reply to Greg Kroah-Hartman

From: Thomas Bogendoerfer <[email protected]>

[ Upstream commit d68cb7cf1fd0ef4287bc0ecd1ed0b6ae8e05fc70 ]

Do skb_put() after a new skb has been successfully allocated otherwise the reused skb leads to skb_panics or incorrect packet sizes.

Fixes f92e1869d74e ("Add Mellanox BlueField Gigabit Ethernet driver")
---
 .../ethernet/mellanox/mlxbf_gige/mlxbf_gige_rx.c    | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)
drivers/net/ethernet/mellanox/mlxbf_gige/mlxbf_gige_rx.c +7 -6
index afa3b92a6905f..0d5a41a2ae010 100644
@@ -245,12 +245,6 @@ static bool mlxbf_gige_rx_packet(struct mlxbf_gige *priv, int *rx_pkts)
 
 		skb = priv->rx_skb[rx_pi_rem];
 
-		skb_put(skb, datalen);
-
-		skb->ip_summed = CHECKSUM_NONE; /* device did not checksum packet */
-
-		skb->protocol = eth_type_trans(skb, netdev);
-
 		/* Alloc another RX SKB for this same index */
 		rx_skb = mlxbf_gige_alloc_skb(priv, MLXBF_GIGE_DEFAULT_BUF_SZ,
 					      &rx_buf_dma, DMA_FROM_DEVICE);
@@ -259,6 +253,13 @@ static bool mlxbf_gige_rx_packet(struct mlxbf_gige *priv, int *rx_pkts)
 		priv->rx_skb[rx_pi_rem] = rx_skb;
 		dma_unmap_single(priv->dev, *rx_wqe_addr,
 				 MLXBF_GIGE_DEFAULT_BUF_SZ, DMA_FROM_DEVICE);
+
+		skb_put(skb, datalen);
+
+		skb->ip_summed = CHECKSUM_NONE; /* device did not checksum packet */
+
+		skb->protocol = eth_type_trans(skb, netdev);
+
 		*rx_wqe_addr = rx_buf_dma;
 	} else if (rx_cqe & MLXBF_GIGE_RX_CQE_PKT_STATUS_MAC_ERR) {
 		priv->stats.rx_mac_errors++;
Signature · 1 line
2.39.2
author maintainer
#36 lore raw

PATCH 36/286 net: stmmac: fix call trace when stmmac_xdp_xmit() is invoked

In reply to Greg Kroah-Hartman

From: Wei Fang <[email protected]>

[ Upstream commit ffb3322181d9e8db880202e4f00991764a35d812 ]

We encountered a kernel call trace issue which was related to ndo_xdp_xmit callback on our i.MX8MP platform. The reproduce steps show as follows.
1. The FEC port (eth0) connects to a PC port, and the PC uses pktgen_sample03_burst_single_flow.sh to generate packets and send these packets to the FEC port. Notice that the script must be executed before step 2.
2. Run the "./xdp_redirect eth0 eth1" command on i.MX8MP, the eth1 interface is the dwmac. Then there will be a call trace issue soon. Please see the log for more details.
The root cause is that the NETDEV_XDP_ACT_NDO_XMIT feature is enabled by default, so when the step 2 command is exexcuted and packets have already been sent to eth0, the stmmac_xdp_xmit() starts running before the stmmac_xdp_set_prog() finishes. To resolve this issue, we disable the NETDEV_XDP_ACT_NDO_XMIT feature by default and turn on/off this feature when the bpf program is installed/uninstalled which just like the other ethernet drivers.

Call Trace log:

[  306.311271] ------------[ cut here ]------------
[  306.315910] WARNING: CPU: 0 PID: 15 at lib/timerqueue.c:55 timerqueue_del+0x68/0x70
[  306.323590] Modules linked in:
[  306.326654] CPU: 0 PID: 15 Comm: ksoftirqd/0 Not tainted 6.4.0-rc1+ #37
[  306.333277] Hardware name: NXP i.MX8MPlus EVK board (DT)
[  306.338591] pstate: 600000c5 (nZCv daIF -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[  306.345561] pc : timerqueue_del+0x68/0x70
[  306.349577] lr : __remove_hrtimer+0x5c/0xa0
[  306.353777] sp : ffff80000b7c3920
[  306.357094] x29: ffff80000b7c3920 x28: 0000000000000000 x27: 0000000000000001
[  306.364244] x26: ffff80000a763a40 x25: ffff0000d0285a00 x24: 0000000000000001
[  306.371390] x23: 0000000000000001 x22: ffff000179389a40 x21: 0000000000000000
[  306.378537] x20: ffff000179389aa0 x19: ffff0000d2951308 x18: 0000000000001000
[  306.385686] x17: f1d3000000000000 x16: 00000000c39c1000 x15: 55e99bbe00001a00
[  306.392835] x14: 09000900120aa8c0 x13: e49af1d300000000 x12: 000000000000c39c
[  306.399987] x11: 100055e99bbe0000 x10: ffff8000090b1048 x9 : ffff8000081603fc
[  306.407133] x8 : 000000000000003c x7 : 000000000000003c x6 : 0000000000000001
[  306.414284] x5 : ffff0000d2950980 x4 : 0000000000000000 x3 : 0000000000000000
[  306.421432] x2 : 0000000000000001 x1 : ffff0000d2951308 x0 : ffff0000d2951308
[  306.428585] Call trace:
[  306.431035]  timerqueue_del+0x68/0x70
[  306.434706]  __remove_hrtimer+0x5c/0xa0
[  306.438549]  hrtimer_start_range_ns+0x2bc/0x370
[  306.443089]  stmmac_xdp_xmit+0x174/0x1b0
[  306.447021]  bq_xmit_all+0x194/0x4b0
[  306.450612]  __dev_flush+0x4c/0x98
[  306.454024]  xdp_do_flush+0x18/0x38
[  306.457522]  fec_enet_rx_napi+0x6c8/0xc68
[  306.461539]  __napi_poll+0x40/0x220
[  306.465038]  net_rx_action+0xf8/0x240
[  306.468707]  __do_softirq+0x128/0x3a8
[  306.472378]  run_ksoftirqd+0x40/0x58
[  306.475961]  smpboot_thread_fn+0x1c4/0x288
[  306.480068]  kthread+0x124/0x138
[  306.483305]  ret_from_fork+0x10/0x20
[  306.486889] ---[ end trace 0000000000000000 ]---
---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 3 +--
 drivers/net/ethernet/stmicro/stmmac/stmmac_xdp.c  | 6 ++++++
 2 files changed, 7 insertions(+), 2 deletions(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +1 -2
index f9cd063f1fe30..71f8f78ce0090 100644
@@ -7176,8 +7176,7 @@ int stmmac_dvr_probe(struct device *device,
 	ndev->hw_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
 			    NETIF_F_RXCSUM;
 	ndev->xdp_features = NETDEV_XDP_ACT_BASIC | NETDEV_XDP_ACT_REDIRECT |
-			     NETDEV_XDP_ACT_XSK_ZEROCOPY |
-			     NETDEV_XDP_ACT_NDO_XMIT;
+			     NETDEV_XDP_ACT_XSK_ZEROCOPY;
 
 	ret = stmmac_tc_init(priv, priv);
 	if (!ret) {
drivers/net/ethernet/stmicro/stmmac/stmmac_xdp.c +6 -0
index 9d4d8c3dad0a3..aa6f16d3df649 100644
@@ -117,6 +117,9 @@ int stmmac_xdp_set_prog(struct stmmac_priv *priv, struct bpf_prog *prog,
 		return -EOPNOTSUPP;
 	}
 
+	if (!prog)
+		xdp_features_clear_redirect_target(dev);
+
 	need_update = !!priv->xdp_prog != !!prog;
 	if (if_running && need_update)
 		stmmac_xdp_release(dev);
@@ -131,5 +134,8 @@ int stmmac_xdp_set_prog(struct stmmac_priv *priv, struct bpf_prog *prog,
 	if (if_running && need_update)
 		stmmac_xdp_open(dev);
 
+	if (prog)
+		xdp_features_set_redirect_target(dev, false);
+
 	return 0;
 }
Signature · 1 line
2.39.2
author maintainer
#37 lore raw

PATCH 37/286 netrom: fix info-leak in nr_write_internal()

In reply to Greg Kroah-Hartman

From: Eric Dumazet <[email protected]>

[ Upstream commit 31642e7089df8fd3f54ca7843f7ee2952978cad1 ]

Simon Kapadia reported the following issue:

<quote>

The Online Amateur Radio Community (OARC) has recently been experimenting with building a nationwide packet network in the UK.
As part of our experimentation, we have been testing out packet on 300bps HF, and playing with net/rom. For HF packet at this baud rate you really need to make sure that your MTU is relatively low; AX.25 suggests a PACLEN of 60, and a net/rom PACLEN of 40 to go with that.
However the Linux net/rom support didn't work with a low PACLEN;
the mkiss module would truncate packets if you set the PACLEN below about 200 or so, e.g.:

Apr 19 14:00:51 radio kernel: [12985.747310] mkiss: ax1: truncating oversized transmit packet!

This didn't make any sense to me (if the packets are smaller why would they be truncated?) so I started investigating.
I looked at the packets using ethereal, and found that many were just huge compared to what I would expect.
A simple net/rom connection request packet had the request and then a bunch of what appeared to be random data following it:

</quote>

Simon provided a patch that I slightly revised:
Not only we must not use skb_tailroom(), we also do not want to count NR_NETWORK_LEN twice.

---
 net/netrom/nr_subr.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
net/netrom/nr_subr.c +4 -3
index 3f99b432ea707..e2d2af924cff4 100644
@@ -123,7 +123,7 @@ void nr_write_internal(struct sock *sk, int frametype)
 	unsigned char  *dptr;
 	int len, timeout;
 
-	len = NR_NETWORK_LEN + NR_TRANSPORT_LEN;
+	len = NR_TRANSPORT_LEN;
 
 	switch (frametype & 0x0F) {
 	case NR_CONNREQ:
@@ -141,7 +141,8 @@ void nr_write_internal(struct sock *sk, int frametype)
 		return;
 	}
 
-	if ((skb = alloc_skb(len, GFP_ATOMIC)) == NULL)
+	skb = alloc_skb(NR_NETWORK_LEN + len, GFP_ATOMIC);
+	if (!skb)
 		return;
 
 	/*
@@ -149,7 +150,7 @@ void nr_write_internal(struct sock *sk, int frametype)
 	 */
 	skb_reserve(skb, NR_NETWORK_LEN);
 
-	dptr = skb_put(skb, skb_tailroom(skb));
+	dptr = skb_put(skb, len);
 
 	switch (frametype & 0x0F) {
 	case NR_CONNREQ:
Signature · 1 line
2.39.2
author maintainer
#38 lore raw

PATCH 38/286 af_packet: Fix data-races of pkt_sk(sk)->num.

In reply to Greg Kroah-Hartman

From: Kuniyuki Iwashima <[email protected]>

[ Upstream commit 822b5a1c17df7e338b9f05d1cfe5764e37c7f74f ]

syzkaller found a data race of pkt_sk(sk)->num.

The value is changed under lock_sock() and po->bind_lock, so we need READ_ONCE() to access pkt_sk(sk)->num without these locks in packet_bind_spkt(), packet_bind(), and sk_diag_fill().

Note that WRITE_ONCE() is already added by commit c7d2ef5dd4b0 ("net/packet: annotate accesses to po->bind").

BUG: KCSAN: data-race in packet_bind / packet_do_bind

write (marked) to 0xffff88802ffd1cee of 2 bytes by task 7322 on cpu 0:

 packet_do_bind+0x446/0x640 net/packet/af_packet.c:3236
 packet_bind+0x99/0xe0 net/packet/af_packet.c:3321
 __sys_bind+0x19b/0x1e0 net/socket.c:1803
 __do_sys_bind net/socket.c:1814 [inline]
 __se_sys_bind net/socket.c:1812 [inline]
 __x64_sys_bind+0x40/0x50 net/socket.c:1812
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x3b/0x90 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x72/0xdc

read to 0xffff88802ffd1cee of 2 bytes by task 7318 on cpu 1:

 packet_bind+0xbf/0xe0 net/packet/af_packet.c:3322
 __sys_bind+0x19b/0x1e0 net/socket.c:1803
 __do_sys_bind net/socket.c:1814 [inline]
 __se_sys_bind net/socket.c:1812 [inline]
 __x64_sys_bind+0x40/0x50 net/socket.c:1812
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x3b/0x90 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x72/0xdc

value changed: 0x0300 -> 0x0000

Reported by Kernel Concurrency Sanitizer on:

CPU: 1 PID: 7318 Comm: syz-executor.4 Not tainted 6.3.0-13380-g7fddb5b5300c #4
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
Fixes 96ec6327144e ("packet: Diag core and basic socket info dumping")
Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
---
 net/packet/af_packet.c | 4 ++--
 net/packet/diag.c      | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)
net/packet/af_packet.c +2 -2
index db9c2fa71c50c..f3513316743ad 100644
@@ -3291,7 +3291,7 @@ static int packet_bind_spkt(struct socket *sock, struct sockaddr *uaddr,
 	memcpy(name, uaddr->sa_data, sizeof(uaddr->sa_data_min));
 	name[sizeof(uaddr->sa_data_min)] = 0;
 
-	return packet_do_bind(sk, name, 0, pkt_sk(sk)->num);
+	return packet_do_bind(sk, name, 0, READ_ONCE(pkt_sk(sk)->num));
 }
 
 static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
@@ -3309,7 +3309,7 @@ static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len
 		return -EINVAL;
 
 	return packet_do_bind(sk, NULL, sll->sll_ifindex,
-			      sll->sll_protocol ? : pkt_sk(sk)->num);
+			      sll->sll_protocol ? : READ_ONCE(pkt_sk(sk)->num));
 }
 
 static struct proto packet_proto = {
net/packet/diag.c +1 -1
index d704c7bf51b20..a68a84574c739 100644
@@ -143,7 +143,7 @@ static int sk_diag_fill(struct sock *sk, struct sk_buff *skb,
 	rp = nlmsg_data(nlh);
 	rp->pdiag_family = AF_PACKET;
 	rp->pdiag_type = sk->sk_type;
-	rp->pdiag_num = ntohs(po->num);
+	rp->pdiag_num = ntohs(READ_ONCE(po->num));
 	rp->pdiag_ino = sk_ino;
 	sock_diag_save_cookie(sk, rp->pdiag_cookie);
 
Signature · 1 line
2.39.2
author maintainer
#39 lore raw

PATCH 39/286 tls: improve lockless access safety of tls_err_abort()

In reply to Greg Kroah-Hartman

From: Jakub Kicinski <[email protected]>

[ Upstream commit 8a0d57df8938e9fd2e99d47a85b7f37d86f91097 ]

Most protos' poll() methods insert a memory barrier between writes to sk_err and sk_error_report(). This dates back to commit a4d258036ed9 ("tcp: Fix race in tcp_poll").

I guess we should do the same thing in TLS, tcp_poll() does not hold the socket lock.

Fixes 3c4d7559159b ("tls: kernel TLS support")
---
 net/tls/tls_strp.c | 4 +++-
 net/tls/tls_sw.c   | 4 +++-
 2 files changed, 6 insertions(+), 2 deletions(-)
net/tls/tls_strp.c +3 -1
index da95abbb7ea32..f37f4a0fcd3c2 100644
@@ -20,7 +20,9 @@ static void tls_strp_abort_strp(struct tls_strparser *strp, int err)
 	strp->stopped = 1;
 
 	/* Report an error on the lower socket */
-	strp->sk->sk_err = -err;
+	WRITE_ONCE(strp->sk->sk_err, -err);
+	/* Paired with smp_rmb() in tcp_poll() */
+	smp_wmb();
 	sk_error_report(strp->sk);
 }
 
net/tls/tls_sw.c +3 -1
index 6e6a7c37d685c..1a53c8f481e9a 100644
@@ -70,7 +70,9 @@ noinline void tls_err_abort(struct sock *sk, int err)
 {
 	WARN_ON_ONCE(err >= 0);
 	/* sk->sk_err should contain a positive error code. */
-	sk->sk_err = -err;
+	WRITE_ONCE(sk->sk_err, -err);
+	/* Paired with smp_rmb() in tcp_poll() */
+	smp_wmb();
 	sk_error_report(sk);
 }
 
Signature · 1 line
2.39.2
author maintainer
#40 lore raw

PATCH 40/286 amd-xgbe: fix the false linkup in xgbe_phy_status

In reply to Greg Kroah-Hartman

From: Raju Rangoju <[email protected]>

[ Upstream commit dc362e20cd6ab7a93d1b09669730c406f0910c35 ]

In the event of a change in XGBE mode, the current auto-negotiation needs to be reset and the AN cycle needs to be re-triggerred. However, the current code ignores the return value of xgbe_set_mode(), leading to false information as the link is declared without checking the status register.

Fix this by propagating the mode switch status information to xgbe_phy_status().

Fixes e57f7a3feaef ("amd-xgbe: Prepare for working with more than one type of phy")
---
 drivers/net/ethernet/amd/xgbe/xgbe-mdio.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)
drivers/net/ethernet/amd/xgbe/xgbe-mdio.c +9 -3
index 33a9574e9e043..32d2c6fac6526 100644
@@ -1329,7 +1329,7 @@ static enum xgbe_mode xgbe_phy_status_aneg(struct xgbe_prv_data *pdata)
 	return pdata->phy_if.phy_impl.an_outcome(pdata);
 }
 
-static void xgbe_phy_status_result(struct xgbe_prv_data *pdata)
+static bool xgbe_phy_status_result(struct xgbe_prv_data *pdata)
 {
 	struct ethtool_link_ksettings *lks = &pdata->phy.lks;
 	enum xgbe_mode mode;
@@ -1367,8 +1367,13 @@ static void xgbe_phy_status_result(struct xgbe_prv_data *pdata)
 
 	pdata->phy.duplex = DUPLEX_FULL;
 
-	if (xgbe_set_mode(pdata, mode) && pdata->an_again)
+	if (!xgbe_set_mode(pdata, mode))
+		return false;
+
+	if (pdata->an_again)
 		xgbe_phy_reconfig_aneg(pdata);
+
+	return true;
 }
 
 static void xgbe_phy_status(struct xgbe_prv_data *pdata)
@@ -1398,7 +1403,8 @@ static void xgbe_phy_status(struct xgbe_prv_data *pdata)
 			return;
 		}
 
-		xgbe_phy_status_result(pdata);
+		if (xgbe_phy_status_result(pdata))
+			return;
 
 		if (test_bit(XGBE_LINK_INIT, &pdata->dev_state))
 			clear_bit(XGBE_LINK_INIT, &pdata->dev_state);
Signature · 1 line
2.39.2
author maintainer
#31 lore raw

PATCH 31/286 net/mlx5e: TC, Remove CT action reordering

In reply to Greg Kroah-Hartman

From: Paul Blakey <[email protected]>

[ Upstream commit 67efaf45930df662111acf7c706d545c83f83999 ]

CT action reordering was done as a workaround when CT misses used to restore the relevant filter's tc chain and continuing sw processing from that chain. As such, there was a need to reorder CT action to be before any packet modifying actions (e.g mac rewrite).

Currently (after patch "net/mlx5e: TC, Set CT miss to the specific ct action instance"), CT misses continues from the relevant ct action in software, and so reordering isn't needed anymore.

Remove the reordering.

Stable-dep-of: 5d862ec631f3 ("net/mlx5: Fix post parse infra to only parse every action once")

---
 .../mellanox/mlx5/core/en/tc/act/act.c        | 20 ------------
 .../mellanox/mlx5/core/en/tc/act/act.h        |  4 ---
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   | 32 ++++++-------------
 3 files changed, 9 insertions(+), 47 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.c +0 -20
index eba0c86989263..fc923a99b6a48 100644
@@ -82,26 +82,6 @@ mlx5e_tc_act_init_parse_state(struct mlx5e_tc_act_parse_state *parse_state,
 	parse_state->flow_action = flow_action;
 }
 
-void
-mlx5e_tc_act_reorder_flow_actions(struct flow_action *flow_action,
-				  struct mlx5e_tc_flow_action *flow_action_reorder)
-{
-	struct flow_action_entry *act;
-	int i, j = 0;
-
-	flow_action_for_each(i, act, flow_action) {
-		/* Add CT action to be first. */
-		if (act->id == FLOW_ACTION_CT)
-			flow_action_reorder->entries[j++] = act;
-	}
-
-	flow_action_for_each(i, act, flow_action) {
-		if (act->id == FLOW_ACTION_CT)
-			continue;
-		flow_action_reorder->entries[j++] = act;
-	}
-}
-
 int
 mlx5e_tc_act_post_parse(struct mlx5e_tc_act_parse_state *parse_state,
 			struct flow_action *flow_action,
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h +0 -4
index cdcddf6e1b08b..d7615e329e6d9 100644
@@ -112,10 +112,6 @@ mlx5e_tc_act_init_parse_state(struct mlx5e_tc_act_parse_state *parse_state,
 			      struct flow_action *flow_action,
 			      struct netlink_ext_ack *extack);
 
-void
-mlx5e_tc_act_reorder_flow_actions(struct flow_action *flow_action,
-				  struct mlx5e_tc_flow_action *flow_action_reorder);
-
 int
 mlx5e_tc_act_post_parse(struct mlx5e_tc_act_parse_state *parse_state,
 			struct flow_action *flow_action,
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +9 -23
index 82b76dcc05ac1..a78e201fd883b 100644
@@ -3967,32 +3967,22 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 		 struct flow_action *flow_action)
 {
 	struct netlink_ext_ack *extack = parse_state->extack;
-	struct mlx5e_tc_flow_action flow_action_reorder;
 	struct mlx5e_tc_flow *flow = parse_state->flow;
 	struct mlx5e_tc_jump_state jump_state = {};
 	struct mlx5_flow_attr *attr = flow->attr;
 	enum mlx5_flow_namespace_type ns_type;
 	struct mlx5e_priv *priv = flow->priv;
-	struct flow_action_entry *act, **_act;
 	struct mlx5_flow_attr *prev_attr;
+	struct flow_action_entry *act;
 	struct mlx5e_tc_act *tc_act;
 	bool is_missable;
 	int err, i;
 
-	flow_action_reorder.num_entries = flow_action->num_entries;
-	flow_action_reorder.entries = kcalloc(flow_action->num_entries,
-					      sizeof(flow_action), GFP_KERNEL);
-	if (!flow_action_reorder.entries)
-		return -ENOMEM;
-
-	mlx5e_tc_act_reorder_flow_actions(flow_action, &flow_action_reorder);
-
 	ns_type = mlx5e_get_flow_namespace(flow);
 	list_add(&attr->list, &flow->attrs);
 
-	flow_action_for_each(i, _act, &flow_action_reorder) {
+	flow_action_for_each(i, act, flow_action) {
 		jump_state.jump_target = false;
-		act = *_act;
 		is_missable = false;
 		prev_attr = attr;
 
@@ -4000,23 +3990,23 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 		if (!tc_act) {
 			NL_SET_ERR_MSG_MOD(extack, "Not implemented offload action");
 			err = -EOPNOTSUPP;
-			goto out_free;
+			goto out_free_post_acts;
 		}
 
 		if (!tc_act->can_offload(parse_state, act, i, attr)) {
 			err = -EOPNOTSUPP;
-			goto out_free;
+			goto out_free_post_acts;
 		}
 
 		err = tc_act->parse_action(parse_state, act, priv, attr);
 		if (err)
-			goto out_free;
+			goto out_free_post_acts;
 
 		dec_jump_count(act, tc_act, attr, priv, &jump_state);
 
 		err = parse_branch_ctrl(act, tc_act, flow, attr, &jump_state, extack);
 		if (err)
-			goto out_free;
+			goto out_free_post_acts;
 
 		parse_state->actions |= attr->action;
 
@@ -4024,17 +4014,17 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 		if (jump_state.jump_target ||
 		    (tc_act->is_multi_table_act &&
 		    tc_act->is_multi_table_act(priv, act, attr) &&
-		    i < flow_action_reorder.num_entries - 1)) {
+		    i < flow_action->num_entries - 1)) {
 			is_missable = tc_act->is_missable ? tc_act->is_missable(act) : false;
 
 			err = mlx5e_tc_act_post_parse(parse_state, flow_action, attr, ns_type);
 			if (err)
-				goto out_free;
+				goto out_free_post_acts;
 
 			attr = mlx5e_clone_flow_attr_for_post_act(flow->attr, ns_type);
 			if (!attr) {
 				err = -ENOMEM;
-				goto out_free;
+				goto out_free_post_acts;
 			}
 
 			list_add(&attr->list, &flow->attrs);
@@ -4051,8 +4041,6 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 		}
 	}
 
-	kfree(flow_action_reorder.entries);
-
 	err = mlx5e_tc_act_post_parse(parse_state, flow_action, attr, ns_type);
 	if (err)
 		goto out_free_post_acts;
@@ -4063,8 +4051,6 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
 
 	return 0;
 
-out_free:
-	kfree(flow_action_reorder.entries);
 out_free_post_acts:
 	free_flow_post_acts(flow);
 
Signature · 1 line
2.39.2
author maintainer
#66 lore raw

PATCH 66/286 net/sched: flower: fix possible OOB write in fl_set_geneve_opt()

In reply to Greg Kroah-Hartman

From: Hangyu Hua <[email protected]>

[ Upstream commit 4d56304e5827c8cc8cc18c75343d283af7c4825c ]

If we send two TCA_FLOWER_KEY_ENC_OPTS_GENEVE packets and their total size is 252 bytes(key->enc_opts.len = 252) then key->enc_opts.len = opt->length = data_len / 4 = 0 when the third TCA_FLOWER_KEY_ENC_OPTS_GENEVE packet enters fl_set_geneve_opt. This bypasses the next bounds check and results in an out-of-bounds.

---
 net/sched/cls_flower.c | 3 +++
 1 file changed, 3 insertions(+)
net/sched/cls_flower.c +3 -0
index a1c4ee2e0be22..fd5dc47cb2134 100644
@@ -1153,6 +1153,9 @@ static int fl_set_geneve_opt(const struct nlattr *nla, struct fl_flow_key *key,
 	if (option_len > sizeof(struct geneve_opt))
 		data_len = option_len - sizeof(struct geneve_opt);
 
+	if (key->enc_opts.len > FLOW_DIS_TUN_OPTS_MAX - 4)
+		return -ERANGE;
+
 	opt = (struct geneve_opt *)&key->enc_opts.data[key->enc_opts.len];
 	memset(opt, 0xff, option_len);
 	opt->length = data_len / 4;
Signature · 1 line
2.39.2
author maintainer
#75 lore raw

PATCH 75/286 rtnetlink: call validate_linkmsg in rtnl_create_link

In reply to Greg Kroah-Hartman

From: Xin Long <[email protected]>

[ Upstream commit b0ad3c179059089d809b477a1d445c1183a7b8fe ]

validate_linkmsg() was introduced by commit 1840bb13c22f5b ("[RTNL]:
Validate hardware and broadcast address attribute for RTM_NEWLINK") to validate tb[IFLA_ADDRESS/BROADCAST] for existing links. The same check should also be done for newly created links.

This patch adds validate_linkmsg() call in rtnl_create_link(), to avoid the invalid address set when creating some devices like:

  # ip link add dummy0 type dummy
  # ip link add link dummy0 name mac0 address 01:02 type macsec
Fixes 0e06877c6fdb ("[RTNETLINK]: rtnl_link: allow specifying initial device address")
---
 net/core/rtnetlink.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
net/core/rtnetlink.c +7 -1
index 6e44e92ebdf5d..f37deb18dd02e 100644
@@ -3282,6 +3282,7 @@ struct net_device *rtnl_create_link(struct net *net, const char *ifname,
 	struct net_device *dev;
 	unsigned int num_tx_queues = 1;
 	unsigned int num_rx_queues = 1;
+	int err;
 
 	if (tb[IFLA_NUM_TX_QUEUES])
 		num_tx_queues = nla_get_u32(tb[IFLA_NUM_TX_QUEUES]);
@@ -3317,13 +3318,18 @@ struct net_device *rtnl_create_link(struct net *net, const char *ifname,
 	if (!dev)
 		return ERR_PTR(-ENOMEM);
 
+	err = validate_linkmsg(dev, tb, extack);
+	if (err < 0) {
+		free_netdev(dev);
+		return ERR_PTR(err);
+	}
+
 	dev_net_set(dev, net);
 	dev->rtnl_link_ops = ops;
 	dev->rtnl_link_state = RTNL_LINK_INITIALIZING;
 
 	if (tb[IFLA_MTU]) {
 		u32 mtu = nla_get_u32(tb[IFLA_MTU]);
-		int err;
 
 		err = dev_validate_mtu(dev, mtu, extack);
 		if (err) {
Signature · 1 line
2.39.2
author maintainer
#76 lore raw

PATCH 76/286 rtnetlink: move IFLA_GSO_ tb check to validate_linkmsg

In reply to Greg Kroah-Hartman

From: Xin Long <[email protected]>

[ Upstream commit fef5b228dd38378148bc850f7e69a7783f3b95a4 ]

These IFLA_GSO_* tb check should also be done for the new created link, otherwise, they can be set to a huge value when creating links:

  # ip link add dummy1 gso_max_size 4294967295 type dummy
  # ip -d link show dummy1
    dummy addrgenmode eui64 ... gso_max_size 4294967295
Fixes 46e6b992c250 ("rtnetlink: allow GSO maximums to be set on device creation")
Fixes 9eefedd58ae1 ("net: add gso_ipv4_max_size and gro_ipv4_max_size per device")
---
 net/core/rtnetlink.c | 34 +++++++++++++++++++---------------
 1 file changed, 19 insertions(+), 15 deletions(-)
net/core/rtnetlink.c +19 -15
index f37deb18dd02e..7094569009b14 100644
@@ -2382,6 +2382,25 @@ static int validate_linkmsg(struct net_device *dev, struct nlattr *tb[],
 		if (tb[IFLA_BROADCAST] &&
 		    nla_len(tb[IFLA_BROADCAST]) < dev->addr_len)
 			return -EINVAL;
+
+		if (tb[IFLA_GSO_MAX_SIZE] &&
+		    nla_get_u32(tb[IFLA_GSO_MAX_SIZE]) > dev->tso_max_size) {
+			NL_SET_ERR_MSG(extack, "too big gso_max_size");
+			return -EINVAL;
+		}
+
+		if (tb[IFLA_GSO_MAX_SEGS] &&
+		    (nla_get_u32(tb[IFLA_GSO_MAX_SEGS]) > GSO_MAX_SEGS ||
+		     nla_get_u32(tb[IFLA_GSO_MAX_SEGS]) > dev->tso_max_segs)) {
+			NL_SET_ERR_MSG(extack, "too big gso_max_segs");
+			return -EINVAL;
+		}
+
+		if (tb[IFLA_GSO_IPV4_MAX_SIZE] &&
+		    nla_get_u32(tb[IFLA_GSO_IPV4_MAX_SIZE]) > dev->tso_max_size) {
+			NL_SET_ERR_MSG(extack, "too big gso_ipv4_max_size");
+			return -EINVAL;
+		}
 	}
 
 	if (tb[IFLA_AF_SPEC]) {
@@ -2855,11 +2874,6 @@ static int do_setlink(const struct sk_buff *skb,
 	if (tb[IFLA_GSO_MAX_SIZE]) {
 		u32 max_size = nla_get_u32(tb[IFLA_GSO_MAX_SIZE]);
 
-		if (max_size > dev->tso_max_size) {
-			err = -EINVAL;
-			goto errout;
-		}
-
 		if (dev->gso_max_size ^ max_size) {
 			netif_set_gso_max_size(dev, max_size);
 			status |= DO_SETLINK_MODIFIED;
@@ -2869,11 +2883,6 @@ static int do_setlink(const struct sk_buff *skb,
 	if (tb[IFLA_GSO_MAX_SEGS]) {
 		u32 max_segs = nla_get_u32(tb[IFLA_GSO_MAX_SEGS]);
 
-		if (max_segs > GSO_MAX_SEGS || max_segs > dev->tso_max_segs) {
-			err = -EINVAL;
-			goto errout;
-		}
-
 		if (dev->gso_max_segs ^ max_segs) {
 			netif_set_gso_max_segs(dev, max_segs);
 			status |= DO_SETLINK_MODIFIED;
@@ -2892,11 +2901,6 @@ static int do_setlink(const struct sk_buff *skb,
 	if (tb[IFLA_GSO_IPV4_MAX_SIZE]) {
 		u32 max_size = nla_get_u32(tb[IFLA_GSO_IPV4_MAX_SIZE]);
 
-		if (max_size > dev->tso_max_size) {
-			err = -EINVAL;
-			goto errout;
-		}
-
 		if (dev->gso_ipv4_max_size ^ max_size) {
 			netif_set_gso_ipv4_max_size(dev, max_size);
 			status |= DO_SETLINK_MODIFIED;
Signature · 1 line
2.39.2
author maintainer
#77 lore raw

PATCH 77/286 rtnetlink: add the missing IFLA_GRO_ tb check in validate_linkmsg

In reply to Greg Kroah-Hartman

From: Xin Long <[email protected]>

[ Upstream commit 65d6914e253f3d83b724a9bbfc889ae95711e512 ]

This fixes the issue that dev gro_max_size and gso_ipv4_max_size can be set to a huge value:

  # ip link add dummy1 type dummy
  # ip link set dummy1 gro_max_size 4294967295
  # ip -d link show dummy1
    dummy addrgenmode eui64 ... gro_max_size 4294967295
Fixes 0fe79f28bfaf ("net: allow gro_max_size to exceed 65536")
Fixes 9eefedd58ae1 ("net: add gso_ipv4_max_size and gro_ipv4_max_size per device")
---
 net/core/rtnetlink.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
net/core/rtnetlink.c +12 -0
index 7094569009b14..f235cc6832767 100644
@@ -2396,11 +2396,23 @@ static int validate_linkmsg(struct net_device *dev, struct nlattr *tb[],
 			return -EINVAL;
 		}
 
+		if (tb[IFLA_GRO_MAX_SIZE] &&
+		    nla_get_u32(tb[IFLA_GRO_MAX_SIZE]) > GRO_MAX_SIZE) {
+			NL_SET_ERR_MSG(extack, "too big gro_max_size");
+			return -EINVAL;
+		}
+
 		if (tb[IFLA_GSO_IPV4_MAX_SIZE] &&
 		    nla_get_u32(tb[IFLA_GSO_IPV4_MAX_SIZE]) > dev->tso_max_size) {
 			NL_SET_ERR_MSG(extack, "too big gso_ipv4_max_size");
 			return -EINVAL;
 		}
+
+		if (tb[IFLA_GRO_IPV4_MAX_SIZE] &&
+		    nla_get_u32(tb[IFLA_GRO_IPV4_MAX_SIZE]) > GRO_MAX_SIZE) {
+			NL_SET_ERR_MSG(extack, "too big gro_ipv4_max_size");
+			return -EINVAL;
+		}
 	}
 
 	if (tb[IFLA_AF_SPEC]) {
Signature · 1 line
2.39.2
author maintainer
#78 lore raw

PATCH 78/286 mptcp: avoid unneeded __mptcp_nmpc_socket() usage

In reply to Greg Kroah-Hartman

From: Paolo Abeni <[email protected]>

[ Upstream commit 617612316953093bc859890e405e1b550c27d840 ]

In a few spots, the mptcp code invokes the __mptcp_nmpc_socket() helper multiple times under the same socket lock scope. Additionally, in such places, the socket status ensures that there is no MP capable handshake running.

Under the above condition we can replace the later __mptcp_nmpc_socket() helper invocation with direct access to the msk->subflow pointer and better document such access is not supposed to fail with WARN().

Stable-dep-of: 5b825727d087 ("mptcp: add annotations around msk->subflow accesses")

---
 net/mptcp/protocol.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
net/mptcp/protocol.c +6 -3
index b998e9df53cef..676a6d24b4b71 100644
@@ -3157,7 +3157,7 @@ static struct sock *mptcp_accept(struct sock *sk, int flags, int *err,
 	struct socket *listener;
 	struct sock *newsk;
 
-	listener = __mptcp_nmpc_socket(msk);
+	listener = msk->subflow;
 	if (WARN_ON_ONCE(!listener)) {
 		*err = -EINVAL;
 		return NULL;
@@ -3377,7 +3377,7 @@ static int mptcp_get_port(struct sock *sk, unsigned short snum)
 	struct mptcp_sock *msk = mptcp_sk(sk);
 	struct socket *ssock;
 
-	ssock = __mptcp_nmpc_socket(msk);
+	ssock = msk->subflow;
 	pr_debug("msk=%p, subflow=%p", msk, ssock);
 	if (WARN_ON_ONCE(!ssock))
 		return -EINVAL;
@@ -3723,7 +3723,10 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 
 	pr_debug("msk=%p", msk);
 
-	ssock = __mptcp_nmpc_socket(msk);
+	/* buggy applications can call accept on socket states other then LISTEN
+	 * but no need to allocate the first subflow just to error out.
+	 */
+	ssock = msk->subflow;
 	if (!ssock)
 		return -EINVAL;
 
Signature · 1 line
2.39.2
author maintainer
#79 lore raw

PATCH 79/286 mptcp: add annotations around msk->subflow accesses

In reply to Greg Kroah-Hartman

From: Paolo Abeni <[email protected]>

[ Upstream commit 5b825727d0871b23e8867f6371183e61628b4a26 ]

The MPTCP can access the first subflow socket in a few spots outside the socket lock scope. That is actually safe, as MPTCP will delete the socket itself only after the msk sock close().

Still the such accesses causes a few KCSAN splats, as reported by Christoph. Silence the harmless warning adding a few annotation around the relevant accesses.

---
 net/mptcp/protocol.c | 18 ++++++++++--------
 net/mptcp/protocol.h |  6 +++++-
 2 files changed, 15 insertions(+), 9 deletions(-)
net/mptcp/protocol.c +10 -8
index 676a6d24b4b71..f5c0a56f0f0ca 100644
@@ -103,7 +103,7 @@ static int __mptcp_socket_create(struct mptcp_sock *msk)
 		return err;
 
 	msk->first = ssock->sk;
-	msk->subflow = ssock;
+	WRITE_ONCE(msk->subflow, ssock);
 	subflow = mptcp_subflow_ctx(ssock->sk);
 	list_add(&subflow->node, &msk->conn_list);
 	sock_hold(ssock->sk);
@@ -2254,7 +2254,7 @@ static void mptcp_dispose_initial_subflow(struct mptcp_sock *msk)
 {
 	if (msk->subflow) {
 		iput(SOCK_INODE(msk->subflow));
-		msk->subflow = NULL;
+		WRITE_ONCE(msk->subflow, NULL);
 	}
 }
 
@@ -3109,7 +3109,7 @@ struct sock *mptcp_sk_clone(const struct sock *sk,
 	msk = mptcp_sk(nsk);
 	msk->local_key = subflow_req->local_key;
 	msk->token = subflow_req->token;
-	msk->subflow = NULL;
+	WRITE_ONCE(msk->subflow, NULL);
 	msk->in_accept_queue = 1;
 	WRITE_ONCE(msk->fully_established, false);
 	if (mp_opt->suboptions & OPTION_MPTCP_CSUMREQD)
@@ -3157,7 +3157,7 @@ static struct sock *mptcp_accept(struct sock *sk, int flags, int *err,
 	struct socket *listener;
 	struct sock *newsk;
 
-	listener = msk->subflow;
+	listener = READ_ONCE(msk->subflow);
 	if (WARN_ON_ONCE(!listener)) {
 		*err = -EINVAL;
 		return NULL;
@@ -3723,10 +3723,10 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 
 	pr_debug("msk=%p", msk);
 
-	/* buggy applications can call accept on socket states other then LISTEN
+	/* Buggy applications can call accept on socket states other then LISTEN
 	 * but no need to allocate the first subflow just to error out.
 	 */
-	ssock = msk->subflow;
+	ssock = READ_ONCE(msk->subflow);
 	if (!ssock)
 		return -EINVAL;
 
@@ -3800,10 +3800,12 @@ static __poll_t mptcp_poll(struct file *file, struct socket *sock,
 	state = inet_sk_state_load(sk);
 	pr_debug("msk=%p state=%d flags=%lx", msk, state, msk->flags);
 	if (state == TCP_LISTEN) {
-		if (WARN_ON_ONCE(!msk->subflow || !msk->subflow->sk))
+		struct socket *ssock = READ_ONCE(msk->subflow);
+
+		if (WARN_ON_ONCE(!ssock || !ssock->sk))
 			return 0;
 
-		return inet_csk_listen_poll(msk->subflow->sk);
+		return inet_csk_listen_poll(ssock->sk);
 	}
 
 	if (state != TCP_SYN_SENT && state != TCP_SYN_RECV) {
net/mptcp/protocol.h +5 -1
index d6469b6ab38e3..84427b3697d89 100644
@@ -306,7 +306,11 @@ struct mptcp_sock {
 	struct list_head rtx_queue;
 	struct mptcp_data_frag *first_pending;
 	struct list_head join_list;
-	struct socket	*subflow; /* outgoing connect/listener/!mp_capable */
+	struct socket	*subflow; /* outgoing connect/listener/!mp_capable
+				   * The mptcp ops can safely dereference, using suitable
+				   * ONCE annotation, the subflow outside the socket
+				   * lock as such sock is freed after close().
+				   */
 	struct sock	*first;
 	struct mptcp_pm_data	pm;
 	struct {
Signature · 1 line
2.39.2
author maintainer
#80 lore raw

PATCH 80/286 mptcp: avoid unneeded address copy

In reply to Greg Kroah-Hartman

From: Paolo Abeni <[email protected]>

[ Upstream commit 2bb9a37f0e194ed95c70603b0efc7898a5a0d9b4 ]

In the syn_recv fallback path, the msk is unused. We can skip setting the socket address.

Stable-dep-of: 7e8b88ec35ee ("mptcp: consolidate passive msk socket initialization")

---
 net/mptcp/subflow.c | 2 --
 1 file changed, 2 deletions(-)
net/mptcp/subflow.c +0 -2
index 281c1cc8dc8dc..8a6bd2782dc97 100644
@@ -823,8 +823,6 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 				goto dispose_child;
 			}
 
-			if (new_msk)
-				mptcp_copy_inaddrs(new_msk, child);
 			mptcp_subflow_drop_ctx(child);
 			goto out;
 		}
Signature · 1 line
2.39.2
author maintainer
#81 lore raw

PATCH 81/286 mptcp: simplify subflow_syn_recv_sock()

In reply to Greg Kroah-Hartman

From: Paolo Abeni <[email protected]>

[ Upstream commit a88d0092b24b8cddce57fe0e88e60a9e29e0b515 ]

Postpone the msk cloning to the child process creation so that we can avoid a bunch of conditionals.

Stable-dep-of: 7e8b88ec35ee ("mptcp: consolidate passive msk socket initialization")

---
 net/mptcp/subflow.c | 41 +++++++++++++----------------------------
 1 file changed, 13 insertions(+), 28 deletions(-)
net/mptcp/subflow.c +13 -28
index 8a6bd2782dc97..3b8abd79bfbd5 100644
@@ -695,14 +695,6 @@ static bool subflow_hmac_valid(const struct request_sock *req,
 	return !crypto_memneq(hmac, mp_opt->hmac, MPTCPOPT_HMAC_LEN);
 }
 
-static void mptcp_force_close(struct sock *sk)
-{
-	/* the msk is not yet exposed to user-space, and refcount is 2 */
-	inet_sk_state_store(sk, TCP_CLOSE);
-	sk_common_release(sk);
-	sock_put(sk);
-}
-
 static void subflow_ulp_fallback(struct sock *sk,
 				 struct mptcp_subflow_context *old_ctx)
 {
@@ -757,7 +749,6 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 	struct mptcp_subflow_request_sock *subflow_req;
 	struct mptcp_options_received mp_opt;
 	bool fallback, fallback_is_fatal;
-	struct sock *new_msk = NULL;
 	struct mptcp_sock *owner;
 	struct sock *child;
 
@@ -786,14 +777,9 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 		 * options.
 		 */
 		mptcp_get_options(skb, &mp_opt);
-		if (!(mp_opt.suboptions & OPTIONS_MPTCP_MPC)) {
+		if (!(mp_opt.suboptions & OPTIONS_MPTCP_MPC))
 			fallback = true;
-			goto create_child;
-		}
 
-		new_msk = mptcp_sk_clone(listener->conn, &mp_opt, req);
-		if (!new_msk)
-			fallback = true;
 	} else if (subflow_req->mp_join) {
 		mptcp_get_options(skb, &mp_opt);
 		if (!(mp_opt.suboptions & OPTIONS_MPTCP_MPJ) ||
@@ -822,21 +808,23 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 				subflow_add_reset_reason(skb, MPTCP_RST_EMPTCP);
 				goto dispose_child;
 			}
-
-			mptcp_subflow_drop_ctx(child);
-			goto out;
+			goto fallback;
 		}
 
 		/* ssk inherits options of listener sk */
 		ctx->setsockopt_seq = listener->setsockopt_seq;
 
 		if (ctx->mp_capable) {
-			owner = mptcp_sk(new_msk);
+			ctx->conn = mptcp_sk_clone(listener->conn, &mp_opt, req);
+			if (!ctx->conn)
+				goto fallback;
+
+			owner = mptcp_sk(ctx->conn);
 
 			/* this can't race with mptcp_close(), as the msk is
 			 * not yet exposted to user-space
 			 */
-			inet_sk_state_store((void *)new_msk, TCP_ESTABLISHED);
+			inet_sk_state_store(ctx->conn, TCP_ESTABLISHED);
 
 			/* record the newly created socket as the first msk
 			 * subflow, but don't link it yet into conn_list
@@ -846,11 +834,9 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 			/* new mpc subflow takes ownership of the newly
 			 * created mptcp socket
 			 */
-			mptcp_sk(new_msk)->setsockopt_seq = ctx->setsockopt_seq;
+			owner->setsockopt_seq = ctx->setsockopt_seq;
 			mptcp_pm_new_connection(owner, child, 1);
 			mptcp_token_accept(subflow_req, owner);
-			ctx->conn = new_msk;
-			new_msk = NULL;
 
 			/* set msk addresses early to ensure mptcp_pm_get_local_id()
 			 * uses the correct data
@@ -900,11 +886,6 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 		}
 	}
 
-out:
-	/* dispose of the left over mptcp master, if any */
-	if (unlikely(new_msk))
-		mptcp_force_close(new_msk);
-
 	/* check for expected invariant - should never trigger, just help
 	 * catching eariler subtle bugs
 	 */
@@ -922,6 +903,10 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 
 	/* The last child reference will be released by the caller */
 	return child;
+
+fallback:
+	mptcp_subflow_drop_ctx(child);
+	return child;
 }
 
 static struct inet_connection_sock_af_ops subflow_specific __ro_after_init;
Signature · 1 line
2.39.2
author maintainer
#82 lore raw

PATCH 82/286 mptcp: consolidate passive msk socket initialization

In reply to Greg Kroah-Hartman

From: Paolo Abeni <[email protected]>

[ Upstream commit 7e8b88ec35eef363040e08d99536d2bebef83774 ]

When the msk socket is cloned at MPC handshake time, a few fields are initialized in a racy way outside mptcp_sk_clone() and the msk socket lock.

The above is due historical reasons: before commit a88d0092b24b ("mptcp: simplify subflow_syn_recv_sock()") as the first subflow socket carrying all the needed date was not available yet at msk creation time

We can now refactor the code moving the missing initialization bit under the socket lock, removing the init race and avoiding some code duplication.

This will also simplify the next patch, as all msk->first write access are now under the msk socket lock.

Fixes 0397c6d85f9c ("mptcp: keep unaccepted MPC subflow into join list")
---
 net/mptcp/protocol.c | 35 ++++++++++++++++++++++++++++-------
 net/mptcp/protocol.h |  8 ++++----
 net/mptcp/subflow.c  | 28 +---------------------------
 3 files changed, 33 insertions(+), 38 deletions(-)
net/mptcp/protocol.c +28 -7
index f5c0a56f0f0ca..8d0c03091c409 100644
@@ -3011,7 +3011,7 @@ static void mptcp_close(struct sock *sk, long timeout)
 	sock_put(sk);
 }
 
-void mptcp_copy_inaddrs(struct sock *msk, const struct sock *ssk)
+static void mptcp_copy_inaddrs(struct sock *msk, const struct sock *ssk)
 {
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 	const struct ipv6_pinfo *ssk6 = inet6_sk(ssk);
@@ -3088,9 +3088,10 @@ static struct ipv6_pinfo *mptcp_inet6_sk(const struct sock *sk)
 }
 #endif
 
-struct sock *mptcp_sk_clone(const struct sock *sk,
-			    const struct mptcp_options_received *mp_opt,
-			    struct request_sock *req)
+struct sock *mptcp_sk_clone_init(const struct sock *sk,
+				 const struct mptcp_options_received *mp_opt,
+				 struct sock *ssk,
+				 struct request_sock *req)
 {
 	struct mptcp_subflow_request_sock *subflow_req = mptcp_subflow_rsk(req);
 	struct sock *nsk = sk_clone_lock(sk, GFP_ATOMIC);
@@ -3122,10 +3123,30 @@ struct sock *mptcp_sk_clone(const struct sock *sk,
 	msk->setsockopt_seq = mptcp_sk(sk)->setsockopt_seq;
 
 	sock_reset_flag(nsk, SOCK_RCU_FREE);
-	/* will be fully established after successful MPC subflow creation */
-	inet_sk_state_store(nsk, TCP_SYN_RECV);
-
 	security_inet_csk_clone(nsk, req);
+
+	/* this can't race with mptcp_close(), as the msk is
+	 * not yet exposted to user-space
+	 */
+	inet_sk_state_store(nsk, TCP_ESTABLISHED);
+
+	/* The msk maintain a ref to each subflow in the connections list */
+	WRITE_ONCE(msk->first, ssk);
+	list_add(&mptcp_subflow_ctx(ssk)->node, &msk->conn_list);
+	sock_hold(ssk);
+
+	/* new mpc subflow takes ownership of the newly
+	 * created mptcp socket
+	 */
+	mptcp_token_accept(subflow_req, msk);
+
+	/* set msk addresses early to ensure mptcp_pm_get_local_id()
+	 * uses the correct data
+	 */
+	mptcp_copy_inaddrs(nsk, ssk);
+	mptcp_propagate_sndbuf(nsk, ssk);
+
+	mptcp_rcv_space_init(msk, ssk);
 	bh_unlock_sock(nsk);
 
 	/* note: the newly allocated socket refcount is 2 now */
net/mptcp/protocol.h +4 -4
index 84427b3697d89..426b43d7ae642 100644
@@ -620,7 +620,6 @@ int mptcp_is_checksum_enabled(const struct net *net);
 int mptcp_allow_join_id0(const struct net *net);
 unsigned int mptcp_stale_loss_cnt(const struct net *net);
 int mptcp_get_pm_type(const struct net *net);
-void mptcp_copy_inaddrs(struct sock *msk, const struct sock *ssk);
 void mptcp_subflow_fully_established(struct mptcp_subflow_context *subflow,
 				     const struct mptcp_options_received *mp_opt);
 bool __mptcp_retransmit_pending_data(struct sock *sk);
@@ -690,9 +689,10 @@ void __init mptcp_proto_init(void);
 int __init mptcp_proto_v6_init(void);
 #endif
 
-struct sock *mptcp_sk_clone(const struct sock *sk,
-			    const struct mptcp_options_received *mp_opt,
-			    struct request_sock *req);
+struct sock *mptcp_sk_clone_init(const struct sock *sk,
+				 const struct mptcp_options_received *mp_opt,
+				 struct sock *ssk,
+				 struct request_sock *req);
 void mptcp_get_options(const struct sk_buff *skb,
 		       struct mptcp_options_received *mp_opt);
 
net/mptcp/subflow.c +1 -27
index 3b8abd79bfbd5..bb0301398d3b4 100644
@@ -815,38 +815,12 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 		ctx->setsockopt_seq = listener->setsockopt_seq;
 
 		if (ctx->mp_capable) {
-			ctx->conn = mptcp_sk_clone(listener->conn, &mp_opt, req);
+			ctx->conn = mptcp_sk_clone_init(listener->conn, &mp_opt, child, req);
 			if (!ctx->conn)
 				goto fallback;
 
 			owner = mptcp_sk(ctx->conn);
-
-			/* this can't race with mptcp_close(), as the msk is
-			 * not yet exposted to user-space
-			 */
-			inet_sk_state_store(ctx->conn, TCP_ESTABLISHED);
-
-			/* record the newly created socket as the first msk
-			 * subflow, but don't link it yet into conn_list
-			 */
-			WRITE_ONCE(owner->first, child);
-
-			/* new mpc subflow takes ownership of the newly
-			 * created mptcp socket
-			 */
-			owner->setsockopt_seq = ctx->setsockopt_seq;
 			mptcp_pm_new_connection(owner, child, 1);
-			mptcp_token_accept(subflow_req, owner);
-
-			/* set msk addresses early to ensure mptcp_pm_get_local_id()
-			 * uses the correct data
-			 */
-			mptcp_copy_inaddrs(ctx->conn, child);
-			mptcp_propagate_sndbuf(ctx->conn, child);
-
-			mptcp_rcv_space_init(owner, child);
-			list_add(&ctx->node, &owner->conn_list);
-			sock_hold(child);
 
 			/* with OoO packets we can reach here without ingress
 			 * mpc option
Signature · 1 line
2.39.2
author maintainer
#83 lore raw

PATCH 83/286 mptcp: fix data race around msk->first access

In reply to Greg Kroah-Hartman

From: Paolo Abeni <[email protected]>

[ Upstream commit 1b1b43ee7a208096ecd79e626f2fc90d4a321111 ]

The first subflow socket is accessed outside the msk socket lock by mptcp_subflow_fail(), we need to annotate each write access with WRITE_ONCE, but a few spots still lacks it.

Fixes 76a13b315709 ("mptcp: invoke MP_FAIL response when needed")
---
 net/mptcp/protocol.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
net/mptcp/protocol.c +3 -3
index 8d0c03091c409..64d4910e337ea 100644
@@ -102,7 +102,7 @@ static int __mptcp_socket_create(struct mptcp_sock *msk)
 	if (err)
 		return err;
 
-	msk->first = ssock->sk;
+	WRITE_ONCE(msk->first, ssock->sk);
 	WRITE_ONCE(msk->subflow, ssock);
 	subflow = mptcp_subflow_ctx(ssock->sk);
 	list_add(&subflow->node, &msk->conn_list);
@@ -2391,7 +2391,7 @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 	sock_put(ssk);
 
 	if (ssk == msk->first)
-		msk->first = NULL;
+		WRITE_ONCE(msk->first, NULL);
 
 out:
 	if (ssk == msk->last_snd)
@@ -2692,7 +2692,7 @@ static int __mptcp_init_sock(struct sock *sk)
 	WRITE_ONCE(msk->rmem_released, 0);
 	msk->timer_ival = TCP_RTO_MIN;
 
-	msk->first = NULL;
+	WRITE_ONCE(msk->first, NULL);
 	inet_csk(sk)->icsk_sync_mss = mptcp_sync_mss;
 	WRITE_ONCE(msk->csum_enabled, mptcp_is_checksum_enabled(sock_net(sk)));
 	WRITE_ONCE(msk->allow_infinite_fallback, true);
Signature · 1 line
2.39.2
author maintainer
#84 lore raw

PATCH 84/286 mptcp: add annotations around sk->sk_shutdown accesses

In reply to Greg Kroah-Hartman

From: Paolo Abeni <[email protected]>

[ Upstream commit 6b9831bfd9322b297eb6d44257808cc055fdc586 ]

Christoph reported the mptcp variant of a recently addressed plain TCP issue. Similar to commit e14cadfd80d7 ("tcp: add annotations around sk->sk_shutdown accesses") add READ/WRITE ONCE annotations to silence KCSAN reports around lockless sk_shutdown access.

---
 net/mptcp/protocol.c | 29 ++++++++++++++++-------------
 1 file changed, 16 insertions(+), 13 deletions(-)
net/mptcp/protocol.c +16 -13
index 64d4910e337ea..17793aa733fd9 100644
@@ -590,7 +590,7 @@ static bool mptcp_check_data_fin(struct sock *sk)
 		WRITE_ONCE(msk->ack_seq, msk->ack_seq + 1);
 		WRITE_ONCE(msk->rcv_data_fin, 0);
 
-		sk->sk_shutdown |= RCV_SHUTDOWN;
+		WRITE_ONCE(sk->sk_shutdown, sk->sk_shutdown | RCV_SHUTDOWN);
 		smp_mb__before_atomic(); /* SHUTDOWN must be visible first */
 
 		switch (sk->sk_state) {
@@ -897,7 +897,7 @@ static void mptcp_check_for_eof(struct mptcp_sock *msk)
 		/* hopefully temporary hack: propagate shutdown status
 		 * to msk, when all subflows agree on it
 		 */
-		sk->sk_shutdown |= RCV_SHUTDOWN;
+		WRITE_ONCE(sk->sk_shutdown, sk->sk_shutdown | RCV_SHUTDOWN);
 
 		smp_mb__before_atomic(); /* SHUTDOWN must be visible first */
 		sk->sk_data_ready(sk);
@@ -2498,7 +2498,7 @@ static void mptcp_check_fastclose(struct mptcp_sock *msk)
 	}
 
 	inet_sk_state_store(sk, TCP_CLOSE);
-	sk->sk_shutdown = SHUTDOWN_MASK;
+	WRITE_ONCE(sk->sk_shutdown, SHUTDOWN_MASK);
 	smp_mb__before_atomic(); /* SHUTDOWN must be visible first */
 	set_bit(MPTCP_WORK_CLOSE_SUBFLOW, &msk->flags);
 
@@ -2934,7 +2934,7 @@ bool __mptcp_close(struct sock *sk, long timeout)
 	bool do_cancel_work = false;
 	int subflows_alive = 0;
 
-	sk->sk_shutdown = SHUTDOWN_MASK;
+	WRITE_ONCE(sk->sk_shutdown, SHUTDOWN_MASK);
 
 	if ((1 << sk->sk_state) & (TCPF_LISTEN | TCPF_CLOSE)) {
 		mptcp_listen_inuse_dec(sk);
@@ -3074,7 +3074,7 @@ static int mptcp_disconnect(struct sock *sk, int flags)
 	mptcp_pm_data_reset(msk);
 	mptcp_ca_reset(sk);
 
-	sk->sk_shutdown = 0;
+	WRITE_ONCE(sk->sk_shutdown, 0);
 	sk_error_report(sk);
 	return 0;
 }
@@ -3793,9 +3793,6 @@ static __poll_t mptcp_check_writeable(struct mptcp_sock *msk)
 {
 	struct sock *sk = (struct sock *)msk;
 
-	if (unlikely(sk->sk_shutdown & SEND_SHUTDOWN))
-		return EPOLLOUT | EPOLLWRNORM;
-
 	if (sk_stream_is_writeable(sk))
 		return EPOLLOUT | EPOLLWRNORM;
 
@@ -3813,6 +3810,7 @@ static __poll_t mptcp_poll(struct file *file, struct socket *sock,
 	struct sock *sk = sock->sk;
 	struct mptcp_sock *msk;
 	__poll_t mask = 0;
+	u8 shutdown;
 	int state;
 
 	msk = mptcp_sk(sk);
@@ -3829,17 +3827,22 @@ static __poll_t mptcp_poll(struct file *file, struct socket *sock,
 		return inet_csk_listen_poll(ssock->sk);
 	}
 
+	shutdown = READ_ONCE(sk->sk_shutdown);
+	if (shutdown == SHUTDOWN_MASK || state == TCP_CLOSE)
+		mask |= EPOLLHUP;
+	if (shutdown & RCV_SHUTDOWN)
+		mask |= EPOLLIN | EPOLLRDNORM | EPOLLRDHUP;
+
 	if (state != TCP_SYN_SENT && state != TCP_SYN_RECV) {
 		mask |= mptcp_check_readable(msk);
-		mask |= mptcp_check_writeable(msk);
+		if (shutdown & SEND_SHUTDOWN)
+			mask |= EPOLLOUT | EPOLLWRNORM;
+		else
+			mask |= mptcp_check_writeable(msk);
 	} else if (state == TCP_SYN_SENT && inet_sk(sk)->defer_connect) {
 		/* cf tcp_poll() note about TFO */
 		mask |= EPOLLOUT | EPOLLWRNORM;
 	}
-	if (sk->sk_shutdown == SHUTDOWN_MASK || state == TCP_CLOSE)
-		mask |= EPOLLHUP;
-	if (sk->sk_shutdown & RCV_SHUTDOWN)
-		mask |= EPOLLIN | EPOLLRDNORM | EPOLLRDHUP;
 
 	/* This barrier is coupled with smp_wmb() in __mptcp_error_report() */
 	smp_rmb();
Signature · 1 line
2.39.2
author maintainer
#67 lore raw

PATCH 67/286 tcp: fix mishandling when the sack compression is deferred.

In reply to Greg Kroah-Hartman

From: fuyuanli <[email protected]>

[ Upstream commit 30c6f0bf9579debce27e45fac34fdc97e46acacc ]

In this patch, we mainly try to handle sending a compressed ack correctly if it's deferred.

Here are more details in the old logic:
When sack compression is triggered in the tcp_compressed_ack_kick(), if the sock is owned by user, it will set TCP_DELACK_TIMER_DEFERRED and then defer to the release cb phrase. Later once user releases the sock, tcp_delack_timer_handler() should send a ack as expected, which, however, cannot happen due to lack of ICSK_ACK_TIMER flag. Therefore, the receiver would not sent an ack until the sender's retransmission timeout. It definitely increases unnecessary latency.

---
 include/net/tcp.h    |  1 +
 net/ipv4/tcp_input.c |  2 +-
 net/ipv4/tcp_timer.c | 16 +++++++++++++---
 3 files changed, 15 insertions(+), 4 deletions(-)
include/net/tcp.h +1 -0
index 76bf0a11bdc77..99c74fc300839 100644
@@ -632,6 +632,7 @@ void tcp_reset(struct sock *sk, struct sk_buff *skb);
 void tcp_skb_mark_lost_uncond_verify(struct tcp_sock *tp, struct sk_buff *skb);
 void tcp_fin(struct sock *sk);
 void tcp_check_space(struct sock *sk);
+void tcp_sack_compress_send_ack(struct sock *sk);
 
 /* tcp_timer.c */
 void tcp_init_xmit_timers(struct sock *);
net/ipv4/tcp_input.c +1 -1
index 10776c54ff784..dee174c40e874 100644
@@ -4530,7 +4530,7 @@ static void tcp_sack_maybe_coalesce(struct tcp_sock *tp)
 	}
 }
 
-static void tcp_sack_compress_send_ack(struct sock *sk)
+void tcp_sack_compress_send_ack(struct sock *sk)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 
net/ipv4/tcp_timer.c +13 -3
index cb79127f45c34..0b5d0a2867a8c 100644
@@ -290,9 +290,19 @@ static int tcp_write_timeout(struct sock *sk)
 void tcp_delack_timer_handler(struct sock *sk)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
+	struct tcp_sock *tp = tcp_sk(sk);
 
-	if (((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN)) ||
-	    !(icsk->icsk_ack.pending & ICSK_ACK_TIMER))
+	if ((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN))
+		return;
+
+	/* Handling the sack compression case */
+	if (tp->compressed_ack) {
+		tcp_mstamp_refresh(tp);
+		tcp_sack_compress_send_ack(sk);
+		return;
+	}
+
+	if (!(icsk->icsk_ack.pending & ICSK_ACK_TIMER))
 		return;
 
 	if (time_after(icsk->icsk_ack.timeout, jiffies)) {
@@ -312,7 +322,7 @@ void tcp_delack_timer_handler(struct sock *sk)
 			inet_csk_exit_pingpong_mode(sk);
 			icsk->icsk_ack.ato      = TCP_ATO_MIN;
 		}
-		tcp_mstamp_refresh(tcp_sk(sk));
+		tcp_mstamp_refresh(tp);
 		tcp_send_ack(sk);
 		__NET_INC_STATS(sock_net(sk), LINUX_MIB_DELAYEDACKS);
 	}
Signature · 1 line
2.39.2
author maintainer
#85 lore raw

PATCH 85/286 drm/amdgpu: release gpu full access after "amdgpu_device_ip_late_init"

In reply to Greg Kroah-Hartman

From: Chong Li <[email protected]>

[ Upstream commit 38eecbe086a4e52f54b2bbda8feba65d44addbef ]

[WHY]
Function "amdgpu_irq_update()" called by "amdgpu_device_ip_late_init()" is an atomic context. We shouldn't access registers through KIQ since "msleep()" may be called in "amdgpu_kiq_rreg()".

[HOW]
Move function "amdgpu_virt_release_full_gpu()" after function "amdgpu_device_ip_late_init()", to ensure that registers be accessed through RLCG instead of KIQ.

Call Trace:
  <TASK>
  show_stack+0x52/0x69
  dump_stack_lvl+0x49/0x6d
  dump_stack+0x10/0x18
  __schedule_bug.cold+0x4f/0x6b
  __schedule+0x473/0x5d0
  ? __wake_up_klogd.part.0+0x40/0x70
  ? vprintk_emit+0xbe/0x1f0
  schedule+0x68/0x110
  schedule_timeout+0x87/0x160
  ? timer_migration_handler+0xa0/0xa0
  msleep+0x2d/0x50
  amdgpu_kiq_rreg+0x18d/0x1f0 [amdgpu]
  amdgpu_device_rreg.part.0+0x59/0xd0 [amdgpu]
  amdgpu_device_rreg+0x3a/0x50 [amdgpu]
  amdgpu_sriov_rreg+0x3c/0xb0 [amdgpu]
  gfx_v10_0_set_gfx_eop_interrupt_state.constprop.0+0x16c/0x190 [amdgpu]
  gfx_v10_0_set_eop_interrupt_state+0xa5/0xb0 [amdgpu]
  amdgpu_irq_update+0x53/0x80 [amdgpu]
  amdgpu_irq_get+0x7c/0xb0 [amdgpu]
  amdgpu_fence_driver_hw_init+0x58/0x90 [amdgpu]
  amdgpu_device_init.cold+0x16b7/0x2022 [amdgpu]
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 32 ++++++++++++----------
 1 file changed, 17 insertions(+), 15 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +17 -15
index aa46726dfdb01..464f0abc855e6 100644
@@ -2523,8 +2523,6 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
 	amdgpu_fru_get_product_info(adev);
 
 init_failed:
-	if (amdgpu_sriov_vf(adev))
-		amdgpu_virt_release_full_gpu(adev, true);
 
 	return r;
 }
@@ -3841,18 +3839,6 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 
 	r = amdgpu_device_ip_init(adev);
 	if (r) {
-		/* failed in exclusive mode due to timeout */
-		if (amdgpu_sriov_vf(adev) &&
-		    !amdgpu_sriov_runtime(adev) &&
-		    amdgpu_virt_mmio_blocked(adev) &&
-		    !amdgpu_virt_wait_reset(adev)) {
-			dev_err(adev->dev, "VF exclusive mode timeout\n");
-			/* Don't send request since VF is inactive. */
-			adev->virt.caps &= ~AMDGPU_SRIOV_CAPS_RUNTIME;
-			adev->virt.ops = NULL;
-			r = -EAGAIN;
-			goto release_ras_con;
-		}
 		dev_err(adev->dev, "amdgpu_device_ip_init failed\n");
 		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, 0, 0);
 		goto release_ras_con;
@@ -3924,8 +3910,10 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 				   msecs_to_jiffies(AMDGPU_RESUME_MS));
 	}
 
-	if (amdgpu_sriov_vf(adev))
+	if (amdgpu_sriov_vf(adev)) {
+		amdgpu_virt_release_full_gpu(adev, true);
 		flush_delayed_work(&adev->delayed_init_work);
+	}
 
 	r = sysfs_create_files(&adev->dev->kobj, amdgpu_dev_attributes);
 	if (r)
@@ -3965,6 +3953,20 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	return 0;
 
 release_ras_con:
+	if (amdgpu_sriov_vf(adev))
+		amdgpu_virt_release_full_gpu(adev, true);
+
+	/* failed in exclusive mode due to timeout */
+	if (amdgpu_sriov_vf(adev) &&
+		!amdgpu_sriov_runtime(adev) &&
+		amdgpu_virt_mmio_blocked(adev) &&
+		!amdgpu_virt_wait_reset(adev)) {
+		dev_err(adev->dev, "VF exclusive mode timeout\n");
+		/* Don't send request since VF is inactive. */
+		adev->virt.caps &= ~AMDGPU_SRIOV_CAPS_RUNTIME;
+		adev->virt.ops = NULL;
+		r = -EAGAIN;
+	}
 	amdgpu_release_ras_context(adev);
 
 failed:
Signature · 1 line
2.39.2
author maintainer
#86 lore raw

PATCH 86/286 watchdog: menz069_wdt: fix watchdog initialisation

In reply to Greg Kroah-Hartman

From: Johannes Thumshirn <[email protected]>

[ Upstream commit 87b22656ca6a896d0378e9e60ffccb0c82f48b08 ]

Doing a 'cat /dev/watchdog0' with menz069_wdt as watchdog0 will result in a NULL pointer dereference.

This happens because we're passing the wrong pointer to watchdog_register_device(). Fix this by getting rid of the static watchdog_device structure and use the one embedded into the driver's per-instance private data.

---
 drivers/watchdog/menz69_wdt.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)
drivers/watchdog/menz69_wdt.c +6 -10
index 8973f98bc6a56..bca0938f3429f 100644
@@ -98,14 +98,6 @@ static const struct watchdog_ops men_z069_ops = {
 	.set_timeout = men_z069_wdt_set_timeout,
 };
 
-static struct watchdog_device men_z069_wdt = {
-	.info = &men_z069_info,
-	.ops = &men_z069_ops,
-	.timeout = MEN_Z069_DEFAULT_TIMEOUT,
-	.min_timeout = 1,
-	.max_timeout = MEN_Z069_WDT_COUNTER_MAX / MEN_Z069_TIMER_FREQ,
-};
-
 static int men_z069_probe(struct mcb_device *dev,
 			  const struct mcb_device_id *id)
 {
@@ -125,15 +117,19 @@ static int men_z069_probe(struct mcb_device *dev,
 		goto release_mem;
 
 	drv->mem = mem;
+	drv->wdt.info = &men_z069_info;
+	drv->wdt.ops = &men_z069_ops;
+	drv->wdt.timeout = MEN_Z069_DEFAULT_TIMEOUT;
+	drv->wdt.min_timeout = 1;
+	drv->wdt.max_timeout = MEN_Z069_WDT_COUNTER_MAX / MEN_Z069_TIMER_FREQ;
 
-	drv->wdt = men_z069_wdt;
 	watchdog_init_timeout(&drv->wdt, 0, &dev->dev);
 	watchdog_set_nowayout(&drv->wdt, nowayout);
 	watchdog_set_drvdata(&drv->wdt, drv);
 	drv->wdt.parent = &dev->dev;
 	mcb_set_drvdata(dev, drv);
 
-	return watchdog_register_device(&men_z069_wdt);
+	return watchdog_register_device(&drv->wdt);
 
 release_mem:
 	mcb_release_mem(mem);
Signature · 1 line
2.39.2
author maintainer
#87 lore raw

PATCH 87/286 ALSA: hda: Glenfly: add HD Audio PCI IDs and HDMI Codec Vendor IDs.

In reply to Greg Kroah-Hartman

From: jasontao <[email protected]>

[ Upstream commit c51e431052e2eacfb23fbf6b39bc6c8770d9827a ]

Add a set of HD Audio PCI IDS, and the HDMI codec vendor IDs for Glenfly Gpus.

- In default_bdl_pos_adj, set bdl to 128 as Glenfly Gpus have hardware limitation, need to increase hdac interrupt interval.
- In azx_first_init, enable polling mode for Glenfly Gpu. When the codec complete the command, it sends interrupt and writes response entries to memory, howerver, the write requests sometimes are not actually synchronized to memory when driver handle hdac interrupt on Glenfly Gpus. If the RIRB status is not updated in the interrupt handler, azx_rirb_get_response keeps trying to recevie a response from rirb until 1s timeout. Enabling polling mode for Glenfly Gpu can fix the issue.
- In patch_gf_hdmi, set Glenlfy Gpu Codec's no_sticky_stream as it need driver to do actual clean-ups for the linked codec when switch from one codec to another.

---
 sound/pci/hda/hda_intel.c  | 21 +++++++++++++++++++++
 sound/pci/hda/patch_hdmi.c | 22 ++++++++++++++++++++++
 2 files changed, 43 insertions(+)
sound/pci/hda/hda_intel.c +21 -0
index 881b2f3a1551f..3226691ac923c 100644
@@ -227,6 +227,7 @@ enum {
 	AZX_DRIVER_ATI,
 	AZX_DRIVER_ATIHDMI,
 	AZX_DRIVER_ATIHDMI_NS,
+	AZX_DRIVER_GFHDMI,
 	AZX_DRIVER_VIA,
 	AZX_DRIVER_SIS,
 	AZX_DRIVER_ULI,
@@ -349,6 +350,7 @@ static const char * const driver_short_names[] = {
 	[AZX_DRIVER_ATI] = "HDA ATI SB",
 	[AZX_DRIVER_ATIHDMI] = "HDA ATI HDMI",
 	[AZX_DRIVER_ATIHDMI_NS] = "HDA ATI HDMI",
+	[AZX_DRIVER_GFHDMI] = "HDA GF HDMI",
 	[AZX_DRIVER_VIA] = "HDA VIA VT82xx",
 	[AZX_DRIVER_SIS] = "HDA SIS966",
 	[AZX_DRIVER_ULI] = "HDA ULI M5461",
@@ -1743,6 +1745,12 @@ static int default_bdl_pos_adj(struct azx *chip)
 	}
 
 	switch (chip->driver_type) {
+	/*
+	 * increase the bdl size for Glenfly Gpus for hardware
+	 * limitation on hdac interrupt interval
+	 */
+	case AZX_DRIVER_GFHDMI:
+		return 128;
 	case AZX_DRIVER_ICH:
 	case AZX_DRIVER_PCH:
 		return 1;
@@ -1858,6 +1866,12 @@ static int azx_first_init(struct azx *chip)
 		pci_write_config_dword(pci, PCI_BASE_ADDRESS_1, 0);
 	}
 #endif
+	/*
+	 * Fix response write request not synced to memory when handle
+	 * hdac interrupt on Glenfly Gpus
+	 */
+	if (chip->driver_type == AZX_DRIVER_GFHDMI)
+		bus->polling_mode = 1;
 
 	err = pcim_iomap_regions(pci, 1 << 0, "ICH HD audio");
 	if (err < 0)
@@ -1959,6 +1973,7 @@ static int azx_first_init(struct azx *chip)
 			chip->playback_streams = ATIHDMI_NUM_PLAYBACK;
 			chip->capture_streams = ATIHDMI_NUM_CAPTURE;
 			break;
+		case AZX_DRIVER_GFHDMI:
 		case AZX_DRIVER_GENERIC:
 		default:
 			chip->playback_streams = ICH6_NUM_PLAYBACK;
@@ -2727,6 +2742,12 @@ static const struct pci_device_id azx_ids[] = {
 	{ PCI_DEVICE(0x1002, 0xab38),
 	  .driver_data = AZX_DRIVER_ATIHDMI_NS | AZX_DCAPS_PRESET_ATI_HDMI_NS |
 	  AZX_DCAPS_PM_RUNTIME },
+	/* GLENFLY */
+	{ PCI_DEVICE(0x6766, PCI_ANY_ID),
+	  .class = PCI_CLASS_MULTIMEDIA_HD_AUDIO << 8,
+	  .class_mask = 0xffffff,
+	  .driver_data = AZX_DRIVER_GFHDMI | AZX_DCAPS_POSFIX_LPIB |
+	  AZX_DCAPS_NO_MSI | AZX_DCAPS_NO_64BIT },
 	/* VIA VT8251/VT8237A */
 	{ PCI_DEVICE(0x1106, 0x3288), .driver_data = AZX_DRIVER_VIA },
 	/* VIA GFX VT7122/VX900 */
sound/pci/hda/patch_hdmi.c +22 -0
index be2c6cff77011..7b5e09070ab9b 100644
@@ -4489,6 +4489,22 @@ static int patch_via_hdmi(struct hda_codec *codec)
 	return patch_simple_hdmi(codec, VIAHDMI_CVT_NID, VIAHDMI_PIN_NID);
 }
 
+static int patch_gf_hdmi(struct hda_codec *codec)
+{
+	int err;
+
+	err = patch_generic_hdmi(codec);
+	if (err)
+		return err;
+
+	/*
+	 * Glenfly GPUs have two codecs, stream switches from one codec to
+	 * another, need to do actual clean-ups in codec_cleanup_stream
+	 */
+	codec->no_sticky_stream = 1;
+	return 0;
+}
+
 /*
  * patch entries
  */
@@ -4584,6 +4600,12 @@ HDA_CODEC_ENTRY(0x10de00a6, "GPU a6 HDMI/DP",	patch_nvhdmi),
 HDA_CODEC_ENTRY(0x10de00a7, "GPU a7 HDMI/DP",	patch_nvhdmi),
 HDA_CODEC_ENTRY(0x10de8001, "MCP73 HDMI",	patch_nvhdmi_2ch),
 HDA_CODEC_ENTRY(0x10de8067, "MCP67/68 HDMI",	patch_nvhdmi_2ch),
+HDA_CODEC_ENTRY(0x67663d82, "Arise 82 HDMI/DP",	patch_gf_hdmi),
+HDA_CODEC_ENTRY(0x67663d83, "Arise 83 HDMI/DP",	patch_gf_hdmi),
+HDA_CODEC_ENTRY(0x67663d84, "Arise 84 HDMI/DP",	patch_gf_hdmi),
+HDA_CODEC_ENTRY(0x67663d85, "Arise 85 HDMI/DP",	patch_gf_hdmi),
+HDA_CODEC_ENTRY(0x67663d86, "Arise 86 HDMI/DP",	patch_gf_hdmi),
+HDA_CODEC_ENTRY(0x67663d87, "Arise 87 HDMI/DP",	patch_gf_hdmi),
 HDA_CODEC_ENTRY(0x11069f80, "VX900 HDMI/DP",	patch_via_hdmi),
 HDA_CODEC_ENTRY(0x11069f81, "VX900 HDMI/DP",	patch_via_hdmi),
 HDA_CODEC_ENTRY(0x11069f84, "VX11 HDMI/DP",	patch_generic_hdmi),
Signature · 1 line
2.39.2
author maintainer
#88 lore raw

PATCH 88/286 drm/amd/display: fix memleak in aconnector->timing_requested

In reply to Greg Kroah-Hartman

From: Hersen Wu <[email protected]>

[ Upstream commit 025ce392b5f213696ca0af3e07735d0fae020694 ]

[Why]
when amdgpu_dm_update_connector_after_detect is called two times successively with valid sink, memory allocated of aconnector->timing_requested for the first call is not free. this causes memeleak.

[How]
allocate memory only when aconnector->timing_requested is null.

---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +6 -3
index 0695c7c3d489d..ce46f3a061c44 100644
@@ -3095,9 +3095,12 @@ void amdgpu_dm_update_connector_after_detect(
 						    aconnector->edid);
 		}
 
-		aconnector->timing_requested = kzalloc(sizeof(struct dc_crtc_timing), GFP_KERNEL);
-		if (!aconnector->timing_requested)
-			dm_error("%s: failed to create aconnector->requested_timing\n", __func__);
+		if (!aconnector->timing_requested) {
+			aconnector->timing_requested =
+				kzalloc(sizeof(struct dc_crtc_timing), GFP_KERNEL);
+			if (!aconnector->timing_requested)
+				dm_error("failed to create aconnector->requested_timing\n");
+		}
 
 		drm_connector_update_edid_property(connector, aconnector->edid);
 		amdgpu_dm_update_freesync_caps(connector, aconnector->edid);
Signature · 1 line
2.39.2
author maintainer
#89 lore raw

PATCH 89/286 LoongArch: Relay BCE exceptions to userland as SIGSEGV with si_code=SEGV_BNDERR

In reply to Greg Kroah-Hartman

From: WANG Xuerui <[email protected]>

[ Upstream commit c23e7f01cf621290770069d968ca4c8356a50d00 ]

SEGV_BNDERR was introduced initially for supporting the Intel MPX, but fell into disuse after the MPX support was removed. The LoongArch bounds-checking instructions behave very differently than MPX, but overall the interface is still kind of suitable for conveying the information to userland when bounds-checking assertions trigger, so we wouldn't have to invent more UAPI. Specifically, when the BCE triggers, a SEGV_BNDERR is sent to userland, with si_addr set to the out-of-bounds address or value (in asrt{gt,le}'s case), and one of si_lower or si_upper set to the configured bound depending on the faulting instruction. The other bound is set to either 0 or ULONG_MAX to resemble a range with both lower and upper bounds.

Note that it is possible to have si_addr == si_lower in case of a failing asrtgt or {ld,st}gt, because those instructions test for strict greater-than relationship. This should not pose a problem for userland, though, because the faulting PC is available for the application to associate back to the exact instruction for figuring out the expectation.

Example exception context generated by a faulting asrtgt.d t0, t1 (assert t0 > t1 or BCE) with t0=100 and t1=200:

Quoted 8 lines

pc 00005555558206a4 ra 00007ffff2d854fc tp 00007ffff2f2f180 sp 00007ffffbf9fb80 a0 0000000000000002 a1 00007ffffbf9fce8 a2 00007ffffbf9fd00 a3 00007ffff2ed4558 a4 0000000000000000 a5 00007ffff2f044c8 a6 00007ffffbf9fce0 a7 fffffffffffff000 t0 0000000000000064 t1 00000000000000c8 t2 00007ffffbfa2d5e t3 00007ffff2f12aa0 t4 00007ffff2ed6158 t5 00007ffff2ed6158 t6 000000000000002e t7 0000000003d8f538 t8 0000000000000005 u0 0000000000000000 s9 0000000000000000 s0 00007ffffbf9fce8 s1 0000000000000002 s2 0000000000000000 s3 00007ffff2f2c038 s4 0000555555820610 s5 00007ffff2ed5000 s6 0000555555827e38 s7 00007ffffbf9fd00 s8 0000555555827e38

Quoted 8 lines

ra: 00007ffff2d854fc
ERA: 00005555558206a4
CRMD: 000000b0 (PLV0 -IE -DA +PG DACF=CC DACM=CC -WE)
PRMD: 00000007 (PPLV3 +PIE -PWE)
EUEN: 00000000 (-FPE -SXE -ASXE -BTE)
ECFG: 0007181c (LIE=2-4,11-12 VS=7)
ESTAT: 000a0000 [BCE] (IS= ECode=10 EsubCode=0)
PRID: 0014c010 (Loongson-64bit, Loongson-3A5000)

---
 arch/loongarch/include/asm/inst.h | 26 +++++++++
 arch/loongarch/kernel/genex.S     |  1 +
 arch/loongarch/kernel/traps.c     | 92 +++++++++++++++++++++++++++++++
 3 files changed, 119 insertions(+)
arch/loongarch/include/asm/inst.h +26 -0
index a04fe755d7193..b09887ffcd15d 100644
@@ -121,6 +121,8 @@ enum reg2bstrd_op {
 };
 
 enum reg3_op {
+	asrtle_op	= 0x02,
+	asrtgt_op	= 0x03,
 	addw_op		= 0x20,
 	addd_op		= 0x21,
 	subw_op		= 0x22,
@@ -176,6 +178,30 @@ enum reg3_op {
 	amord_op	= 0x70c7,
 	amxorw_op	= 0x70c8,
 	amxord_op	= 0x70c9,
+	fldgts_op	= 0x70e8,
+	fldgtd_op	= 0x70e9,
+	fldles_op	= 0x70ea,
+	fldled_op	= 0x70eb,
+	fstgts_op	= 0x70ec,
+	fstgtd_op	= 0x70ed,
+	fstles_op	= 0x70ee,
+	fstled_op	= 0x70ef,
+	ldgtb_op	= 0x70f0,
+	ldgth_op	= 0x70f1,
+	ldgtw_op	= 0x70f2,
+	ldgtd_op	= 0x70f3,
+	ldleb_op	= 0x70f4,
+	ldleh_op	= 0x70f5,
+	ldlew_op	= 0x70f6,
+	ldled_op	= 0x70f7,
+	stgtb_op	= 0x70f8,
+	stgth_op	= 0x70f9,
+	stgtw_op	= 0x70fa,
+	stgtd_op	= 0x70fb,
+	stleb_op	= 0x70fc,
+	stleh_op	= 0x70fd,
+	stlew_op	= 0x70fe,
+	stled_op	= 0x70ff,
 };
 
 enum reg3sa2_op {
arch/loongarch/kernel/genex.S +1 -0
index 44ff1ff642601..78f0663846575 100644
@@ -82,6 +82,7 @@ SYM_FUNC_END(except_vec_cex)
 
 	BUILD_HANDLER ade ade badv
 	BUILD_HANDLER ale ale badv
+	BUILD_HANDLER bce bce none
 	BUILD_HANDLER bp bp none
 	BUILD_HANDLER fpe fpe fcsr
 	BUILD_HANDLER fpu fpu none
arch/loongarch/kernel/traps.c +92 -0

116 lines not on this page · raw message

Signature · 1 line
2.39.2
author maintainer
#90 lore raw

PATCH 90/286 LoongArch: Add ARCH_HAS_FORTIFY_SOURCE selection

In reply to Greg Kroah-Hartman

From: Qing Zhang <[email protected]>

[ Upstream commit d4c937c2a57bbba24790be6fe7a791456f5fbb60 ]

FORTIFY_SOURCE could detect various overflows at compile and run time. ARCH_HAS_FORTIFY_SOURCE means that the architecture can be built and run with CONFIG_FORTIFY_SOURCE. So select it in LoongArch.

See more about this feature from commit 6974f0c4555e285 ("include/linux/ string.h: add the option of fortified string.h functions").

---
 arch/loongarch/Kconfig | 1 +
 1 file changed, 1 insertion(+)
arch/loongarch/Kconfig +1 -0
index 3ddde336e6a56..3e5d6acbf2409 100644
@@ -10,6 +10,7 @@ config LOONGARCH
 	select ARCH_ENABLE_MEMORY_HOTPLUG
 	select ARCH_ENABLE_MEMORY_HOTREMOVE
 	select ARCH_HAS_ACPI_TABLE_UPGRADE	if ACPI
+	select ARCH_HAS_FORTIFY_SOURCE
 	select ARCH_HAS_NMI_SAFE_THIS_CPU_OPS
 	select ARCH_HAS_PTE_SPECIAL
 	select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
Signature · 1 line
2.39.2
author maintainer
#91 lore raw

PATCH 91/286 ASoC: Intel: soc-acpi-cht: Add quirk for Nextbook Ares 8A tablet

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

[ Upstream commit ec6f82b4c63cc68f8dc03316e725106d242706be ]

The Nextbook Ares 8A tablet which has Android as factory OS, has a buggy DSDT with both ESSX8316 and 10EC5651 ACPI devices.

This tablet actually uses an rt5651 codec, but the matching code ends up picking the ESSX8316 device, add a quirk to ignote the ESSX8316 device on this tablet.

Message-Id: <[email protected]>

---
 .../intel/common/soc-acpi-intel-cht-match.c   | 26 +++++++++++++++++++
 1 file changed, 26 insertions(+)
sound/soc/intel/common/soc-acpi-intel-cht-match.c +26 -0
index 6beb00858c33f..cdcbf04b8832f 100644
@@ -50,6 +50,31 @@ static struct snd_soc_acpi_mach *cht_quirk(void *arg)
 		return mach;
 }
 
+/*
+ * Some tablets with Android factory OS have buggy DSDTs with an ESSX8316 device
+ * in the ACPI tables. While they are not using an ESS8316 codec. These DSDTs
+ * also have an ACPI device for the correct codec, ignore the ESSX8316.
+ */
+static const struct dmi_system_id cht_ess8316_not_present_table[] = {
+	{
+		/* Nextbook Ares 8A */
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Insyde"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "CherryTrail"),
+			DMI_MATCH(DMI_BIOS_VERSION, "M882"),
+		},
+	},
+	{ }
+};
+
+static struct snd_soc_acpi_mach *cht_ess8316_quirk(void *arg)
+{
+	if (dmi_check_system(cht_ess8316_not_present_table))
+		return NULL;
+
+	return arg;
+}
+
 static const struct snd_soc_acpi_codecs rt5640_comp_ids = {
 	.num_codecs = 2,
 	.codecs = { "10EC5640", "10EC3276" },
@@ -113,6 +138,7 @@ struct snd_soc_acpi_mach  snd_soc_acpi_intel_cherrytrail_machines[] = {
 		.drv_name = "bytcht_es8316",
 		.fw_filename = "intel/fw_sst_22a8.bin",
 		.board = "bytcht_es8316",
+		.machine_quirk = cht_ess8316_quirk,
 		.sof_tplg_filename = "sof-cht-es8316.tplg",
 	},
 	/* some CHT-T platforms rely on RT5640, use Baytrail machine driver */
Signature · 1 line
2.39.2
author maintainer
#92 lore raw

PATCH 92/286 drm/amdgpu: Use the default reset when loading or reloading the driver

In reply to Greg Kroah-Hartman

From: lyndonli <[email protected]>

[ Upstream commit 4eea7fb980dc44545a32eec92e2662053b34cd9d ]

Below call trace and errors are observed when reloading amdgpu driver with the module parameter reset_method=3.

It should do a default reset when loading or reloading the driver, regardless of the module parameter reset_method.

v2: add comments inside and modify commit messages.

[ +2.180243] [drm] psp gfx command ID_LOAD_TOC(0x20) failed and response status is (0x0)

[  +0.000011] [drm:psp_hw_start [amdgpu]] *ERROR* Failed to load toc
[  +0.000890] [drm:psp_hw_start [amdgpu]] *ERROR* PSP tmr init failed!
[  +0.020683] [drm:amdgpu_fill_buffer [amdgpu]] *ERROR* Trying to
clear memory with ring turned off.
[  +0.000003] RIP: 0010:amdgpu_bo_release_notify+0x1ef/0x210 [amdgpu]
[  +0.000004] Call Trace:
[  +0.000003]  <TASK>
[  +0.000008]  ttm_bo_release+0x2c4/0x330 [amdttm]
[  +0.000026]  amdttm_bo_put+0x3c/0x70 [amdttm]
[  +0.000020]  amdgpu_bo_free_kernel+0xe6/0x140 [amdgpu]
[  +0.000728]  psp_v11_0_ring_destroy+0x34/0x60 [amdgpu]
[  +0.000826]  psp_hw_init+0xe7/0x2f0 [amdgpu]
[  +0.000813]  amdgpu_device_fw_loading+0x1ad/0x2d0 [amdgpu]
[  +0.000731]  amdgpu_device_init.cold+0x108e/0x2002 [amdgpu]
[  +0.001071]  ? do_pci_enable_device+0xe1/0x110
[  +0.000011]  amdgpu_driver_load_kms+0x1a/0x160 [amdgpu]
[  +0.000729]  amdgpu_pci_probe+0x179/0x3a0 [amdgpu]
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +7 -0
index 464f0abc855e6..412cb3f1f8826 100644
@@ -3558,6 +3558,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	int r, i;
 	bool px = false;
 	u32 max_MBps;
+	int tmp;
 
 	adev->shutdown = false;
 	adev->flags = flags;
@@ -3779,7 +3780,13 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 				}
 			}
 		} else {
+			tmp = amdgpu_reset_method;
+			/* It should do a default reset when loading or reloading the driver,
+			 * regardless of the module parameter reset_method.
+			 */
+			amdgpu_reset_method = AMD_RESET_METHOD_NONE;
 			r = amdgpu_asic_reset(adev);
+			amdgpu_reset_method = tmp;
 			if (r) {
 				dev_err(adev->dev, "asic reset on init failed\n");
 				goto failed;
Signature · 1 line
2.39.2
author maintainer
#93 lore raw

PATCH 93/286 mailbox: mailbox-test: Fix potential double-free in mbox_test_message_write()

In reply to Greg Kroah-Hartman

From: Lee Jones <[email protected]>

[ Upstream commit 2d1e952a2b8e5e92d8d55ac88a7cf7ca5ea591ad ]

If a user can make copy_from_user() fail, there is a potential for UAF/DF due to a lack of locking around the allocation, use and freeing of the data buffers.

This issue is not theoretical. I managed to author a POC for it:

    BUG: KASAN: double-free in kfree+0x5c/0xac
    Free of addr ffff29280be5de00 by task poc/356
    CPU: 1 PID: 356 Comm: poc Not tainted 6.1.0-00001-g961aa6552c04-dirty #20
    Hardware name: linux,dummy-virt (DT)
    Call trace:
     dump_backtrace.part.0+0xe0/0xf0
     show_stack+0x18/0x40
     dump_stack_lvl+0x64/0x80
     print_report+0x188/0x48c
     kasan_report_invalid_free+0xa0/0xc0
     ____kasan_slab_free+0x174/0x1b0
     __kasan_slab_free+0x18/0x24
     __kmem_cache_free+0x130/0x2e0
     kfree+0x5c/0xac
     mbox_test_message_write+0x208/0x29c
     full_proxy_write+0x90/0xf0
     vfs_write+0x154/0x440
     ksys_write+0xcc/0x180
     __arm64_sys_write+0x44/0x60
     invoke_syscall+0x60/0x190
     el0_svc_common.constprop.0+0x7c/0x160
     do_el0_svc+0x40/0xf0
     el0_svc+0x2c/0x6c
     el0t_64_sync_handler+0xf4/0x120
     el0t_64_sync+0x18c/0x190

    Allocated by task 356:
     kasan_save_stack+0x3c/0x70
     kasan_set_track+0x2c/0x40
     kasan_save_alloc_info+0x24/0x34
     __kasan_kmalloc+0xb8/0xc0
     kmalloc_trace+0x58/0x70
     mbox_test_message_write+0x6c/0x29c
     full_proxy_write+0x90/0xf0
     vfs_write+0x154/0x440
     ksys_write+0xcc/0x180
     __arm64_sys_write+0x44/0x60
     invoke_syscall+0x60/0x190
     el0_svc_common.constprop.0+0x7c/0x160
     do_el0_svc+0x40/0xf0
     el0_svc+0x2c/0x6c
     el0t_64_sync_handler+0xf4/0x120
     el0t_64_sync+0x18c/0x190

    Freed by task 357:
     kasan_save_stack+0x3c/0x70
     kasan_set_track+0x2c/0x40
     kasan_save_free_info+0x38/0x5c
     ____kasan_slab_free+0x13c/0x1b0
     __kasan_slab_free+0x18/0x24
     __kmem_cache_free+0x130/0x2e0
     kfree+0x5c/0xac
     mbox_test_message_write+0x208/0x29c
     full_proxy_write+0x90/0xf0
     vfs_write+0x154/0x440
     ksys_write+0xcc/0x180
     __arm64_sys_write+0x44/0x60
     invoke_syscall+0x60/0x190
     el0_svc_common.constprop.0+0x7c/0x160
     do_el0_svc+0x40/0xf0
     el0_svc+0x2c/0x6c
     el0t_64_sync_handler+0xf4/0x120
     el0t_64_sync+0x18c/0x190
---
 drivers/mailbox/mailbox-test.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/mailbox/mailbox-test.c +7 -0
index 4555d678fadda..6dd5b9614452b 100644
@@ -12,6 +12,7 @@
 #include <linux/kernel.h>
 #include <linux/mailbox_client.h>
 #include <linux/module.h>
+#include <linux/mutex.h>
 #include <linux/of.h>
 #include <linux/platform_device.h>
 #include <linux/poll.h>
@@ -38,6 +39,7 @@ struct mbox_test_device {
 	char			*signal;
 	char			*message;
 	spinlock_t		lock;
+	struct mutex		mutex;
 	wait_queue_head_t	waitq;
 	struct fasync_struct	*async_queue;
 	struct dentry		*root_debugfs_dir;
@@ -110,6 +112,8 @@ static ssize_t mbox_test_message_write(struct file *filp,
 		return -EINVAL;
 	}
 
+	mutex_lock(&tdev->mutex);
+
 	tdev->message = kzalloc(MBOX_MAX_MSG_LEN, GFP_KERNEL);
 	if (!tdev->message)
 		return -ENOMEM;
@@ -144,6 +148,8 @@ static ssize_t mbox_test_message_write(struct file *filp,
 	kfree(tdev->message);
 	tdev->signal = NULL;
 
+	mutex_unlock(&tdev->mutex);
+
 	return ret < 0 ? ret : count;
 }
 
@@ -392,6 +398,7 @@ static int mbox_test_probe(struct platform_device *pdev)
 	platform_set_drvdata(pdev, tdev);
 
 	spin_lock_init(&tdev->lock);
+	mutex_init(&tdev->mutex);
 
 	if (tdev->rx_channel) {
 		tdev->rx_buffer = devm_kzalloc(&pdev->dev,
Signature · 1 line
2.39.2
author maintainer
#94 lore raw

PATCH 94/286 drm/ast: Fix ARM compatibility

In reply to Greg Kroah-Hartman

From: Jammy Huang <[email protected]>

[ Upstream commit 4327a6137ed43a091d900b1ac833345d60f32228 ]

ARM architecture only has 'memory', so all devices are accessed by MMIO if possible.

---
 drivers/gpu/drm/ast/ast_main.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/gpu/drm/ast/ast_main.c +5 -4
index f83ce77127cb4..a6d0ee4da2b88 100644
@@ -425,11 +425,12 @@ struct ast_private *ast_device_create(const struct drm_driver *drv,
 		return ERR_PTR(-EIO);
 
 	/*
-	 * If we don't have IO space at all, use MMIO now and
-	 * assume the chip has MMIO enabled by default (rev 0x20
-	 * and higher).
+	 * After AST2500, MMIO is enabled by default, and it should be adopted
+	 * to be compatible with Arm.
 	 */
-	if (!(pci_resource_flags(pdev, 2) & IORESOURCE_IO)) {
+	if (pdev->revision >= 0x40) {
+		ast->ioregs = ast->regs + AST_IO_MM_OFFSET;
+	} else if (!(pci_resource_flags(pdev, 2) & IORESOURCE_IO)) {
 		drm_info(dev, "platform has no IO space, trying MMIO\n");
 		ast->ioregs = ast->regs + AST_IO_MM_OFFSET;
 	}
Signature · 1 line
2.39.2
author maintainer
#68 lore raw

PATCH 68/286 net: ipa: Use correct value for IPA_STATUS_SIZE

In reply to Greg Kroah-Hartman

From: Bert Karwatzki <[email protected]>

[ Upstream commit be7f8012a513f5099916ee2da28420156cbb8cf3 ]

IPA_STATUS_SIZE was introduced in commit b8dc7d0eea5a as a replacement for the size of the removed struct ipa_status which had size sizeof(__le32[8]). Use this value as IPA_STATUS_SIZE.

---
 drivers/net/ipa/ipa_endpoint.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ipa/ipa_endpoint.c +1 -1
index 2ee80ed140b72..afa1d56d9095c 100644
@@ -119,7 +119,7 @@ enum ipa_status_field_id {
 };
 
 /* Size in bytes of an IPA packet status structure */
-#define IPA_STATUS_SIZE			sizeof(__le32[4])
+#define IPA_STATUS_SIZE			sizeof(__le32[8])
 
 /* IPA status structure decoder; looks up field values for a structure */
 static u32 ipa_status_extract(struct ipa *ipa, const void *data,
Signature · 1 line
2.39.2
author maintainer
#95 lore raw

PATCH 95/286 btrfs: abort transaction when sibling keys check fails for leaves

In reply to Greg Kroah-Hartman

From: Filipe Manana <[email protected]>

[ Upstream commit 9ae5afd02a03d4e22a17a9609b19400b77c36273 ]

If the sibling keys check fails before we move keys from one sibling leaf to another, we are not aborting the transaction - we leave that to some higher level caller of btrfs_search_slot() (or anything else that uses it to insert items into a b+tree).

This means that the transaction abort will provide a stack trace that omits the b+tree modification call chain. So change this to immediately abort the transaction and therefore get a more useful stack trace that shows us the call chain in the bt+tree modification code.

It's also important to immediately abort the transaction just in case some higher level caller is not doing it, as this indicates a very serious corruption and we should stop the possibility of doing further damage.

---
 fs/btrfs/ctree.c | 2 ++
 1 file changed, 2 insertions(+)
fs/btrfs/ctree.c +2 -0
index 26bb10b6ca85d..986827370d8e1 100644
@@ -3222,6 +3222,7 @@ static int push_leaf_right(struct btrfs_trans_handle *trans, struct btrfs_root
 
 	if (check_sibling_keys(left, right)) {
 		ret = -EUCLEAN;
+		btrfs_abort_transaction(trans, ret);
 		btrfs_tree_unlock(right);
 		free_extent_buffer(right);
 		return ret;
@@ -3444,6 +3445,7 @@ static int push_leaf_left(struct btrfs_trans_handle *trans, struct btrfs_root
 
 	if (check_sibling_keys(left, right)) {
 		ret = -EUCLEAN;
+		btrfs_abort_transaction(trans, ret);
 		goto out;
 	}
 	return __push_leaf_left(trans, path, min_data_size, empty, left,
Signature · 1 line
2.39.2
author maintainer
#96 lore raw

PATCH 96/286 ARM: 9295/1: unwind:fix unwind abort for uleb128 case

In reply to Greg Kroah-Hartman

From: Haibo Li <[email protected]>

[ Upstream commit fa3eeb638de0c1a9d2d860e5b48259facdd65176 ]

When unwind instruction is 0xb2,the subsequent instructions are uleb128 bytes.
For now,it uses only the first uleb128 byte in code.

For vsp increments of 0x204~0x400,use one uleb128 byte like below:
0xc06a00e4 <unwind_test_work>: 0x80b27fac
Compact model index: 0
0xb2 0x7f vsp = vsp + 1024
0xac pop {r4, r5, r6, r7, r8, r14}

For vsp increments larger than 0x400,use two uleb128 bytes like below:
0xc06a00e4 <unwind_test_work>: @0xc0cc9e0c
Compact model index: 1
0xb2 0x81 0x01 vsp = vsp + 1032
0xac pop {r4, r5, r6, r7, r8, r14}
The unwind works well since the decoded uleb128 byte is also 0x81.

For vsp increments larger than 0x600,use two uleb128 bytes like below:
0xc06a00e4 <unwind_test_work>: @0xc0cc9e0c
Compact model index: 1
0xb2 0x81 0x02 vsp = vsp + 1544
0xac pop {r4, r5, r6, r7, r8, r14}
In this case,the decoded uleb128 result is 0x101(vsp=0x204+(0x101<<2)). While the uleb128 used in code is 0x81(vsp=0x204+(0x81<<2)). The unwind aborts at this frame since it gets incorrect vsp.

To fix this,add uleb128 decode to cover all the above case.

---
 arch/arm/kernel/unwind.c | 25 ++++++++++++++++++++++++-
 1 file changed, 24 insertions(+), 1 deletion(-)
arch/arm/kernel/unwind.c +24 -1
index 53be7ea6181b3..9d2192156087b 100644
@@ -308,6 +308,29 @@ static int unwind_exec_pop_subset_r0_to_r3(struct unwind_ctrl_block *ctrl,
 	return URC_OK;
 }
 
+static unsigned long unwind_decode_uleb128(struct unwind_ctrl_block *ctrl)
+{
+	unsigned long bytes = 0;
+	unsigned long insn;
+	unsigned long result = 0;
+
+	/*
+	 * unwind_get_byte() will advance `ctrl` one instruction at a time, so
+	 * loop until we get an instruction byte where bit 7 is not set.
+	 *
+	 * Note: This decodes a maximum of 4 bytes to output 28 bits data where
+	 * max is 0xfffffff: that will cover a vsp increment of 1073742336, hence
+	 * it is sufficient for unwinding the stack.
+	 */
+	do {
+		insn = unwind_get_byte(ctrl);
+		result |= (insn & 0x7f) << (bytes * 7);
+		bytes++;
+	} while (!!(insn & 0x80) && (bytes != sizeof(result)));
+
+	return result;
+}
+
 /*
  * Execute the current unwind instruction.
  */
@@ -361,7 +384,7 @@ static int unwind_exec_insn(struct unwind_ctrl_block *ctrl)
 		if (ret)
 			goto error;
 	} else if (insn == 0xb2) {
-		unsigned long uleb128 = unwind_get_byte(ctrl);
+		unsigned long uleb128 = unwind_decode_uleb128(ctrl);
 
 		ctrl->vrs[SP] += 0x204 + (uleb128 << 2);
 	} else {
Signature · 1 line
2.39.2
author maintainer
#69 lore raw

PATCH 69/286 net: dsa: mv88e6xxx: Increase wait after reset deactivation

In reply to Greg Kroah-Hartman

From: Andreas Svensson <[email protected]>

[ Upstream commit 3c27f3d53d588618d81d30d6712459a3cc9489b8 ]

A switch held in reset by default needs to wait longer until we can reliably detect it.

An issue was observed when testing on the Marvell 88E6393X (Link Street). The driver failed to detect the switch on some upstarts. Increasing the wait time after reset deactivation solves this issue.

The updated wait time is now also the same as the wait time in the mv88e6xxx_hardware_reset function.

Fixes 7b75e49de424 ("net: dsa: mv88e6xxx: wait after reset deactivation")
---
 drivers/net/dsa/mv88e6xxx/chip.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/dsa/mv88e6xxx/chip.c +1 -1
index 902f407213404..39770a5b74e2e 100644
@@ -7158,7 +7158,7 @@ static int mv88e6xxx_probe(struct mdio_device *mdiodev)
 		goto out;
 	}
 	if (chip->reset)
-		usleep_range(1000, 2000);
+		usleep_range(10000, 20000);
 
 	/* Detect if the device is configured in single chip addressing mode,
 	 * otherwise continue with address specific smi init/detection.
Signature · 1 line
2.39.2
author maintainer
#70 lore raw

PATCH 70/286 mtd: rawnand: marvell: ensure timing values are written

In reply to Greg Kroah-Hartman

From: Chris Packham <[email protected]>

[ Upstream commit 8a6f4d346f3bad9c68b4a87701eb3f7978542d57 ]

When new timing values are calculated in marvell_nfc_setup_interface() ensure that they will be applied in marvell_nfc_select_target() by clearing the selected_chip pointer.

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
---
 drivers/mtd/nand/raw/marvell_nand.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +6 -0
index 3034916d2e252..52342f2c126da 100644
@@ -2457,6 +2457,12 @@ static int marvell_nfc_setup_interface(struct nand_chip *chip, int chipnr,
 			NDTR1_WAIT_MODE;
 	}
 
+	/*
+	 * Reset nfc->selected_chip so the next command will cause the timing
+	 * registers to be updated in marvell_nfc_select_target().
+	 */
+	nfc->selected_chip = NULL;
+
 	return 0;
 }
 
Signature · 1 line
2.39.2
author maintainer
#71 lore raw

PATCH 71/286 mtd: rawnand: marvell: dont set the NAND frequency select

In reply to Greg Kroah-Hartman

From: Chris Packham <[email protected]>

[ Upstream commit c4d28e30a8d0b979e4029465ab8f312ab6ce2644 ]

marvell_nfc_setup_interface() uses the frequency retrieved from the clock associated with the nand interface to determine the timings that will be used. By changing the NAND frequency select without reflecting this in the clock configuration this means that the timings calculated don't correctly meet the requirements of the NAND chip. This hasn't been an issue up to now because of a different bug that was stopping the timings being updated after they were initially set.

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
---
 drivers/mtd/nand/raw/marvell_nand.c | 4 ----
 1 file changed, 4 deletions(-)
drivers/mtd/nand/raw/marvell_nand.c +0 -4
index 52342f2c126da..2d34c2cb0f7ae 100644
@@ -2900,10 +2900,6 @@ static int marvell_nfc_init(struct marvell_nfc *nfc)
 		regmap_update_bits(sysctrl_base, GENCONF_CLK_GATING_CTRL,
 				   GENCONF_CLK_GATING_CTRL_ND_GATE,
 				   GENCONF_CLK_GATING_CTRL_ND_GATE);
-
-		regmap_update_bits(sysctrl_base, GENCONF_ND_CLK_CTRL,
-				   GENCONF_ND_CLK_CTRL_EN,
-				   GENCONF_ND_CLK_CTRL_EN);
 	}
 
 	/* Configure the DMA if appropriate */
Signature · 1 line
2.39.2
author maintainer
#72 lore raw

PATCH 72/286 net: renesas: rswitch: Fix return value in error path of xmit

In reply to Greg Kroah-Hartman

From: Yoshihiro Shimoda <[email protected]>

[ Upstream commit a60caf039e96d806b1ced893242bae82ba3ccf0d ]

Fix return value in the error path of rswitch_start_xmit(). If TX queues are full, this function should return NETDEV_TX_BUSY.

Fixes 3590918b5d07 ("net: ethernet: renesas: Add support for "Ethernet Switch"")
---
 drivers/net/ethernet/renesas/rswitch.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/renesas/rswitch.c +1 -1
index c4f93d24c6a42..7855d9ef81eb1 100644
@@ -1487,7 +1487,7 @@ static netdev_tx_t rswitch_start_xmit(struct sk_buff *skb, struct net_device *nd
 
 	if (rswitch_get_num_cur_queues(gq) >= gq->ring_size - 1) {
 		netif_stop_subqueue(ndev, 0);
-		return ret;
+		return NETDEV_TX_BUSY;
 	}
 
 	if (skb_put_padto(skb, ETH_ZLEN))
Signature · 1 line
2.39.2
author maintainer
#73 lore raw

PATCH 73/286 net: phy: mxl-gpy: extend interrupt fix to all impacted variants

In reply to Greg Kroah-Hartman

From: Xu Liang <[email protected]>

[ Upstream commit 519d6487640835d19461817c75907e6308074a73 ]

The interrupt fix in commit 97a89ed101bb should be applied on all variants of GPY2xx PHY and GPY115C.

Fixes 97a89ed101bb ("net: phy: mxl-gpy: disable interrupts on GPY215 by default")
---
 drivers/net/phy/mxl-gpy.c | 16 +++-------------
 1 file changed, 3 insertions(+), 13 deletions(-)
drivers/net/phy/mxl-gpy.c +3 -13
index e5972b4ef6e8f..4041ebd7ad9b3 100644
@@ -267,13 +267,6 @@ static int gpy_config_init(struct phy_device *phydev)
 	return ret < 0 ? ret : 0;
 }
 
-static bool gpy_has_broken_mdint(struct phy_device *phydev)
-{
-	/* At least these PHYs are known to have broken interrupt handling */
-	return phydev->drv->phy_id == PHY_ID_GPY215B ||
-	       phydev->drv->phy_id == PHY_ID_GPY215C;
-}
-
 static int gpy_probe(struct phy_device *phydev)
 {
 	struct device *dev = &phydev->mdio.dev;
@@ -293,8 +286,7 @@ static int gpy_probe(struct phy_device *phydev)
 	phydev->priv = priv;
 	mutex_init(&priv->mbox_lock);
 
-	if (gpy_has_broken_mdint(phydev) &&
-	    !device_property_present(dev, "maxlinear,use-broken-interrupts"))
+	if (!device_property_present(dev, "maxlinear,use-broken-interrupts"))
 		phydev->dev_flags |= PHY_F_NO_IRQ;
 
 	fw_version = phy_read(phydev, PHY_FWV);
@@ -652,11 +644,9 @@ static irqreturn_t gpy_handle_interrupt(struct phy_device *phydev)
 	 * frame. Therefore, polling is the best we can do and won't do any more
 	 * harm.
 	 * It was observed that this bug happens on link state and link speed
-	 * changes on a GPY215B and GYP215C independent of the firmware version
-	 * (which doesn't mean that this list is exhaustive).
+	 * changes independent of the firmware version.
 	 */
-	if (gpy_has_broken_mdint(phydev) &&
-	    (reg & (PHY_IMASK_LSTC | PHY_IMASK_LSPC))) {
+	if (reg & (PHY_IMASK_LSTC | PHY_IMASK_LSPC)) {
 		reg = gpy_mbox_read(phydev, REG_GPIO0_OUT);
 		if (reg < 0) {
 			phy_error(phydev);
Signature · 1 line
2.39.2
author maintainer
#74 lore raw

PATCH 74/286 ice: recycle/free all of the fragments from multi-buffer frame

In reply to Greg Kroah-Hartman

From: Maciej Fijalkowski <[email protected]>

[ Upstream commit abaf8d51b0cedb16af51fb6b2189370d7515977c ]

The ice driver caches next_to_clean value at the beginning of ice_clean_rx_irq() in order to remember the first buffer that has to be freed/recycled after main Rx processing loop. The end boundary is indicated by first descriptor of frame that Rx processing loop has ended its duties. Note that if mentioned loop ended in the middle of gathering multi-buffer frame, next_to_clean would be pointing to the descriptor in the middle of the frame BUT freeing/recycling stage will stop at the first descriptor. This means that next iteration of ice_clean_rx_irq() will miss the (first_desc, next_to_clean - 1) entries.

When running various 9K MTU workloads, such splats were observed:

[  540.780716] BUG: kernel NULL pointer dereference, address: 0000000000000000
[  540.787787] #PF: supervisor read access in kernel mode
[  540.793002] #PF: error_code(0x0000) - not-present page
[  540.798218] PGD 0 P4D 0
[  540.800801] Oops: 0000 [#1] PREEMPT SMP NOPTI
[  540.805231] CPU: 18 PID: 3984 Comm: xskxceiver Tainted: G        W          6.3.0-rc7+ #96
[  540.813619] Hardware name: Intel Corporation S2600WFT/S2600WFT, BIOS SE5C620.86B.02.01.0008.031920191559 03/19/2019
[  540.824209] RIP: 0010:ice_clean_rx_irq+0x2b6/0xf00 [ice]
[  540.829678] Code: 74 24 10 e9 aa 00 00 00 8b 55 78 41 31 57 10 41 09 c4 4d 85 ff 0f 84 83 00 00 00 49 8b 57 08 41 8b 4f 1c 65 8b 35 1a fa 4b 3f <48> 8b 02 48 c1 e8 3a 39 c6 0f 85 a2 00 00 00 f6 42 08 02 0f 85 98
[  540.848717] RSP: 0018:ffffc9000f42fc50 EFLAGS: 00010282
[  540.854029] RAX: 0000000000000004 RBX: 0000000000000002 RCX: 000000000000fffe
[  540.861272] RDX: 0000000000000000 RSI: 0000000000000001 RDI: 00000000ffffffff
[  540.868519] RBP: ffff88984a05ac00 R08: 0000000000000000 R09: dead000000000100
[  540.875760] R10: ffff88983fffcd00 R11: 000000000010f2b8 R12: 0000000000000004
[  540.883008] R13: 0000000000000003 R14: 0000000000000800 R15: ffff889847a10040
[  540.890253] FS:  00007f6ddf7fe640(0000) GS:ffff88afdf800000(0000) knlGS:0000000000000000
[  540.898465] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  540.904299] CR2: 0000000000000000 CR3: 000000010d3da001 CR4: 00000000007706e0
[  540.911542] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  540.918789] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  540.926032] PKRU: 55555554
[  540.928790] Call Trace:
[  540.931276]  <TASK>
[  540.933418]  ice_napi_poll+0x4ca/0x6d0 [ice]
[  540.937804]  ? __pfx_ice_napi_poll+0x10/0x10 [ice]
[  540.942716]  napi_busy_loop+0xd7/0x320
[  540.946537]  xsk_recvmsg+0x143/0x170
[  540.950178]  sock_recvmsg+0x99/0xa0
[  540.953729]  __sys_recvfrom+0xa8/0x120
[  540.957543]  ? do_futex+0xbd/0x1d0
[  540.961008]  ? __x64_sys_futex+0x73/0x1d0
[  540.965083]  __x64_sys_recvfrom+0x20/0x30
[  540.969155]  do_syscall_64+0x38/0x90
[  540.972796]  entry_SYSCALL_64_after_hwframe+0x72/0xdc
[  540.977934] RIP: 0033:0x7f6de5f27934

To fix this, set cached_ntc to first_desc so that at the end, when freeing/recycling buffers, descriptors from first to ntc are not missed.

---
 drivers/net/ethernet/intel/ice/ice_txrx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/intel/ice/ice_txrx.c +1 -1
index 059bd911c51d8..52d0a126eb616 100644
@@ -1152,11 +1152,11 @@ int ice_clean_rx_irq(struct ice_rx_ring *rx_ring, int budget)
 	unsigned int total_rx_bytes = 0, total_rx_pkts = 0;
 	unsigned int offset = rx_ring->rx_offset;
 	struct xdp_buff *xdp = &rx_ring->xdp;
+	u32 cached_ntc = rx_ring->first_desc;
 	struct ice_tx_ring *xdp_ring = NULL;
 	struct bpf_prog *xdp_prog = NULL;
 	u32 ntc = rx_ring->next_to_clean;
 	u32 cnt = rx_ring->count;
-	u32 cached_ntc = ntc;
 	u32 xdp_xmit = 0;
 	u32 cached_ntu;
 	bool failure;
Signature · 1 line
2.39.2
author maintainer
#65 lore raw

PATCH 65/286 iommu/mediatek: Flush IOTLB completely only if domain has been attached

In reply to Greg Kroah-Hartman

From: Chen-Yu Tsai <[email protected]>

[ Upstream commit b3fc95709c54ffbe80f16801e0a792a4d2b3d55e ]

If an IOMMU domain was never attached, it lacks any linkage to the actual IOMMU hardware. Attempting to do flush_iotlb_all() on it will result in a NULL pointer dereference. This seems to happen after the recent IOMMU core rework in v6.4-rc1.

    Unable to handle kernel read from unreadable memory at virtual address 0000000000000018
    Call trace:
     mtk_iommu_flush_iotlb_all+0x20/0x80
     iommu_create_device_direct_mappings.part.0+0x13c/0x230
     iommu_setup_default_domain+0x29c/0x4d0
     iommu_probe_device+0x12c/0x190
     of_iommu_configure+0x140/0x208
     of_dma_configure_id+0x19c/0x3c0
     platform_dma_configure+0x38/0x88
     really_probe+0x78/0x2c0

Check if the "bank" field has been filled in before actually attempting the IOTLB flush to avoid it. The IOTLB is also flushed when the device comes out of runtime suspend, so it should have a clean initial state.

---
 drivers/iommu/mtk_iommu.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/iommu/mtk_iommu.c +2 -1
index 6a00ce208dc2b..248b8c2bc4071 100644
@@ -738,7 +738,8 @@ static void mtk_iommu_flush_iotlb_all(struct iommu_domain *domain)
 {
 	struct mtk_iommu_domain *dom = to_mtk_domain(domain);
 
-	mtk_iommu_tlb_flush_all(dom->bank->parent_data);
+	if (dom->bank)
+		mtk_iommu_tlb_flush_all(dom->bank->parent_data);
 }
 
 static void mtk_iommu_iotlb_sync(struct iommu_domain *domain,
Signature · 1 line
2.39.2
author maintainer
#98 lore raw

PATCH 98/286 hwmon: (k10temp) Add PCI ID for family 19, model 78h

In reply to Greg Kroah-Hartman

From: Mario Limonciello <[email protected]>

[ Upstream commit 7d8accfaa0ab65e4282c8e58950f7d688342cd86 ]

Enable k10temp on this system.

[ bp: Massage. ]

---
 drivers/hwmon/k10temp.c | 1 +
 1 file changed, 1 insertion(+)
drivers/hwmon/k10temp.c +1 -0
index be8bbb1c3a02d..823d0ca1d6059 100644
@@ -507,6 +507,7 @@ static const struct pci_device_id k10temp_id_table[] = {
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_19H_M50H_DF_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_19H_M60H_DF_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_19H_M70H_DF_F3) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_19H_M78H_DF_F3) },
 	{ PCI_VDEVICE(HYGON, PCI_DEVICE_ID_AMD_17H_DF_F3) },
 	{}
 };
Signature · 1 line
2.39.2
author maintainer
#107 lore raw

PATCH 107/286 fbdev: imsttfb: Fix use after free bug in imsttfb_probe

In reply to Greg Kroah-Hartman

From: Zheng Wang <[email protected]>

[ Upstream commit c75f5a55061091030a13fef71b9995b89bc86213 ]

A use-after-free bug may occur if init_imstt invokes framebuffer_release and free the info ptr. The caller, imsttfb_probe didn't notice that and still keep the ptr as private data in pdev.

If we remove the driver which will call imsttfb_remove to make cleanup, UAF happens.

Fix it by return error code if bad case happens in init_imstt.

---
 drivers/video/fbdev/imsttfb.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
drivers/video/fbdev/imsttfb.c +8 -7
index bea45647184e1..975dd682fae4b 100644
@@ -1347,7 +1347,7 @@ static const struct fb_ops imsttfb_ops = {
 	.fb_ioctl 	= imsttfb_ioctl,
 };
 
-static void init_imstt(struct fb_info *info)
+static int init_imstt(struct fb_info *info)
 {
 	struct imstt_par *par = info->par;
 	__u32 i, tmp, *ip, *end;
@@ -1420,7 +1420,7 @@ static void init_imstt(struct fb_info *info)
 	    || !(compute_imstt_regvals(par, info->var.xres, info->var.yres))) {
 		printk("imsttfb: %ux%ux%u not supported\n", info->var.xres, info->var.yres, info->var.bits_per_pixel);
 		framebuffer_release(info);
-		return;
+		return -ENODEV;
 	}
 
 	sprintf(info->fix.id, "IMS TT (%s)", par->ramdac == IBM ? "IBM" : "TVP");
@@ -1456,12 +1456,13 @@ static void init_imstt(struct fb_info *info)
 
 	if (register_framebuffer(info) < 0) {
 		framebuffer_release(info);
-		return;
+		return -ENODEV;
 	}
 
 	tmp = (read_reg_le32(par->dc_regs, SSTATUS) & 0x0f00) >> 8;
 	fb_info(info, "%s frame buffer; %uMB vram; chip version %u\n",
 		info->fix.id, info->fix.smem_len >> 20, tmp);
+	return 0;
 }
 
 static int imsttfb_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
@@ -1529,10 +1530,10 @@ static int imsttfb_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (!par->cmap_regs)
 		goto error;
 	info->pseudo_palette = par->palette;
-	init_imstt(info);
-
-	pci_set_drvdata(pdev, info);
-	return 0;
+	ret = init_imstt(info);
+	if (!ret)
+		pci_set_drvdata(pdev, info);
+	return ret;
 
 error:
 	if (par->dc_regs)
Signature · 1 line
2.39.2
author maintainer
#108 lore raw

PATCH 108/286 fbdev: modedb: Add 1920x1080 at 60 Hz video mode

In reply to Greg Kroah-Hartman

From: Helge Deller <[email protected]>

[ Upstream commit c8902258b2b8ecaa1b8d88c312853c5b14c2553d ]

Add typical resolution for Full-HD monitors.

---
 drivers/video/fbdev/core/modedb.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/video/fbdev/core/modedb.c +5 -0
index 6473e0dfe1464..e78ec7f728463 100644
@@ -257,6 +257,11 @@ static const struct fb_videomode modedb[] = {
 	{ NULL, 72, 480, 300, 33386, 40, 24, 11, 19, 80, 3, 0,
 		FB_VMODE_DOUBLE },
 
+	/* 1920x1080 @ 60 Hz, 67.3 kHz hsync */
+	{ NULL, 60, 1920, 1080, 6734, 148, 88, 36, 4, 44, 5, 0,
+		FB_SYNC_HOR_HIGH_ACT | FB_SYNC_VERT_HIGH_ACT,
+		FB_VMODE_NONINTERLACED },
+
 	/* 1920x1200 @ 60 Hz, 74.5 Khz hsync */
 	{ NULL, 60, 1920, 1200, 5177, 128, 336, 1, 38, 208, 3,
 		FB_SYNC_HOR_HIGH_ACT | FB_SYNC_VERT_HIGH_ACT,
Signature · 1 line
2.39.2
author maintainer
#109 lore raw

PATCH 109/286 fbdev: stifb: Fix info entry in sti_struct on error path

In reply to Greg Kroah-Hartman

From: Helge Deller <[email protected]>

[ Upstream commit 0bdf1ad8d10bd4e50a8b1a2c53d15984165f7fea ]

Minor fix to reset the info field to NULL in case of error.

---
 drivers/video/fbdev/stifb.c | 1 +
 1 file changed, 1 insertion(+)
drivers/video/fbdev/stifb.c +1 -0
index ef8a4c5fc6875..63f51783352dc 100644
@@ -1413,6 +1413,7 @@ static int __init stifb_init_fb(struct sti_struct *sti, int bpp_pref)
 	iounmap(info->screen_base);
 out_err0:
 	kfree(fb);
+	sti->info = NULL;
 	return -ENXIO;
 }
 
Signature · 1 line
2.39.2
author maintainer
#110 lore raw

PATCH 110/286 nbd: Fix debugfs_create_dir error checking

In reply to Greg Kroah-Hartman

From: Ivan Orlov <[email protected]>

[ Upstream commit 4913cfcf014c95f0437db2df1734472fd3e15098 ]

The debugfs_create_dir function returns ERR_PTR in case of error, and the only correct way to check if an error occurred is 'IS_ERR' inline function. This patch will replace the null-comparison with IS_ERR.

---
 drivers/block/nbd.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/block/nbd.c +2 -2
index e1c954094b6c0..dd0adcf745ff5 100644
@@ -1666,7 +1666,7 @@ static int nbd_dev_dbg_init(struct nbd_device *nbd)
 		return -EIO;
 
 	dir = debugfs_create_dir(nbd_name(nbd), nbd_dbg_dir);
-	if (!dir) {
+	if (IS_ERR(dir)) {
 		dev_err(nbd_to_dev(nbd), "Failed to create debugfs dir for '%s'\n",
 			nbd_name(nbd));
 		return -EIO;
@@ -1692,7 +1692,7 @@ static int nbd_dbg_init(void)
 	struct dentry *dbg_dir;
 
 	dbg_dir = debugfs_create_dir("nbd", NULL);
-	if (!dbg_dir)
+	if (IS_ERR(dbg_dir))
 		return -EIO;
 
 	nbd_dbg_dir = dbg_dir;
Signature · 1 line
2.39.2
author maintainer
#111 lore raw

PATCH 111/286 block/rnbd: replace REQ_OP_FLUSH with REQ_OP_WRITE

In reply to Greg Kroah-Hartman

From: Guoqing Jiang <[email protected]>

[ Upstream commit 5e6e08087a4acb4ee3574cea32dbff0f63c7f608 ]

Since flush bios are implemented as writes with no data and the preflush flag per Christoph's comment [1].

And we need to change it in rnbd accordingly. Otherwise, I got splatting when create fs from rnbd client.

[  464.028545] ------------[ cut here ]------------
[  464.028553] WARNING: CPU: 0 PID: 65 at block/blk-core.c:751 submit_bio_noacct+0x32c/0x5d0
[ ... ]
[  464.028668] CPU: 0 PID: 65 Comm: kworker/0:1H Tainted: G           OE      6.4.0-rc1 #9
[  464.028671] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.15.0-0-g2dd4b9b-rebuilt.opensuse.org 04/01/2014
[  464.028673] Workqueue: ib-comp-wq ib_cq_poll_work [ib_core]
[  464.028717] RIP: 0010:submit_bio_noacct+0x32c/0x5d0
[  464.028720] Code: 03 0f 85 51 fe ff ff 48 8b 43 18 8b 88 04 03 00 00 85 c9 0f 85 3f fe ff ff e9 be fd ff ff 0f b6 d0 3c 0d 74 26 83 fa 01 74 21 <0f> 0b b8 0a 00 00 00 e9 56 fd ff ff 4c 89 e7 e8 70 a1 03 00 84 c0
[  464.028722] RSP: 0018:ffffaf3680b57c68 EFLAGS: 00010202
[  464.028724] RAX: 0000000000060802 RBX: ffffa09dcc18bf00 RCX: 0000000000000000
[  464.028726] RDX: 0000000000000002 RSI: 0000000000000000 RDI: ffffa09dde081d00
[  464.028727] RBP: ffffaf3680b57c98 R08: ffffa09dde081d00 R09: ffffa09e38327200
[  464.028729] R10: 0000000000000000 R11: 0000000000000000 R12: ffffa09dde081d00
[  464.028730] R13: ffffa09dcb06e1e8 R14: 0000000000000000 R15: 0000000000200000
[  464.028733] FS:  0000000000000000(0000) GS:ffffa09e3bc00000(0000) knlGS:0000000000000000
[  464.028735] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  464.028736] CR2: 000055a4e8206c40 CR3: 0000000119f06000 CR4: 00000000003506f0
[  464.028738] Call Trace:
[  464.028740]  <TASK>
[  464.028746]  submit_bio+0x1b/0x80
[  464.028748]  rnbd_srv_rdma_ev+0x50d/0x10c0 [rnbd_server]
[  464.028754]  ? percpu_ref_get_many.constprop.0+0x55/0x140 [rtrs_server]
[  464.028760]  ? __this_cpu_preempt_check+0x13/0x20
[  464.028769]  process_io_req+0x1dc/0x450 [rtrs_server]
[  464.028775]  rtrs_srv_inv_rkey_done+0x67/0xb0 [rtrs_server]
[  464.028780]  __ib_process_cq+0xbc/0x1f0 [ib_core]
[  464.028793]  ib_cq_poll_work+0x2b/0xa0 [ib_core]
[  464.028804]  process_one_work+0x2a9/0x580

[1]. https://lore.kernel.org/all/[email protected]/

---
 drivers/block/rnbd/rnbd-proto.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/block/rnbd/rnbd-proto.h +1 -1
index ea7ac8bca63cf..da1d0542d7e2c 100644
@@ -241,7 +241,7 @@ static inline blk_opf_t rnbd_to_bio_flags(u32 rnbd_opf)
 		bio_opf = REQ_OP_WRITE;
 		break;
 	case RNBD_OP_FLUSH:
-		bio_opf = REQ_OP_FLUSH | REQ_PREFLUSH;
+		bio_opf = REQ_OP_WRITE | REQ_PREFLUSH;
 		break;
 	case RNBD_OP_DISCARD:
 		bio_opf = REQ_OP_DISCARD;
Signature · 1 line
2.39.2
author maintainer
#112 lore raw

PATCH 112/286 nvme-pci: add NVME_QUIRK_BOGUS_NID for HS-SSD-FUTURE 2048G

In reply to Greg Kroah-Hartman

From: Sagi Grimberg <[email protected]>

[ Upstream commit 1616d6c3717bae9041a4240d381ec56ccdaafedc ]

Add a quirk to fix HS-SSD-FUTURE 2048G SSD drives reporting duplicate nsids.

---
 drivers/nvme/host/pci.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/nvme/host/pci.c +2 -0
index cd7873de31215..bbf96567365cd 100644
@@ -3445,6 +3445,8 @@ static const struct pci_device_id nvme_id_table[] = {
 				NVME_QUIRK_IGNORE_DEV_SUBNQN, },
 	{ PCI_DEVICE(0x10ec, 0x5763), /* TEAMGROUP T-FORCE CARDEA ZERO Z330 SSD */
 		.driver_data = NVME_QUIRK_BOGUS_NID, },
+	{ PCI_DEVICE(0x1e4b, 0x1602), /* HS-SSD-FUTURE 2048G  */
+		.driver_data = NVME_QUIRK_BOGUS_NID, },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMAZON, 0x0061),
 		.driver_data = NVME_QUIRK_DMA_ADDRESS_BITS_48, },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMAZON, 0x0065),
Signature · 1 line
2.39.2
author maintainer
#113 lore raw

PATCH 113/286 nvme-pci: add quirk for missing secondary temperature thresholds

In reply to Greg Kroah-Hartman

From: Hristo Venev <[email protected]>

[ Upstream commit bd375feeaf3408ed00e08c3bc918d6be15f691ad ]

On Kingston KC3000 and Kingston FURY Renegade (both have the same PCI IDs) accessing temp3_{min,max} fails with an invalid field error (note that there is no problem setting the thresholds for temp1).

This contradicts the NVM Express Base Specification 2.0b, page 292:

The over temperature threshold and under temperature threshold features shall be implemented for all implemented temperature sensors (i.e., all Temperature Sensor fields that report a non-zero value).

Define NVME_QUIRK_NO_SECONDARY_TEMP_THRESH that disables the thresholds for all but the composite temperature and set it for this device.

---
 drivers/nvme/host/hwmon.c | 4 +++-
 drivers/nvme/host/nvme.h  | 5 +++++
 drivers/nvme/host/pci.c   | 2 ++
 3 files changed, 10 insertions(+), 1 deletion(-)
drivers/nvme/host/hwmon.c +3 -1
index 9e6e56c20ec99..316f3e4ca7cc6 100644
@@ -163,7 +163,9 @@ static umode_t nvme_hwmon_is_visible(const void *_data,
 	case hwmon_temp_max:
 	case hwmon_temp_min:
 		if ((!channel && data->ctrl->wctemp) ||
-		    (channel && data->log->temp_sensor[channel - 1])) {
+		    (channel && data->log->temp_sensor[channel - 1] &&
+		     !(data->ctrl->quirks &
+		       NVME_QUIRK_NO_SECONDARY_TEMP_THRESH))) {
 			if (data->ctrl->quirks &
 			    NVME_QUIRK_NO_TEMP_THRESH_CHANGE)
 				return 0444;
drivers/nvme/host/nvme.h +5 -0
index bf46f122e9e1e..a2d4f59e0535a 100644
@@ -149,6 +149,11 @@ enum nvme_quirks {
 	 * Reports garbage in the namespace identifiers (eui64, nguid, uuid).
 	 */
 	NVME_QUIRK_BOGUS_NID			= (1 << 18),
+
+	/*
+	 * No temperature thresholds for channels other than 0 (Composite).
+	 */
+	NVME_QUIRK_NO_SECONDARY_TEMP_THRESH	= (1 << 19),
 };
 
 /*
drivers/nvme/host/pci.c +2 -0
index bbf96567365cd..a7772c0194d5a 100644
@@ -3406,6 +3406,8 @@ static const struct pci_device_id nvme_id_table[] = {
 		.driver_data = NVME_QUIRK_NO_DEEPEST_PS, },
 	{ PCI_DEVICE(0x2646, 0x2263),   /* KINGSTON A2000 NVMe SSD  */
 		.driver_data = NVME_QUIRK_NO_DEEPEST_PS, },
+	{ PCI_DEVICE(0x2646, 0x5013),   /* Kingston KC3000, Kingston FURY Renegade */
+		.driver_data = NVME_QUIRK_NO_SECONDARY_TEMP_THRESH, },
 	{ PCI_DEVICE(0x2646, 0x5018),   /* KINGSTON OM8SFP4xxxxP OS21012 NVMe SSD */
 		.driver_data = NVME_QUIRK_DISABLE_WRITE_ZEROES, },
 	{ PCI_DEVICE(0x2646, 0x5016),   /* KINGSTON OM3PGP4xxxxP OS21011 NVMe SSD */
Signature · 1 line
2.39.2
author maintainer
#114 lore raw

PATCH 114/286 nvme-pci: clamp max_hw_sectors based on DMA optimized limitation

In reply to Greg Kroah-Hartman

From: Adrian Huang <[email protected]>

[ Upstream commit 3710e2b056cb92ad816e4d79fa54a6a5b6ad8cbd ]

When running the fio test on a 448-core AMD server + a NVME disk, a soft lockup or a hard lockup call trace is shown:

[soft lockup]

watchdog: BUG: soft lockup - CPU#126 stuck for 23s! [swapper/126:0]
RIP: 0010:_raw_spin_unlock_irqrestore+0x21/0x50
...
Call Trace:
 <IRQ>
 fq_flush_timeout+0x7d/0xd0
 ? __pfx_fq_flush_timeout+0x10/0x10
 call_timer_fn+0x2e/0x150
 run_timer_softirq+0x48a/0x560
 ? __pfx_fq_flush_timeout+0x10/0x10
 ? clockevents_program_event+0xaf/0x130
 __do_softirq+0xf1/0x335
 irq_exit_rcu+0x9f/0xd0
 sysvec_apic_timer_interrupt+0xb4/0xd0
 </IRQ>
 <TASK>
 asm_sysvec_apic_timer_interrupt+0x1f/0x30

...

Obvisouly, fq_flush_timeout spends over 20 seconds. Here is ftrace log:

               |  fq_flush_timeout() {
               |    fq_ring_free() {
               |      put_pages_list() {
   0.170 us    |        free_unref_page_list();
   0.810 us    |      }
               |      free_iova_fast() {
               |        free_iova() {
 * 85622.66 us |          _raw_spin_lock_irqsave();
   2.860 us    |          remove_iova();
   0.600 us    |          _raw_spin_unlock_irqrestore();
   0.470 us    |          lock_info_report();
   2.420 us    |          free_iova_mem.part.0();
 * 85638.27 us |        }
 * 85638.84 us |      }
               |      put_pages_list() {
   0.230 us    |        free_unref_page_list();
   0.470 us    |      }
   ...            ...
 $ 31017069 us |  }

Most of cores are under lock contention for acquiring iova_rbtree_lock due to the iova flush queue mechanism.

[hard lockup]
NMI watchdog: Watchdog detected hard LOCKUP on cpu 351

RIP: 0010:native_queued_spin_lock_slowpath+0x2d8/0x330

Call Trace:
 <IRQ>
 _raw_spin_lock_irqsave+0x4f/0x60
 free_iova+0x27/0xd0
 free_iova_fast+0x4d/0x1d0
 fq_ring_free+0x9b/0x150
 iommu_dma_free_iova+0xb4/0x2e0
 __iommu_dma_unmap+0x10b/0x140
 iommu_dma_unmap_sg+0x90/0x110
 dma_unmap_sg_attrs+0x4a/0x50
 nvme_unmap_data+0x5d/0x120 [nvme]
 nvme_pci_complete_batch+0x77/0xc0 [nvme]
 nvme_irq+0x2ee/0x350 [nvme]
 ? __pfx_nvme_pci_complete_batch+0x10/0x10 [nvme]
 __handle_irq_event_percpu+0x53/0x1a0
 handle_irq_event_percpu+0x19/0x60
 handle_irq_event+0x3d/0x60
 handle_edge_irq+0xb3/0x210
 __common_interrupt+0x7f/0x150
 common_interrupt+0xc5/0xf0
 </IRQ>
 <TASK>
 asm_common_interrupt+0x2b/0x40

...

ftrace shows fq_ring_free spends over 10 seconds [1]. Again, most of cores are under lock contention for acquiring iova_rbtree_lock due to the iova flush queue mechanism.

[Root Cause]
The root cause is that the max_hw_sectors_kb of nvme disk (mdts=10) is 4096kb, which streaming DMA mappings cannot benefit from the scalable IOVA mechanism introduced by the commit 9257b4a206fc ("iommu/iova: introduce per-cpu caching to iova allocation") if the length is greater than 128kb.

To fix the lock contention issue, clamp max_hw_sectors based on DMA optimized limitation in order to leverage scalable IOVA mechanism.

Note: The issue does not happen with another NVME disk (mdts = 5 and max_hw_sectors_kb = 128)

[1] https://gist.github.com/AdrianHuang/bf8ec7338204837631fbdaed25d19cc4

---
 drivers/nvme/host/pci.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/nvme/host/pci.c +1 -1
index a7772c0194d5a..a389f1ea0b151 100644
@@ -2960,7 +2960,7 @@ static struct nvme_dev *nvme_pci_alloc_dev(struct pci_dev *pdev,
 	 * over a single page.
 	 */
 	dev->ctrl.max_hw_sectors = min_t(u32,
-		NVME_MAX_KB_SZ << 1, dma_max_mapping_size(&pdev->dev) >> 9);
+		NVME_MAX_KB_SZ << 1, dma_opt_mapping_size(&pdev->dev) >> 9);
 	dev->ctrl.max_segments = NVME_MAX_SEGS;
 
 	/*
Signature · 1 line
2.39.2
author maintainer
#115 lore raw

PATCH 115/286 ASoC: amd: yc: Add DMI entry to support System76 Pangolin 12

In reply to Greg Kroah-Hartman

From: Jeremy Soller <[email protected]>

[ Upstream commit 7b9891ad25246b18b5ccc19518da7abc7763aa0a ]

Add pang12 quirk to enable the internal microphone.

---
 sound/soc/amd/yc/acp6x-mach.c | 7 +++++++
 1 file changed, 7 insertions(+)
sound/soc/amd/yc/acp6x-mach.c +7 -0
index b9958e5553674..84b401b685f7f 100644
@@ -297,6 +297,13 @@ static const struct dmi_system_id yc_acp_quirk_table[] = {
 			DMI_MATCH(DMI_BOARD_NAME, "8A22"),
 		}
 	},
+	{
+		.driver_data = &acp6x_card,
+		.matches = {
+			DMI_MATCH(DMI_BOARD_VENDOR, "System76"),
+			DMI_MATCH(DMI_PRODUCT_VERSION, "pang12"),
+		}
+	},
 	{}
 };
 
Signature · 1 line
2.39.2
author maintainer
#116 lore raw

PATCH 116/286 ASoC: dwc: limit the number of overrun messages

In reply to Greg Kroah-Hartman

From: Maxim Kochetkov <[email protected]>

[ Upstream commit ab6ecfbf40fccf74b6ec2ba7ed6dd2fc024c3af2 ]

On slow CPU (FPGA/QEMU emulated) printing overrun messages from interrupt handler to uart console may leads to more overrun errors. So use dev_err_ratelimited to limit the number of error messages.

---
 sound/soc/dwc/dwc-i2s.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
sound/soc/dwc/dwc-i2s.c +2 -2
index 7f7dd07c63b2f..3496301582b22 100644
@@ -132,13 +132,13 @@ static irqreturn_t i2s_irq_handler(int irq, void *dev_id)
 
 		/* Error Handling: TX */
 		if (isr[i] & ISR_TXFO) {
-			dev_err(dev->dev, "TX overrun (ch_id=%d)\n", i);
+			dev_err_ratelimited(dev->dev, "TX overrun (ch_id=%d)\n", i);
 			irq_valid = true;
 		}
 
 		/* Error Handling: TX */
 		if (isr[i] & ISR_RXFO) {
-			dev_err(dev->dev, "RX overrun (ch_id=%d)\n", i);
+			dev_err_ratelimited(dev->dev, "RX overrun (ch_id=%d)\n", i);
 			irq_valid = true;
 		}
 	}
Signature · 1 line
2.39.2
author maintainer
#99 lore raw

PATCH 99/286 media: rcar-vin: Gen3 can not scale NV12

In reply to Greg Kroah-Hartman

From: Niklas Söderlund <[email protected]>

[ Upstream commit 879c5a458e532b95783ce27f704d1b21573066f7 ]

The VIN modules on Gen3 can not scale NV12, fail format validation if the user tries. Currently no frames are produced if this is attempted.

---
 drivers/media/platform/renesas/rcar-vin/rcar-dma.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/media/platform/renesas/rcar-vin/rcar-dma.c +5 -0
index 98bfd445a649b..cc6b59e5621ae 100644
@@ -1312,6 +1312,11 @@ static int rvin_mc_validate_format(struct rvin_dev *vin, struct v4l2_subdev *sd,
 	}
 
 	if (rvin_scaler_needed(vin)) {
+		/* Gen3 can't scale NV12 */
+		if (vin->info->model == RCAR_GEN3 &&
+		    vin->format.pixelformat == V4L2_PIX_FMT_NV12)
+			return -EPIPE;
+
 		if (!vin->scaler)
 			return -EPIPE;
 	} else {
Signature · 1 line
2.39.2
author maintainer
#117 lore raw

PATCH 117/286 cpupower:Fix resource leaks in sysfs_get_enabled()

In reply to Greg Kroah-Hartman

From: Hao Zeng <[email protected]>

[ Upstream commit e652be0f59d4ba4d5c636b1f7f4dcb73aae049fa ]

The sysfs_get_enabled() opened file processor not closed, may cause a file handle leak.
Putting error handling and resource cleanup code together makes the code easy to maintain and read.
Removed the unnecessary else if branch from the original function, as it should return an error in cases other than '0'.

---
 tools/power/cpupower/lib/powercap.c | 23 ++++++++++++++++-------
 1 file changed, 16 insertions(+), 7 deletions(-)
tools/power/cpupower/lib/powercap.c +16 -7
index 0ce29ee4c2e46..a7a59c6bacda8 100644
@@ -40,25 +40,34 @@ static int sysfs_get_enabled(char *path, int *mode)
 {
 	int fd;
 	char yes_no;
+	int ret = 0;
 
 	*mode = 0;
 
 	fd = open(path, O_RDONLY);
-	if (fd == -1)
-		return -1;
+	if (fd == -1) {
+		ret = -1;
+		goto out;
+	}
 
 	if (read(fd, &yes_no, 1) != 1) {
-		close(fd);
-		return -1;
+		ret = -1;
+		goto out_close;
 	}
 
 	if (yes_no == '1') {
 		*mode = 1;
-		return 0;
+		goto out_close;
 	} else if (yes_no == '0') {
-		return 0;
+		goto out_close;
+	} else {
+		ret = -1;
+		goto out_close;
 	}
-	return -1;
+out_close:
+	close(fd);
+out:
+	return ret;
 }
 
 int powercap_get_enabled(int *mode)
Signature · 1 line
2.39.2
author maintainer
#118 lore raw

PATCH 118/286 ASoC: SOF: amd: Fix NULL pointer crash in acp_sof_ipc_msg_data function

In reply to Greg Kroah-Hartman

From: V sujith kumar Reddy <[email protected]>

[ Upstream commit 051d71e073614a72ad423d6dacba37a7eeff274d ]

Check substream and runtime variables before assigning.

---
 sound/soc/sof/amd/acp-ipc.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
sound/soc/sof/amd/acp-ipc.c +6 -1
index 4e0c48a361599..749e856dc6011 100644
@@ -209,7 +209,12 @@ int acp_sof_ipc_msg_data(struct snd_sof_dev *sdev, struct snd_sof_pcm_stream *sp
 		acp_mailbox_read(sdev, offset, p, sz);
 	} else {
 		struct snd_pcm_substream *substream = sps->substream;
-		struct acp_dsp_stream *stream = substream->runtime->private_data;
+		struct acp_dsp_stream *stream;
+
+		if (!substream || !substream->runtime)
+			return -ESTRPIPE;
+
+		stream = substream->runtime->private_data;
 
 		if (!stream)
 			return -ESTRPIPE;
Signature · 1 line
2.39.2
author maintainer
#119 lore raw

PATCH 119/286 um: harddog: fix modular build

In reply to Greg Kroah-Hartman

From: Johannes Berg <[email protected]>

[ Upstream commit 73a23d7710331a530e972903318528b75e5a5f58 ]

Since we no longer (want to) export any libc symbols the _user portions of any drivers need to be built into image rather than the module. I missed this for the watchdog. Fix the watchdog accordingly.

---
 arch/um/drivers/Makefile           | 4 +++-
 arch/um/drivers/harddog.h          | 9 +++++++++
 arch/um/drivers/harddog_kern.c     | 7 +------
 arch/um/drivers/harddog_user.c     | 1 +
 arch/um/drivers/harddog_user_exp.c | 9 +++++++++
 5 files changed, 23 insertions(+), 7 deletions(-)
 create mode 100644 arch/um/drivers/harddog.h
 create mode 100644 arch/um/drivers/harddog_user_exp.c
arch/um/drivers/Makefile +3 -1
index dee6f66353b33..a461a950f0518 100644
@@ -16,7 +16,8 @@ mconsole-objs := mconsole_kern.o mconsole_user.o
 hostaudio-objs := hostaudio_kern.o
 ubd-objs := ubd_kern.o ubd_user.o
 port-objs := port_kern.o port_user.o
-harddog-objs := harddog_kern.o harddog_user.o
+harddog-objs := harddog_kern.o
+harddog-builtin-$(CONFIG_UML_WATCHDOG) := harddog_user.o harddog_user_exp.o
 rtc-objs := rtc_kern.o rtc_user.o
 
 LDFLAGS_pcap.o = $(shell $(CC) $(KBUILD_CFLAGS) -print-file-name=libpcap.a)
@@ -60,6 +61,7 @@ obj-$(CONFIG_PTY_CHAN) += pty.o
 obj-$(CONFIG_TTY_CHAN) += tty.o 
 obj-$(CONFIG_XTERM_CHAN) += xterm.o xterm_kern.o
 obj-$(CONFIG_UML_WATCHDOG) += harddog.o
+obj-y += $(harddog-builtin-y) $(harddog-builtin-m)
 obj-$(CONFIG_BLK_DEV_COW_COMMON) += cow_user.o
 obj-$(CONFIG_UML_RANDOM) += random.o
 obj-$(CONFIG_VIRTIO_UML) += virtio_uml.o
arch/um/drivers/harddog.h new file +9 -0
new file mode 100644
index 0000000000000..6d9ea60e7133e
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef UM_WATCHDOG_H
+#define UM_WATCHDOG_H
+
+int start_watchdog(int *in_fd_ret, int *out_fd_ret, char *sock);
+void stop_watchdog(int in_fd, int out_fd);
+int ping_watchdog(int fd);
+
+#endif /* UM_WATCHDOG_H */
arch/um/drivers/harddog_kern.c +1 -6
index e6d4f43deba82..60d1c6cab8a95 100644
@@ -47,6 +47,7 @@
 #include <linux/spinlock.h>
 #include <linux/uaccess.h>
 #include "mconsole.h"
+#include "harddog.h"
 
 MODULE_LICENSE("GPL");
 
@@ -60,8 +61,6 @@ static int harddog_out_fd = -1;
  *	Allow only one person to hold it open
  */
 
-extern int start_watchdog(int *in_fd_ret, int *out_fd_ret, char *sock);
-
 static int harddog_open(struct inode *inode, struct file *file)
 {
 	int err = -EBUSY;
@@ -92,8 +91,6 @@ static int harddog_open(struct inode *inode, struct file *file)
 	return err;
 }
 
-extern void stop_watchdog(int in_fd, int out_fd);
-
 static int harddog_release(struct inode *inode, struct file *file)
 {
 	/*
@@ -112,8 +109,6 @@ static int harddog_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-extern int ping_watchdog(int fd);
-
 static ssize_t harddog_write(struct file *file, const char __user *data, size_t len,
 			     loff_t *ppos)
 {
arch/um/drivers/harddog_user.c +1 -0
index 070468d22e394..9ed89304975ed 100644
@@ -7,6 +7,7 @@
 #include <unistd.h>
 #include <errno.h>
 #include <os.h>
+#include "harddog.h"
 
 struct dog_data {
 	int stdin_fd;
arch/um/drivers/harddog_user_exp.c new file +9 -0
new file mode 100644
index 0000000000000..c74d4b815d143
@@ -0,0 +1,9 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/export.h>
+#include "harddog.h"
+
+#if IS_MODULE(CONFIG_UML_WATCHDOG)
+EXPORT_SYMBOL(start_watchdog);
+EXPORT_SYMBOL(stop_watchdog);
+EXPORT_SYMBOL(ping_watchdog);
+#endif
Signature · 1 line
2.39.2
author maintainer
#120 lore raw

PATCH 120/286 xfrm: Check if_id in inbound policy/secpath match

In reply to Greg Kroah-Hartman

From: Benedict Wong <[email protected]>

[ Upstream commit 8680407b6f8f5fba59e8f1d63c869abc280f04df ]

This change ensures that if configured in the policy, the if_id set in the policy and secpath states match during the inbound policy check. Without this, there is potential for ambiguity where entries in the secpath differing by only the if_id could be mismatched.

Notably, this is checked in the outbound direction when resolving templates to SAs, but not on the inbound path when matching SAs and policies.

Test: Tested against Android kernel unit tests & CTS

---
 net/xfrm/xfrm_policy.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
net/xfrm/xfrm_policy.c +6 -5
index 21a3a1cd3d6de..6d15788b51231 100644
@@ -3312,7 +3312,7 @@ xfrm_secpath_reject(int idx, struct sk_buff *skb, const struct flowi *fl)
 
 static inline int
 xfrm_state_ok(const struct xfrm_tmpl *tmpl, const struct xfrm_state *x,
-	      unsigned short family)
+	      unsigned short family, u32 if_id)
 {
 	if (xfrm_state_kern(x))
 		return tmpl->optional && !xfrm_state_addr_cmp(tmpl, x, tmpl->encap_family);
@@ -3323,7 +3323,8 @@ xfrm_state_ok(const struct xfrm_tmpl *tmpl, const struct xfrm_state *x,
 		(tmpl->allalgs || (tmpl->aalgos & (1<<x->props.aalgo)) ||
 		 !(xfrm_id_proto_match(tmpl->id.proto, IPSEC_PROTO_ANY))) &&
 		!(x->props.mode != XFRM_MODE_TRANSPORT &&
-		  xfrm_state_addr_cmp(tmpl, x, family));
+		  xfrm_state_addr_cmp(tmpl, x, family)) &&
+		(if_id == 0 || if_id == x->if_id);
 }
 
 /*
@@ -3335,7 +3336,7 @@ xfrm_state_ok(const struct xfrm_tmpl *tmpl, const struct xfrm_state *x,
  */
 static inline int
 xfrm_policy_ok(const struct xfrm_tmpl *tmpl, const struct sec_path *sp, int start,
-	       unsigned short family)
+	       unsigned short family, u32 if_id)
 {
 	int idx = start;
 
@@ -3345,7 +3346,7 @@ xfrm_policy_ok(const struct xfrm_tmpl *tmpl, const struct sec_path *sp, int star
 	} else
 		start = -1;
 	for (; idx < sp->len; idx++) {
-		if (xfrm_state_ok(tmpl, sp->xvec[idx], family))
+		if (xfrm_state_ok(tmpl, sp->xvec[idx], family, if_id))
 			return ++idx;
 		if (sp->xvec[idx]->props.mode != XFRM_MODE_TRANSPORT) {
 			if (start == -1)
@@ -3724,7 +3725,7 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 		 * are implied between each two transformations.
 		 */
 		for (i = xfrm_nr-1, k = 0; i >= 0; i--) {
-			k = xfrm_policy_ok(tpp[i], sp, k, family);
+			k = xfrm_policy_ok(tpp[i], sp, k, family, if_id);
 			if (k < 0) {
 				if (k < -1)
 					/* "-2 - errored_index" returned */
Signature · 1 line
2.39.2
author maintainer
#121 lore raw

PATCH 121/286 ASoC: jz4740-i2s: Make I2S divider calculations more robust

In reply to Greg Kroah-Hartman

From: Aidan MacDonald <[email protected]>

[ Upstream commit ad721bc919edfd8b4b06977458a412011e2f0c50 ]

When the CPU supplies bit/frame clocks, the system clock (clk_i2s) is divided to produce the bit clock. This is a simple 1/N divider with a fairly limited range, so for a given system clock frequency only a few sample rates can be produced. Usually a wider range of sample rates is supported by varying the system clock frequency.

The old calculation method was not very robust and could easily produce the wrong clock rate, especially with non-standard rates. For example, if the system clock is 1.99x the target bit clock rate, the divider would be calculated as 1 instead of the more accurate 2.

Instead, use a more accurate method that considers two adjacent divider settings and selects the one that produces the least error versus the requested rate. If the error is 5% or higher then the rate setting is rejected to prevent garbled audio.

Skip divider calculation when the codec is supplying both the bit and frame clock; in that case, the divider outputs are unused and we don't want to constrain the sample rate.

---
 sound/soc/jz4740/jz4740-i2s.c | 54 ++++++++++++++++++++++++++++++++---
 1 file changed, 50 insertions(+), 4 deletions(-)
sound/soc/jz4740/jz4740-i2s.c +50 -4
index 6d9cfe0a50411..d0f6c945d9aee 100644
@@ -218,18 +218,48 @@ static int jz4740_i2s_set_fmt(struct snd_soc_dai *dai, unsigned int fmt)
 	return 0;
 }
 
+static int jz4740_i2s_get_i2sdiv(unsigned long mclk, unsigned long rate,
+				 unsigned long i2sdiv_max)
+{
+	unsigned long div, rate1, rate2, err1, err2;
+
+	div = mclk / (64 * rate);
+	if (div == 0)
+		div = 1;
+
+	rate1 = mclk / (64 * div);
+	rate2 = mclk / (64 * (div + 1));
+
+	err1 = abs(rate1 - rate);
+	err2 = abs(rate2 - rate);
+
+	/*
+	 * Choose the divider that produces the smallest error in the
+	 * output rate and reject dividers with a 5% or higher error.
+	 * In the event that both dividers are outside the acceptable
+	 * error margin, reject the rate to prevent distorted audio.
+	 * (The number 5% is arbitrary.)
+	 */
+	if (div <= i2sdiv_max && err1 <= err2 && err1 < rate/20)
+		return div;
+	if (div < i2sdiv_max && err2 < rate/20)
+		return div + 1;
+
+	return -EINVAL;
+}
+
 static int jz4740_i2s_hw_params(struct snd_pcm_substream *substream,
 	struct snd_pcm_hw_params *params, struct snd_soc_dai *dai)
 {
 	struct jz4740_i2s *i2s = snd_soc_dai_get_drvdata(dai);
 	struct regmap_field *div_field;
+	unsigned long i2sdiv_max;
 	unsigned int sample_size;
-	uint32_t ctrl;
-	int div;
+	uint32_t ctrl, conf;
+	int div = 1;
 
 	regmap_read(i2s->regmap, JZ_REG_AIC_CTRL, &ctrl);
-
-	div = clk_get_rate(i2s->clk_i2s) / (64 * params_rate(params));
+	regmap_read(i2s->regmap, JZ_REG_AIC_CONF, &conf);
 
 	switch (params_format(params)) {
 	case SNDRV_PCM_FORMAT_S8:
@@ -258,11 +288,27 @@ static int jz4740_i2s_hw_params(struct snd_pcm_substream *substream,
 			ctrl &= ~JZ_AIC_CTRL_MONO_TO_STEREO;
 
 		div_field = i2s->field_i2sdiv_playback;
+		i2sdiv_max = GENMASK(i2s->soc_info->field_i2sdiv_playback.msb,
+				     i2s->soc_info->field_i2sdiv_playback.lsb);
 	} else {
 		ctrl &= ~JZ_AIC_CTRL_INPUT_SAMPLE_SIZE;
 		ctrl |= FIELD_PREP(JZ_AIC_CTRL_INPUT_SAMPLE_SIZE, sample_size);
 
 		div_field = i2s->field_i2sdiv_capture;
+		i2sdiv_max = GENMASK(i2s->soc_info->field_i2sdiv_capture.msb,
+				     i2s->soc_info->field_i2sdiv_capture.lsb);
+	}
+
+	/*
+	 * Only calculate I2SDIV if we're supplying the bit or frame clock.
+	 * If the codec is supplying both clocks then the divider output is
+	 * unused, and we don't want it to limit the allowed sample rates.
+	 */
+	if (conf & (JZ_AIC_CONF_BIT_CLK_MASTER | JZ_AIC_CONF_SYNC_CLK_MASTER)) {
+		div = jz4740_i2s_get_i2sdiv(clk_get_rate(i2s->clk_i2s),
+					    params_rate(params), i2sdiv_max);
+		if (div < 0)
+			return div;
 	}
 
 	regmap_write(i2s->regmap, JZ_REG_AIC_CTRL, ctrl);
Signature · 1 line
2.39.2
author maintainer
#122 lore raw

PATCH 122/286 ASoC: dt-bindings: Adjust #sound-dai-cells on TIs single-DAI codecs

In reply to Greg Kroah-Hartman

From: Martin Povišer <[email protected]>

[ Upstream commit efb2bfd7b3d210c479b9361c176d7426e5eb8663 ]

A bunch of TI's codecs have binding schemas which force #sound-dai-cells to one despite those codecs only having a single DAI. Allow for bindings with zero DAI cells and deprecate the former non-zero value.

---
 Documentation/devicetree/bindings/sound/tas2562.yaml | 6 ++++--
 Documentation/devicetree/bindings/sound/tas2770.yaml | 6 ++++--
 Documentation/devicetree/bindings/sound/tas27xx.yaml | 6 ++++--
 3 files changed, 12 insertions(+), 6 deletions(-)
Documentation/devicetree/bindings/sound/tas2562.yaml +4 -2
index 1085592cefccc..81218c07079a8 100644
@@ -55,7 +55,9 @@ properties:
     description: TDM TX current sense time slot.
 
   '#sound-dai-cells':
-    const: 1
+    # The codec has a single DAI, the #sound-dai-cells=<1>; case is left in for backward
+    # compatibility but is deprecated.
+    enum: [0, 1]
 
 required:
   - compatible
@@ -72,7 +74,7 @@ examples:
      codec: codec@4c {
        compatible = "ti,tas2562";
        reg = <0x4c>;
-       #sound-dai-cells = <1>;
+       #sound-dai-cells = <0>;
        interrupt-parent = <&gpio1>;
        interrupts = <14>;
        shutdown-gpios = <&gpio1 15 0>;
Documentation/devicetree/bindings/sound/tas2770.yaml +4 -2
index 982949ba8a4be..cdb493db47f9b 100644
@@ -57,7 +57,9 @@ properties:
       - 1 # Falling edge
 
   '#sound-dai-cells':
-    const: 1
+    # The codec has a single DAI, the #sound-dai-cells=<1>; case is left in for backward
+    # compatibility but is deprecated.
+    enum: [0, 1]
 
 required:
   - compatible
@@ -74,7 +76,7 @@ examples:
      codec: codec@41 {
        compatible = "ti,tas2770";
        reg = <0x41>;
-       #sound-dai-cells = <1>;
+       #sound-dai-cells = <0>;
        interrupt-parent = <&gpio1>;
        interrupts = <14>;
        reset-gpio = <&gpio1 15 0>;
Documentation/devicetree/bindings/sound/tas27xx.yaml +4 -2
index 0957dd435bb4b..2ef05aacc167a 100644
@@ -50,7 +50,9 @@ properties:
     description: TDM TX voltage sense time slot.
 
   '#sound-dai-cells':
-    const: 1
+    # The codec has a single DAI, the #sound-dai-cells=<1>; case is left in for backward
+    # compatibility but is deprecated.
+    enum: [0, 1]
 
 required:
   - compatible
@@ -67,7 +69,7 @@ examples:
      codec: codec@38 {
        compatible = "ti,tas2764";
        reg = <0x38>;
-       #sound-dai-cells = <1>;
+       #sound-dai-cells = <0>;
        interrupt-parent = <&gpio1>;
        interrupts = <14>;
        reset-gpios = <&gpio1 15 0>;
Signature · 1 line
2.39.2
author maintainer
#123 lore raw

PATCH 123/286 ALSA: hda/realtek: Add quirks for ASUS GU604V and GU603V

In reply to Greg Kroah-Hartman

From: Alexandru Sorodoc <[email protected]>

[ Upstream commit 4b963ae1df6426f0e51de64133d379d9bde50c48 ]

These models use 2 CS35L41 amplifiers using SPI for down-facing speakers.

alc285_fixup_speaker2_to_dac1 is needed to fix volume control of the down-facing speakers.

Pin configs are needed to enable headset mic detection.

Note that these models lack the ACPI _DSD properties needed to initialize the amplifiers. They can be added during boot to get working sound out of the speakers:
https://gist.github.com/lamperez/862763881c0e1c812392b5574727f6ff

---
 sound/pci/hda/patch_realtek.c | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)
sound/pci/hda/patch_realtek.c +20 -0
index 379f216158ab4..7b5f194513c7b 100644
@@ -7063,6 +7063,8 @@ enum {
 	ALC225_FIXUP_DELL1_MIC_NO_PRESENCE,
 	ALC295_FIXUP_DISABLE_DAC3,
 	ALC285_FIXUP_SPEAKER2_TO_DAC1,
+	ALC285_FIXUP_ASUS_SPEAKER2_TO_DAC1,
+	ALC285_FIXUP_ASUS_HEADSET_MIC,
 	ALC280_FIXUP_HP_HEADSET_MIC,
 	ALC221_FIXUP_HP_FRONT_MIC,
 	ALC292_FIXUP_TPT460,
@@ -8033,6 +8035,22 @@ static const struct hda_fixup alc269_fixups[] = {
 		.chained = true,
 		.chain_id = ALC269_FIXUP_THINKPAD_ACPI
 	},
+	[ALC285_FIXUP_ASUS_SPEAKER2_TO_DAC1] = {
+		.type = HDA_FIXUP_FUNC,
+		.v.func = alc285_fixup_speaker2_to_dac1,
+		.chained = true,
+		.chain_id = ALC245_FIXUP_CS35L41_SPI_2
+	},
+	[ALC285_FIXUP_ASUS_HEADSET_MIC] = {
+		.type = HDA_FIXUP_PINS,
+		.v.pins = (const struct hda_pintbl[]) {
+			{ 0x19, 0x03a11050 },
+			{ 0x1b, 0x03a11c30 },
+			{ }
+		},
+		.chained = true,
+		.chain_id = ALC285_FIXUP_ASUS_SPEAKER2_TO_DAC1
+	},
 	[ALC256_FIXUP_DELL_INSPIRON_7559_SUBWOOFER] = {
 		.type = HDA_FIXUP_PINS,
 		.v.pins = (const struct hda_pintbl[]) {
@@ -9507,6 +9525,8 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = {
 	SND_PCI_QUIRK(0x1043, 0x1313, "Asus K42JZ", ALC269VB_FIXUP_ASUS_MIC_NO_PRESENCE),
 	SND_PCI_QUIRK(0x1043, 0x13b0, "ASUS Z550SA", ALC256_FIXUP_ASUS_MIC),
 	SND_PCI_QUIRK(0x1043, 0x1427, "Asus Zenbook UX31E", ALC269VB_FIXUP_ASUS_ZENBOOK),
+	SND_PCI_QUIRK(0x1043, 0x1473, "ASUS GU604V", ALC285_FIXUP_ASUS_HEADSET_MIC),
+	SND_PCI_QUIRK(0x1043, 0x1483, "ASUS GU603V", ALC285_FIXUP_ASUS_HEADSET_MIC),
 	SND_PCI_QUIRK(0x1043, 0x1517, "Asus Zenbook UX31A", ALC269VB_FIXUP_ASUS_ZENBOOK_UX31A),
 	SND_PCI_QUIRK(0x1043, 0x1662, "ASUS GV301QH", ALC294_FIXUP_ASUS_DUAL_SPK),
 	SND_PCI_QUIRK(0x1043, 0x1683, "ASUS UM3402YAR", ALC287_FIXUP_CS35L41_I2C_2),
Signature · 1 line
2.39.2
author maintainer
#124 lore raw

PATCH 124/286 ASoC: ssm2602: Add workaround for playback distortions

In reply to Greg Kroah-Hartman

From: Paweł Anikiel <[email protected]>

[ Upstream commit f63550e2b165208a2f382afcaf5551df9569e1d4 ]

Apply a workaround for what appears to be a hardware quirk.

The problem seems to happen when enabling "whole chip power" (bit D7 register R6) for the very first time after the chip receives power. If either "output" (D4) or "DAC" (D3) aren't powered on at that time, playback becomes very distorted later on.

This happens on the Google Chameleon v3, as well as on a ZYBO Z7-10:
https://ez.analog.com/audio/f/q-a/543726/solved-ssm2603-right-output-offset-issue/480229
I suspect this happens only when using an external MCLK signal (which is the case for both of these boards).

Here are some experiments run on a Google Chameleon v3. These were run in userspace using a wrapper around the i2cset utility:

ssmset() {
        i2cset -y 0 0x1a $(($1*2)) $2
}

For each of the following sequences, we apply power to the ssm2603 chip, set the configuration registers R0-R5 and R7-R8, run the selected sequence, and check for distortions on playback.

ssmset 0x09 0x01 # core
ssmset 0x06 0x07 # chip, out, dac
OK

ssmset 0x09 0x01 # core
ssmset 0x06 0x87 # out, dac
ssmset 0x06 0x07 # chip
OK

(disable MCLK)
ssmset 0x09 0x01 # core
ssmset 0x06 0x1f # chip
ssmset 0x06 0x07 # out, dac
(enable MCLK)
OK

ssmset 0x09 0x01 # core
ssmset 0x06 0x1f # chip
ssmset 0x06 0x07 # out, dac
NOT OK

ssmset 0x06 0x1f # chip
ssmset 0x09 0x01 # core
ssmset 0x06 0x07 # out, dac
NOT OK

ssmset 0x09 0x01 # core
ssmset 0x06 0x0f # chip, out
ssmset 0x06 0x07 # dac
NOT OK

ssmset 0x09 0x01 # core
ssmset 0x06 0x17 # chip, dac
ssmset 0x06 0x07 # out
NOT OK

For each of the following sequences, we apply power to the ssm2603 chip, run the selected sequence, issue a reset with R15, configure R0-R5 and R7-R8, run one of the NOT OK sequences from above, and check for distortions.

ssmset 0x09 0x01 # core
ssmset 0x06 0x07 # chip, out, dac
OK

(disable MCLK)
ssmset 0x09 0x01 # core
ssmset 0x06 0x07 # chip, out, dac
(enable MCLK after reset)
NOT OK

ssmset 0x09 0x01 # core
ssmset 0x06 0x17 # chip, dac
NOT OK

ssmset 0x09 0x01 # core
ssmset 0x06 0x0f # chip, out
NOT OK

ssmset 0x06 0x07 # chip, out, dac
NOT OK

---
 sound/soc/codecs/ssm2602.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)
sound/soc/codecs/ssm2602.c +15 -0
index cbbe83b85adaf..cf7927222be15 100644
@@ -53,6 +53,18 @@ static const struct reg_default ssm2602_reg[SSM2602_CACHEREGNUM] = {
 	{ .reg = 0x09, .def = 0x0000 }
 };
 
+/*
+ * ssm2602 register patch
+ * Workaround for playback distortions after power up: activates digital
+ * core, and then powers on output, DAC, and whole chip at the same time
+ */
+
+static const struct reg_sequence ssm2602_patch[] = {
+	{ SSM2602_ACTIVE, 0x01 },
+	{ SSM2602_PWR,    0x07 },
+	{ SSM2602_RESET,  0x00 },
+};
+
 
 /*Appending several "None"s just for OSS mixer use*/
 static const char *ssm2602_input_select[] = {
@@ -589,6 +601,9 @@ static int ssm260x_component_probe(struct snd_soc_component *component)
 		return ret;
 	}
 
+	regmap_register_patch(ssm2602->regmap, ssm2602_patch,
+			      ARRAY_SIZE(ssm2602_patch));
+
 	/* set the update bits */
 	regmap_update_bits(ssm2602->regmap, SSM2602_LINVOL,
 			    LINVOL_LRIN_BOTH, LINVOL_LRIN_BOTH);
Signature · 1 line
2.39.2
author maintainer
#125 lore raw

PATCH 125/286 media: dvb_demux: fix a bug for the continuity counter

In reply to Greg Kroah-Hartman

From: YongSu Yoo <[email protected]>

[ Upstream commit 7efb10d8dc70ea3000cc70dca53407c52488acd1 ]

In dvb_demux.c, some logics exist which compare the expected continuity counter and the real continuity counter. If they are not matched each other, both of the expected continuity counter and the real continuity counter should be printed. But there exists a bug that the expected continuity counter is not correctly printed. The expected continuity counter is replaced with the real countinuity counter + 1 so that the epected continuity counter is not correclty printed. This is wrong. This bug is fixed.

---
 drivers/media/dvb-core/dvb_demux.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/media/dvb-core/dvb_demux.c +2 -2
index 398c86279b5b0..7c4d86bfdd6c9 100644
@@ -115,12 +115,12 @@ static inline int dvb_dmx_swfilter_payload(struct dvb_demux_feed *feed,
 
 	cc = buf[3] & 0x0f;
 	ccok = ((feed->cc + 1) & 0x0f) == cc;
-	feed->cc = cc;
 	if (!ccok) {
 		set_buf_flags(feed, DMX_BUFFER_FLAG_DISCONTINUITY_DETECTED);
 		dprintk_sect_loss("missed packet: %d instead of %d!\n",
 				  cc, (feed->cc + 1) & 0x0f);
 	}
+	feed->cc = cc;
 
 	if (buf[1] & 0x40)	// PUSI ?
 		feed->peslen = 0xfffa;
@@ -300,7 +300,6 @@ static int dvb_dmx_swfilter_section_packet(struct dvb_demux_feed *feed,
 
 	cc = buf[3] & 0x0f;
 	ccok = ((feed->cc + 1) & 0x0f) == cc;
-	feed->cc = cc;
 
 	if (buf[3] & 0x20) {
 		/* adaption field present, check for discontinuity_indicator */
@@ -336,6 +335,7 @@ static int dvb_dmx_swfilter_section_packet(struct dvb_demux_feed *feed,
 		feed->pusi_seen = false;
 		dvb_dmx_swfilter_section_new(feed);
 	}
+	feed->cc = cc;
 
 	if (buf[1] & 0x40) {
 		/* PUSI=1 (is set), section boundary is here */
Signature · 1 line
2.39.2
author maintainer
#126 lore raw

PATCH 126/286 media: dvb-usb: az6027: fix three null-ptr-deref in az6027_i2c_xfer()

In reply to Greg Kroah-Hartman

From: Wei Chen <[email protected]>

[ Upstream commit 858e97d7956d17a2cb56a9413468704a4d5abfe1 ]

In az6027_i2c_xfer, msg is controlled by user. When msg[i].buf is null, commit 0ed554fd769a ("media: dvb-usb: az6027: fix null-ptr-deref in az6027_i2c_xfer()") fix the null-ptr-deref bug when msg[i].addr is 0x99. However, null-ptr-deref also happens when msg[i].addr is 0xd0 and 0xc0. We add check on msg[i].len to prevent null-ptr-deref.

---
 drivers/media/usb/dvb-usb/az6027.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
drivers/media/usb/dvb-usb/az6027.c +12 -0
index 7d78ee09be5e1..a31c6f82f4e90 100644
@@ -988,6 +988,10 @@ static int az6027_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msg[], int n
 			/* write/read request */
 			if (i + 1 < num && (msg[i + 1].flags & I2C_M_RD)) {
 				req = 0xB9;
+				if (msg[i].len < 1) {
+					i = -EOPNOTSUPP;
+					break;
+				}
 				index = (((msg[i].buf[0] << 8) & 0xff00) | (msg[i].buf[1] & 0x00ff));
 				value = msg[i].addr + (msg[i].len << 8);
 				length = msg[i + 1].len + 6;
@@ -1001,6 +1005,10 @@ static int az6027_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msg[], int n
 
 				/* demod 16bit addr */
 				req = 0xBD;
+				if (msg[i].len < 1) {
+					i = -EOPNOTSUPP;
+					break;
+				}
 				index = (((msg[i].buf[0] << 8) & 0xff00) | (msg[i].buf[1] & 0x00ff));
 				value = msg[i].addr + (2 << 8);
 				length = msg[i].len - 2;
@@ -1026,6 +1034,10 @@ static int az6027_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msg[], int n
 			} else {
 
 				req = 0xBD;
+				if (msg[i].len < 1) {
+					i = -EOPNOTSUPP;
+					break;
+				}
 				index = msg[i].buf[0] & 0x00FF;
 				value = msg[i].addr + (1 << 8);
 				length = msg[i].len - 1;
Signature · 1 line
2.39.2
author maintainer
#100 lore raw

PATCH 100/286 media: rcar-vin: Fix NV12 size alignment

In reply to Greg Kroah-Hartman

From: Niklas Söderlund <[email protected]>

[ Upstream commit cb88d8289fc222bd21b7a7f99b055e7e73e316f4 ]

When doing format validation for NV12 the width and height should be aligned to 32 pixels.

---
 drivers/media/platform/renesas/rcar-vin/rcar-dma.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)
drivers/media/platform/renesas/rcar-vin/rcar-dma.c +9 -3
index cc6b59e5621ae..23598e22adc72 100644
@@ -1320,9 +1320,15 @@ static int rvin_mc_validate_format(struct rvin_dev *vin, struct v4l2_subdev *sd,
 		if (!vin->scaler)
 			return -EPIPE;
 	} else {
-		if (fmt.format.width != vin->format.width ||
-		    fmt.format.height != vin->format.height)
-			return -EPIPE;
+		if (vin->format.pixelformat == V4L2_PIX_FMT_NV12) {
+			if (ALIGN(fmt.format.width, 32) != vin->format.width ||
+			    ALIGN(fmt.format.height, 32) != vin->format.height)
+				return -EPIPE;
+		} else {
+			if (fmt.format.width != vin->format.width ||
+			    fmt.format.height != vin->format.height)
+				return -EPIPE;
+		}
 	}
 
 	if (fmt.format.code != vin->mbus_code)
Signature · 1 line
2.39.2
author maintainer
#127 lore raw

PATCH 127/286 media: dvb-usb-v2: ec168: fix null-ptr-deref in ec168_i2c_xfer()

In reply to Greg Kroah-Hartman

From: Wei Chen <[email protected]>

[ Upstream commit a6dcefcc08eca1bf4e3d213c97c3cfb75f377935 ]

In ec168_i2c_xfer, msg is controlled by user. When msg[i].buf is null and msg[i].len is zero, former checks on msg[i].buf would be passed. If accessing msg[i].buf[0] without sanity check, null pointer deref would happen. We add check on msg[i].len to prevent crash.

Similar commit:
commit 0ed554fd769a ("media: dvb-usb: az6027: fix null-ptr-deref in az6027_i2c_xfer()")

---
 drivers/media/usb/dvb-usb-v2/ec168.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
drivers/media/usb/dvb-usb-v2/ec168.c +12 -0
index 7ed0ab9e429b1..0e4773fc025c9 100644
@@ -115,6 +115,10 @@ static int ec168_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msg[],
 	while (i < num) {
 		if (num > i + 1 && (msg[i+1].flags & I2C_M_RD)) {
 			if (msg[i].addr == ec168_ec100_config.demod_address) {
+				if (msg[i].len < 1) {
+					i = -EOPNOTSUPP;
+					break;
+				}
 				req.cmd = READ_DEMOD;
 				req.value = 0;
 				req.index = 0xff00 + msg[i].buf[0]; /* reg */
@@ -131,6 +135,10 @@ static int ec168_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msg[],
 			}
 		} else {
 			if (msg[i].addr == ec168_ec100_config.demod_address) {
+				if (msg[i].len < 1) {
+					i = -EOPNOTSUPP;
+					break;
+				}
 				req.cmd = WRITE_DEMOD;
 				req.value = msg[i].buf[1]; /* val */
 				req.index = 0xff00 + msg[i].buf[0]; /* reg */
@@ -139,6 +147,10 @@ static int ec168_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msg[],
 				ret = ec168_ctrl_msg(d, &req);
 				i += 1;
 			} else {
+				if (msg[i].len < 1) {
+					i = -EOPNOTSUPP;
+					break;
+				}
 				req.cmd = WRITE_I2C;
 				req.value = msg[i].buf[0]; /* val */
 				req.index = 0x0100 + msg[i].addr; /* I2C addr */
Signature · 1 line
2.39.2
author maintainer
#128 lore raw

PATCH 128/286 media: dvb-usb-v2: ce6230: fix null-ptr-deref in ce6230_i2c_master_xfer()

In reply to Greg Kroah-Hartman

From: Wei Chen <[email protected]>

[ Upstream commit dff919090155fb22679869e8469168f270dcd97f ]

In ce6230_i2c_master_xfer, msg is controlled by user. When msg[i].buf is null and msg[i].len is zero, former checks on msg[i].buf would be passed. Malicious data finally reach ce6230_i2c_master_xfer. If accessing msg[i].buf[0] without sanity check, null ptr deref would happen. We add check on msg[i].len to prevent crash.

Similar commit:
commit 0ed554fd769a ("media: dvb-usb: az6027: fix null-ptr-deref in az6027_i2c_xfer()")

---
 drivers/media/usb/dvb-usb-v2/ce6230.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/media/usb/dvb-usb-v2/ce6230.c +8 -0
index 44540de1a2066..d3b5cb4a24daf 100644
@@ -101,6 +101,10 @@ static int ce6230_i2c_master_xfer(struct i2c_adapter *adap,
 		if (num > i + 1 && (msg[i+1].flags & I2C_M_RD)) {
 			if (msg[i].addr ==
 				ce6230_zl10353_config.demod_address) {
+				if (msg[i].len < 1) {
+					i = -EOPNOTSUPP;
+					break;
+				}
 				req.cmd = DEMOD_READ;
 				req.value = msg[i].addr >> 1;
 				req.index = msg[i].buf[0];
@@ -117,6 +121,10 @@ static int ce6230_i2c_master_xfer(struct i2c_adapter *adap,
 		} else {
 			if (msg[i].addr ==
 				ce6230_zl10353_config.demod_address) {
+				if (msg[i].len < 1) {
+					i = -EOPNOTSUPP;
+					break;
+				}
 				req.cmd = DEMOD_WRITE;
 				req.value = msg[i].addr >> 1;
 				req.index = msg[i].buf[0];
Signature · 1 line
2.39.2
author maintainer
#101 lore raw

PATCH 101/286 media: rcar-vin: Select correct interrupt mode for V4L2_FIELD_ALTERNATE

In reply to Greg Kroah-Hartman

From: Niklas Söderlund <[email protected]>

[ Upstream commit e10707d5865c90d3dfe4ef589ce02ff4287fef85 ]

When adding proper support for V4L2_FIELD_ALTERNATE it was missed that this field format should trigger an interrupt for each field, not just for the whole frame. Fix this by marking it as progressive in the capture setup, which will then select the correct interrupt mode.

Tested on both Gen2 and Gen3 with the result of a doubling of the frame rate for V4L2_FIELD_ALTERNATE. From a PAL video source the frame rate is now 50, which is expected for alternate field capture.

---
 drivers/media/platform/renesas/rcar-vin/rcar-dma.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/media/platform/renesas/rcar-vin/rcar-dma.c +1 -3
index 23598e22adc72..2a77353f10b59 100644
@@ -728,11 +728,9 @@ static int rvin_setup(struct rvin_dev *vin)
 	case V4L2_FIELD_SEQ_TB:
 	case V4L2_FIELD_SEQ_BT:
 	case V4L2_FIELD_NONE:
-		vnmc = VNMC_IM_ODD_EVEN;
-		progressive = true;
-		break;
 	case V4L2_FIELD_ALTERNATE:
 		vnmc = VNMC_IM_ODD_EVEN;
+		progressive = true;
 		break;
 	default:
 		vnmc = VNMC_IM_ODD;
Signature · 1 line
2.39.2
author maintainer
#102 lore raw

PATCH 102/286 platform/x86: intel_scu_pcidrv: Add back PCI ID for Medfield

In reply to Greg Kroah-Hartman

From: Julian Winkler <[email protected]>

[ Upstream commit 4a9b6850c794e4394cad99e2b863d75f5bc8e92f ]

This id was removed in commit b47018a778c1 ("platform/x86: intel_scu_ipc:
Remove Lincroft support"), saying it is only used on Moorestown, but apparently the same id is also used on Medfield.

Tested on the Medfield based Motorola RAZR i smartphone.

---
 drivers/platform/x86/intel_scu_pcidrv.c | 1 +
 1 file changed, 1 insertion(+)
drivers/platform/x86/intel_scu_pcidrv.c +1 -0
index 80abc708e4f2f..d904fad499aa5 100644
@@ -34,6 +34,7 @@ static int intel_scu_pci_probe(struct pci_dev *pdev,
 
 static const struct pci_device_id pci_ids[] = {
 	{ PCI_VDEVICE(INTEL, 0x080e) },
+	{ PCI_VDEVICE(INTEL, 0x082a) },
 	{ PCI_VDEVICE(INTEL, 0x08ea) },
 	{ PCI_VDEVICE(INTEL, 0x0a94) },
 	{ PCI_VDEVICE(INTEL, 0x11a0) },
Signature · 1 line
2.39.2
author maintainer
#103 lore raw

PATCH 103/286 platform/mellanox: fix potential race in mlxbf-tmfifo driver

In reply to Greg Kroah-Hartman

From: Liming Sun <[email protected]>

[ Upstream commit 3d43f9f639542fadfb28f40b509bf147a6624d48 ]

This commit adds memory barrier for the 'vq' update in function mlxbf_tmfifo_virtio_find_vqs() to avoid potential race due to out-of-order memory write. It also adds barrier for the 'is_ready' flag to make sure the initializations are visible before this flag is checked.

---
 drivers/platform/mellanox/mlxbf-tmfifo.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
drivers/platform/mellanox/mlxbf-tmfifo.c +9 -2
index 91a077c35b8b8..a79318e90a139 100644
@@ -784,7 +784,7 @@ static void mlxbf_tmfifo_rxtx(struct mlxbf_tmfifo_vring *vring, bool is_rx)
 	fifo = vring->fifo;
 
 	/* Return if vdev is not ready. */
-	if (!fifo->vdev[devid])
+	if (!fifo || !fifo->vdev[devid])
 		return;
 
 	/* Return if another vring is running. */
@@ -980,9 +980,13 @@ static int mlxbf_tmfifo_virtio_find_vqs(struct virtio_device *vdev,
 
 		vq->num_max = vring->num;
 
+		vq->priv = vring;
+
+		/* Make vq update visible before using it. */
+		virtio_mb(false);
+
 		vqs[i] = vq;
 		vring->vq = vq;
-		vq->priv = vring;
 	}
 
 	return 0;
@@ -1302,6 +1306,9 @@ static int mlxbf_tmfifo_probe(struct platform_device *pdev)
 
 	mod_timer(&fifo->timer, jiffies + MLXBF_TMFIFO_TIMER_INTERVAL);
 
+	/* Make all updates visible before setting the 'is_ready' flag. */
+	virtio_mb(false);
+
 	fifo->is_ready = true;
 	return 0;
 
Signature · 1 line
2.39.2
author maintainer
#104 lore raw

PATCH 104/286 gfs2: Dont deref jdesc in evict

In reply to Greg Kroah-Hartman

From: Bob Peterson <[email protected]>

[ Upstream commit 504a10d9e46bc37b23d0a1ae2f28973c8516e636 ]

On corrupt gfs2 file systems the evict code can try to reference the journal descriptor structure, jdesc, after it has been freed and set to NULL. The sequence of events is:

init_journal()
...
fail_jindex:
gfs2_jindex_free(sdp); <------frees journals, sets jdesc = NULL

      if (gfs2_holder_initialized(&ji_gh))
         gfs2_glock_dq_uninit(&ji_gh);
fail:
   iput(sdp->sd_jindex); <--references jdesc in evict_linked_inode
      evict()
         gfs2_evict_inode()
            evict_linked_inode()
               ret = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks);

<------references the now freed/zeroed sd_jdesc pointer.

The call to gfs2_trans_begin is done because the truncate_inode_pages call can cause gfs2 events that require a transaction, such as removing journaled data (jdata) blocks from the journal.

This patch fixes the problem by adding a check for sdp->sd_jdesc to function gfs2_evict_inode. In theory, this should only happen to corrupt gfs2 file systems, when gfs2 detects the problem, reports it, then tries to evict all the system inodes it has read in up to that point.

---
 fs/gfs2/super.c | 8 ++++++++
 1 file changed, 8 insertions(+)
fs/gfs2/super.c +8 -0
index a83fa62106f0e..7891f331082aa 100644
@@ -1410,6 +1410,14 @@ static void gfs2_evict_inode(struct inode *inode)
 	if (inode->i_nlink || sb_rdonly(sb) || !ip->i_no_addr)
 		goto out;
 
+	/*
+	 * In case of an incomplete mount, gfs2_evict_inode() may be called for
+	 * system files without having an active journal to write to.  In that
+	 * case, skip the filesystem evict.
+	 */
+	if (!sdp->sd_jdesc)
+		goto out;
+
 	gfs2_holder_mark_uninitialized(&gh);
 	ret = evict_should_delete(inode, &gh);
 	if (ret == SHOULD_DEFER_EVICTION)
Signature · 1 line
2.39.2
author maintainer
#105 lore raw

PATCH 105/286 drm/amdgpu/nv: update VCN 3 max HEVC encoding resolution

In reply to Greg Kroah-Hartman

From: Thong Thai <[email protected]>

[ Upstream commit 476ac50fc30540e29191615a26aaf5f9dee91c49 ]

Update the maximum resolution reported for HEVC encoding on VCN 3 devices to reflect its 8K encoding capability.

v2: Also update the max height for H.264 encoding to match spec. (Ruijing)

---
 drivers/gpu/drm/amd/amdgpu/nv.c | 22 ++++++++++++++++------
 1 file changed, 16 insertions(+), 6 deletions(-)
drivers/gpu/drm/amd/amdgpu/nv.c +16 -6
index ebe0e2d7dbd1b..aa7f82b3fd6a9 100644
@@ -98,6 +98,16 @@ static const struct amdgpu_video_codecs nv_video_codecs_decode =
 };
 
 /* Sienna Cichlid */
+static const struct amdgpu_video_codec_info sc_video_codecs_encode_array[] = {
+	{codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 2160, 0)},
+	{codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 7680, 4352, 0)},
+};
+
+static const struct amdgpu_video_codecs sc_video_codecs_encode = {
+	.codec_count = ARRAY_SIZE(sc_video_codecs_encode_array),
+	.codec_array = sc_video_codecs_encode_array,
+};
+
 static const struct amdgpu_video_codec_info sc_video_codecs_decode_array_vcn0[] =
 {
 	{codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG2, 4096, 4096, 3)},
@@ -136,8 +146,8 @@ static const struct amdgpu_video_codecs sc_video_codecs_decode_vcn1 =
 /* SRIOV Sienna Cichlid, not const since data is controlled by host */
 static struct amdgpu_video_codec_info sriov_sc_video_codecs_encode_array[] =
 {
-	{codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 2304, 0)},
-	{codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 4096, 2304, 0)},
+	{codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 2160, 0)},
+	{codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 7680, 4352, 0)},
 };
 
 static struct amdgpu_video_codec_info sriov_sc_video_codecs_decode_array_vcn0[] =
@@ -237,12 +247,12 @@ static int nv_query_video_codecs(struct amdgpu_device *adev, bool encode,
 		} else {
 			if (adev->vcn.harvest_config & AMDGPU_VCN_HARVEST_VCN0) {
 				if (encode)
-					*codecs = &nv_video_codecs_encode;
+					*codecs = &sc_video_codecs_encode;
 				else
 					*codecs = &sc_video_codecs_decode_vcn1;
 			} else {
 				if (encode)
-					*codecs = &nv_video_codecs_encode;
+					*codecs = &sc_video_codecs_encode;
 				else
 					*codecs = &sc_video_codecs_decode_vcn0;
 			}
@@ -251,14 +261,14 @@ static int nv_query_video_codecs(struct amdgpu_device *adev, bool encode,
 	case IP_VERSION(3, 0, 16):
 	case IP_VERSION(3, 0, 2):
 		if (encode)
-			*codecs = &nv_video_codecs_encode;
+			*codecs = &sc_video_codecs_encode;
 		else
 			*codecs = &sc_video_codecs_decode_vcn0;
 		return 0;
 	case IP_VERSION(3, 1, 1):
 	case IP_VERSION(3, 1, 2):
 		if (encode)
-			*codecs = &nv_video_codecs_encode;
+			*codecs = &sc_video_codecs_encode;
 		else
 			*codecs = &yc_video_codecs_decode;
 		return 0;
Signature · 1 line
2.39.2
author maintainer
#106 lore raw

PATCH 106/286 drm/amdgpu: set gfx9 onwards APU atomics support to be true

In reply to Greg Kroah-Hartman

From: Yifan Zhang <[email protected]>

[ Upstream commit af7828fbceed4f9e503034111066a0adef3db383 ]

APUs w/ gfx9 onwards doesn't reply on PCIe atomics, rather it is internal path w/ native atomic support. Set have_atomics_support to true.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +6 -0
index 412cb3f1f8826..31413a604d0ae 100644
@@ -3737,6 +3737,12 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 		adev->have_atomics_support = ((struct amd_sriov_msg_pf2vf_info *)
 			adev->virt.fw_reserve.p_pf2vf)->pcie_atomic_ops_support_flags ==
 			(PCI_EXP_DEVCAP2_ATOMIC_COMP32 | PCI_EXP_DEVCAP2_ATOMIC_COMP64);
+	/* APUs w/ gfx9 onwards doesn't reply on PCIe atomics, rather it is a
+	 * internal path natively support atomics, set have_atomics_support to true.
+	 */
+	else if ((adev->flags & AMD_IS_APU) &&
+		(adev->ip_versions[GC_HWIP][0] > IP_VERSION(9, 0, 0)))
+		adev->have_atomics_support = true;
 	else
 		adev->have_atomics_support =
 			!pci_enable_atomic_ops_to_root(adev->pdev,
Signature · 1 line
2.39.2
author maintainer
#97 lore raw

PATCH 97/286 perf/x86/intel/ds: Flush PEBS DS when changing PEBS_DATA_CFG

In reply to Greg Kroah-Hartman

From: Kan Liang <[email protected]>

[ Upstream commit b752ea0c28e3f7f0aaaad6abf84f735eebc37a60 ]

Several similar kernel warnings can be triggered,

  [56605.607840] CPU0 PEBS record size 0, expected 32, config 0 cpuc->record_size=208

when the below commands are running in parallel for a while on SPR.

while true;
do

	perf record --no-buildid -a --intr-regs=AX  \
		    -e cpu/event=0xd0,umask=0x81/pp \
		    -c 10003 -o /dev/null ./triad;

done &

while true;
do

	perf record -o /tmp/out -W -d \
		    -e '{ld_blocks.store_forward:period=1000000, \
                         MEM_TRANS_RETIRED.LOAD_LATENCY:u:precise=2:ldlat=4}' \
		    -c 1037 ./triad;

done

The triad program is just the generation of loads/stores.

The warnings are triggered when an unexpected PEBS record (with a different config and size) is found.

A system-wide PEBS event with the large PEBS config may be enabled during a context switch. Some PEBS records for the system-wide PEBS may be generated while the old task is sched out but the new one hasn't been sched in yet. When the new task is sched in, the cpuc->pebs_record_size may be updated for the per-task PEBS events. So the existing system-wide PEBS records have a different size from the later PEBS records.

The PEBS buffer should be flushed right before the hardware is reprogrammed. The new size and threshold should be updated after the old buffer has been flushed.

---
 arch/x86/events/intel/ds.c        | 56 ++++++++++++++++++-------------
 arch/x86/include/asm/perf_event.h |  3 ++
 2 files changed, 35 insertions(+), 24 deletions(-)
arch/x86/events/intel/ds.c +32 -24
index a2e566e53076e..df88576d6b2a5 100644
@@ -1229,12 +1229,14 @@ pebs_update_state(bool needed_cb, struct cpu_hw_events *cpuc,
 		  struct perf_event *event, bool add)
 {
 	struct pmu *pmu = event->pmu;
+
 	/*
 	 * Make sure we get updated with the first PEBS
 	 * event. It will trigger also during removal, but
 	 * that does not hurt:
 	 */
-	bool update = cpuc->n_pebs == 1;
+	if (cpuc->n_pebs == 1)
+		cpuc->pebs_data_cfg = PEBS_UPDATE_DS_SW;
 
 	if (needed_cb != pebs_needs_sched_cb(cpuc)) {
 		if (!needed_cb)
@@ -1242,7 +1244,7 @@ pebs_update_state(bool needed_cb, struct cpu_hw_events *cpuc,
 		else
 			perf_sched_cb_dec(pmu);
 
-		update = true;
+		cpuc->pebs_data_cfg |= PEBS_UPDATE_DS_SW;
 	}
 
 	/*
@@ -1252,24 +1254,13 @@ pebs_update_state(bool needed_cb, struct cpu_hw_events *cpuc,
 	if (x86_pmu.intel_cap.pebs_baseline && add) {
 		u64 pebs_data_cfg;
 
-		/* Clear pebs_data_cfg and pebs_record_size for first PEBS. */
-		if (cpuc->n_pebs == 1) {
-			cpuc->pebs_data_cfg = 0;
-			cpuc->pebs_record_size = sizeof(struct pebs_basic);
-		}
-
 		pebs_data_cfg = pebs_update_adaptive_cfg(event);
-
-		/* Update pebs_record_size if new event requires more data. */
-		if (pebs_data_cfg & ~cpuc->pebs_data_cfg) {
-			cpuc->pebs_data_cfg |= pebs_data_cfg;
-			adaptive_pebs_record_size_update();
-			update = true;
-		}
+		/*
+		 * Be sure to update the thresholds when we change the record.
+		 */
+		if (pebs_data_cfg & ~cpuc->pebs_data_cfg)
+			cpuc->pebs_data_cfg |= pebs_data_cfg | PEBS_UPDATE_DS_SW;
 	}
-
-	if (update)
-		pebs_update_threshold(cpuc);
 }
 
 void intel_pmu_pebs_add(struct perf_event *event)
@@ -1326,9 +1317,17 @@ static void intel_pmu_pebs_via_pt_enable(struct perf_event *event)
 	wrmsrl(base + idx, value);
 }
 
+static inline void intel_pmu_drain_large_pebs(struct cpu_hw_events *cpuc)
+{
+	if (cpuc->n_pebs == cpuc->n_large_pebs &&
+	    cpuc->n_pebs != cpuc->n_pebs_via_pt)
+		intel_pmu_drain_pebs_buffer();
+}
+
 void intel_pmu_pebs_enable(struct perf_event *event)
 {
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
+	u64 pebs_data_cfg = cpuc->pebs_data_cfg & ~PEBS_UPDATE_DS_SW;
 	struct hw_perf_event *hwc = &event->hw;
 	struct debug_store *ds = cpuc->ds;
 	unsigned int idx = hwc->idx;
@@ -1344,11 +1343,22 @@ void intel_pmu_pebs_enable(struct perf_event *event)
 
 	if (x86_pmu.intel_cap.pebs_baseline) {
 		hwc->config |= ICL_EVENTSEL_ADAPTIVE;
-		if (cpuc->pebs_data_cfg != cpuc->active_pebs_data_cfg) {
-			wrmsrl(MSR_PEBS_DATA_CFG, cpuc->pebs_data_cfg);
-			cpuc->active_pebs_data_cfg = cpuc->pebs_data_cfg;
+		if (pebs_data_cfg != cpuc->active_pebs_data_cfg) {
+			/*
+			 * drain_pebs() assumes uniform record size;
+			 * hence we need to drain when changing said
+			 * size.
+			 */
+			intel_pmu_drain_large_pebs(cpuc);
+			adaptive_pebs_record_size_update();
+			wrmsrl(MSR_PEBS_DATA_CFG, pebs_data_cfg);
+			cpuc->active_pebs_data_cfg = pebs_data_cfg;
 		}
 	}
+	if (cpuc->pebs_data_cfg & PEBS_UPDATE_DS_SW) {
+		cpuc->pebs_data_cfg = pebs_data_cfg;
+		pebs_update_threshold(cpuc);
+	}
 
 	if (idx >= INTEL_PMC_IDX_FIXED) {
 		if (x86_pmu.intel_cap.pebs_format < 5)
@@ -1391,9 +1401,7 @@ void intel_pmu_pebs_disable(struct perf_event *event)
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
 	struct hw_perf_event *hwc = &event->hw;
 
-	if (cpuc->n_pebs == cpuc->n_large_pebs &&
-	    cpuc->n_pebs != cpuc->n_pebs_via_pt)
-		intel_pmu_drain_pebs_buffer();
+	intel_pmu_drain_large_pebs(cpuc);
 
 	cpuc->pebs_enabled &= ~(1ULL << hwc->idx);
 
arch/x86/include/asm/perf_event.h +3 -0
index 8fc15ed5e60bb..abf09882f58b6 100644
@@ -121,6 +121,9 @@
 #define PEBS_DATACFG_LBRS	BIT_ULL(3)
 #define PEBS_DATACFG_LBR_SHIFT	24
 
+/* Steal the highest bit of pebs_data_cfg for SW usage */
+#define PEBS_UPDATE_DS_SW	BIT_ULL(63)
+
 /*
  * Intel "Architectural Performance Monitoring" CPUID
  * detection/enumeration details:
Signature · 1 line
2.39.2
author maintainer
#130 lore raw

PATCH 130/286 media: dvb-usb: digitv: fix null-ptr-deref in digitv_i2c_xfer()

In reply to Greg Kroah-Hartman

From: Wei Chen <[email protected]>

[ Upstream commit 9ded5bd2a49ce3015b7c936743eec0a0e6e11f0c ]

In digitv_i2c_xfer, msg is controlled by user. When msg[i].buf is null and msg[i].len is zero, former checks on msg[i].buf would be passed. Malicious data finally reach digitv_i2c_xfer. If accessing msg[i].buf[0] without sanity check, null ptr deref would happen. We add check on msg[i].len to prevent crash.

Similar commit:
commit 0ed554fd769a ("media: dvb-usb: az6027: fix null-ptr-deref in az6027_i2c_xfer()")

---
 drivers/media/usb/dvb-usb/digitv.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/media/usb/dvb-usb/digitv.c +4 -0
index 2756815a780bc..32134be169148 100644
@@ -63,6 +63,10 @@ static int digitv_i2c_xfer(struct i2c_adapter *adap,struct i2c_msg msg[],int num
 		warn("more than 2 i2c messages at a time is not handled yet. TODO.");
 
 	for (i = 0; i < num; i++) {
+		if (msg[i].len < 1) {
+			i = -EOPNOTSUPP;
+			break;
+		}
 		/* write/read request */
 		if (i+1 < num && (msg[i+1].flags & I2C_M_RD)) {
 			if (digitv_ctrl_msg(d, USB_READ_COFDM, msg[i].buf[0], NULL, 0,
Signature · 1 line
2.39.2
author maintainer
#142 lore raw

PATCH 139/286 media: dvb-core: Fix kernel WARNING for blocking operation in wait_event*()

In reply to Greg Kroah-Hartman

From: Takashi Iwai <[email protected]>

[ Upstream commit b8c75e4a1b325ea0a9433fa8834be97b5836b946 ]

Using a semaphore in the wait_event*() condition is no good idea. It hits a kernel WARN_ON() at prepare_to_wait_event() like:
do not call blocking ops when !TASK_RUNNING; state=1 set at prepare_to_wait_event+0x6d/0x690

For avoiding the potential deadlock, rewrite to an open-coded loop instead. Unlike the loop in wait_event*(), this uses wait_woken() after the condition check, hence the task state stays consistent.

CVE-2023-31084 was assigned to this bug.

---
 drivers/media/dvb-core/dvb_frontend.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)
drivers/media/dvb-core/dvb_frontend.c +12 -4
index 947b61959b2b8..bc6950a5740f6 100644
@@ -293,14 +293,22 @@ static int dvb_frontend_get_event(struct dvb_frontend *fe,
 	}
 
 	if (events->eventw == events->eventr) {
-		int ret;
+		struct wait_queue_entry wait;
+		int ret = 0;
 
 		if (flags & O_NONBLOCK)
 			return -EWOULDBLOCK;
 
-		ret = wait_event_interruptible(events->wait_queue,
-					       dvb_frontend_test_event(fepriv, events));
-
+		init_waitqueue_entry(&wait, current);
+		add_wait_queue(&events->wait_queue, &wait);
+		while (!dvb_frontend_test_event(fepriv, events)) {
+			wait_woken(&wait, TASK_INTERRUPTIBLE, 0);
+			if (signal_pending(current)) {
+				ret = -ERESTARTSYS;
+				break;
+			}
+		}
+		remove_wait_queue(&events->wait_queue, &wait);
 		if (ret < 0)
 			return ret;
 	}
Signature · 1 line
2.39.2
author maintainer
#143 lore raw

PATCH 140/286 media: dvb-core: Fix use-after-free due to race condition at dvb_ca_en50221

In reply to Greg Kroah-Hartman

From: Hyunwoo Kim <[email protected]>

[ Upstream commit 280a8ab81733da8bc442253c700a52c4c0886ffd ]

If the device node of dvb_ca_en50221 is open() and the device is disconnected, a UAF may occur when calling close() on the device node.

The root cause is that wake_up() and wait_event() for dvbdev->wait_queue are not implemented.

So implement wait_event() function in dvb_ca_en50221_release() and add 'remove_mutex' which prevents race condition for 'ca->exit'.

[mchehab: fix a checkpatch warning]

---
 drivers/media/dvb-core/dvb_ca_en50221.c | 37 ++++++++++++++++++++++++-
 1 file changed, 36 insertions(+), 1 deletion(-)
drivers/media/dvb-core/dvb_ca_en50221.c +36 -1
index b6ca29dfb184a..baf64540dc00a 100644
@@ -151,6 +151,12 @@ struct dvb_ca_private {
 
 	/* mutex serializing ioctls */
 	struct mutex ioctl_mutex;
+
+	/* A mutex used when a device is disconnected */
+	struct mutex remove_mutex;
+
+	/* Whether the device is disconnected */
+	int exit;
 };
 
 static void dvb_ca_private_free(struct dvb_ca_private *ca)
@@ -1711,12 +1717,22 @@ static int dvb_ca_en50221_io_open(struct inode *inode, struct file *file)
 
 	dprintk("%s\n", __func__);
 
-	if (!try_module_get(ca->pub->owner))
+	mutex_lock(&ca->remove_mutex);
+
+	if (ca->exit) {
+		mutex_unlock(&ca->remove_mutex);
+		return -ENODEV;
+	}
+
+	if (!try_module_get(ca->pub->owner)) {
+		mutex_unlock(&ca->remove_mutex);
 		return -EIO;
+	}
 
 	err = dvb_generic_open(inode, file);
 	if (err < 0) {
 		module_put(ca->pub->owner);
+		mutex_unlock(&ca->remove_mutex);
 		return err;
 	}
 
@@ -1741,6 +1757,7 @@ static int dvb_ca_en50221_io_open(struct inode *inode, struct file *file)
 
 	dvb_ca_private_get(ca);
 
+	mutex_unlock(&ca->remove_mutex);
 	return 0;
 }
 
@@ -1760,6 +1777,8 @@ static int dvb_ca_en50221_io_release(struct inode *inode, struct file *file)
 
 	dprintk("%s\n", __func__);
 
+	mutex_lock(&ca->remove_mutex);
+
 	/* mark the CA device as closed */
 	ca->open = 0;
 	dvb_ca_en50221_thread_update_delay(ca);
@@ -1770,6 +1789,13 @@ static int dvb_ca_en50221_io_release(struct inode *inode, struct file *file)
 
 	dvb_ca_private_put(ca);
 
+	if (dvbdev->users == 1 && ca->exit == 1) {
+		mutex_unlock(&ca->remove_mutex);
+		wake_up(&dvbdev->wait_queue);
+	} else {
+		mutex_unlock(&ca->remove_mutex);
+	}
+
 	return err;
 }
 
@@ -1893,6 +1919,7 @@ int dvb_ca_en50221_init(struct dvb_adapter *dvb_adapter,
 	}
 
 	mutex_init(&ca->ioctl_mutex);
+	mutex_init(&ca->remove_mutex);
 
 	if (signal_pending(current)) {
 		ret = -EINTR;
@@ -1935,6 +1962,14 @@ void dvb_ca_en50221_release(struct dvb_ca_en50221 *pubca)
 
 	dprintk("%s\n", __func__);
 
+	mutex_lock(&ca->remove_mutex);
+	ca->exit = 1;
+	mutex_unlock(&ca->remove_mutex);
+
+	if (ca->dvbdev->users < 1)
+		wait_event(ca->dvbdev->wait_queue,
+				ca->dvbdev->users == 1);
+
 	/* shutdown the thread if there was one */
 	kthread_stop(ca->thread);
 
Signature · 1 line
2.39.2
author maintainer
#144 lore raw

PATCH 141/286 ASoC: SOF: debug: conditionally bump runtime_pm counter on exceptions

In reply to Greg Kroah-Hartman

From: Pierre-Louis Bossart <[email protected]>

[ Upstream commit 3de975862f985f1c9e225a0d13aa3d501373f7c3 ]

When a firmware IPC error happens during a pm_runtime suspend, we ignore the error and suspend anyways. However, the code unconditionally increases the runtime_pm counter. This results in a confusing configuration where the code will suspend, resume but never suspend again due to the use of pm_runtime_get_noresume().

The intent of the counter increase was to prevent entry in D3, but if that transition to D3 is already started it cannot be stopped. In addition, there's no point in that case in trying to prevent anything, the firmware error is handled and the next resume will re-initialize the firmware completely.

This patch changes the logic to prevent suspend when the device is pm_runtime active and has a use_count > 0.

---
 sound/soc/sof/debug.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
sound/soc/sof/debug.c +2 -2
index ade0507328af4..5042312b1b98d 100644
@@ -437,8 +437,8 @@ void snd_sof_handle_fw_exception(struct snd_sof_dev *sdev, const char *msg)
 		/* should we prevent DSP entering D3 ? */
 		if (!sdev->ipc_dump_printed)
 			dev_info(sdev->dev,
-				 "preventing DSP entering D3 state to preserve context\n");
-		pm_runtime_get_noresume(sdev->dev);
+				 "Attempting to prevent DSP from entering D3 state to preserve context\n");
+		pm_runtime_get_if_in_use(sdev->dev);
 	}
 
 	/* dump vital information to the logs */
Signature · 1 line
2.39.2
author maintainer
#145 lore raw

PATCH 142/286 ASoC: SOF: pcm: fix pm_runtime imbalance in error handling

In reply to Greg Kroah-Hartman

From: Pierre-Louis Bossart <[email protected]>

[ Upstream commit da0fe8fd515a471d373acc3682bfb5522cca4d55 ]

When an error occurs, we need to make sure the device can pm_runtime suspend instead of keeping it active.

---
 sound/soc/sof/pcm.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
sound/soc/sof/pcm.c +6 -5
index 445acb5c3a21b..2570f33db9f3e 100644
@@ -616,16 +616,17 @@ static int sof_pcm_probe(struct snd_soc_component *component)
 				       "%s/%s",
 				       plat_data->tplg_filename_prefix,
 				       plat_data->tplg_filename);
-	if (!tplg_filename)
-		return -ENOMEM;
+	if (!tplg_filename) {
+		ret = -ENOMEM;
+		goto pm_error;
+	}
 
 	ret = snd_sof_load_topology(component, tplg_filename);
-	if (ret < 0) {
+	if (ret < 0)
 		dev_err(component->dev, "error: failed to load DSP topology %d\n",
 			ret);
-		return ret;
-	}
 
+pm_error:
 	pm_runtime_mark_last_busy(component->dev);
 	pm_runtime_put_autosuspend(component->dev);
 
Signature · 1 line
2.39.2
author maintainer
#146 lore raw

PATCH 143/286 ASoC: SOF: sof-client-probes: fix pm_runtime imbalance in error handling

In reply to Greg Kroah-Hartman

From: Pierre-Louis Bossart <[email protected]>

[ Upstream commit bc424273c74c1565c459c8f2a6ed95caee368d0a ]

When an error occurs, we need to make sure the device can pm_runtime suspend instead of keeping it active.

---
 sound/soc/sof/sof-client-probes.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
sound/soc/sof/sof-client-probes.c +8 -6
index fff126808bc04..8d9e9d5f40e45 100644
@@ -218,12 +218,7 @@ static ssize_t sof_probes_dfs_points_read(struct file *file, char __user *to,
 
 	ret = ipc->points_info(cdev, &desc, &num_desc);
 	if (ret < 0)
-		goto exit;
-
-	pm_runtime_mark_last_busy(dev);
-	err = pm_runtime_put_autosuspend(dev);
-	if (err < 0)
-		dev_err_ratelimited(dev, "debugfs read failed to idle %d\n", err);
+		goto pm_error;
 
 	for (i = 0; i < num_desc; i++) {
 		offset = strlen(buf);
@@ -241,6 +236,13 @@ static ssize_t sof_probes_dfs_points_read(struct file *file, char __user *to,
 	ret = simple_read_from_buffer(to, count, ppos, buf, strlen(buf));
 
 	kfree(desc);
+
+pm_error:
+	pm_runtime_mark_last_busy(dev);
+	err = pm_runtime_put_autosuspend(dev);
+	if (err < 0)
+		dev_err_ratelimited(dev, "debugfs read failed to idle %d\n", err);
+
 exit:
 	kfree(buf);
 	return ret;
Signature · 1 line
2.39.2
author maintainer
#147 lore raw

PATCH 144/286 ASoC: SOF: pm: save io region state in case of errors in resume

In reply to Greg Kroah-Hartman

From: Kai Vehmanen <[email protected]>

[ Upstream commit 171b53be635ac15d4feafeb33946035649b1ca14 ]

If there are failures in DSP runtime resume, the device state will not reach active and this makes it impossible e.g. to retrieve a possible DSP panic dump via "exception" debugfs node. If CONFIG_SND_SOC_SOF_DEBUG_ENABLE_DEBUGFS_CACHE=y is set, the data in cache is stale. If debugfs cache is not used, the region simply cannot be read.

To allow debugging these scenarios, update the debugfs cache contents in resume error handler. User-space can then later retrieve DSP panic and other state via debugfs (requires SOF debugfs cache to be enabled in build).

---
 sound/soc/sof/pm.c | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
sound/soc/sof/pm.c +13 -1
index 85412aeb1ca16..40f392efd8246 100644
@@ -159,7 +159,7 @@ static int sof_resume(struct device *dev, bool runtime_resume)
 		ret = tplg_ops->set_up_all_pipelines(sdev, false);
 		if (ret < 0) {
 			dev_err(sdev->dev, "Failed to restore pipeline after resume %d\n", ret);
-			return ret;
+			goto setup_fail;
 		}
 	}
 
@@ -173,6 +173,18 @@ static int sof_resume(struct device *dev, bool runtime_resume)
 			dev_err(sdev->dev, "ctx_restore IPC error during resume: %d\n", ret);
 	}
 
+setup_fail:
+#if IS_ENABLED(CONFIG_SND_SOC_SOF_DEBUG_ENABLE_DEBUGFS_CACHE)
+	if (ret < 0) {
+		/*
+		 * Debugfs cannot be read in runtime suspend, so cache
+		 * the contents upon failure. This allows to capture
+		 * possible DSP coredump information.
+		 */
+		sof_cache_debugfs(sdev);
+	}
+#endif
+
 	return ret;
 }
 
Signature · 1 line
2.39.2
author maintainer
#148 lore raw

PATCH 145/286 s390/pkey: zeroize key blobs

In reply to Greg Kroah-Hartman

From: Holger Dengler <[email protected]>

[ Upstream commit 844cf829e5f33e00b279230470c8c93b58b8c16f ]

Key blobs for the IOCTLs PKEY_KBLOB2PROTK[23] may contain clear key material. Zeroize the copies of these keys in kernel memory after creating the protected key.

---
 drivers/s390/crypto/pkey_api.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/s390/crypto/pkey_api.c +3 -0
index 5a05d1cdfec20..a8def50c149bd 100644
@@ -1293,6 +1293,7 @@ static long pkey_unlocked_ioctl(struct file *filp, unsigned int cmd,
 			return PTR_ERR(kkey);
 		rc = pkey_keyblob2pkey(kkey, ktp.keylen, &ktp.protkey);
 		DEBUG_DBG("%s pkey_keyblob2pkey()=%d\n", __func__, rc);
+		memzero_explicit(kkey, ktp.keylen);
 		kfree(kkey);
 		if (rc)
 			break;
@@ -1426,6 +1427,7 @@ static long pkey_unlocked_ioctl(struct file *filp, unsigned int cmd,
 					kkey, ktp.keylen, &ktp.protkey);
 		DEBUG_DBG("%s pkey_keyblob2pkey2()=%d\n", __func__, rc);
 		kfree(apqns);
+		memzero_explicit(kkey, ktp.keylen);
 		kfree(kkey);
 		if (rc)
 			break;
@@ -1552,6 +1554,7 @@ static long pkey_unlocked_ioctl(struct file *filp, unsigned int cmd,
 					protkey, &protkeylen);
 		DEBUG_DBG("%s pkey_keyblob2pkey3()=%d\n", __func__, rc);
 		kfree(apqns);
+		memzero_explicit(kkey, ktp.keylen);
 		kfree(kkey);
 		if (rc) {
 			kfree(protkey);
Signature · 1 line
2.39.2
author maintainer
#149 lore raw

PATCH 146/286 s390/topology: honour nr_cpu_ids when adding CPUs

In reply to Greg Kroah-Hartman

From: Alexander Gordeev <[email protected]>

[ Upstream commit a33239be2d38ff5a44427db1707c08787508d34a ]

When SMT thread CPUs are added to CPU masks the nr_cpu_ids limit is not checked and could be exceeded. This leads to a warning for example if CONFIG_DEBUG_PER_CPU_MAPS is set and the command line parameter nr_cpus is set to 1.

---
 arch/s390/kernel/topology.c | 32 +++++++++++++++++---------------
 1 file changed, 17 insertions(+), 15 deletions(-)
arch/s390/kernel/topology.c +17 -15
index c6eecd4a5302d..10b20aeb27d3b 100644
@@ -95,7 +95,7 @@ static void cpu_group_map(cpumask_t *dst, struct mask_info *info, unsigned int c
 static void cpu_thread_map(cpumask_t *dst, unsigned int cpu)
 {
 	static cpumask_t mask;
-	int i;
+	unsigned int max_cpu;
 
 	cpumask_clear(&mask);
 	if (!cpumask_test_cpu(cpu, &cpu_setup_mask))
@@ -104,9 +104,10 @@ static void cpu_thread_map(cpumask_t *dst, unsigned int cpu)
 	if (topology_mode != TOPOLOGY_MODE_HW)
 		goto out;
 	cpu -= cpu % (smp_cpu_mtid + 1);
-	for (i = 0; i <= smp_cpu_mtid; i++) {
-		if (cpumask_test_cpu(cpu + i, &cpu_setup_mask))
-			cpumask_set_cpu(cpu + i, &mask);
+	max_cpu = min(cpu + smp_cpu_mtid, nr_cpu_ids - 1);
+	for (; cpu <= max_cpu; cpu++) {
+		if (cpumask_test_cpu(cpu, &cpu_setup_mask))
+			cpumask_set_cpu(cpu, &mask);
 	}
 out:
 	cpumask_copy(dst, &mask);
@@ -123,25 +124,26 @@ static void add_cpus_to_mask(struct topology_core *tl_core,
 	unsigned int core;
 
 	for_each_set_bit(core, &tl_core->mask, TOPOLOGY_CORE_BITS) {
-		unsigned int rcore;
-		int lcpu, i;
+		unsigned int max_cpu, rcore;
+		int cpu;
 
 		rcore = TOPOLOGY_CORE_BITS - 1 - core + tl_core->origin;
-		lcpu = smp_find_processor_id(rcore << smp_cpu_mt_shift);
-		if (lcpu < 0)
+		cpu = smp_find_processor_id(rcore << smp_cpu_mt_shift);
+		if (cpu < 0)
 			continue;
-		for (i = 0; i <= smp_cpu_mtid; i++) {
-			topo = &cpu_topology[lcpu + i];
+		max_cpu = min(cpu + smp_cpu_mtid, nr_cpu_ids - 1);
+		for (; cpu <= max_cpu; cpu++) {
+			topo = &cpu_topology[cpu];
 			topo->drawer_id = drawer->id;
 			topo->book_id = book->id;
 			topo->socket_id = socket->id;
 			topo->core_id = rcore;
-			topo->thread_id = lcpu + i;
+			topo->thread_id = cpu;
 			topo->dedicated = tl_core->d;
-			cpumask_set_cpu(lcpu + i, &drawer->mask);
-			cpumask_set_cpu(lcpu + i, &book->mask);
-			cpumask_set_cpu(lcpu + i, &socket->mask);
-			smp_cpu_set_polarization(lcpu + i, tl_core->pp);
+			cpumask_set_cpu(cpu, &drawer->mask);
+			cpumask_set_cpu(cpu, &book->mask);
+			cpumask_set_cpu(cpu, &socket->mask);
+			smp_cpu_set_polarization(cpu, tl_core->pp);
 		}
 	}
 }
Signature · 1 line
2.39.2
author maintainer
#150 lore raw

PATCH 147/286 s390/ipl: fix IPIB virtual vs physical address confusion

In reply to Greg Kroah-Hartman

From: Alexander Gordeev <[email protected]>

[ Upstream commit 2facd5d3980f3a26c04fe6ec8689a1d019a5812c ]

The pointer to IPL Parameter Information Block is stored in the absolute lowcore for later use by dump tools. That pointer is a virtual address, though it should be physical instead.

Note, this does not fix a real issue, since virtual and physical addresses are currently the same.

---
 arch/s390/kernel/ipl.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
arch/s390/kernel/ipl.c +1 -2
index 5f0f5c86963a9..e43ee9becbbb9 100644
@@ -1936,14 +1936,13 @@ static struct shutdown_action __refdata dump_action = {
 
 static void dump_reipl_run(struct shutdown_trigger *trigger)
 {
-	unsigned long ipib = (unsigned long) reipl_block_actual;
 	struct lowcore *abs_lc;
 	unsigned int csum;
 
 	csum = (__force unsigned int)
 	       csum_partial(reipl_block_actual, reipl_block_actual->hdr.len, 0);
 	abs_lc = get_abs_lowcore();
-	abs_lc->ipib = ipib;
+	abs_lc->ipib = __pa(reipl_block_actual);
 	abs_lc->ipib_checksum = csum;
 	put_abs_lowcore(abs_lc);
 	dump_run(trigger);
Signature · 1 line
2.39.2
author maintainer
#151 lore raw

PATCH 148/286 ACPI: resource: Add IRQ override quirk for LG UltraPC 17U70P

In reply to Greg Kroah-Hartman

From: Rubén Gómez <[email protected]>

[ Upstream commit 71a485624c4cbb144169852d7bb8ca8c0667d7a3 ]

Add an ACPI IRQ override quirk for LG UltraPC 17U70P to address the internal keyboard problem on it.

---
 drivers/acpi/resource.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
drivers/acpi/resource.c +12 -0
index e8492b3a393ab..0800a9d775580 100644
@@ -516,6 +516,17 @@ static const struct dmi_system_id maingear_laptop[] = {
 	{ }
 };
 
+static const struct dmi_system_id lg_laptop[] = {
+	{
+		.ident = "LG Electronics 17U70P",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LG Electronics"),
+			DMI_MATCH(DMI_BOARD_NAME, "17U70P"),
+		},
+	},
+	{ }
+};
+
 struct irq_override_cmp {
 	const struct dmi_system_id *system;
 	unsigned char irq;
@@ -532,6 +543,7 @@ static const struct irq_override_cmp override_table[] = {
 	{ lenovo_laptop, 10, ACPI_LEVEL_SENSITIVE, ACPI_ACTIVE_LOW, 0, true },
 	{ tongfang_gm_rg, 1, ACPI_EDGE_SENSITIVE, ACPI_ACTIVE_LOW, 1, true },
 	{ maingear_laptop, 1, ACPI_EDGE_SENSITIVE, ACPI_ACTIVE_LOW, 1, true },
+	{ lg_laptop, 1, ACPI_LEVEL_SENSITIVE, ACPI_ACTIVE_LOW, 0, false },
 };
 
 static bool acpi_dev_irq_override(u32 gsi, u8 triggering, u8 polarity,
Signature · 1 line
2.39.2
author maintainer
#131 lore raw

PATCH 131/286 media: dvb-usb: dw2102: fix uninit-value in su3000_read_mac_address

In reply to Greg Kroah-Hartman

From: Wei Chen <[email protected]>

[ Upstream commit a3fd1ef27aa686d871cefe207bd6168c4b0cd29e ]

In su3000_read_mac_address, if i2c_transfer fails to execute two messages, array mac address will not be initialized. Without handling such error, later in function dvb_usb_adapter_dvb_init, proposed_mac is accessed before initialization.

Fix this error by returning a negative value if message execution fails.

---
 drivers/media/usb/dvb-usb/dw2102.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/usb/dvb-usb/dw2102.c +1 -1
index 0ca764282c767..8747960e61461 100644
@@ -946,7 +946,7 @@ static int su3000_read_mac_address(struct dvb_usb_device *d, u8 mac[6])
 	for (i = 0; i < 6; i++) {
 		obuf[1] = 0xf0 + i;
 		if (i2c_transfer(&d->i2c_adap, msg, 2) != 2)
-			break;
+			return -1;
 		else
 			mac[i] = ibuf[0];
 	}
Signature · 1 line
2.39.2
author maintainer
#152 lore raw

PATCH 149/286 wifi: rtl8xxxu: fix authentication timeout due to incorrect RCR value

In reply to Greg Kroah-Hartman

From: Yun Lu <[email protected]>

[ Upstream commit 20429444e653ee8242dfbf815c0c37866beb371b ]

When using rtl8192cu with rtl8xxxu driver to connect wifi, there is a probability of failure, which shows "authentication with ... timed out". Through debugging, it was found that the RCR register has been inexplicably modified to an incorrect value, resulting in the nic not being able to receive authenticated frames.

To fix this problem, add regrcr in rtl8xxxu_priv struct, and store the RCR value every time the register is written, and use it the next time the register need to be modified.

---
 drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu.h      | 1 +
 drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu_core.c | 4 +++-
 2 files changed, 4 insertions(+), 1 deletion(-)
drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu.h +1 -0
index c8cee4a247551..4088aaa1c618d 100644
@@ -1518,6 +1518,7 @@ struct rtl8xxxu_priv {
 	u32 rege9c;
 	u32 regeb4;
 	u32 regebc;
+	u32 regrcr;
 	int next_mbox;
 	int nr_out_eps;
 
drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu_core.c +3 -1
index 54ca6f2ced3f3..74ff5130971e2 100644
@@ -4049,6 +4049,7 @@ static int rtl8xxxu_init_device(struct ieee80211_hw *hw)
 		RCR_ACCEPT_MGMT_FRAME | RCR_HTC_LOC_CTRL |
 		RCR_APPEND_PHYSTAT | RCR_APPEND_ICV | RCR_APPEND_MIC;
 	rtl8xxxu_write32(priv, REG_RCR, val32);
+	priv->regrcr = val32;
 
 	if (priv->rtl_chip == RTL8188F) {
 		/* Accept all data frames */
@@ -6269,7 +6270,7 @@ static void rtl8xxxu_configure_filter(struct ieee80211_hw *hw,
 				      unsigned int *total_flags, u64 multicast)
 {
 	struct rtl8xxxu_priv *priv = hw->priv;
-	u32 rcr = rtl8xxxu_read32(priv, REG_RCR);
+	u32 rcr = priv->regrcr;
 
 	dev_dbg(&priv->udev->dev, "%s: changed_flags %08x, total_flags %08x\n",
 		__func__, changed_flags, *total_flags);
@@ -6315,6 +6316,7 @@ static void rtl8xxxu_configure_filter(struct ieee80211_hw *hw,
 	 */
 
 	rtl8xxxu_write32(priv, REG_RCR, rcr);
+	priv->regrcr = rcr;
 
 	*total_flags &= (FIF_ALLMULTI | FIF_FCSFAIL | FIF_BCN_PRBRESP_PROMISC |
 			 FIF_CONTROL | FIF_OTHER_BSS | FIF_PSPOLL |
Signature · 1 line
2.39.2
author maintainer
#153 lore raw

PATCH 150/286 ARM: dts: stm32: add pin map for CAN controller on stm32f7

In reply to Greg Kroah-Hartman

From: Dario Binacchi <[email protected]>

[ Upstream commit 011644249686f2675e142519cd59e81e04cfc231 ]

Add pin configurations for using CAN controller on stm32f7.

---
 arch/arm/boot/dts/stm32f7-pinctrl.dtsi | 82 ++++++++++++++++++++++++++
 1 file changed, 82 insertions(+)
arch/arm/boot/dts/stm32f7-pinctrl.dtsi +82 -0
index c8e6c52fb248e..9f65403295ca0 100644
@@ -283,6 +283,88 @@
 					slew-rate = <2>;
 				};
 			};
+
+			can1_pins_a: can1-0 {
+				pins1 {
+					pinmux = <STM32_PINMUX('A', 12, AF9)>; /* CAN1_TX */
+				};
+				pins2 {
+					pinmux = <STM32_PINMUX('A', 11, AF9)>; /* CAN1_RX */
+					bias-pull-up;
+				};
+			};
+
+			can1_pins_b: can1-1 {
+				pins1 {
+					pinmux = <STM32_PINMUX('B', 9, AF9)>; /* CAN1_TX */
+				};
+				pins2 {
+					pinmux = <STM32_PINMUX('B', 8, AF9)>; /* CAN1_RX */
+					bias-pull-up;
+				};
+			};
+
+			can1_pins_c: can1-2 {
+				pins1 {
+					pinmux = <STM32_PINMUX('D', 1, AF9)>; /* CAN1_TX */
+				};
+				pins2 {
+					pinmux = <STM32_PINMUX('D', 0, AF9)>; /* CAN1_RX */
+					bias-pull-up;
+
+				};
+			};
+
+			can1_pins_d: can1-3 {
+				pins1 {
+					pinmux = <STM32_PINMUX('H', 13, AF9)>; /* CAN1_TX */
+				};
+				pins2 {
+					pinmux = <STM32_PINMUX('H', 14, AF9)>; /* CAN1_RX */
+					bias-pull-up;
+
+				};
+			};
+
+			can2_pins_a: can2-0 {
+				pins1 {
+					pinmux = <STM32_PINMUX('B', 6, AF9)>; /* CAN2_TX */
+				};
+				pins2 {
+					pinmux = <STM32_PINMUX('B', 5, AF9)>; /* CAN2_RX */
+					bias-pull-up;
+				};
+			};
+
+			can2_pins_b: can2-1 {
+				pins1 {
+					pinmux = <STM32_PINMUX('B', 13, AF9)>; /* CAN2_TX */
+				};
+				pins2 {
+					pinmux = <STM32_PINMUX('B', 12, AF9)>; /* CAN2_RX */
+					bias-pull-up;
+				};
+			};
+
+			can3_pins_a: can3-0 {
+				pins1 {
+					pinmux = <STM32_PINMUX('A', 15, AF11)>; /* CAN3_TX */
+				};
+				pins2 {
+					pinmux = <STM32_PINMUX('A', 8, AF11)>; /* CAN3_RX */
+					bias-pull-up;
+				};
+			};
+
+			can3_pins_b: can3-1 {
+				pins1 {
+					pinmux = <STM32_PINMUX('B', 4, AF11)>;  /* CAN3_TX */
+				};
+				pins2 {
+					pinmux = <STM32_PINMUX('B', 3, AF11)>; /* CAN3_RX */
+					bias-pull-up;
+				};
+			};
 		};
 	};
 };
Signature · 1 line
2.39.2
author maintainer
#154 lore raw

PATCH 151/286 arm64/mm: mark private VM_FAULT_X defines as vm_fault_t

In reply to Greg Kroah-Hartman

From: Min-Hua Chen <[email protected]>

[ Upstream commit d91d580878064b880f3574ac35b98d8b70ee8620 ]

This patch fixes several sparse warnings for fault.c:

arch/arm64/mm/fault.c:493:24: sparse: warning: incorrect type in return expression (different base types)
arch/arm64/mm/fault.c:493:24: sparse: expected restricted vm_fault_t
arch/arm64/mm/fault.c:493:24: sparse: got int
arch/arm64/mm/fault.c:501:32: sparse: warning: incorrect type in return expression (different base types)
arch/arm64/mm/fault.c:501:32: sparse: expected restricted vm_fault_t
arch/arm64/mm/fault.c:501:32: sparse: got int
arch/arm64/mm/fault.c:503:32: sparse: warning: incorrect type in return expression (different base types)
arch/arm64/mm/fault.c:503:32: sparse: expected restricted vm_fault_t
arch/arm64/mm/fault.c:503:32: sparse: got int
arch/arm64/mm/fault.c:511:24: sparse: warning: incorrect type in return expression (different base types)
arch/arm64/mm/fault.c:511:24: sparse: expected restricted vm_fault_t
arch/arm64/mm/fault.c:511:24: sparse: got int
arch/arm64/mm/fault.c:670:13: sparse: warning: restricted vm_fault_t degrades to integer arch/arm64/mm/fault.c:670:13: sparse: warning: restricted vm_fault_t degrades to integer arch/arm64/mm/fault.c:713:39: sparse: warning: restricted vm_fault_t degrades to integer

---
 arch/arm64/mm/fault.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/arm64/mm/fault.c +2 -2
index f4cb0f85ccf49..d1136259b7b85 100644
@@ -480,8 +480,8 @@ static void do_bad_area(unsigned long far, unsigned long esr,
 	}
 }
 
-#define VM_FAULT_BADMAP		0x010000
-#define VM_FAULT_BADACCESS	0x020000
+#define VM_FAULT_BADMAP		((__force vm_fault_t)0x010000)
+#define VM_FAULT_BADACCESS	((__force vm_fault_t)0x020000)
 
 static vm_fault_t __do_page_fault(struct mm_struct *mm, unsigned long addr,
 				  unsigned int mm_flags, unsigned long vm_flags,
Signature · 1 line
2.39.2
author maintainer
#155 lore raw

PATCH 152/286 arm64: vdso: Pass (void *) to virt_to_page()

In reply to Greg Kroah-Hartman

From: Linus Walleij <[email protected]>

[ Upstream commit b0abde80620f42d1ceb3de5e4c1a49cdd5628229 ]

Like the other calls in this function virt_to_page() expects a pointer, not an integer.

However since many architectures implement virt_to_pfn() as a macro, this function becomes polymorphic and accepts both a (unsigned long) and a (void *).

Fix this up with an explicit cast.

---
 arch/arm64/kernel/vdso.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/kernel/vdso.c +1 -1
index 0119dc91abb5d..d9e1355730ef5 100644
@@ -288,7 +288,7 @@ static int aarch32_alloc_kuser_vdso_page(void)
 
 	memcpy((void *)(vdso_page + 0x1000 - kuser_sz), __kuser_helper_start,
 	       kuser_sz);
-	aarch32_vectors_page = virt_to_page(vdso_page);
+	aarch32_vectors_page = virt_to_page((void *)vdso_page);
 	return 0;
 }
 
Signature · 1 line
2.39.2
author maintainer
#156 lore raw

PATCH 153/286 wifi: mac80211: simplify chanctx allocation

In reply to Greg Kroah-Hartman

From: Johannes Berg <[email protected]>

[ Upstream commit 860e1b43da94551cd1e73adc36b3c64cc3e5dc01 ]

There's no need to call ieee80211_recalc_chanctx_min_def() since it cannot and won't call the driver anyway; just use _ieee80211_recalc_chanctx_min_def() instead.

---
 net/mac80211/chan.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/mac80211/chan.c +1 -1
index dbc34fbe7c8f4..d23d1a7b4cc39 100644
@@ -638,7 +638,7 @@ ieee80211_alloc_chanctx(struct ieee80211_local *local,
 	ctx->conf.rx_chains_dynamic = 1;
 	ctx->mode = mode;
 	ctx->conf.radar_enabled = false;
-	ieee80211_recalc_chanctx_min_def(local, ctx);
+	_ieee80211_recalc_chanctx_min_def(local, ctx);
 
 	return ctx;
 }
Signature · 1 line
2.39.2