5.6.11-rc1 review

This is the start of the stable review cycle for the 5.6.11 release. There are 73 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 Wed, 06 May 2020 16:52:55 +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/v5.x/stable-review/patch-5.6.11-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-5.6.y
and the diffstat can be found below.

thanks,

greg k-h

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

Greg Kroah-Hartman <[email protected]>

    Linux 5.6.11-rc1

Jens Axboe <[email protected]>
io_uring: statx must grab the file table for valid fd

Vincenzo Frascino <[email protected]>
arm64: vdso: Add -fasynchronous-unwind-tables to cflags

Andy Shevchenko <[email protected]>
dmaengine: dmatest: Fix process hang when reading 'wait' parameter

Matt Roper <[email protected]>
drm/i915: Use proper fault mask in interrupt postinstall too

Andy Shevchenko <[email protected]>
dmaengine: dmatest: Fix iteration non-stop logic

Andreas Gruenbacher <[email protected]>
nfs: Fix potential posix_acl refcnt leak in nfs3_set_acl

Xiyu Yang <[email protected]>
drm/i915/selftests: Fix i915_address_space refcnt leak

Niklas Cassel <[email protected]>
nvme: prevent double free in nvme_alloc_ns() error handling

David Howells <[email protected]>
Fix use after free in get_tree_bdev()

Arnd Bergmann <[email protected]>
ALSA: opti9xx: shut up gcc-10 range warning

ryan_chen <[email protected]>
i2c: aspeed: Avoid i2c interrupt status clear race condition.

Suravee Suthikulpanit <[email protected]>
iommu/amd: Fix legacy interrupt remapping for x2APIC-enabled system

Lu Baolu <[email protected]>
iommu/vt-d: Use right Kconfig option name

Greg Kroah-Hartman <[email protected]>
iommu: Properly export iommu_group_get_for_dev()

David Disseldorp <[email protected]>
scsi: target/iblock: fix WRITE SAME zeroing

Dave Jiang <[email protected]>
dmaengine: fix channel index enumeration

Grygorii Strashko <[email protected]>
dmaengine: ti: k3-psil: fix deadlock on error path

Tang Bin <[email protected]>
iommu/qcom: Fix local_base status check

Sean Christopherson <[email protected]>

    vfio/type1: Fix VA->PA translation for PFNMAP VMAs in vaddr_get_pfn()

Yan Zhao <[email protected]>
vfio: avoid possible overflow in vfio_iommu_type1_pin_pages

YueHaibing <[email protected]>
dmaengine: hisilicon: Fix build error without PCI_MSI

Rayagonda Kokatanur <[email protected]>
i2c: iproc: generate stop event for slave writes

Dan Carpenter <[email protected]>
RDMA/cm: Fix an error check in cm_alloc_id_priv()

Jason Gunthorpe <[email protected]>
RDMA/cm: Fix ordering of xa_alloc_cyclic() in ib_create_cm_id()

Leon Romanovsky <[email protected]>
RDMA/core: Fix race between destroy and release FD object

Leon Romanovsky <[email protected]>
RDMA/core: Fix overwriting of uobj in case of error

Leon Romanovsky <[email protected]>
RDMA/core: Prevent mixed use of FDs between shared ufiles

Jason Gunthorpe <[email protected]>
RDMA/siw: Fix potential siw_mem refcnt leak in siw_fastreg_mr()

Alaa Hleihel <[email protected]>
RDMA/mlx4: Initialize ib_spec on the stack

Jason Gunthorpe <[email protected]>
RDMA/uverbs: Fix a race with disassociate and exit_mmap()

Aharon Landau <[email protected]>
RDMA/mlx5: Set GRH fields in query QP on RoCE

Martin Wilck <[email protected]>
scsi: qla2xxx: check UNLOADING before posting async work

Martin Wilck <[email protected]>
scsi: qla2xxx: set UNLOADING before waiting for session deletion

Christoph Hellwig <[email protected]>
block: remove the bd_openers checks in blk_drop_partitions

Russell King <[email protected]>
ARM: dts: imx6qdl-sr-som-ti: indicate powering off wifi is safe

Gabriel Krisman Bertazi <[email protected]>
dm multipath: use updated MPATHF_QUEUE_IO on mapping for bio-based mpath

Mikulas Patocka <[email protected]>
dm writecache: fix data corruption when reloading the target

Sunwook Eom <[email protected]>
dm verity fec: fix hash block number in verity_fec_decode

Dexuan Cui <[email protected]>
PM: hibernate: Freeze kernel threads in software_resume()

Kai-Heng Feng <[email protected]>
PM: ACPI: Output correct message on target power state

Sudip Mukherjee <[email protected]>
IB/rdmavt: Always return ERR_PTR from rvt_create_mmap_info()

Paul Moore <[email protected]>
selinux: properly handle multiple messages in selinux_netlink_send()

Al Viro <[email protected]>
dlmfs_file_write(): fix the bogosity in handling non-zero *ppos

Dexuan Cui <[email protected]>
Drivers: hv: vmbus: Fix Suspend-to-Idle for Generation-2 VM

Dexuan Cui <[email protected]>
x86/hyperv: Suspend/resume the VP assist page for hibernation

Dan Carpenter <[email protected]>
i2c: amd-mp2-pci: Fix Oops in amd_mp2_pci_init() error handling

Takashi Iwai <[email protected]>
ALSA: pcm: oss: Place the plugin buffer overflow checks correctly

Vasily Khoruzhick <[email protected]>
ALSA: line6: Fix POD HD500 audio playback

Wu Bo <[email protected]>
ALSA: hda/hdmi: fix without unlocked before return

Takashi Iwai <[email protected]>
ALSA: usb-audio: Correct a typo of NuPrime DAC-10 USB ID

Hui Wang <[email protected]>
ALSA: hda/realtek - Two front mics on a Lenovo ThinkCenter

Iuliana Prodan <[email protected]>
crypto: caam - fix the address of the last entry of S/G

Martin Blumenstingl <[email protected]>

    mmc: meson-mx-sdio: remove the broken ->card_busy() op

Martin Blumenstingl <[email protected]>
mmc: meson-mx-sdio: Set MMC_CAP_WAIT_WHILE_BUSY

Veerabhadrarao Badiganti <[email protected]>
mmc: sdhci-msm: Enable host capabilities pertains to R1b response

Adrian Hunter <[email protected]>
mmc: sdhci-pci: Fix eMMC driver strength for BYT-based controllers

Marek Behún <[email protected]>
mmc: sdhci-xenon: fix annoying 1.8V regulator warning

Douglas Anderson <[email protected]>
mmc: cqhci: Avoid false "cqhci: CQE stuck on" by not open-coding timeout loop

Qu Wenruo <[email protected]>
btrfs: transaction: Avoid deadlock due to bad initialization timing of fs_info::journal_info

Filipe Manana <[email protected]>
btrfs: fix partial loss of prealloc extent past i_size after fsync

Xiyu Yang <[email protected]>
btrfs: fix block group leak when removing fails

Xiyu Yang <[email protected]>
btrfs: fix transaction leak in btrfs_recover_relocation

Olga Kornievskaia <[email protected]>
NFSv4.1: fix handling of backchannel binding in BIND_CONN_TO_SESSION

Vasily Averin <[email protected]>
drm/qxl: qxl_release use after free

Vasily Averin <[email protected]>
drm/qxl: qxl_release leak in qxl_hw_surface_alloc()

Vasily Averin <[email protected]>
drm/qxl: qxl_release leak in qxl_draw_dirty_fb()

Chris Wilson <[email protected]>
drm/i915/gt: Check cacheline is valid before acquiring

Chris Wilson <[email protected]>

    drm/i915/gem: Hold obj->vma.lock over for_each_ggtt_vma()

Rodrigo Siqueira <[email protected]>
drm/amd/display: Fix green screen issue after suspend

Ville Syrjälä <[email protected]>
drm/edid: Fix off-by-one in DispID DTD pixel clock

Marek Olšák <[email protected]>
drm/amdgpu: invalidate L2 before SDMA IBs (v2)

Daniel Vetter <[email protected]>
dma-buf: Fix SET_NAME ioctl uapi

Christian König <[email protected]>
drm/scheduler: fix drm_sched_get_cleanup_job

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

Diffstat:

 Makefile                                          |  4 +-
 arch/arm/boot/dts/imx6qdl-sr-som-ti.dtsi          |  1 +
 arch/arm64/kernel/vdso/Makefile                   |  2 +-
 arch/x86/hyperv/hv_init.c                         | 12 +++-
 block/partition-generic.c                         |  2 +-
 drivers/acpi/device_pm.c                          |  4 +-
 drivers/crypto/caam/caamalg.c                     |  2 +-
 drivers/dma-buf/dma-buf.c                         |  3 +-
 drivers/dma/Kconfig                               |  3 +-
 drivers/dma/dmaengine.c                           | 60 +++++++++----------
 drivers/dma/dmatest.c                             |  6 +-
 drivers/dma/ti/k3-psil.c                          |  1 +
 drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h | 16 ++++++
 drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c            | 14 ++++-
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 38 +++++++++---
 drivers/gpu/drm/drm_edid.c                        |  2 +-
 drivers/gpu/drm/i915/gem/i915_gem_tiling.c        | 20 ++++++-
 drivers/gpu/drm/i915/gem/selftests/huge_pages.c   | 12 ++--
 drivers/gpu/drm/i915/gt/intel_timeline.c          |  2 +
 drivers/gpu/drm/i915/i915_irq.c                   |  6 +-
 drivers/gpu/drm/i915/i915_vma.c                   | 10 ++--
 drivers/gpu/drm/qxl/qxl_cmd.c                     | 10 ++--
 drivers/gpu/drm/qxl/qxl_display.c                 |  6 +-
 drivers/gpu/drm/qxl/qxl_draw.c                    |  7 ++-
 drivers/gpu/drm/qxl/qxl_ioctl.c                   |  5 +-
 drivers/gpu/drm/scheduler/sched_main.c            |  2 +-
 drivers/hv/vmbus_drv.c                            | 43 +++++++++++---
 drivers/i2c/busses/i2c-amd-mp2-pci.c              |  2 +-
 drivers/i2c/busses/i2c-aspeed.c                   |  5 +-
 drivers/i2c/busses/i2c-bcm-iproc.c                |  3 +
 drivers/infiniband/core/cm.c                      | 27 ++++-----
 drivers/infiniband/core/rdma_core.c               |  9 ++-
 drivers/infiniband/core/uverbs_main.c             |  4 ++
 drivers/infiniband/hw/mlx4/main.c                 |  3 +-
 drivers/infiniband/hw/mlx5/qp.c                   |  4 +-
 drivers/infiniband/sw/rdmavt/cq.c                 |  4 +-
 drivers/infiniband/sw/rdmavt/mmap.c               |  4 +-
 drivers/infiniband/sw/rdmavt/qp.c                 |  4 +-
 drivers/infiniband/sw/rdmavt/srq.c                |  4 +-
 drivers/infiniband/sw/siw/siw_qp_tx.c             | 15 +++--
 drivers/iommu/amd_iommu_init.c                    |  2 +-
 drivers/iommu/intel-iommu.c                       |  4 +-
 drivers/iommu/iommu.c                             |  2 +-
 drivers/iommu/qcom_iommu.c                        |  5 +-
 drivers/md/dm-mpath.c                             |  6 +-
 drivers/md/dm-verity-fec.c                        |  2 +-
 drivers/md/dm-writecache.c                        | 52 ++++++++++++-----
 drivers/mmc/host/cqhci.c                          | 21 ++++---
 drivers/mmc/host/meson-mx-sdio.c                  | 11 +---
 drivers/mmc/host/sdhci-msm.c                      |  2 +
 drivers/mmc/host/sdhci-pci-core.c                 |  3 +
 drivers/mmc/host/sdhci-xenon.c                    | 10 ++++
 drivers/nvme/host/core.c                          |  2 +
 drivers/scsi/qla2xxx/qla_os.c                     | 35 ++++++------
 drivers/target/target_core_iblock.c               |  2 +-
 drivers/vfio/vfio_iommu_type1.c                   |  6 +-
 fs/btrfs/block-group.c                            | 16 ++++--
 fs/btrfs/relocation.c                             |  1 +
 fs/btrfs/transaction.c                            | 13 ++++-
 fs/btrfs/tree-log.c                               | 43 +++++++++++++-
 fs/io_uring.c                                     | 12 +++-
 fs/nfs/nfs3acl.c                                  | 22 ++++---
 fs/nfs/nfs4proc.c                                 |  8 +++
 fs/ocfs2/dlmfs/dlmfs.c                            | 27 ++++-----
 fs/super.c                                        |  2 +-
 include/linux/dmaengine.h                         |  4 +-
 include/linux/nfs_xdr.h                           |  2 +
 include/linux/sunrpc/clnt.h                       |  5 ++
 include/uapi/linux/dma-buf.h                      |  6 ++
 kernel/power/hibernate.c                          |  7 +++
 security/selinux/hooks.c                          | 70 +++++++++++++++--------
 sound/core/oss/pcm_plugin.c                       | 20 ++++---
 sound/isa/opti9xx/miro.c                          |  9 ++-
 sound/isa/opti9xx/opti92x-ad1848.c                |  9 ++-
 sound/pci/hda/patch_hdmi.c                        |  4 +-
 sound/pci/hda/patch_realtek.c                     |  1 +
 sound/usb/line6/podhd.c                           | 22 ++-----
 sound/usb/quirks.c                                |  2 +-
 78 files changed, 554 insertions(+), 297 deletions(-)

From: Christian König <[email protected]>

commit 8623b5255ae7ccaf276aac3920787bf575fa6b37 upstream.

We are racing to initialize sched->thread here, just always check the current thread.

---
 drivers/gpu/drm/scheduler/sched_main.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -687,7 +687,7 @@ drm_sched_get_cleanup_job(struct drm_gpu
 	 */
 	if ((sched->timeout != MAX_SCHEDULE_TIMEOUT &&
 	    !cancel_delayed_work(&sched->work_tdr)) ||
-	    __kthread_should_park(sched->thread))
+	    kthread_should_park())
 		return NULL;
 
 	spin_lock_irqsave(&sched->job_list_lock, flags);

From: Daniel Vetter <[email protected]>

commit a5bff92eaac45bdf6221badf9505c26792fdf99e upstream.

The uapi is the same on 32 and 64 bit, but the number isn't. Everyone who botched this please re-read:

https://www.kernel.org/doc/html/v5.4-preprc-cpu/ioctl/botching-up-ioctls.html

Also, the type argument for the ioctl macros is for the type the void __user *arg pointer points at, which in this case would be the variable-sized char[] of a 0 terminated string. So this was botched in more than just the usual ways.

---
 drivers/dma-buf/dma-buf.c    |    3 ++-
 include/uapi/linux/dma-buf.h |    6 ++++++
 2 files changed, 8 insertions(+), 1 deletion(-)

@@ -388,7 +388,8 @@ static long dma_buf_ioctl(struct file *f
 
 		return ret;
 
-	case DMA_BUF_SET_NAME:
+	case DMA_BUF_SET_NAME_A:
+	case DMA_BUF_SET_NAME_B:
 		return dma_buf_set_name(dmabuf, (const char __user *)arg);
 
 	default:

@@ -39,6 +39,12 @@ struct dma_buf_sync {
 
 #define DMA_BUF_BASE		'b'
 #define DMA_BUF_IOCTL_SYNC	_IOW(DMA_BUF_BASE, 0, struct dma_buf_sync)
+
+/* 32/64bitness of this uapi was botched in android, there's no difference
+ * between them in actual uapi, they're just different numbers.
+ */
 #define DMA_BUF_SET_NAME	_IOW(DMA_BUF_BASE, 1, const char *)
+#define DMA_BUF_SET_NAME_A	_IOW(DMA_BUF_BASE, 1, u32)
+#define DMA_BUF_SET_NAME_B	_IOW(DMA_BUF_BASE, 1, u64)
 
 #endif

From: Marek Olšák <[email protected]>

commit fdf83646c0542ecfb9adc4db8f741a1f43dca058 upstream.

This fixes GPU hangs due to cache coherency issues.

v2: Split the version bump to a separate patch

---
 drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h |   16 ++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c            |   14 +++++++++++++-
 2 files changed, 29 insertions(+), 1 deletion(-)

@@ -73,6 +73,22 @@
 #define SDMA_OP_AQL_COPY  0
 #define SDMA_OP_AQL_BARRIER_OR  0
 
+#define SDMA_GCR_RANGE_IS_PA		(1 << 18)
+#define SDMA_GCR_SEQ(x)			(((x) & 0x3) << 16)
+#define SDMA_GCR_GL2_WB			(1 << 15)
+#define SDMA_GCR_GL2_INV		(1 << 14)
+#define SDMA_GCR_GL2_DISCARD		(1 << 13)
+#define SDMA_GCR_GL2_RANGE(x)		(((x) & 0x3) << 11)
+#define SDMA_GCR_GL2_US			(1 << 10)
+#define SDMA_GCR_GL1_INV		(1 << 9)
+#define SDMA_GCR_GLV_INV		(1 << 8)
+#define SDMA_GCR_GLK_INV		(1 << 7)
+#define SDMA_GCR_GLK_WB			(1 << 6)
+#define SDMA_GCR_GLM_INV		(1 << 5)
+#define SDMA_GCR_GLM_WB			(1 << 4)
+#define SDMA_GCR_GL1_RANGE(x)		(((x) & 0x3) << 2)
+#define SDMA_GCR_GLI_INV(x)		(((x) & 0x3) << 0)
+
 /*define for op field*/
 #define SDMA_PKT_HEADER_op_offset 0
 #define SDMA_PKT_HEADER_op_mask   0x000000FF

@@ -382,6 +382,18 @@ static void sdma_v5_0_ring_emit_ib(struc
 	unsigned vmid = AMDGPU_JOB_GET_VMID(job);
 	uint64_t csa_mc_addr = amdgpu_sdma_get_csa_mc_addr(ring, vmid);
 
+	/* Invalidate L2, because if we don't do it, we might get stale cache
+	 * lines from previous IBs.
+	 */
+	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_GCR_REQ));
+	amdgpu_ring_write(ring, 0);
+	amdgpu_ring_write(ring, (SDMA_GCR_GL2_INV |
+				 SDMA_GCR_GL2_WB |
+				 SDMA_GCR_GLM_INV |
+				 SDMA_GCR_GLM_WB) << 16);
+	amdgpu_ring_write(ring, 0xffffff80);
+	amdgpu_ring_write(ring, 0xffff);
+
 	/* An IB packet must end on a 8 DW boundary--the next dword
 	 * must be on a 8-dword boundary. Our IB packet below is 6
 	 * dwords long, thus add x number of NOPs, such that, in

@@ -1597,7 +1609,7 @@ static const struct amdgpu_ring_funcs sd
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 6 * 2 +
 		10 + 10 + 10, /* sdma_v5_0_ring_emit_fence x3 for user fence, vm fence */
-	.emit_ib_size = 7 + 6, /* sdma_v5_0_ring_emit_ib */
+	.emit_ib_size = 5 + 7 + 6, /* sdma_v5_0_ring_emit_ib */
 	.emit_ib = sdma_v5_0_ring_emit_ib,
 	.emit_fence = sdma_v5_0_ring_emit_fence,
 	.emit_pipeline_sync = sdma_v5_0_ring_emit_pipeline_sync,

From: Ville Syrjälä <[email protected]>

commit 6292b8efe32e6be408af364132f09572aed14382 upstream.

The DispID DTD pixel clock is documented as:
"00 00 00 h → FF FF FF h | Pixel clock ÷ 10,000 0.01 → 167,772.16 Mega Pixels per Sec" Which seems to imply that we to add one to the raw value.

Reality seems to agree as there are tiled displays in the wild which currently show a 10kHz difference in the pixel clock between the tiles (one tile gets its mode from the base EDID, the other from the DispID block).

---
 drivers/gpu/drm/drm_edid.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -5009,7 +5009,7 @@ static struct drm_display_mode *drm_mode
 	struct drm_display_mode *mode;
 	unsigned pixel_clock = (timings->pixel_clock[0] |
 				(timings->pixel_clock[1] << 8) |
-				(timings->pixel_clock[2] << 16));
+				(timings->pixel_clock[2] << 16)) + 1;
 	unsigned hactive = (timings->hactive[0] | timings->hactive[1] << 8) + 1;
 	unsigned hblank = (timings->hblank[0] | timings->hblank[1] << 8) + 1;
 	unsigned hsync = (timings->hsync[0] | (timings->hsync[1] & 0x7f) << 8) + 1;

From: Rodrigo Siqueira <[email protected]>

commit 87b7ebc2e16c14d32a912f18206a4d6cc9abc3e8 upstream.

[why]
We have seen a green screen after resume from suspend in a Raven system connected with two displays (HDMI and DP) on X based system. We noticed that this issue is related to bad DCC metadata from user space which may generate hangs and consequently an underflow on HUBP. After taking a deep look at the code path we realized that after resume we try to restore the commit with the DCC enabled framebuffer but the framebuffer is no longer valid.

[how]
This problem was only reported on Raven based system and after suspend, for this reason, this commit adds a new parameter on fill_plane_dcc_attributes() to give the option of disabling DCC programmatically. In summary, for disabling DCC we first verify if is a Raven system and if it is in suspend; if both conditions are true we disable DCC temporarily, otherwise, it is enabled.

---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |   38 ++++++++++++++++------
 1 file changed, 29 insertions(+), 9 deletions(-)

@@ -3212,7 +3212,8 @@ fill_plane_dcc_attributes(struct amdgpu_
 			  const union dc_tiling_info *tiling_info,
 			  const uint64_t info,
 			  struct dc_plane_dcc_param *dcc,
-			  struct dc_plane_address *address)
+			  struct dc_plane_address *address,
+			  bool force_disable_dcc)
 {
 	struct dc *dc = adev->dm.dc;
 	struct dc_dcc_surface_param input;

@@ -3224,6 +3225,9 @@ fill_plane_dcc_attributes(struct amdgpu_
 	memset(&input, 0, sizeof(input));
 	memset(&output, 0, sizeof(output));
 
+	if (force_disable_dcc)
+		return 0;
+
 	if (!offset)
 		return 0;
 

@@ -3273,7 +3277,8 @@ fill_plane_buffer_attributes(struct amdg
 			     union dc_tiling_info *tiling_info,
 			     struct plane_size *plane_size,
 			     struct dc_plane_dcc_param *dcc,
-			     struct dc_plane_address *address)
+			     struct dc_plane_address *address,
+			     bool force_disable_dcc)
 {
 	const struct drm_framebuffer *fb = &afb->base;
 	int ret;

@@ -3379,7 +3384,8 @@ fill_plane_buffer_attributes(struct amdg
 
 		ret = fill_plane_dcc_attributes(adev, afb, format, rotation,
 						plane_size, tiling_info,
-						tiling_flags, dcc, address);
+						tiling_flags, dcc, address,
+						force_disable_dcc);
 		if (ret)
 			return ret;
 	}

@@ -3471,7 +3477,8 @@ fill_dc_plane_info_and_addr(struct amdgp
 			    const struct drm_plane_state *plane_state,
 			    const uint64_t tiling_flags,
 			    struct dc_plane_info *plane_info,
-			    struct dc_plane_address *address)
+			    struct dc_plane_address *address,
+			    bool force_disable_dcc)
 {
 	const struct drm_framebuffer *fb = plane_state->fb;
 	const struct amdgpu_framebuffer *afb =

@@ -3550,7 +3557,8 @@ fill_dc_plane_info_and_addr(struct amdgp
 					   plane_info->rotation, tiling_flags,
 					   &plane_info->tiling_info,
 					   &plane_info->plane_size,
-					   &plane_info->dcc, address);
+					   &plane_info->dcc, address,
+					   force_disable_dcc);
 	if (ret)
 		return ret;
 

@@ -3573,6 +3581,7 @@ static int fill_dc_plane_attributes(stru
 	struct dc_plane_info plane_info;
 	uint64_t tiling_flags;
 	int ret;
+	bool force_disable_dcc = false;
 
 	ret = fill_dc_scaling_info(plane_state, &scaling_info);
 	if (ret)

@@ -3587,9 +3596,11 @@ static int fill_dc_plane_attributes(stru
 	if (ret)
 		return ret;
 
+	force_disable_dcc = adev->asic_type == CHIP_RAVEN && adev->in_suspend;
 	ret = fill_dc_plane_info_and_addr(adev, plane_state, tiling_flags,
 					  &plane_info,
-					  &dc_plane_state->address);
+					  &dc_plane_state->address,
+					  force_disable_dcc);
 	if (ret)
 		return ret;
 

@@ -5171,6 +5182,7 @@ static int dm_plane_helper_prepare_fb(st
 	uint64_t tiling_flags;
 	uint32_t domain;
 	int r;
+	bool force_disable_dcc = false;
 
 	dm_plane_state_old = to_dm_plane_state(plane->state);
 	dm_plane_state_new = to_dm_plane_state(new_state);

@@ -5229,11 +5241,13 @@ static int dm_plane_helper_prepare_fb(st
 			dm_plane_state_old->dc_state != dm_plane_state_new->dc_state) {
 		struct dc_plane_state *plane_state = dm_plane_state_new->dc_state;
 
+		force_disable_dcc = adev->asic_type == CHIP_RAVEN && adev->in_suspend;
 		fill_plane_buffer_attributes(
 			adev, afb, plane_state->format, plane_state->rotation,
 			tiling_flags, &plane_state->tiling_info,
 			&plane_state->plane_size, &plane_state->dcc,
-			&plane_state->address);
+			&plane_state->address,
+			force_disable_dcc);
 	}
 
 	return 0;

@@ -6514,7 +6528,12 @@ static void amdgpu_dm_commit_planes(stru
 		fill_dc_plane_info_and_addr(
 			dm->adev, new_plane_state, tiling_flags,
 			&bundle->plane_infos[planes_count],
-			&bundle->flip_addrs[planes_count].address);
+			&bundle->flip_addrs[planes_count].address,
+			false);
+
+		DRM_DEBUG_DRIVER("plane: id=%d dcc_en=%d\n",
+				 new_plane_state->plane->index,
+				 bundle->plane_infos[planes_count].dcc.enable);
 
 		bundle->surface_updates[planes_count].plane_info =
 			&bundle->plane_infos[planes_count];

@@ -7935,7 +7954,8 @@ dm_determine_update_type_for_commit(stru
 				ret = fill_dc_plane_info_and_addr(
 					dm->adev, new_plane_state, tiling_flags,
 					plane_info,
-					&flip_addr->address);
+					&flip_addr->address,
+					false);
 				if (ret)
 					goto cleanup;
 

From: Chris Wilson <[email protected]>

commit f524a774a4ff702bdfbfc094c9dd463ee623252b upstream.

While the ggtt vma are protected by their object lifetime, the list continues until it hits a non-ggtt vma, and that vma is not protected and may be freed as we inspect it. Hence, we require the obj->vma.lock to protect the list as we iterate.

An example of forgetting to hold the obj->vma.lock is

[1642834.464973] general protection fault, probably for non-canonical address 0xdead000000000122: 0000 [#1] SMP PTI
[1642834.464977] CPU: 3 PID: 1954 Comm: Xorg Not tainted 5.6.0-300.fc32.x86_64 #1
[1642834.464979] Hardware name: LENOVO 20ARS25701/20ARS25701, BIOS GJET94WW (2.44 ) 09/14/2017
[1642834.465021] RIP: 0010:i915_gem_object_set_tiling+0x2c0/0x3e0 [i915]
[1642834.465024] Code: 8b 84 24 18 01 00 00 f6 c4 80 74 59 49 8b 94 24 a0 00 00 00 49 8b 84 24 e0 00 00 00 49 8b 74 24 10 48 8b 92 30 01 00 00 89 c7 <80> ba 0a 06 00 00 03 0f 87 86 00 00 00 ba 00 00 08 00 b9 00 00 10
[1642834.465025] RSP: 0018:ffffa98780c77d60 EFLAGS: 00010282
[1642834.465028] RAX: ffff8d232bfb2578 RBX: 0000000000000002 RCX: ffff8d25873a0000
[1642834.465029] RDX: dead000000000122 RSI: fffff0af8ac6e408 RDI: 000000002bfb2578
[1642834.465030] RBP: ffff8d25873a0000 R08: ffff8d252bfb5638 R09: 0000000000000000
[1642834.465031] R10: 0000000000000000 R11: ffff8d252bfb5640 R12: ffffa987801cb8f8
[1642834.465032] R13: 0000000000001000 R14: ffff8d233e972e50 R15: ffff8d233e972d00
[1642834.465034] FS:  00007f6a3d327f00(0000) GS:ffff8d25926c0000(0000) knlGS:0000000000000000
[1642834.465036] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[1642834.465037] CR2: 00007f6a2064d000 CR3: 00000002fb57c001 CR4: 00000000001606e0
[1642834.465038] Call Trace:
[1642834.465083]  i915_gem_set_tiling_ioctl+0x122/0x230 [i915]
[1642834.465121]  ? i915_gem_object_set_tiling+0x3e0/0x3e0 [i915]
[1642834.465151]  drm_ioctl_kernel+0x86/0xd0 [drm]
[1642834.465156]  ? avc_has_perm+0x3b/0x160
[1642834.465178]  drm_ioctl+0x206/0x390 [drm]
[1642834.465216]  ? i915_gem_object_set_tiling+0x3e0/0x3e0 [i915]
[1642834.465221]  ? selinux_file_ioctl+0x122/0x1c0
[1642834.465226]  ? __do_munmap+0x24b/0x4d0
[1642834.465231]  ksys_ioctl+0x82/0xc0
[1642834.465235]  __x64_sys_ioctl+0x16/0x20
[1642834.465238]  do_syscall_64+0x5b/0xf0
[1642834.465243]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
[1642834.465245] RIP: 0033:0x7f6a3d7b047b
[1642834.465247] Code: 0f 1e fa 48 8b 05 1d aa 0c 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 0f 1f 44 00 00 f3 0f 1e fa b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d ed a9 0c 00 f7 d8 64 89 01 48
[1642834.465249] RSP: 002b:00007ffe71adba28 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
[1642834.465251] RAX: ffffffffffffffda RBX: 000055f99048fa40 RCX: 00007f6a3d7b047b
[1642834.465253] RDX: 00007ffe71adba30 RSI: 00000000c0106461 RDI: 000000000000000e
[1642834.465254] RBP: 0000000000000002 R08: 000055f98f3f1798 R09: 0000000000000002
[1642834.465255] R10: 0000000000001000 R11: 0000000000000246 R12: 0000000000000080
[1642834.465257] R13: 000055f98f3f1690 R14: 00000000c0106461 R15: 00007ffe71adba30

Now to take the spinlock during the list iteration, we need to break it down into two phases. In the first phase under the lock, we cannot sleep and so must defer the actual work to a second list, protected by the ggtt->mutex.

We also need to hold the spinlock during creation of a new vma to serialise with updates of the tiling on the object.

---
 drivers/gpu/drm/i915/gem/i915_gem_tiling.c |   20 ++++++++++++++++++--
 drivers/gpu/drm/i915/i915_vma.c            |   10 ++++++----
 2 files changed, 24 insertions(+), 6 deletions(-)

@@ -183,21 +183,35 @@ i915_gem_object_fence_prepare(struct drm
 			      int tiling_mode, unsigned int stride)
 {
 	struct i915_ggtt *ggtt = &to_i915(obj->base.dev)->ggtt;
-	struct i915_vma *vma;
+	struct i915_vma *vma, *vn;
+	LIST_HEAD(unbind);
 	int ret = 0;
 
 	if (tiling_mode == I915_TILING_NONE)
 		return 0;
 
 	mutex_lock(&ggtt->vm.mutex);
+
+	spin_lock(&obj->vma.lock);
 	for_each_ggtt_vma(vma, obj) {
+		GEM_BUG_ON(vma->vm != &ggtt->vm);
+
 		if (i915_vma_fence_prepare(vma, tiling_mode, stride))
 			continue;
 
+		list_move(&vma->vm_link, &unbind);
+	}
+	spin_unlock(&obj->vma.lock);
+
+	list_for_each_entry_safe(vma, vn, &unbind, vm_link) {
 		ret = __i915_vma_unbind(vma);
-		if (ret)
+		if (ret) {
+			/* Restore the remaining vma on an error */
+			list_splice(&unbind, &ggtt->vm.bound_list);
 			break;
+		}
 	}
+
 	mutex_unlock(&ggtt->vm.mutex);
 
 	return ret;

@@ -269,6 +283,7 @@ i915_gem_object_set_tiling(struct drm_i9
 	}
 	mutex_unlock(&obj->mm.lock);
 
+	spin_lock(&obj->vma.lock);
 	for_each_ggtt_vma(vma, obj) {
 		vma->fence_size =
 			i915_gem_fence_size(i915, vma->size, tiling, stride);

@@ -279,6 +294,7 @@ i915_gem_object_set_tiling(struct drm_i9
 		if (vma->fence)
 			vma->fence->dirty = true;
 	}
+	spin_unlock(&obj->vma.lock);
 
 	obj->tiling_and_stride = tiling | stride;
 	i915_gem_object_unlock(obj);

@@ -158,16 +158,18 @@ vma_create(struct drm_i915_gem_object *o
 
 	GEM_BUG_ON(!IS_ALIGNED(vma->size, I915_GTT_PAGE_SIZE));
 
+	spin_lock(&obj->vma.lock);
+
 	if (i915_is_ggtt(vm)) {
 		if (unlikely(overflows_type(vma->size, u32)))
-			goto err_vma;
+			goto err_unlock;
 
 		vma->fence_size = i915_gem_fence_size(vm->i915, vma->size,
 						      i915_gem_object_get_tiling(obj),
 						      i915_gem_object_get_stride(obj));
 		if (unlikely(vma->fence_size < vma->size || /* overflow */
 			     vma->fence_size > vm->total))
-			goto err_vma;
+			goto err_unlock;
 
 		GEM_BUG_ON(!IS_ALIGNED(vma->fence_size, I915_GTT_MIN_ALIGNMENT));
 

@@ -179,8 +181,6 @@ vma_create(struct drm_i915_gem_object *o
 		__set_bit(I915_VMA_GGTT_BIT, __i915_vma_flags(vma));
 	}
 
-	spin_lock(&obj->vma.lock);
-
 	rb = NULL;
 	p = &obj->vma.tree.rb_node;
 	while (*p) {

@@ -225,6 +225,8 @@ vma_create(struct drm_i915_gem_object *o
 
 	return vma;
 
+err_unlock:
+	spin_unlock(&obj->vma.lock);
 err_vma:
 	i915_vma_free(vma);
 	return ERR_PTR(-E2BIG);

From: Chris Wilson <[email protected]>

commit 2abaad4eb59d1cdc903ea84c06acb406e2fbb263 upstream.

The hwsp_cacheline pointer from i915_request is very, very flimsy. The i915_request.timeline (and the hwsp_cacheline) are lost upon retiring (after an RCU grace). Therefore we need to confirm that once we have the right pointer for the cacheline, it is not in the process of being retired and disposed of before we attempt to acquire a reference to the cacheline.

<3>[  547.208237] BUG: KASAN: use-after-free in active_debug_hint+0x6a/0x70 [i915]
<3>[  547.208366] Read of size 8 at addr ffff88822a0d2710 by task gem_exec_parall/2536

<4>[  547.208547] CPU: 3 PID: 2536 Comm: gem_exec_parall Tainted: G     U            5.7.0-rc2-ged7a286b5d02d-kasan_117+ #1
<4>[  547.208556] Hardware name: Dell Inc. XPS 13 9350/, BIOS 1.4.12 11/30/2016
<4>[  547.208564] Call Trace:
<4>[  547.208579]  dump_stack+0x96/0xdb
<4>[  547.208707]  ? active_debug_hint+0x6a/0x70 [i915]
<4>[  547.208719]  print_address_description.constprop.6+0x16/0x310
<4>[  547.208841]  ? active_debug_hint+0x6a/0x70 [i915]
<4>[  547.208963]  ? active_debug_hint+0x6a/0x70 [i915]
<4>[  547.208975]  __kasan_report+0x137/0x190
<4>[  547.209106]  ? active_debug_hint+0x6a/0x70 [i915]
<4>[  547.209127]  kasan_report+0x32/0x50
<4>[  547.209257]  ? i915_gemfs_fini+0x40/0x40 [i915]
<4>[  547.209376]  active_debug_hint+0x6a/0x70 [i915]
<4>[  547.209389]  debug_print_object+0xa7/0x220
<4>[  547.209405]  ? lockdep_hardirqs_on+0x348/0x5f0
<4>[  547.209426]  debug_object_assert_init+0x297/0x430
<4>[  547.209449]  ? debug_object_free+0x360/0x360
<4>[  547.209472]  ? lock_acquire+0x1ac/0x8a0
<4>[  547.209592]  ? intel_timeline_read_hwsp+0x4f/0x840 [i915]
<4>[  547.209737]  ? i915_active_acquire_if_busy+0x66/0x120 [i915]
<4>[  547.209861]  i915_active_acquire_if_busy+0x66/0x120 [i915]
<4>[  547.209990]  ? __live_alloc.isra.15+0xc0/0xc0 [i915]
<4>[  547.210005]  ? rcu_read_lock_sched_held+0xd0/0xd0
<4>[  547.210017]  ? print_usage_bug+0x580/0x580
<4>[  547.210153]  intel_timeline_read_hwsp+0xbc/0x840 [i915]
<4>[  547.210284]  __emit_semaphore_wait+0xd5/0x480 [i915]
<4>[  547.210415]  ? i915_fence_get_timeline_name+0x110/0x110 [i915]
<4>[  547.210428]  ? lockdep_hardirqs_on+0x348/0x5f0
<4>[  547.210442]  ? _raw_spin_unlock_irq+0x2a/0x40
<4>[  547.210567]  ? __await_execution.constprop.51+0x2e0/0x570 [i915]
<4>[  547.210706]  i915_request_await_dma_fence+0x8f7/0xc70 [i915]

---
 drivers/gpu/drm/i915/gt/intel_timeline.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -519,6 +519,8 @@ int intel_timeline_read_hwsp(struct i915
 
 	rcu_read_lock();
 	cl = rcu_dereference(from->hwsp_cacheline);
+	if (i915_request_completed(from)) /* confirm cacheline is valid */
+		goto unlock;
 	if (unlikely(!i915_active_acquire_if_busy(&cl->active)))
 		goto unlock; /* seqno wrapped and completed! */
 	if (unlikely(i915_request_completed(from)))

From: Vasily Averin <[email protected]>

commit 85e9b88af1e6164f19ec71381efd5e2bcfc17620 upstream.

ret should be changed to release allocated struct qxl_release

---
 drivers/gpu/drm/qxl/qxl_draw.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

@@ -209,9 +209,10 @@ void qxl_draw_dirty_fb(struct qxl_device
 		goto out_release_backoff;
 
 	rects = drawable_set_clipping(qdev, num_clips, clips_bo);
-	if (!rects)
+	if (!rects) {
+		ret = -EINVAL;
 		goto out_release_backoff;
-
+	}
 	drawable = (struct qxl_drawable *)qxl_release_map(qdev, release);
 
 	drawable->clip.type = SPICE_CLIP_TYPE_RECTS;

From: Vasily Averin <[email protected]>

commit a65aa9c3676ffccb21361d52fcfedd5b5ff387d7 upstream.

---
 drivers/gpu/drm/qxl/qxl_cmd.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

@@ -480,9 +480,10 @@ int qxl_hw_surface_alloc(struct qxl_devi
 		return ret;
 
 	ret = qxl_release_reserve_list(release, true);
-	if (ret)
+	if (ret) {
+		qxl_release_free(qdev, release);
 		return ret;
-
+	}
 	cmd = (struct qxl_surface_cmd *)qxl_release_map(qdev, release);
 	cmd->type = QXL_SURFACE_CMD_CREATE;
 	cmd->flags = QXL_SURF_FLAG_KEEP_DATA;

From: Vasily Averin <[email protected]>

commit 933db73351d359f74b14f4af095808260aff11f9 upstream.

qxl_release should not be accesses after qxl_push_*_ring_release() calls:
userspace driver can process submitted command quickly, move qxl_release into release_ring, generate interrupt and trigger garbage collector.

It can lead to crashes in qxl driver or trigger memory corruption in some kmalloc-192 slab object

Gerd Hoffmann proposes to swap the qxl_release_fence_buffer_objects() + qxl_push_{cursor,command}_ring_release() calls to close that race window.

---
 drivers/gpu/drm/qxl/qxl_cmd.c     |    5 ++---
 drivers/gpu/drm/qxl/qxl_display.c |    6 +++---
 drivers/gpu/drm/qxl/qxl_draw.c    |    2 +-
 drivers/gpu/drm/qxl/qxl_ioctl.c   |    5 +----
 4 files changed, 7 insertions(+), 11 deletions(-)

@@ -500,8 +500,8 @@ int qxl_hw_surface_alloc(struct qxl_devi
 	/* no need to add a release to the fence for this surface bo,
 	   since it is only released when we ask to destroy the surface
 	   and it would never signal otherwise */
-	qxl_push_command_ring_release(qdev, release, QXL_CMD_SURFACE, false);
 	qxl_release_fence_buffer_objects(release);
+	qxl_push_command_ring_release(qdev, release, QXL_CMD_SURFACE, false);
 
 	surf->hw_surf_alloc = true;
 	spin_lock(&qdev->surf_id_idr_lock);

@@ -543,9 +543,8 @@ int qxl_hw_surface_dealloc(struct qxl_de
 	cmd->surface_id = id;
 	qxl_release_unmap(qdev, release, &cmd->release_info);
 
-	qxl_push_command_ring_release(qdev, release, QXL_CMD_SURFACE, false);
-
 	qxl_release_fence_buffer_objects(release);
+	qxl_push_command_ring_release(qdev, release, QXL_CMD_SURFACE, false);
 
 	return 0;
 }

@@ -523,8 +523,8 @@ static int qxl_primary_apply_cursor(stru
 	cmd->u.set.visible = 1;
 	qxl_release_unmap(qdev, release, &cmd->release_info);
 
-	qxl_push_cursor_ring_release(qdev, release, QXL_CMD_CURSOR, false);
 	qxl_release_fence_buffer_objects(release);
+	qxl_push_cursor_ring_release(qdev, release, QXL_CMD_CURSOR, false);
 
 	return ret;
 

@@ -665,8 +665,8 @@ static void qxl_cursor_atomic_update(str
 	cmd->u.position.y = plane->state->crtc_y + fb->hot_y;
 
 	qxl_release_unmap(qdev, release, &cmd->release_info);
-	qxl_push_cursor_ring_release(qdev, release, QXL_CMD_CURSOR, false);
 	qxl_release_fence_buffer_objects(release);
+	qxl_push_cursor_ring_release(qdev, release, QXL_CMD_CURSOR, false);
 
 	if (old_cursor_bo != NULL)
 		qxl_bo_unpin(old_cursor_bo);

@@ -713,8 +713,8 @@ static void qxl_cursor_atomic_disable(st
 	cmd->type = QXL_CURSOR_HIDE;
 	qxl_release_unmap(qdev, release, &cmd->release_info);
 
-	qxl_push_cursor_ring_release(qdev, release, QXL_CMD_CURSOR, false);
 	qxl_release_fence_buffer_objects(release);
+	qxl_push_cursor_ring_release(qdev, release, QXL_CMD_CURSOR, false);
 }
 
 static void qxl_update_dumb_head(struct qxl_device *qdev,

@@ -243,8 +243,8 @@ void qxl_draw_dirty_fb(struct qxl_device
 	}
 	qxl_bo_kunmap(clips_bo);
 
-	qxl_push_command_ring_release(qdev, release, QXL_CMD_DRAW, false);
 	qxl_release_fence_buffer_objects(release);
+	qxl_push_command_ring_release(qdev, release, QXL_CMD_DRAW, false);
 
 out_release_backoff:
 	if (ret)

@@ -261,11 +261,8 @@ static int qxl_process_single_command(st
 			apply_surf_reloc(qdev, &reloc_info[i]);
 	}
 
+	qxl_release_fence_buffer_objects(release);
 	ret = qxl_push_command_ring_release(qdev, release, cmd->type, true);
-	if (ret)
-		qxl_release_backoff_reserve_list(release);
-	else
-		qxl_release_fence_buffer_objects(release);
 
 out_free_bos:
 out_free_release:

From: Olga Kornievskaia <[email protected]>

commit dff58530c4ca8ce7ee5a74db431c6e35362cf682 upstream.

Currently, if the client sends BIND_CONN_TO_SESSION with NFS4_CDFC4_FORE_OR_BOTH but only gets NFS4_CDFS4_FORE back it ignores that it wasn't able to enable a backchannel.

To make sure, the client sends BIND_CONN_TO_SESSION as the first operation on the connections (ie., no other session compounds haven't been sent before), and if the client's request to bind the backchannel is not satisfied, then reset the connection and retry.

---
 fs/nfs/nfs4proc.c           |    8 ++++++++
 include/linux/nfs_xdr.h     |    2 ++
 include/linux/sunrpc/clnt.h |    5 +++++
 3 files changed, 15 insertions(+)

@@ -7893,6 +7893,7 @@ static void
 nfs4_bind_one_conn_to_session_done(struct rpc_task *task, void *calldata)
 {
 	struct nfs41_bind_conn_to_session_args *args = task->tk_msg.rpc_argp;
+	struct nfs41_bind_conn_to_session_res *res = task->tk_msg.rpc_resp;
 	struct nfs_client *clp = args->client;
 
 	switch (task->tk_status) {

@@ -7901,6 +7902,12 @@ nfs4_bind_one_conn_to_session_done(struc
 		nfs4_schedule_session_recovery(clp->cl_session,
 				task->tk_status);
 	}
+	if (args->dir == NFS4_CDFC4_FORE_OR_BOTH &&
+			res->dir != NFS4_CDFS4_BOTH) {
+		rpc_task_close_connection(task);
+		if (args->retries++ < MAX_BIND_CONN_TO_SESSION_RETRIES)
+			rpc_restart_call(task);
+	}
 }
 
 static const struct rpc_call_ops nfs4_bind_one_conn_to_session_ops = {

@@ -7923,6 +7930,7 @@ int nfs4_proc_bind_one_conn_to_session(s
 	struct nfs41_bind_conn_to_session_args args = {
 		.client = clp,
 		.dir = NFS4_CDFC4_FORE_OR_BOTH,
+		.retries = 0,
 	};
 	struct nfs41_bind_conn_to_session_res res;
 	struct rpc_message msg = {

@@ -1307,11 +1307,13 @@ struct nfs41_impl_id {
 	struct nfstime4			date;
 };
 
+#define MAX_BIND_CONN_TO_SESSION_RETRIES 3
 struct nfs41_bind_conn_to_session_args {
 	struct nfs_client		*client;
 	struct nfs4_sessionid		sessionid;
 	u32				dir;
 	bool				use_conn_in_rdma_mode;
+	int				retries;
 };
 
 struct nfs41_bind_conn_to_session_res {

@@ -236,4 +236,9 @@ static inline int rpc_reply_expected(str
 		(task->tk_msg.rpc_proc->p_decode != NULL);
 }
 
+static inline void rpc_task_close_connection(struct rpc_task *task)
+{
+	if (task->tk_xprt)
+		xprt_force_disconnect(task->tk_xprt);
+}
 #endif /* _LINUX_SUNRPC_CLNT_H */

From: Xiyu Yang <[email protected]>

commit 1402d17dfd9657be0da8458b2079d03c2d61c86a upstream.

btrfs_recover_relocation() invokes btrfs_join_transaction(), which joins a btrfs_trans_handle object into transactions and returns a reference of it with increased refcount to "trans".

When btrfs_recover_relocation() returns, "trans" becomes invalid, so the refcount should be decreased to keep refcount balanced.

The reference counting issue happens in one exception handling path of btrfs_recover_relocation(). When read_fs_root() failed, the refcnt increased by btrfs_join_transaction() is not decreased, causing a refcnt leak.

Fix this issue by calling btrfs_end_transaction() on this error path when read_fs_root() failed.

---
 fs/btrfs/relocation.c |    1 +
 1 file changed, 1 insertion(+)

@@ -4614,6 +4614,7 @@ int btrfs_recover_relocation(struct btrf
 		if (IS_ERR(fs_root)) {
 			err = PTR_ERR(fs_root);
 			list_add_tail(&reloc_root->root_list, &reloc_roots);
+			btrfs_end_transaction(trans);
 			goto out_unset;
 		}
 

From: Xiyu Yang <[email protected]>

commit f6033c5e333238f299c3ae03fac8cc1365b23b77 upstream.

btrfs_remove_block_group() invokes btrfs_lookup_block_group(), which returns a local reference of the block group that contains the given bytenr to "block_group" with increased refcount.

When btrfs_remove_block_group() returns, "block_group" becomes invalid, so the refcount should be decreased to keep refcount balanced.

The reference counting issue happens in several exception handling paths of btrfs_remove_block_group(). When those error scenarios occur such as btrfs_alloc_path() returns NULL, the function forgets to decrease its refcnt increased by btrfs_lookup_block_group() and will cause a refcnt leak.

Fix this issue by jumping to "out_put_group" label and calling btrfs_put_block_group() when those error scenarios occur.

---
 fs/btrfs/block-group.c |   16 ++++++++++------
 1 file changed, 10 insertions(+), 6 deletions(-)

@@ -916,7 +916,7 @@ int btrfs_remove_block_group(struct btrf
 	path = btrfs_alloc_path();
 	if (!path) {
 		ret = -ENOMEM;
-		goto out;
+		goto out_put_group;
 	}
 
 	/*

@@ -954,7 +954,7 @@ int btrfs_remove_block_group(struct btrf
 		ret = btrfs_orphan_add(trans, BTRFS_I(inode));
 		if (ret) {
 			btrfs_add_delayed_iput(inode);
-			goto out;
+			goto out_put_group;
 		}
 		clear_nlink(inode);
 		/* One for the block groups ref */

@@ -977,13 +977,13 @@ int btrfs_remove_block_group(struct btrf
 
 	ret = btrfs_search_slot(trans, tree_root, &key, path, -1, 1);
 	if (ret < 0)
-		goto out;
+		goto out_put_group;
 	if (ret > 0)
 		btrfs_release_path(path);
 	if (ret == 0) {
 		ret = btrfs_del_item(trans, tree_root, path);
 		if (ret)
-			goto out;
+			goto out_put_group;
 		btrfs_release_path(path);
 	}
 

@@ -1102,9 +1102,9 @@ int btrfs_remove_block_group(struct btrf
 
 	ret = remove_block_group_free_space(trans, block_group);
 	if (ret)
-		goto out;
+		goto out_put_group;
 
-	btrfs_put_block_group(block_group);
+	/* Once for the block groups rbtree */
 	btrfs_put_block_group(block_group);
 
 	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);

@@ -1127,6 +1127,10 @@ int btrfs_remove_block_group(struct btrf
 		/* once for the tree */
 		free_extent_map(em);
 	}
+
+out_put_group:
+	/* Once for the lookup reference */
+	btrfs_put_block_group(block_group);
 out:
 	if (remove_rsv)
 		btrfs_delayed_refs_rsv_release(fs_info, 1);

From: Filipe Manana <[email protected]>

commit f135cea30de5f74d5bfb5116682073841fb4af8f upstream.

When we have an inode with a prealloc extent that starts at an offset lower than the i_size and there is another prealloc extent that starts at an offset beyond i_size, we can end up losing part of the first prealloc extent (the part that starts at i_size) and have an implicit hole if we fsync the file and then have a power failure.

Consider the following example with comments explaining how and why it happens.

  $ mkfs.btrfs -f /dev/sdb
  $ mount /dev/sdb /mnt

# Create our test file with 2 consecutive prealloc extents, each with a

  # size of 128Kb, and covering the range from 0 to 256Kb, with a file
  # size of 0.
  $ xfs_io -f -c "falloc -k 0 128K" /mnt/foo
  $ xfs_io -c "falloc -k 128K 128K" /mnt/foo

# Fsync the file to record both extents in the log tree.

  $ xfs_io -c "fsync" /mnt/foo

# Now do a redudant extent allocation for the range from 0 to 64Kb. # This will merely increase the file size from 0 to 64Kb. Instead we

  # could also do a truncate to set the file size to 64Kb.
  $ xfs_io -c "falloc 0 64K" /mnt/foo

# Fsync the file, so we update the inode item in the log tree with the

  # new file size (64Kb). This also ends up setting the number of bytes
  # for the first prealloc extent to 64Kb. This is done by the truncation
  # at btrfs_log_prealloc_extents().
  # This means that if a power failure happens after this, a write into
  # the file range 64Kb to 128Kb will not use the prealloc extent and
  # will result in allocation of a new extent.
  $ xfs_io -c "fsync" /mnt/foo

# Now set the file size to 256K with a truncate and then fsync the file. # Since no changes happened to the extents, the fsync only updates the

  # i_size in the inode item at the log tree. This results in an implicit
  # hole for the file range from 64Kb to 128Kb, something which fsck will
  # complain when not using the NO_HOLES feature if we replay the log
  # after a power failure.
  $ xfs_io -c "truncate 256K" -c "fsync" /mnt/foo

So instead of always truncating the log to the inode's current i_size at btrfs_log_prealloc_extents(), check first if there's a prealloc extent that starts at an offset lower than the i_size and with a length that crosses the i_size - if there is one, just make sure we truncate to a size that corresponds to the end offset of that prealloc extent, so that we don't lose the part of that extent that starts at i_size if a power failure happens.

A test case for fstests follows soon.

---
 fs/btrfs/tree-log.c |   43 ++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 40 insertions(+), 3 deletions(-)

@@ -4211,6 +4211,9 @@ static int btrfs_log_prealloc_extents(st
 	const u64 ino = btrfs_ino(inode);
 	struct btrfs_path *dst_path = NULL;
 	bool dropped_extents = false;
+	u64 truncate_offset = i_size;
+	struct extent_buffer *leaf;
+	int slot;
 	int ins_nr = 0;
 	int start_slot;
 	int ret;

@@ -4225,9 +4228,43 @@ static int btrfs_log_prealloc_extents(st
 	if (ret < 0)
 		goto out;
 
+	/*
+	 * We must check if there is a prealloc extent that starts before the
+	 * i_size and crosses the i_size boundary. This is to ensure later we
+	 * truncate down to the end of that extent and not to the i_size, as
+	 * otherwise we end up losing part of the prealloc extent after a log
+	 * replay and with an implicit hole if there is another prealloc extent
+	 * that starts at an offset beyond i_size.
+	 */
+	ret = btrfs_previous_item(root, path, ino, BTRFS_EXTENT_DATA_KEY);
+	if (ret < 0)
+		goto out;
+
+	if (ret == 0) {
+		struct btrfs_file_extent_item *ei;
+
+		leaf = path->nodes[0];
+		slot = path->slots[0];
+		ei = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
+
+		if (btrfs_file_extent_type(leaf, ei) ==
+		    BTRFS_FILE_EXTENT_PREALLOC) {
+			u64 extent_end;
+
+			btrfs_item_key_to_cpu(leaf, &key, slot);
+			extent_end = key.offset +
+				btrfs_file_extent_num_bytes(leaf, ei);
+
+			if (extent_end > i_size)
+				truncate_offset = extent_end;
+		}
+	} else {
+		ret = 0;
+	}
+
 	while (true) {
-		struct extent_buffer *leaf = path->nodes[0];
-		int slot = path->slots[0];
+		leaf = path->nodes[0];
+		slot = path->slots[0];
 
 		if (slot >= btrfs_header_nritems(leaf)) {
 			if (ins_nr > 0) {

@@ -4265,7 +4302,7 @@ static int btrfs_log_prealloc_extents(st
 				ret = btrfs_truncate_inode_items(trans,
 							 root->log_root,
 							 &inode->vfs_inode,
-							 i_size,
+							 truncate_offset,
 							 BTRFS_EXTENT_DATA_KEY);
 			} while (ret == -EAGAIN);
 			if (ret)

From: Qu Wenruo <[email protected]>

commit fcc99734d1d4ced30167eb02e17f656735cb9928 upstream.

[BUG]
One run of btrfs/063 triggered the following lockdep warning:
============================================
WARNING: possible recursive locking detected
5.6.0-rc7-custom+ #48 Not tainted
--------------------------------------------
kworker/u24:0/7 is trying to acquire lock:
ffff88817d3a46e0 (sb_internal#2){.+.+}, at: start_transaction+0x66c/0x890 [btrfs]

but task is already holding lock:
ffff88817d3a46e0 (sb_internal#2){.+.+}, at: start_transaction+0x66c/0x890 [btrfs]

other info that might help us debug this:

   Possible unsafe locking scenario:

         CPU0
         ----
    lock(sb_internal#2);
    lock(sb_internal#2);

*** DEADLOCK ***

May be due to missing lock nesting notation

4 locks held by kworker/u24:0/7:
#0: ffff88817b495948 ((wq_completion)btrfs-endio-write){+.+.}, at: process_one_work+0x557/0xb80 #1: ffff888189ea7db8 ((work_completion)(&work->normal_work)){+.+.}, at: process_one_work+0x557/0xb80 #2: ffff88817d3a46e0 (sb_internal#2){.+.+}, at: start_transaction+0x66c/0x890 [btrfs] #3: ffff888174ca4da8 (&fs_info->reloc_mutex){+.+.}, at: btrfs_record_root_in_trans+0x83/0xd0 [btrfs]

  stack backtrace:
  CPU: 0 PID: 7 Comm: kworker/u24:0 Not tainted 5.6.0-rc7-custom+ #48
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
  Workqueue: btrfs-endio-write btrfs_work_helper [btrfs]
  Call Trace:
   dump_stack+0xc2/0x11a
   __lock_acquire.cold+0xce/0x214
   lock_acquire+0xe6/0x210
   __sb_start_write+0x14e/0x290
   start_transaction+0x66c/0x890 [btrfs]
   btrfs_join_transaction+0x1d/0x20 [btrfs]
   find_free_extent+0x1504/0x1a50 [btrfs]
   btrfs_reserve_extent+0xd5/0x1f0 [btrfs]
   btrfs_alloc_tree_block+0x1ac/0x570 [btrfs]
   btrfs_copy_root+0x213/0x580 [btrfs]
   create_reloc_root+0x3bd/0x470 [btrfs]
   btrfs_init_reloc_root+0x2d2/0x310 [btrfs]
   record_root_in_trans+0x191/0x1d0 [btrfs]
   btrfs_record_root_in_trans+0x90/0xd0 [btrfs]
   start_transaction+0x16e/0x890 [btrfs]
   btrfs_join_transaction+0x1d/0x20 [btrfs]
   btrfs_finish_ordered_io+0x55d/0xcd0 [btrfs]
   finish_ordered_fn+0x15/0x20 [btrfs]
   btrfs_work_helper+0x116/0x9a0 [btrfs]
   process_one_work+0x632/0xb80
   worker_thread+0x80/0x690
   kthread+0x1a3/0x1f0
   ret_from_fork+0x27/0x50

It's pretty hard to reproduce, only one hit so far.

[CAUSE]
This is because we're calling btrfs_join_transaction() without re-using the current running one:

btrfs_finish_ordered_io()
|- btrfs_join_transaction() <<< Call #1
|- btrfs_record_root_in_trans()

      |- btrfs_reserve_extent()
	 |- btrfs_join_transaction()	<<< Call #2

Normally such btrfs_join_transaction() call should re-use the existing one, without trying to re-start a transaction.

But the problem is, in btrfs_join_transaction() call #1, we call btrfs_record_root_in_trans() before initializing current::journal_info.

And in btrfs_join_transaction() call #2, we're relying on current::journal_info to avoid such deadlock.

[FIX]
Call btrfs_record_root_in_trans() after we have initialized current::journal_info.

---
 fs/btrfs/transaction.c |   13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

@@ -660,10 +660,19 @@ again:
 	}
 
 got_it:
-	btrfs_record_root_in_trans(h, root);
-
 	if (!current->journal_info)
 		current->journal_info = h;
+
+	/*
+	 * btrfs_record_root_in_trans() needs to alloc new extents, and may
+	 * call btrfs_join_transaction() while we're also starting a
+	 * transaction.
+	 *
+	 * Thus it need to be called after current->journal_info initialized,
+	 * or we can deadlock.
+	 */
+	btrfs_record_root_in_trans(h, root);
+
 	return h;
 
 join_fail:

From: Douglas Anderson <[email protected]>

commit b1ac62a7ac386d76968af5f374a4a7a82a35fe31 upstream.

Open-coding a timeout loop invariably leads to errors with handling the timeout properly in one corner case or another. In the case of cqhci we might report "CQE stuck on" even if it wasn't stuck on. You'd just need this sequence of events to happen in cqhci_off():

1. Call ktime_get().
2. Something happens to interrupt the CPU for > 100 us (context switch or interrupt).
3. Check time and; set "timed_out" to true since > 100 us.
4. Read CQHCI_CTL.
5. Both "reg & CQHCI_HALT" and "timed_out" are true, so break.
6. Since "timed_out" is true, falsely print the error message.

Rather than fixing the polling loop, use readx_poll_timeout() like many people do. This has been time tested to handle the corner cases.

---
 drivers/mmc/host/cqhci.c |   21 ++++++++++-----------
 1 file changed, 10 insertions(+), 11 deletions(-)

@@ -5,6 +5,7 @@
 #include <linux/delay.h>
 #include <linux/highmem.h>
 #include <linux/io.h>
+#include <linux/iopoll.h>
 #include <linux/module.h>
 #include <linux/dma-mapping.h>
 #include <linux/slab.h>

@@ -343,12 +344,16 @@ static int cqhci_enable(struct mmc_host
 /* CQHCI is idle and should halt immediately, so set a small timeout */
 #define CQHCI_OFF_TIMEOUT 100
 
+static u32 cqhci_read_ctl(struct cqhci_host *cq_host)
+{
+	return cqhci_readl(cq_host, CQHCI_CTL);
+}
+
 static void cqhci_off(struct mmc_host *mmc)
 {
 	struct cqhci_host *cq_host = mmc->cqe_private;
-	ktime_t timeout;
-	bool timed_out;
 	u32 reg;
+	int err;
 
 	if (!cq_host->enabled || !mmc->cqe_on || cq_host->recovery_halt)
 		return;

@@ -358,15 +363,9 @@ static void cqhci_off(struct mmc_host *m
 
 	cqhci_writel(cq_host, CQHCI_HALT, CQHCI_CTL);
 
-	timeout = ktime_add_us(ktime_get(), CQHCI_OFF_TIMEOUT);
-	while (1) {
-		timed_out = ktime_compare(ktime_get(), timeout) > 0;
-		reg = cqhci_readl(cq_host, CQHCI_CTL);
-		if ((reg & CQHCI_HALT) || timed_out)
-			break;
-	}
-
-	if (timed_out)
+	err = readx_poll_timeout(cqhci_read_ctl, cq_host, reg,
+				 reg & CQHCI_HALT, 0, CQHCI_OFF_TIMEOUT);
+	if (err < 0)
 		pr_err("%s: cqhci: CQE stuck on\n", mmc_hostname(mmc));
 	else
 		pr_debug("%s: cqhci: CQE off\n", mmc_hostname(mmc));

From: Marek Behún <[email protected]>

commit bb32e1987bc55ce1db400faf47d85891da3c9b9f upstream.

For some reason the Host Control2 register of the Xenon SDHCI controller sometimes reports the bit representing 1.8V signaling as 0 when read after it was written as 1. Subsequent read reports 1.

This causes the sdhci_start_signal_voltage_switch function to report
1.8V regulator output did not become stable

When CONFIG_PM is enabled, the host is suspended and resumend many times, and in each resume the switch to 1.8V is called, and so the kernel log reports this message annoyingly often.

Do an empty read of the Host Control2 register in Xenon's .voltage_switch method to circumvent this.

This patch fixes this particular problem on Turris MOX.

---
 drivers/mmc/host/sdhci-xenon.c |   10 ++++++++++
 1 file changed, 10 insertions(+)

@@ -235,6 +235,16 @@ static void xenon_voltage_switch(struct
 {
 	/* Wait for 5ms after set 1.8V signal enable bit */
 	usleep_range(5000, 5500);
+
+	/*
+	 * For some reason the controller's Host Control2 register reports
+	 * the bit representing 1.8V signaling as 0 when read after it was
+	 * written as 1. Subsequent read reports 1.
+	 *
+	 * Since this may cause some issues, do an empty read of the Host
+	 * Control2 register here to circumvent this.
+	 */
+	sdhci_readw(host, SDHCI_HOST_CONTROL2);
 }
 
 static const struct sdhci_ops sdhci_xenon_ops = {

From: Adrian Hunter <[email protected]>

commit 1a8eb6b373c2af6533c13d1ea11f504e5010ed9a upstream.

BIOS writers have begun the practice of setting 40 ohm eMMC driver strength even though the eMMC may not support it, on the assumption that the kernel will validate the value against the eMMC (Extended CSD DRIVER_STRENGTH [offset 197]) and revert to the default 50 ohm value if 40 ohm is invalid.

This is done to avoid changing the value for different boards.

Putting aside the merits of this approach, it is clear the eMMC's mask of supported driver strengths is more reliable than the value provided by BIOS. Add validation accordingly.

---
 drivers/mmc/host/sdhci-pci-core.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -601,6 +601,9 @@ static int intel_select_drive_strength(s
 	struct sdhci_pci_slot *slot = sdhci_priv(host);
 	struct intel_host *intel_host = sdhci_pci_priv(slot);
 
+	if (!(mmc_driver_type_mask(intel_host->drv_strength) & card_drv))
+		return 0;
+
 	return intel_host->drv_strength;
 }
 

From: Veerabhadrarao Badiganti <[email protected]>

commit 9d8cb58691f85cef687512262acb2c7109ee4868 upstream.

MSM sd host controller is capable of HW busy detection of device busy signaling over DAT0 line. And it requires the R1B response for commands that have this response associated with them.

So set the below two host capabilities for qcom SDHC.
- MMC_CAP_WAIT_WHILE_BUSY
- MMC_CAP_NEED_RSP_BUSY

Recent development of the mmc core in regards to this, revealed this as being a potential bug, hence the stable tag.

---
 drivers/mmc/host/sdhci-msm.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -2068,6 +2068,8 @@ static int sdhci_msm_probe(struct platfo
 		goto clk_disable;
 	}
 
+	msm_host->mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY | MMC_CAP_NEED_RSP_BUSY;
+
 	pm_runtime_get_noresume(&pdev->dev);
 	pm_runtime_set_active(&pdev->dev);
 	pm_runtime_enable(&pdev->dev);

From: Martin Blumenstingl <[email protected]>

commit e53b868b3cf5beeaa2f851ec6740112bf4d6a8cb upstream.

The Meson SDIO controller uses the DAT0 lane for hardware busy detection. Set MMC_CAP_WAIT_WHILE_BUSY accordingly. This fixes the following error observed with Linux 5.7 (pre-rc-1):
mmc1: Card stuck being busy! __mmc_poll_for_busy
blk_update_request: I/O error, dev mmcblk1, sector 17111080 op
0x3:(DISCARD) flags 0x0 phys_seg 1 prio class 0

---
 drivers/mmc/host/meson-mx-sdio.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -570,7 +570,7 @@ static int meson_mx_mmc_add_host(struct
 	mmc->f_max = clk_round_rate(host->cfg_div_clk,
 				    clk_get_rate(host->parent_clk));
 
-	mmc->caps |= MMC_CAP_ERASE | MMC_CAP_CMD23;
+	mmc->caps |= MMC_CAP_ERASE | MMC_CAP_CMD23 | MMC_CAP_WAIT_WHILE_BUSY;
 	mmc->ops = &meson_mx_mmc_ops;
 
 	ret = mmc_of_parse(mmc);

From: Martin Blumenstingl <[email protected]>

commit ddca1092c4324c89cf692b5efe655aa251864b51 upstream.

The recent commit 0d84c3e6a5b2 ("mmc: core: Convert to mmc_poll_for_busy() for erase/trim/discard") makes use of the ->card_busy() op for SD cards. This uncovered that the ->card_busy() op in the Meson SDIO driver was never working right:
while polling the busy status with ->card_busy() meson_mx_mmc_card_busy() reads only one of the two MESON_MX_SDIO_IRQC register values 0x1f001f10 or 0x1f003f10. This translates to "three out of four DAT lines are HIGH" and "all four DAT lines are HIGH", which is interpreted as "the card is busy".

It turns out that no situation can be observed where all four DAT lines are LOW, meaning the card is not busy anymore. Upon further research the 3.10 vendor driver for this controller does not implement the ->card_busy() op.

Remove the ->card_busy() op from the meson-mx-sdio driver since it is not working. At the time of writing this patch it is not clear what's needed to make the ->card_busy() implementation work with this specific controller hardware. For all use-cases which have previously worked the MMC_CAP_WAIT_WHILE_BUSY flag is now taking over, even if we don't have a ->card_busy() op anymore.

---
 drivers/mmc/host/meson-mx-sdio.c |    9 ---------
 1 file changed, 9 deletions(-)

@@ -357,14 +357,6 @@ static void meson_mx_mmc_request(struct
 		meson_mx_mmc_start_cmd(mmc, mrq->cmd);
 }
 
-static int meson_mx_mmc_card_busy(struct mmc_host *mmc)
-{
-	struct meson_mx_mmc_host *host = mmc_priv(mmc);
-	u32 irqc = readl(host->base + MESON_MX_SDIO_IRQC);
-
-	return !!(irqc & MESON_MX_SDIO_IRQC_FORCE_DATA_DAT_MASK);
-}
-
 static void meson_mx_mmc_read_response(struct mmc_host *mmc,
 				       struct mmc_command *cmd)
 {

@@ -506,7 +498,6 @@ static void meson_mx_mmc_timeout(struct
 static struct mmc_host_ops meson_mx_mmc_ops = {
 	.request		= meson_mx_mmc_request,
 	.set_ios		= meson_mx_mmc_set_ios,
-	.card_busy		= meson_mx_mmc_card_busy,
 	.get_cd			= mmc_gpio_get_cd,
 	.get_ro			= mmc_gpio_get_ro,
 };

From: Iuliana Prodan <[email protected]>

commit 55b3209acbb01cb02b1ee6b1afe80d83b1aab36d upstream.

For skcipher algorithms, the input, output HW S/G tables look like this: [IV, src][dst, IV]
Now, we can have 2 conditions here:
- there is no IV;
- src and dst are equal (in-place encryption) and scattered and the error is an "off-by-one" in the HW S/G table.

This issue was seen with KASAN:

BUG: KASAN: slab-out-of-bounds in skcipher_edesc_alloc+0x95c/0x1018

Read of size 4 at addr ffff000022a02958 by task cryptomgr_test/321

CPU: 2 PID: 321 Comm: cryptomgr_test Not tainted
5.6.0-rc1-00165-ge4ef8383-dirty #4
Hardware name: LS1046A RDB Board (DT)
Call trace:
 dump_backtrace+0x0/0x260
 show_stack+0x14/0x20
 dump_stack+0xe8/0x144
 print_address_description.isra.11+0x64/0x348
 __kasan_report+0x11c/0x230
 kasan_report+0xc/0x18
 __asan_load4+0x90/0xb0
 skcipher_edesc_alloc+0x95c/0x1018
 skcipher_encrypt+0x84/0x150
 crypto_skcipher_encrypt+0x50/0x68
 test_skcipher_vec_cfg+0x4d4/0xc10
 test_skcipher_vec+0x178/0x1d8
 alg_test_skcipher+0xec/0x230
 alg_test.part.44+0x114/0x4a0
 alg_test+0x1c/0x60
 cryptomgr_test+0x34/0x58
 kthread+0x1b8/0x1c0
 ret_from_fork+0x10/0x18

Allocated by task 321:
 save_stack+0x24/0xb0
 __kasan_kmalloc.isra.10+0xc4/0xe0
 kasan_kmalloc+0xc/0x18
 __kmalloc+0x178/0x2b8
 skcipher_edesc_alloc+0x21c/0x1018
 skcipher_encrypt+0x84/0x150
 crypto_skcipher_encrypt+0x50/0x68
 test_skcipher_vec_cfg+0x4d4/0xc10
 test_skcipher_vec+0x178/0x1d8
 alg_test_skcipher+0xec/0x230
 alg_test.part.44+0x114/0x4a0
 alg_test+0x1c/0x60
 cryptomgr_test+0x34/0x58
 kthread+0x1b8/0x1c0
 ret_from_fork+0x10/0x18

Freed by task 0:

(stack is not available)

The buggy address belongs to the object at ffff000022a02800
 which belongs to the cache dma-kmalloc-512 of size 512
The buggy address is located 344 bytes inside of
 512-byte region [ffff000022a02800, ffff000022a02a00)
The buggy address belongs to the page:
page:fffffe00006a8000 refcount:1 mapcount:0 mapping:ffff00093200c400
index:0x0 compound_mapcount: 0
flags: 0xffff00000010200(slab|head)
raw: 0ffff00000010200 dead000000000100 dead000000000122 ffff00093200c400
raw: 0000000000000000 0000000080100010 00000001ffffffff 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
 ffff000022a02800: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 ffff000022a02880: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>ffff000022a02900: 00 00 00 00 00 00 00 00 00 00 fc fc fc fc fc fc
                                                    ^
 ffff000022a02980: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
 ffff000022a02a00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc

---
 drivers/crypto/caam/caamalg.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1791,7 +1791,7 @@ static struct skcipher_edesc *skcipher_e
 
 	if (ivsize || mapped_dst_nents > 1)
 		sg_to_sec4_set_last(edesc->sec4_sg + dst_sg_idx +
-				    mapped_dst_nents);
+				    mapped_dst_nents - 1 + !!ivsize);
 
 	if (sec4_sg_bytes) {
 		edesc->sec4_sg_dma = dma_map_single(jrdev, edesc->sec4_sg,

From: Hui Wang <[email protected]>

commit ef0b3203c758b6b8abdb5dca651880347eae6b8c upstream.

This new Lenovo ThinkCenter has two front mics which can't be handled by PA so far, so apply the fixup ALC283_FIXUP_HEADSET_MIC to change the location for one of the mics.

---
 sound/pci/hda/patch_realtek.c |    1 +
 1 file changed, 1 insertion(+)

@@ -7295,6 +7295,7 @@ static const struct snd_pci_quirk alc269
 	SND_PCI_QUIRK(0x1558, 0x8560, "System76 Gazelle (gaze14)", ALC269_FIXUP_HEADSET_MIC),
 	SND_PCI_QUIRK(0x1558, 0x8561, "System76 Gazelle (gaze14)", ALC269_FIXUP_HEADSET_MIC),
 	SND_PCI_QUIRK(0x17aa, 0x1036, "Lenovo P520", ALC233_FIXUP_LENOVO_MULTI_CODECS),
+	SND_PCI_QUIRK(0x17aa, 0x1048, "ThinkCentre Station", ALC283_FIXUP_HEADSET_MIC),
 	SND_PCI_QUIRK(0x17aa, 0x20f2, "Thinkpad SL410/510", ALC269_FIXUP_SKU_IGNORE),
 	SND_PCI_QUIRK(0x17aa, 0x215e, "Thinkpad L512", ALC269_FIXUP_SKU_IGNORE),
 	SND_PCI_QUIRK(0x17aa, 0x21b8, "Thinkpad Edge 14", ALC269_FIXUP_SKU_IGNORE),

From: Takashi Iwai <[email protected]>

commit 547d2c9cf4f1f72adfecacbd5b093681fb0e8b3e upstream.

The USB vendor ID of NuPrime DAC-10 is not 16b0 but 16d0.

---
 sound/usb/quirks.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1643,7 +1643,7 @@ u64 snd_usb_interface_dsd_format_quirks(
 
 	case USB_ID(0x0d8c, 0x0316): /* Hegel HD12 DSD */
 	case USB_ID(0x10cb, 0x0103): /* The Bit Opus #3; with fp->dsd_raw */
-	case USB_ID(0x16b0, 0x06b2): /* NuPrime DAC-10 */
+	case USB_ID(0x16d0, 0x06b2): /* NuPrime DAC-10 */
 	case USB_ID(0x16d0, 0x09dd): /* Encore mDSD */
 	case USB_ID(0x16d0, 0x0733): /* Furutech ADL Stratos */
 	case USB_ID(0x16d0, 0x09db): /* NuPrime Audio DAC-9 */

From: Wu Bo <[email protected]>

commit a2f647240998aa49632fb09b01388fdf2b87acfc upstream.

Fix the following coccicheck warning:
sound/pci/hda/patch_hdmi.c:1852:2-8: preceding lock on line 1846

After add sanity check to pass klockwork check, The spdif_mutex should be unlock before return true in check_non_pcm_per_cvt().

---
 sound/pci/hda/patch_hdmi.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -1934,8 +1934,10 @@ static bool check_non_pcm_per_cvt(struct
 	/* Add sanity check to pass klockwork check.
 	 * This should never happen.
 	 */
-	if (WARN_ON(spdif == NULL))
+	if (WARN_ON(spdif == NULL)) {
+		mutex_unlock(&codec->spdif_mutex);
 		return true;
+	}
 	non_pcm = !!(spdif->status & IEC958_AES0_NONAUDIO);
 	mutex_unlock(&codec->spdif_mutex);
 	return non_pcm;

From: Vasily Khoruzhick <[email protected]>

commit cc18b2f4f3f1d7ed3125ac1840794f9feab0325c upstream.

Apparently interface 1 is control interface akin to HD500X, setting LINE6_CAP_CONTROL and choosing it as ctrl_if fixes audio playback on POD HD500.

---
 sound/usb/line6/podhd.c |   22 +++++-----------------
 1 file changed, 5 insertions(+), 17 deletions(-)

@@ -21,8 +21,7 @@
 enum {
 	LINE6_PODHD300,
 	LINE6_PODHD400,
-	LINE6_PODHD500_0,
-	LINE6_PODHD500_1,
+	LINE6_PODHD500,
 	LINE6_PODX3,
 	LINE6_PODX3LIVE,
 	LINE6_PODHD500X,

@@ -318,8 +317,7 @@ static const struct usb_device_id podhd_
 	/* TODO: no need to alloc data interfaces when only audio is used */
 	{ LINE6_DEVICE(0x5057),    .driver_info = LINE6_PODHD300 },
 	{ LINE6_DEVICE(0x5058),    .driver_info = LINE6_PODHD400 },
-	{ LINE6_IF_NUM(0x414D, 0), .driver_info = LINE6_PODHD500_0 },
-	{ LINE6_IF_NUM(0x414D, 1), .driver_info = LINE6_PODHD500_1 },
+	{ LINE6_IF_NUM(0x414D, 0), .driver_info = LINE6_PODHD500 },
 	{ LINE6_IF_NUM(0x414A, 0), .driver_info = LINE6_PODX3 },
 	{ LINE6_IF_NUM(0x414B, 0), .driver_info = LINE6_PODX3LIVE },
 	{ LINE6_IF_NUM(0x4159, 0), .driver_info = LINE6_PODHD500X },

@@ -352,23 +350,13 @@ static const struct line6_properties pod
 		.ep_audio_r = 0x82,
 		.ep_audio_w = 0x01,
 	},
-	[LINE6_PODHD500_0] = {
+	[LINE6_PODHD500] = {
 		.id = "PODHD500",
 		.name = "POD HD500",
-		.capabilities	= LINE6_CAP_PCM
+		.capabilities	= LINE6_CAP_PCM | LINE6_CAP_CONTROL
 				| LINE6_CAP_HWMON,
 		.altsetting = 1,
-		.ep_ctrl_r = 0x81,
-		.ep_ctrl_w = 0x01,
-		.ep_audio_r = 0x86,
-		.ep_audio_w = 0x02,
-	},
-	[LINE6_PODHD500_1] = {
-		.id = "PODHD500",
-		.name = "POD HD500",
-		.capabilities	= LINE6_CAP_PCM
-				| LINE6_CAP_HWMON,
-		.altsetting = 0,
+		.ctrl_if = 1,
 		.ep_ctrl_r = 0x81,
 		.ep_ctrl_w = 0x01,
 		.ep_audio_r = 0x86,

From: Takashi Iwai <[email protected]>

commit 4285de0725b1bf73608abbcd35ad7fd3ddc0b61e upstream.

The checks of the plugin buffer overflow in the previous fix by commit
f2ecf903ef06 ("ALSA: pcm: oss: Avoid plugin buffer overflow")
are put in the wrong places mistakenly, which leads to the expected (repeated) sound when the rate plugin is involved. Fix in the right places.

Also, at those right places, the zero check is needed for the termination node, so added there as well, and let's get it done, finally.

---
 sound/core/oss/pcm_plugin.c |   20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)

@@ -211,21 +211,23 @@ static snd_pcm_sframes_t plug_client_siz
 	if (stream == SNDRV_PCM_STREAM_PLAYBACK) {
 		plugin = snd_pcm_plug_last(plug);
 		while (plugin && drv_frames > 0) {
-			if (check_size && drv_frames > plugin->buf_frames)
-				drv_frames = plugin->buf_frames;
 			plugin_prev = plugin->prev;
 			if (plugin->src_frames)
 				drv_frames = plugin->src_frames(plugin, drv_frames);
+			if (check_size && plugin->buf_frames &&
+			    drv_frames > plugin->buf_frames)
+				drv_frames = plugin->buf_frames;
 			plugin = plugin_prev;
 		}
 	} else if (stream == SNDRV_PCM_STREAM_CAPTURE) {
 		plugin = snd_pcm_plug_first(plug);
 		while (plugin && drv_frames > 0) {
 			plugin_next = plugin->next;
+			if (check_size && plugin->buf_frames &&
+			    drv_frames > plugin->buf_frames)
+				drv_frames = plugin->buf_frames;
 			if (plugin->dst_frames)
 				drv_frames = plugin->dst_frames(plugin, drv_frames);
-			if (check_size && drv_frames > plugin->buf_frames)
-				drv_frames = plugin->buf_frames;
 			plugin = plugin_next;
 		}
 	} else

@@ -251,26 +253,28 @@ static snd_pcm_sframes_t plug_slave_size
 		plugin = snd_pcm_plug_first(plug);
 		while (plugin && frames > 0) {
 			plugin_next = plugin->next;
+			if (check_size && plugin->buf_frames &&
+			    frames > plugin->buf_frames)
+				frames = plugin->buf_frames;
 			if (plugin->dst_frames) {
 				frames = plugin->dst_frames(plugin, frames);
 				if (frames < 0)
 					return frames;
 			}
-			if (check_size && frames > plugin->buf_frames)
-				frames = plugin->buf_frames;
 			plugin = plugin_next;
 		}
 	} else if (stream == SNDRV_PCM_STREAM_CAPTURE) {
 		plugin = snd_pcm_plug_last(plug);
 		while (plugin) {
-			if (check_size && frames > plugin->buf_frames)
-				frames = plugin->buf_frames;
 			plugin_prev = plugin->prev;
 			if (plugin->src_frames) {
 				frames = plugin->src_frames(plugin, frames);
 				if (frames < 0)
 					return frames;
 			}
+			if (check_size && plugin->buf_frames &&
+			    frames > plugin->buf_frames)
+				frames = plugin->buf_frames;
 			plugin = plugin_prev;
 		}
 	} else

From: Dan Carpenter <[email protected]>

commit ac2b0813fceaf7cb3d8d46c7b33c90bae9fa49db upstream.

The problem is that we dereference "privdata->pci_dev" when we print the error messages in amd_mp2_pci_init():

	dev_err(ndev_dev(privdata), "Failed to enable MP2 PCI device\n");
		^^^^^^^^^^^^^^^^^

---
 drivers/i2c/busses/i2c-amd-mp2-pci.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -349,12 +349,12 @@ static int amd_mp2_pci_probe(struct pci_
 	if (!privdata)
 		return -ENOMEM;
 
+	privdata->pci_dev = pci_dev;
 	rc = amd_mp2_pci_init(privdata, pci_dev);
 	if (rc)
 		return rc;
 
 	mutex_init(&privdata->c2p_lock);
-	privdata->pci_dev = pci_dev;
 
 	pm_runtime_set_autosuspend_delay(&pci_dev->dev, 1000);
 	pm_runtime_use_autosuspend(&pci_dev->dev);

From: Dexuan Cui <[email protected]>

commit 421f090c819d695942a470051cd624dc43deaf95 upstream.

Unlike the other CPUs, CPU0 is never offlined during hibernation, so in the resume path, the "new" kernel's VP assist page is not suspended (i.e. not disabled), and later when we jump to the "old" kernel, the page is not properly re-enabled for CPU0 with the allocated page from the old kernel.

So far, the VP assist page is used by hv_apic_eoi_write(), and is also used in the case of nested virtualization (running KVM atop Hyper-V).

For hv_apic_eoi_write(), when the page is not properly re-enabled, hvp->apic_assist is always 0, so the HV_X64_MSR_EOI MSR is always written. This is not ideal with respect to performance, but Hyper-V can still correctly handle this according to the Hyper-V spec; nevertheless, Linux still must update the Hyper-V hypervisor with the correct VP assist page to prevent Hyper-V from writing to the stale page, which causes guest memory corruption and consequently may have caused the hangs and triple faults seen during non-boot CPUs resume.

Fix the issue by calling hv_cpu_die()/hv_cpu_init() in the syscore ops. Without the fix, hibernation can fail at a rate of 1/300 ~ 1/500. With the fix, hibernation can pass a long-haul test of 2000 runs.

In the case of nested virtualization, disabling/reenabling the assist page upon hibernation may be unsafe if there are active L2 guests. It looks KVM should be enhanced to abort the hibernation request if there is any active L2 guest.

---
 arch/x86/hyperv/hv_init.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

@@ -73,7 +73,8 @@ static int hv_cpu_init(unsigned int cpu)
 	struct page *pg;
 
 	input_arg = (void **)this_cpu_ptr(hyperv_pcpu_input_arg);
-	pg = alloc_page(GFP_KERNEL);
+	/* hv_cpu_init() can be called with IRQs disabled from hv_resume() */
+	pg = alloc_page(irqs_disabled() ? GFP_ATOMIC : GFP_KERNEL);
 	if (unlikely(!pg))
 		return -ENOMEM;
 	*input_arg = page_address(pg);

@@ -254,6 +255,7 @@ static int __init hv_pci_init(void)
 static int hv_suspend(void)
 {
 	union hv_x64_msr_hypercall_contents hypercall_msr;
+	int ret;
 
 	/*
 	 * Reset the hypercall page as it is going to be invalidated

@@ -270,12 +272,17 @@ static int hv_suspend(void)
 	hypercall_msr.enable = 0;
 	wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64);
 
-	return 0;
+	ret = hv_cpu_die(0);
+	return ret;
 }
 
 static void hv_resume(void)
 {
 	union hv_x64_msr_hypercall_contents hypercall_msr;
+	int ret;
+
+	ret = hv_cpu_init(0);
+	WARN_ON(ret);
 
 	/* Re-enable the hypercall page */
 	rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64);

@@ -288,6 +295,7 @@ static void hv_resume(void)
 	hv_hypercall_pg_saved = NULL;
 }
 
+/* Note: when the ops are called, only CPU0 is online and IRQs are disabled. */
 static struct syscore_ops hv_syscore_ops = {
 	.suspend	= hv_suspend,
 	.resume		= hv_resume,

From: Dexuan Cui <[email protected]>

commit 1a06d017fb3f388734ffbe5dedee6f8c3af5f2db upstream.

Before the hibernation patchset (e.g. f53335e3289f), in a Generation-2 Linux VM on Hyper-V, the user can run "echo freeze > /sys/power/state" to freeze the system, i.e. Suspend-to-Idle. The user can press the keyboard or move the mouse to wake up the VM.

With the hibernation patchset, Linux VM on Hyper-V can hibernate to disk, but Suspend-to-Idle is broken: when the synthetic keyboard/mouse are suspended, there is no way to wake up the VM.

Fix the issue by not suspending and resuming the vmbus devices upon Suspend-to-Idle.

---
 drivers/hv/vmbus_drv.c |   43 ++++++++++++++++++++++++++++++++++---------
 1 file changed, 34 insertions(+), 9 deletions(-)

@@ -978,6 +978,9 @@ static int vmbus_resume(struct device *c
 
 	return drv->resume(dev);
 }
+#else
+#define vmbus_suspend NULL
+#define vmbus_resume NULL
 #endif /* CONFIG_PM_SLEEP */
 
 /*

@@ -997,11 +1000,22 @@ static void vmbus_device_release(struct
 }
 
 /*
- * Note: we must use SET_NOIRQ_SYSTEM_SLEEP_PM_OPS rather than
- * SET_SYSTEM_SLEEP_PM_OPS: see the comment before vmbus_bus_pm.
+ * Note: we must use the "noirq" ops: see the comment before vmbus_bus_pm.
+ *
+ * suspend_noirq/resume_noirq are set to NULL to support Suspend-to-Idle: we
+ * shouldn't suspend the vmbus devices upon Suspend-to-Idle, otherwise there
+ * is no way to wake up a Generation-2 VM.
+ *
+ * The other 4 ops are for hibernation.
  */
+
 static const struct dev_pm_ops vmbus_pm = {
-	SET_NOIRQ_SYSTEM_SLEEP_PM_OPS(vmbus_suspend, vmbus_resume)
+	.suspend_noirq	= NULL,
+	.resume_noirq	= NULL,
+	.freeze_noirq	= vmbus_suspend,
+	.thaw_noirq	= vmbus_resume,
+	.poweroff_noirq	= vmbus_suspend,
+	.restore_noirq	= vmbus_resume,
 };
 
 /* The one and only one */

@@ -2281,6 +2295,9 @@ static int vmbus_bus_resume(struct devic
 
 	return 0;
 }
+#else
+#define vmbus_bus_suspend NULL
+#define vmbus_bus_resume NULL
 #endif /* CONFIG_PM_SLEEP */
 
 static const struct acpi_device_id vmbus_acpi_device_ids[] = {

@@ -2291,16 +2308,24 @@ static const struct acpi_device_id vmbus
 MODULE_DEVICE_TABLE(acpi, vmbus_acpi_device_ids);
 
 /*
- * Note: we must use SET_NOIRQ_SYSTEM_SLEEP_PM_OPS rather than
- * SET_SYSTEM_SLEEP_PM_OPS, otherwise NIC SR-IOV can not work, because the
- * "pci_dev_pm_ops" uses the "noirq" callbacks: in the resume path, the
- * pci "noirq" restore callback runs before "non-noirq" callbacks (see
+ * Note: we must use the "no_irq" ops, otherwise hibernation can not work with
+ * PCI device assignment, because "pci_dev_pm_ops" uses the "noirq" ops: in
+ * the resume path, the pci "noirq" restore op runs before "non-noirq" op (see
  * resume_target_kernel() -> dpm_resume_start(), and hibernation_restore() ->
  * dpm_resume_end()). This means vmbus_bus_resume() and the pci-hyperv's
- * resume callback must also run via the "noirq" callbacks.
+ * resume callback must also run via the "noirq" ops.
+ *
+ * Set suspend_noirq/resume_noirq to NULL for Suspend-to-Idle: see the comment
+ * earlier in this file before vmbus_pm.
  */
+
 static const struct dev_pm_ops vmbus_bus_pm = {
-	SET_NOIRQ_SYSTEM_SLEEP_PM_OPS(vmbus_bus_suspend, vmbus_bus_resume)
+	.suspend_noirq	= NULL,
+	.resume_noirq	= NULL,
+	.freeze_noirq	= vmbus_bus_suspend,
+	.thaw_noirq	= vmbus_bus_resume,
+	.poweroff_noirq	= vmbus_bus_suspend,
+	.restore_noirq	= vmbus_bus_resume
 };
 
 static struct acpi_driver vmbus_acpi_driver = {

From: Al Viro <[email protected]>

commit 3815f1be546e752327b5868af103ccdddcc4db77 upstream.

'count' is how much you want written, not the final position. Moreover, it can legitimately be less than the current position...

---
 fs/ocfs2/dlmfs/dlmfs.c |   27 ++++++++++++---------------
 1 file changed, 12 insertions(+), 15 deletions(-)

@@ -275,7 +275,6 @@ static ssize_t dlmfs_file_write(struct f
 				loff_t *ppos)
 {
 	int bytes_left;
-	ssize_t writelen;
 	char *lvb_buf;
 	struct inode *inode = file_inode(filp);
 

@@ -285,32 +284,30 @@ static ssize_t dlmfs_file_write(struct f
 	if (*ppos >= i_size_read(inode))
 		return -ENOSPC;
 
+	/* don't write past the lvb */
+	if (count > i_size_read(inode) - *ppos)
+		count = i_size_read(inode) - *ppos;
+
 	if (!count)
 		return 0;
 
 	if (!access_ok(buf, count))
 		return -EFAULT;
 
-	/* don't write past the lvb */
-	if ((count + *ppos) > i_size_read(inode))
-		writelen = i_size_read(inode) - *ppos;
-	else
-		writelen = count - *ppos;
-
-	lvb_buf = kmalloc(writelen, GFP_NOFS);
+	lvb_buf = kmalloc(count, GFP_NOFS);
 	if (!lvb_buf)
 		return -ENOMEM;
 
-	bytes_left = copy_from_user(lvb_buf, buf, writelen);
-	writelen -= bytes_left;
-	if (writelen)
-		user_dlm_write_lvb(inode, lvb_buf, writelen);
+	bytes_left = copy_from_user(lvb_buf, buf, count);
+	count -= bytes_left;
+	if (count)
+		user_dlm_write_lvb(inode, lvb_buf, count);
 
 	kfree(lvb_buf);
 
-	*ppos = *ppos + writelen;
-	mlog(0, "wrote %zd bytes\n", writelen);
-	return writelen;
+	*ppos = *ppos + count;
+	mlog(0, "wrote %zu bytes\n", count);
+	return count;
 }
 
 static void dlmfs_init_once(void *foo)

From: Paul Moore <[email protected]>

commit fb73974172ffaaf57a7c42f35424d9aece1a5af6 upstream.

Fix the SELinux netlink_send hook to properly handle multiple netlink messages in a single sk_buff; each message is parsed and subject to SELinux access control. Prior to this patch, SELinux only inspected the first message in the sk_buff.

---
 security/selinux/hooks.c |   70 ++++++++++++++++++++++++++++++-----------------
 1 file changed, 45 insertions(+), 25 deletions(-)

@@ -5829,40 +5829,60 @@ static unsigned int selinux_ipv6_postrou
 
 static int selinux_netlink_send(struct sock *sk, struct sk_buff *skb)
 {
-	int err = 0;
-	u32 perm;
+	int rc = 0;
+	unsigned int msg_len;
+	unsigned int data_len = skb->len;
+	unsigned char *data = skb->data;
 	struct nlmsghdr *nlh;
 	struct sk_security_struct *sksec = sk->sk_security;
+	u16 sclass = sksec->sclass;
+	u32 perm;
 
-	if (skb->len < NLMSG_HDRLEN) {
-		err = -EINVAL;
-		goto out;
-	}
-	nlh = nlmsg_hdr(skb);
+	while (data_len >= nlmsg_total_size(0)) {
+		nlh = (struct nlmsghdr *)data;
 
-	err = selinux_nlmsg_lookup(sksec->sclass, nlh->nlmsg_type, &perm);
-	if (err) {
-		if (err == -EINVAL) {
+		/* NOTE: the nlmsg_len field isn't reliably set by some netlink
+		 *       users which means we can't reject skb's with bogus
+		 *       length fields; our solution is to follow what
+		 *       netlink_rcv_skb() does and simply skip processing at
+		 *       messages with length fields that are clearly junk
+		 */
+		if (nlh->nlmsg_len < NLMSG_HDRLEN || nlh->nlmsg_len > data_len)
+			return 0;
+
+		rc = selinux_nlmsg_lookup(sclass, nlh->nlmsg_type, &perm);
+		if (rc == 0) {
+			rc = sock_has_perm(sk, perm);
+			if (rc)
+				return rc;
+		} else if (rc == -EINVAL) {
+			/* -EINVAL is a missing msg/perm mapping */
 			pr_warn_ratelimited("SELinux: unrecognized netlink"
-			       " message: protocol=%hu nlmsg_type=%hu sclass=%s"
-			       " pid=%d comm=%s\n",
-			       sk->sk_protocol, nlh->nlmsg_type,
-			       secclass_map[sksec->sclass - 1].name,
-			       task_pid_nr(current), current->comm);
-			if (!enforcing_enabled(&selinux_state) ||
-			    security_get_allow_unknown(&selinux_state))
-				err = 0;
+				" message: protocol=%hu nlmsg_type=%hu sclass=%s"
+				" pid=%d comm=%s\n",
+				sk->sk_protocol, nlh->nlmsg_type,
+				secclass_map[sclass - 1].name,
+				task_pid_nr(current), current->comm);
+			if (enforcing_enabled(&selinux_state) &&
+			    !security_get_allow_unknown(&selinux_state))
+				return rc;
+			rc = 0;
+		} else if (rc == -ENOENT) {
+			/* -ENOENT is a missing socket/class mapping, ignore */
+			rc = 0;
+		} else {
+			return rc;
 		}
 
-		/* Ignore */
-		if (err == -ENOENT)
-			err = 0;
-		goto out;
+		/* move to the next message after applying netlink padding */
+		msg_len = NLMSG_ALIGN(nlh->nlmsg_len);
+		if (msg_len >= data_len)
+			return 0;
+		data_len -= msg_len;
+		data += msg_len;
 	}
 
-	err = sock_has_perm(sk, perm);
-out:
-	return err;
+	return rc;
 }
 
 static void ipc_init_security(struct ipc_security_struct *isec, u16 sclass)

From: Sudip Mukherjee <[email protected]>

commit 47c370c1a5eea9b2f6f026d49e060c3748c89667 upstream.

The commit below modified rvt_create_mmap_info() to return ERR_PTR's but didn't update the callers to handle them. Modify rvt_create_mmap_info() to only return ERR_PTR and fix all error checking after rvt_create_mmap_info() was called.

---
 drivers/infiniband/sw/rdmavt/cq.c   |    4 ++--
 drivers/infiniband/sw/rdmavt/mmap.c |    4 ++--
 drivers/infiniband/sw/rdmavt/qp.c   |    4 ++--
 drivers/infiniband/sw/rdmavt/srq.c  |    4 ++--
 4 files changed, 8 insertions(+), 8 deletions(-)

@@ -248,8 +248,8 @@ int rvt_create_cq(struct ib_cq *ibcq, co
 	 */
 	if (udata && udata->outlen >= sizeof(__u64)) {
 		cq->ip = rvt_create_mmap_info(rdi, sz, udata, u_wc);
-		if (!cq->ip) {
-			err = -ENOMEM;
+		if (IS_ERR(cq->ip)) {
+			err = PTR_ERR(cq->ip);
 			goto bail_wc;
 		}
 

@@ -154,7 +154,7 @@ done:
  * @udata: user data (must be valid!)
  * @obj: opaque pointer to a cq, wq etc
  *
- * Return: rvt_mmap struct on success
+ * Return: rvt_mmap struct on success, ERR_PTR on failure
  */
 struct rvt_mmap_info *rvt_create_mmap_info(struct rvt_dev_info *rdi, u32 size,
 					   struct ib_udata *udata, void *obj)

@@ -166,7 +166,7 @@ struct rvt_mmap_info *rvt_create_mmap_in
 
 	ip = kmalloc_node(sizeof(*ip), GFP_KERNEL, rdi->dparms.node);
 	if (!ip)
-		return ip;
+		return ERR_PTR(-ENOMEM);
 
 	size = PAGE_ALIGN(size);
 

@@ -1244,8 +1244,8 @@ struct ib_qp *rvt_create_qp(struct ib_pd
 
 			qp->ip = rvt_create_mmap_info(rdi, s, udata,
 						      qp->r_rq.wq);
-			if (!qp->ip) {
-				ret = ERR_PTR(-ENOMEM);
+			if (IS_ERR(qp->ip)) {
+				ret = ERR_CAST(qp->ip);
 				goto bail_qpn;
 			}
 

@@ -111,8 +111,8 @@ int rvt_create_srq(struct ib_srq *ibsrq,
 		u32 s = sizeof(struct rvt_rwq) + srq->rq.size * sz;
 
 		srq->ip = rvt_create_mmap_info(dev, s, udata, srq->rq.wq);
-		if (!srq->ip) {
-			ret = -ENOMEM;
+		if (IS_ERR(srq->ip)) {
+			ret = PTR_ERR(srq->ip);
 			goto bail_wq;
 		}
 

From: Kai-Heng Feng <[email protected]>

commit a9b760b0266f563b4784f695bbd0e717610dc10a upstream.

Transitioned power state logged at the end of setting ACPI power.

However, D3cold won't be in the message because state can only be D3hot at most.

Use target_state to corretly report when power state is D3cold.

---
 drivers/acpi/device_pm.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -273,13 +273,13 @@ int acpi_device_set_power(struct acpi_de
  end:
 	if (result) {
 		dev_warn(&device->dev, "Failed to change power state to %s\n",
-			 acpi_power_state_string(state));
+			 acpi_power_state_string(target_state));
 	} else {
 		device->power.state = target_state;
 		ACPI_DEBUG_PRINT((ACPI_DB_INFO,
 				  "Device [%s] transitioned to %s\n",
 				  device->pnp.bus_id,
-				  acpi_power_state_string(state)));
+				  acpi_power_state_string(target_state)));
 	}
 
 	return result;

From: Dexuan Cui <[email protected]>

commit 2351f8d295ed63393190e39c2f7c1fee1a80578f upstream.

Currently the kernel threads are not frozen in software_resume(), so between dpm_suspend_start(PMSG_QUIESCE) and resume_target_kernel(), system_freezable_power_efficient_wq can still try to submit SCSI commands and this can cause a panic since the low level SCSI driver (e.g. hv_storvsc) has quiesced the SCSI adapter and can not accept any SCSI commands: https://lkml.org/lkml/2020/4/10/47

At first I posted a fix (https://lkml.org/lkml/2020/4/21/1318) trying to resolve the issue from hv_storvsc, but with the help of Bart Van Assche, I realized it's better to fix software_resume(), since this looks like a generic issue, not only pertaining to SCSI.

---
 kernel/power/hibernate.c |    7 +++++++
 1 file changed, 7 insertions(+)

@@ -898,6 +898,13 @@ static int software_resume(void)
 	error = freeze_processes();
 	if (error)
 		goto Close_Finish;
+
+	error = freeze_kernel_threads();
+	if (error) {
+		thaw_processes();
+		goto Close_Finish;
+	}
+
 	error = load_image_and_restore();
 	thaw_processes();
  Finish:

From: Sunwook Eom <[email protected]>

commit ad4e80a639fc61d5ecebb03caa5cdbfb91fcebfc upstream.

The error correction data is computed as if data and hash blocks were concatenated. But hash block number starts from v->hash_start. So, we have to calculate hash block number based on that.

---
 drivers/md/dm-verity-fec.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -435,7 +435,7 @@ int verity_fec_decode(struct dm_verity *
 	fio->level++;
 
 	if (type == DM_VERITY_BLOCK_TYPE_METADATA)
-		block += v->data_blocks;
+		block = block - v->hash_start + v->data_blocks;
 
 	/*
 	 * For RS(M, N), the continuous FEC data is divided into blocks of N

From: Mikulas Patocka <[email protected]>

commit 31b22120194b5c0d460f59e0c98504de1d3f1f14 upstream.

The dm-writecache reads metadata in the target constructor. However, when we reload the target, there could be another active instance running on the same device. This is the sequence of operations when doing a reload:

1. construct new target
2. suspend old target
3. resume new target
4. destroy old target

Metadata that were written by the old target between steps 1 and 2 would not be visible by the new target.

Fix the data corruption by loading the metadata in the resume handler.

Also, validate block_size is at least as large as both the devices' logical block size and only read 1 block from the metadata during target constructor -- no need to read entirety of metadata now that it is done during resume.

---
 drivers/md/dm-writecache.c |   52 ++++++++++++++++++++++++++++++++-------------
 1 file changed, 37 insertions(+), 15 deletions(-)

@@ -882,6 +882,24 @@ static int writecache_alloc_entries(stru
 	return 0;
 }
 
+static int writecache_read_metadata(struct dm_writecache *wc, sector_t n_sectors)
+{
+	struct dm_io_region region;
+	struct dm_io_request req;
+
+	region.bdev = wc->ssd_dev->bdev;
+	region.sector = wc->start_sector;
+	region.count = n_sectors;
+	req.bi_op = REQ_OP_READ;
+	req.bi_op_flags = REQ_SYNC;
+	req.mem.type = DM_IO_VMA;
+	req.mem.ptr.vma = (char *)wc->memory_map;
+	req.client = wc->dm_io;
+	req.notify.fn = NULL;
+
+	return dm_io(&req, 1, &region, NULL);
+}
+
 static void writecache_resume(struct dm_target *ti)
 {
 	struct dm_writecache *wc = ti->private;

@@ -892,8 +910,18 @@ static void writecache_resume(struct dm_
 
 	wc_lock(wc);
 
-	if (WC_MODE_PMEM(wc))
+	if (WC_MODE_PMEM(wc)) {
 		persistent_memory_invalidate_cache(wc->memory_map, wc->memory_map_size);
+	} else {
+		r = writecache_read_metadata(wc, wc->metadata_sectors);
+		if (r) {
+			size_t sb_entries_offset;
+			writecache_error(wc, r, "unable to read metadata: %d", r);
+			sb_entries_offset = offsetof(struct wc_memory_superblock, entries);
+			memset((char *)wc->memory_map + sb_entries_offset, -1,
+			       (wc->metadata_sectors << SECTOR_SHIFT) - sb_entries_offset);
+		}
+	}
 
 	wc->tree = RB_ROOT;
 	INIT_LIST_HEAD(&wc->lru);

@@ -2005,6 +2033,12 @@ static int writecache_ctr(struct dm_targ
 		ti->error = "Invalid block size";
 		goto bad;
 	}
+	if (wc->block_size < bdev_logical_block_size(wc->dev->bdev) ||
+	    wc->block_size < bdev_logical_block_size(wc->ssd_dev->bdev)) {
+		r = -EINVAL;
+		ti->error = "Block size is smaller than device logical block size";
+		goto bad;
+	}
 	wc->block_size_bits = __ffs(wc->block_size);
 
 	wc->max_writeback_jobs = MAX_WRITEBACK_JOBS;

@@ -2093,8 +2127,6 @@ invalid_optional:
 			goto bad;
 		}
 	} else {
-		struct dm_io_region region;
-		struct dm_io_request req;
 		size_t n_blocks, n_metadata_blocks;
 		uint64_t n_bitmap_bits;
 

@@ -2151,19 +2183,9 @@ invalid_optional:
 			goto bad;
 		}
 
-		region.bdev = wc->ssd_dev->bdev;
-		region.sector = wc->start_sector;
-		region.count = wc->metadata_sectors;
-		req.bi_op = REQ_OP_READ;
-		req.bi_op_flags = REQ_SYNC;
-		req.mem.type = DM_IO_VMA;
-		req.mem.ptr.vma = (char *)wc->memory_map;
-		req.client = wc->dm_io;
-		req.notify.fn = NULL;
-
-		r = dm_io(&req, 1, &region, NULL);
+		r = writecache_read_metadata(wc, wc->block_size >> SECTOR_SHIFT);
 		if (r) {
-			ti->error = "Unable to read metadata";
+			ti->error = "Unable to read first block of metadata";
 			goto bad;
 		}
 	}

From: Gabriel Krisman Bertazi <[email protected]>

commit 5686dee34dbfe0238c0274e0454fa0174ac0a57a upstream.

When adding devices that don't have a scsi_dh on a BIO based multipath, I was able to consistently hit the warning below and lock-up the system.

The problem is that __map_bio reads the flag before it potentially being modified by choose_pgpath, and ends up using the older value.

The WARN_ON below is not trivially linked to the issue. It goes like
this: The activate_path delayed_work is not initialized for non-scsi_dh devices, but we always set MPATHF_QUEUE_IO, asking for initialization. That is fine, since MPATHF_QUEUE_IO would be cleared in choose_pgpath. Nevertheless, only for BIO-based mpath, we cache the flag before calling choose_pgpath, and use the older version when deciding if we should initialize the path. Therefore, we end up trying to initialize the paths, and calling the non-initialized activate_path work.

[   82.437100] ------------[ cut here ]------------
[   82.437659] WARNING: CPU: 3 PID: 602 at kernel/workqueue.c:1624
  __queue_delayed_work+0x71/0x90
[   82.438436] Modules linked in:
[   82.438911] CPU: 3 PID: 602 Comm: systemd-udevd Not tainted 5.6.0-rc6+ #339
[   82.439680] RIP: 0010:__queue_delayed_work+0x71/0x90
[   82.440287] Code: c1 48 89 4a 50 81 ff 00 02 00 00 75 2a 4c 89 cf e9
94 d6 07 00 e9 7f e9 ff ff 0f 0b eb c7 0f 0b 48 81 7a 58 40 74 a8 94 74
a7 <0f> 0b 48 83 7a 48 00 74 a5 0f 0b eb a1 89 fe 4c 89 cf e9 c8 c4 07
[   82.441719] RSP: 0018:ffffb738803977c0 EFLAGS: 00010007
[   82.442121] RAX: ffffa086389f9740 RBX: 0000000000000002 RCX: 0000000000000000
[   82.442718] RDX: ffffa086350dd930 RSI: ffffa0863d76f600 RDI: 0000000000000200
[   82.443484] RBP: 0000000000000200 R08: 0000000000000000 R09: ffffa086350dd970
[   82.444128] R10: 0000000000000000 R11: 0000000000000000 R12: ffffa086350dd930
[   82.444773] R13: ffffa0863d76f600 R14: 0000000000000000 R15: ffffa08636738008
[   82.445427] FS:  00007f6abfe9dd40(0000) GS:ffffa0863dd80000(0000) knlGS:00000
[   82.446040] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   82.446478] CR2: 0000557d288db4e8 CR3: 0000000078b36000 CR4: 00000000000006e0
[   82.447104] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[   82.447561] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[   82.448012] Call Trace:
[   82.448164]  queue_delayed_work_on+0x6d/0x80
[   82.448472]  __pg_init_all_paths+0x7b/0xf0
[   82.448714]  pg_init_all_paths+0x26/0x40
[   82.448980]  __multipath_map_bio.isra.0+0x84/0x210
[   82.449267]  __map_bio+0x3c/0x1f0
[   82.449468]  __split_and_process_non_flush+0x14a/0x1b0
[   82.449775]  __split_and_process_bio+0xde/0x340
[   82.450045]  ? dm_get_live_table+0x5/0xb0
[   82.450278]  dm_process_bio+0x98/0x290
[   82.450518]  dm_make_request+0x54/0x120
[   82.450778]  generic_make_request+0xd2/0x3e0
[   82.451038]  ? submit_bio+0x3c/0x150
[   82.451278]  submit_bio+0x3c/0x150
[   82.451492]  mpage_readpages+0x129/0x160
[   82.451756]  ? bdev_evict_inode+0x1d0/0x1d0
[   82.452033]  read_pages+0x72/0x170
[   82.452260]  __do_page_cache_readahead+0x1ba/0x1d0
[   82.452624]  force_page_cache_readahead+0x96/0x110
[   82.452903]  generic_file_read_iter+0x84f/0xae0
[   82.453192]  ? __seccomp_filter+0x7c/0x670
[   82.453547]  new_sync_read+0x10e/0x190
[   82.453883]  vfs_read+0x9d/0x150
[   82.454172]  ksys_read+0x65/0xe0
[   82.454466]  do_syscall_64+0x4e/0x210
[   82.454828]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
[...]
[   82.462501] ---[ end trace bb39975e9cf45daa ]---

---
 drivers/md/dm-mpath.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

@@ -585,10 +585,12 @@ static struct pgpath *__map_bio(struct m
 
 	/* Do we need to select a new pgpath? */
 	pgpath = READ_ONCE(m->current_pgpath);
-	queue_io = test_bit(MPATHF_QUEUE_IO, &m->flags);
-	if (!pgpath || !queue_io)
+	if (!pgpath || !test_bit(MPATHF_QUEUE_IO, &m->flags))
 		pgpath = choose_pgpath(m, bio->bi_iter.bi_size);
 
+	/* MPATHF_QUEUE_IO might have been cleared by choose_pgpath. */
+	queue_io = test_bit(MPATHF_QUEUE_IO, &m->flags);
+
 	if ((pgpath && queue_io) ||
 	    (!pgpath && test_bit(MPATHF_QUEUE_IF_NO_PATH, &m->flags))) {
 		/* Queue for the daemon to resubmit */

From: Russell King <[email protected]>

commit b7dc7205b2ae6b6c9d9cfc3e47d6f08da8647b10 upstream.

We need to indicate that powering off the TI WiFi is safe, to avoid:

wl18xx_driver wl18xx.2.auto: Unbalanced pm_runtime_enable! wl1271_sdio mmc0:0001:2: wl12xx_sdio_power_on: failed to get_sync(-13)

which prevents the WiFi being functional.

---
 arch/arm/boot/dts/imx6qdl-sr-som-ti.dtsi |    1 +
 1 file changed, 1 insertion(+)

@@ -153,6 +153,7 @@
 	bus-width = <4>;
 	keep-power-in-suspend;
 	mmc-pwrseq = <&pwrseq_ti_wifi>;
+	cap-power-off-card;
 	non-removable;
 	vmmc-supply = <&vcc_3v3>;
 	/* vqmmc-supply = <&nvcc_sd1>; - MMC layer doesn't like it! */

From: Christoph Hellwig <[email protected]>

commit 10c70d95c0f2f9a6f52d0e33243d2877370cef51 upstream.

When replacing the bd_super check with a bd_openers I followed a logical conclusion, which turns out to be utterly wrong. When a block device has bd_super sets it has a mount file system on it (although not every mounted file system sets bd_super), but that also implies it doesn't even have partitions to start with.

So instead of trying to come up with a logical check for all openers, just remove the check entirely.

---
 block/partition-generic.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -468,7 +468,7 @@ int blk_drop_partitions(struct gendisk *
 
 	if (!disk_part_scan_enabled(disk))
 		return 0;
-	if (bdev->bd_part_count || bdev->bd_openers > 1)
+	if (bdev->bd_part_count)
 		return -EBUSY;
 	res = invalidate_partition(disk, 0);
 	if (res)

From: Martin Wilck <[email protected]>

commit 856e152a3c08bf7987cbd41900741d83d9cddc8e upstream.

The purpose of the UNLOADING flag is to avoid port login procedures to continue when a controller is in the process of shutting down. It makes sense to set this flag before starting session teardown.

Furthermore, use atomic test_and_set_bit() to avoid the shutdown being run multiple times in parallel. In qla2x00_disable_board_on_pci_error(), the test for UNLOADING is postponed until after the check for an already disabled PCI board.

---
 drivers/scsi/qla2xxx/qla_os.c |   32 ++++++++++++++------------------
 1 file changed, 14 insertions(+), 18 deletions(-)

@@ -3720,6 +3720,13 @@ qla2x00_remove_one(struct pci_dev *pdev)
 	}
 	qla2x00_wait_for_hba_ready(base_vha);
 
+	/*
+	 * if UNLOADING flag is already set, then continue unload,
+	 * where it was set first.
+	 */
+	if (test_and_set_bit(UNLOADING, &base_vha->dpc_flags))
+		return;
+
 	if (IS_QLA25XX(ha) || IS_QLA2031(ha) || IS_QLA27XX(ha) ||
 	    IS_QLA28XX(ha)) {
 		if (ha->flags.fw_started)

@@ -3738,15 +3745,6 @@ qla2x00_remove_one(struct pci_dev *pdev)
 
 	qla2x00_wait_for_sess_deletion(base_vha);
 
-	/*
-	 * if UNLOAD flag is already set, then continue unload,
-	 * where it was set first.
-	 */
-	if (test_bit(UNLOADING, &base_vha->dpc_flags))
-		return;
-
-	set_bit(UNLOADING, &base_vha->dpc_flags);
-
 	qla_nvme_delete(base_vha);
 
 	dma_free_coherent(&ha->pdev->dev,

@@ -6044,13 +6042,6 @@ qla2x00_disable_board_on_pci_error(struc
 	struct pci_dev *pdev = ha->pdev;
 	scsi_qla_host_t *base_vha = pci_get_drvdata(ha->pdev);
 
-	/*
-	 * if UNLOAD flag is already set, then continue unload,
-	 * where it was set first.
-	 */
-	if (test_bit(UNLOADING, &base_vha->dpc_flags))
-		return;
-
 	ql_log(ql_log_warn, base_vha, 0x015b,
 	    "Disabling adapter.\n");
 

@@ -6061,9 +6052,14 @@ qla2x00_disable_board_on_pci_error(struc
 		return;
 	}
 
-	qla2x00_wait_for_sess_deletion(base_vha);
+	/*
+	 * if UNLOADING flag is already set, then continue unload,
+	 * where it was set first.
+	 */
+	if (test_and_set_bit(UNLOADING, &base_vha->dpc_flags))
+		return;
 
-	set_bit(UNLOADING, &base_vha->dpc_flags);
+	qla2x00_wait_for_sess_deletion(base_vha);
 
 	qla2x00_delete_all_vps(ha, base_vha);
 

From: Martin Wilck <[email protected]>

commit 5a263892d7d0b4fe351363f8d1a14c6a75955475 upstream.

qlt_free_session_done() tries to post async PRLO / LOGO, and waits for the completion of these async commands. If UNLOADING is set, this is doomed to timeout, because the async logout command will never complete.

The only way to avoid waiting pointlessly is to fail posting these commands in the first place if the driver is in UNLOADING state. In general, posting any command should be avoided when the driver is UNLOADING.

With this patch, "rmmod qla2xxx" completes without noticeable delay.

---
 drivers/scsi/qla2xxx/qla_os.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -4854,6 +4854,9 @@ qla2x00_alloc_work(struct scsi_qla_host
 	struct qla_work_evt *e;
 	uint8_t bail;
 
+	if (test_bit(UNLOADING, &vha->dpc_flags))
+		return NULL;
+
 	QLA_VHA_MARK_BUSY(vha, bail);
 	if (bail)
 		return NULL;

From: Aharon Landau <[email protected]>

commit 2d7e3ff7b6f2c614eb21d0dc348957a47eaffb57 upstream.

GRH fields such as sgid_index, hop limit, et. are set in the QP context when QP is created/modified.

Currently, when query QP is performed, we fill the GRH fields only if the GRH bit is set in the QP context, but this bit is not set for RoCE. Adjust the check so we will set all relevant data for the RoCE too.

Since this data is returned to userspace, the below is an ABI regression.

---
 drivers/infiniband/hw/mlx5/qp.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -5545,7 +5545,9 @@ static void to_rdma_ah_attr(struct mlx5_
 	rdma_ah_set_path_bits(ah_attr, path->grh_mlid & 0x7f);
 	rdma_ah_set_static_rate(ah_attr,
 				path->static_rate ? path->static_rate - 5 : 0);
-	if (path->grh_mlid & (1 << 7)) {
+
+	if (path->grh_mlid & (1 << 7) ||
+	    ah_attr->type == RDMA_AH_ATTR_TYPE_ROCE) {
 		u32 tc_fl = be32_to_cpu(path->tclass_flowlabel);
 
 		rdma_ah_set_grh(ah_attr, NULL,

From: Jason Gunthorpe <[email protected]>

commit 39c011a538272589b9eb02ff1228af528522a22c upstream.

If uverbs_user_mmap_disassociate() is called while the mmap is concurrently doing exit_mmap then the ordering of the rdma_user_mmap_entry_put() is not reliable.

The put must be done before uvers_user_mmap_disassociate() returns, otherwise there can be a use after free on the ucontext, and a left over entry in the xarray. If the put is not done here then it is done during rdma_umap_close() later.

Add the missing put to the error exit path.

  WARNING: CPU: 7 PID: 7111 at drivers/infiniband/core/rdma_core.c:810 uverbs_destroy_ufile_hw+0x2a5/0x340 [ib_uverbs]
  Modules linked in: bonding ipip tunnel4 geneve ip6_udp_tunnel udp_tunnel ip6_gre ip6_tunnel tunnel6 ip_gre ip_tunnel gre mlx5_ib mlx5_core mlxfw pci_hyperv_intf act_ct nf_flow_table ptp pps_core rdma_ucm ib_uverbs ib_ipoib ib_umad 8021q garp mrp openvswitch nsh nf_conncount nfsv3 nfs_acl xt_MASQUERADE nf_conntrack_netlink nfnetlink iptable_nat xt_addrtype iptable_filter xt_conntrack br_netfilter bridge stp llc rpcsec_gss_krb5 auth_rpcgss nfsv4 dns_resolver nfs lockd grace fscache overlay rpcrdma ib_isert iscsi_target_mod ib_iser kvm_intel ib_srpt iTCO_wdt target_core_mod iTCO_vendor_support kvm ib_srp nf_nat irqbypass crc32_pclmul crc32c_intel nf_conntrack rfkill nf_defrag_ipv6 virtio_net nf_defrag_ipv4 pcspkr ghash_clmulni_intel i2c_i801 net_failover failover i2c_core lpc_ich mfd_core rdma_cm ib_cm iw_cm button ib_core sunrpc sch_fq_codel ip_tables serio_raw [last unloaded: tunnel4]
  CPU: 7 PID: 7111 Comm: python3 Tainted: G        W         5.6.0-rc6-for-upstream-dbg-2020-03-21_06-41-26-18 #1
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.12.1-0-ga5cab58e9a3f-prebuilt.qemu.org 04/01/2014
  RIP: 0010:uverbs_destroy_ufile_hw+0x2a5/0x340 [ib_uverbs]
  Code: ff df 48 89 fa 48 c1 ea 03 80 3c 02 00 75 74 49 8b 84 24 08 01 00 00 48 85 c0 0f 84 13 ff ff ff 48 89 ef ff d0 e9 09 ff ff ff <0f> 0b e9 77 ff ff ff e8 0f d8 fa e0 e9 c5 fd ff ff e8 05 d8 fa e0
  RSP: 0018:ffff88840e0779a0 EFLAGS: 00010286
  RAX: dffffc0000000000 RBX: ffff8882a7721c00 RCX: 0000000000000000
  RDX: 1ffff11054ee469f RSI: ffffffff8446d7e0 RDI: ffff8882a77234f8
  RBP: ffff8882a7723400 R08: ffffed1085c0112c R09: 0000000000000001
  R10: 0000000000000001 R11: ffffed1085c0112b R12: ffff888403c30000
  R13: 0000000000000002 R14: ffff8882a7721cb0 R15: ffff8882a7721cd0
  FS:  00007f2046089700(0000) GS:ffff88842de00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007f7cfe9a6e20 CR3: 000000040b8ac006 CR4: 0000000000360ee0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  Call Trace:
   ib_uverbs_remove_one+0x273/0x480 [ib_uverbs]
   ? up_write+0x15c/0x4a0
   remove_client_context+0xa6/0xf0 [ib_core]
   disable_device+0x12d/0x200 [ib_core]
   ? remove_client_context+0xf0/0xf0 [ib_core]
   ? mnt_get_count+0x1d0/0x1d0
   __ib_unregister_device+0x79/0x150 [ib_core]
   ib_unregister_device+0x21/0x30 [ib_core]
   __mlx5_ib_remove+0x91/0x110 [mlx5_ib]
   ? __mlx5_ib_remove+0x110/0x110 [mlx5_ib]
   mlx5_remove_device+0x241/0x310 [mlx5_core]
   mlx5_unregister_device+0x4d/0x1e0 [mlx5_core]
   mlx5_unload_one+0xc0/0x260 [mlx5_core]
   remove_one+0x5c/0x160 [mlx5_core]
   pci_device_remove+0xef/0x2a0
   ? pcibios_free_irq+0x10/0x10
   device_release_driver_internal+0x1d8/0x470
   unbind_store+0x152/0x200
   ? sysfs_kf_write+0x3b/0x180
   ? sysfs_file_ops+0x160/0x160
   kernfs_fop_write+0x284/0x460
   ? __sb_start_write+0x243/0x3a0
   vfs_write+0x197/0x4a0
   ksys_write+0x156/0x1e0
   ? __x64_sys_read+0xb0/0xb0
   ? do_syscall_64+0x73/0x1330
   ? do_syscall_64+0x73/0x1330
   do_syscall_64+0xe7/0x1330
   ? down_write_nested+0x3e0/0x3e0
   ? syscall_return_slowpath+0x970/0x970
   ? entry_SYSCALL_64_after_hwframe+0x3e/0xbe
   ? lockdep_hardirqs_off+0x1de/0x2d0
   ? trace_hardirqs_off_thunk+0x1a/0x1c
   entry_SYSCALL_64_after_hwframe+0x49/0xbe
  RIP: 0033:0x7f20a3ff0cdb
  Code: 53 48 89 d5 48 89 f3 48 83 ec 18 48 89 7c 24 08 e8 5a fd ff ff 48 89 ea 41 89 c0 48 89 de 48 8b 7c 24 08 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 31 44 89 c7 48 89 44 24 08 e8 90 fd ff ff 48
  RSP: 002b:00007f2046087040 EFLAGS: 00000293 ORIG_RAX: 0000000000000001
  RAX: ffffffffffffffda RBX: 00007f2038016df0 RCX: 00007f20a3ff0cdb
  RDX: 000000000000000d RSI: 00007f2038016df0 RDI: 0000000000000018
  RBP: 000000000000000d R08: 0000000000000000 R09: 0000000000000000
  R10: 0000000000000100 R11: 0000000000000293 R12: 00007f2046e29630
  R13: 00007f20280035a0 R14: 0000000000000018 R15: 00007f2038016df0

---
 drivers/infiniband/core/uverbs_main.c |    4 ++++
 1 file changed, 4 insertions(+)

@@ -820,6 +820,10 @@ void uverbs_user_mmap_disassociate(struc
 			ret = mmget_not_zero(mm);
 			if (!ret) {
 				list_del_init(&priv->list);
+				if (priv->entry) {
+					rdma_user_mmap_entry_put(priv->entry);
+					priv->entry = NULL;
+				}
 				mm = NULL;
 				continue;
 			}

From: Alaa Hleihel <[email protected]>

commit c08cfb2d8d78bfe81b37cc6ba84f0875bddd0d5c upstream.

Initialize ib_spec on the stack before using it, otherwise we will have garbage values that will break creating default rules with invalid parsing error.

---
 drivers/infiniband/hw/mlx4/main.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -1502,8 +1502,9 @@ static int __mlx4_ib_create_default_rule
 	int i;
 
 	for (i = 0; i < ARRAY_SIZE(pdefault_rules->rules_create_list); i++) {
+		union ib_flow_spec ib_spec = {};
 		int ret;
-		union ib_flow_spec ib_spec;
+
 		switch (pdefault_rules->rules_create_list[i]) {
 		case 0:
 			/* no rule */

From: Jason Gunthorpe <[email protected]>

commit 6e051971b0e2eeb0ce7ec65d3cc8180450512d36 upstream.

siw_fastreg_mr() invokes siw_mem_id2obj(), which returns a local reference of the siw_mem object to "mem" with increased refcnt. When siw_fastreg_mr() returns, "mem" becomes invalid, so the refcount should be decreased to keep refcount balanced.

The issue happens in one error path of siw_fastreg_mr(). When "base_mr" equals to NULL but "mem" is not NULL, the function forgets to decrease the refcnt increased by siw_mem_id2obj() and causes a refcnt leak.

Reorganize the flow so that the goto unwind can be used as expected.

---
 drivers/infiniband/sw/siw/siw_qp_tx.c |   15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)

@@ -920,20 +920,27 @@ static int siw_fastreg_mr(struct ib_pd *
 {
 	struct ib_mr *base_mr = (struct ib_mr *)(uintptr_t)sqe->base_mr;
 	struct siw_device *sdev = to_siw_dev(pd->device);
-	struct siw_mem *mem = siw_mem_id2obj(sdev, sqe->rkey  >> 8);
+	struct siw_mem *mem;
 	int rv = 0;
 
 	siw_dbg_pd(pd, "STag 0x%08x\n", sqe->rkey);
 
-	if (unlikely(!mem || !base_mr)) {
+	if (unlikely(!base_mr)) {
 		pr_warn("siw: fastreg: STag 0x%08x unknown\n", sqe->rkey);
 		return -EINVAL;
 	}
+
 	if (unlikely(base_mr->rkey >> 8 != sqe->rkey  >> 8)) {
 		pr_warn("siw: fastreg: STag 0x%08x: bad MR\n", sqe->rkey);
-		rv = -EINVAL;
-		goto out;
+		return -EINVAL;
+	}
+
+	mem = siw_mem_id2obj(sdev, sqe->rkey  >> 8);
+	if (unlikely(!mem)) {
+		pr_warn("siw: fastreg: STag 0x%08x unknown\n", sqe->rkey);
+		return -EINVAL;
 	}
+
 	if (unlikely(mem->pd != pd)) {
 		pr_warn("siw: fastreg: PD mismatch\n");
 		rv = -EINVAL;

From: Leon Romanovsky <[email protected]>

commit 0fb00941dc63990a10951146df216fc7b0e20bc2 upstream.

FDs can only be used on the ufile that created them, they cannot be mixed to other ufiles. We are lacking a check to prevent it.

  BUG: KASAN: null-ptr-deref in atomic64_sub_and_test include/asm-generic/atomic-instrumented.h:1547 [inline]
  BUG: KASAN: null-ptr-deref in atomic_long_sub_and_test include/asm-generic/atomic-long.h:460 [inline]
  BUG: KASAN: null-ptr-deref in fput_many+0x1a/0x140 fs/file_table.c:336
  Write of size 8 at addr 0000000000000038 by task syz-executor179/284

  CPU: 0 PID: 284 Comm: syz-executor179 Not tainted 5.5.0-rc5+ #1
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.1-0-ga5cab58e9a3f-prebuilt.qemu.org 04/01/2014
  Call Trace:
   __dump_stack lib/dump_stack.c:77 [inline]
   dump_stack+0x94/0xce lib/dump_stack.c:118
   __kasan_report+0x18f/0x1b7 mm/kasan/report.c:510
   kasan_report+0xe/0x20 mm/kasan/common.c:639
   check_memory_region_inline mm/kasan/generic.c:185 [inline]
   check_memory_region+0x15d/0x1b0 mm/kasan/generic.c:192
   atomic64_sub_and_test include/asm-generic/atomic-instrumented.h:1547 [inline]
   atomic_long_sub_and_test include/asm-generic/atomic-long.h:460 [inline]
   fput_many+0x1a/0x140 fs/file_table.c:336
   rdma_lookup_put_uobject+0x85/0x130 drivers/infiniband/core/rdma_core.c:692
   uobj_put_read include/rdma/uverbs_std_types.h:96 [inline]
   _ib_uverbs_lookup_comp_file drivers/infiniband/core/uverbs_cmd.c:198 [inline]
   create_cq+0x375/0xba0 drivers/infiniband/core/uverbs_cmd.c:1006
   ib_uverbs_create_cq+0x114/0x140 drivers/infiniband/core/uverbs_cmd.c:1089
   ib_uverbs_write+0xaa5/0xdf0 drivers/infiniband/core/uverbs_main.c:769
   __vfs_write+0x7c/0x100 fs/read_write.c:494
   vfs_write+0x168/0x4a0 fs/read_write.c:558
   ksys_write+0xc8/0x200 fs/read_write.c:611
   do_syscall_64+0x9c/0x390 arch/x86/entry/common.c:294
   entry_SYSCALL_64_after_hwframe+0x44/0xa9
  RIP: 0033:0x44ef99
  Code: 00 b8 00 01 00 00 eb e1 e8 74 1c 00 00 0f 1f 40 00 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 c4 ff ff ff f7 d8 64 89 01 48
  RSP: 002b:00007ffc0b74c028 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
  RAX: ffffffffffffffda RBX: 00007ffc0b74c030 RCX: 000000000044ef99
  RDX: 0000000000000040 RSI: 0000000020000040 RDI: 0000000000000005
  RBP: 00007ffc0b74c038 R08: 0000000000401830 R09: 0000000000401830
  R10: 00007ffc0b74c038 R11: 0000000000000246 R12: 0000000000000000
  R13: 0000000000000000 R14: 00000000006be018 R15: 0000000000000000

---
 drivers/infiniband/core/rdma_core.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -360,7 +360,7 @@ lookup_get_fd_uobject(const struct uverb
 	 * uverbs_uobject_fd_release(), and the caller is expected to ensure
 	 * that release is never done while a call to lookup is possible.
 	 */
-	if (f->f_op != fd_type->fops) {
+	if (f->f_op != fd_type->fops || uobject->ufile != ufile) {
 		fput(f);
 		return ERR_PTR(-EBADF);
 	}

From: Leon Romanovsky <[email protected]>

commit 83a2670212215a569ed133efc10c92055c96cc8c upstream.

In case of failure to get file, the uobj is overwritten and causes to supply bad pointer as an input to uverbs_uobject_put().

  BUG: KASAN: null-ptr-deref in atomic_fetch_sub include/asm-generic/atomic-instrumented.h:199 [inline]
  BUG: KASAN: null-ptr-deref in refcount_sub_and_test include/linux/refcount.h:253 [inline]
  BUG: KASAN: null-ptr-deref in refcount_dec_and_test include/linux/refcount.h:281 [inline]
  BUG: KASAN: null-ptr-deref in kref_put include/linux/kref.h:64 [inline]
  BUG: KASAN: null-ptr-deref in uverbs_uobject_put+0x22/0x90 drivers/infiniband/core/rdma_core.c:57
  Write of size 4 at addr 0000000000000030 by task syz-executor.4/1691

  CPU: 1 PID: 1691 Comm: syz-executor.4 Not tainted 5.6.0 #17
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.1-0-ga5cab58e9a3f-prebuilt.qemu.org 04/01/2014
  Call Trace:
   __dump_stack lib/dump_stack.c:77 [inline]
   dump_stack+0x94/0xce lib/dump_stack.c:118
   __kasan_report+0x10c/0x190 mm/kasan/report.c:515
   kasan_report+0x32/0x50 mm/kasan/common.c:625
   check_memory_region_inline mm/kasan/generic.c:187 [inline]
   check_memory_region+0x16d/0x1c0 mm/kasan/generic.c:193
   atomic_fetch_sub include/asm-generic/atomic-instrumented.h:199 [inline]
   refcount_sub_and_test include/linux/refcount.h:253 [inline]
   refcount_dec_and_test include/linux/refcount.h:281 [inline]
   kref_put include/linux/kref.h:64 [inline]
   uverbs_uobject_put+0x22/0x90 drivers/infiniband/core/rdma_core.c:57
   alloc_begin_fd_uobject+0x1d0/0x250 drivers/infiniband/core/rdma_core.c:486
   rdma_alloc_begin_uobject+0xa8/0xf0 drivers/infiniband/core/rdma_core.c:509
   __uobj_alloc include/rdma/uverbs_std_types.h:117 [inline]
   ib_uverbs_create_comp_channel+0x16d/0x230 drivers/infiniband/core/uverbs_cmd.c:982
   ib_uverbs_write+0xaa5/0xdf0 drivers/infiniband/core/uverbs_main.c:665
   __vfs_write+0x7c/0x100 fs/read_write.c:494
   vfs_write+0x168/0x4a0 fs/read_write.c:558
   ksys_write+0xc8/0x200 fs/read_write.c:611
   do_syscall_64+0x9c/0x390 arch/x86/entry/common.c:295
   entry_SYSCALL_64_after_hwframe+0x44/0xa9
  RIP: 0033:0x466479
  Code: f7 d8 64 89 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 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 bc ff ff ff f7 d8 64 89 01 48
  RSP: 002b:00007efe9f6a7c48 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
  RAX: ffffffffffffffda RBX: 000000000073bf00 RCX: 0000000000466479
  RDX: 0000000000000018 RSI: 0000000020000040 RDI: 0000000000000003
  RBP: 00007efe9f6a86bc R08: 0000000000000000 R09: 0000000000000000
  R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000005
  R13: 0000000000000bf2 R14: 00000000004cb80a R15: 00000000006fefc0

---
 drivers/infiniband/core/rdma_core.c |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

@@ -474,16 +474,15 @@ alloc_begin_fd_uobject(const struct uver
 	filp = anon_inode_getfile(fd_type->name, fd_type->fops, NULL,
 				  fd_type->flags);
 	if (IS_ERR(filp)) {
+		uverbs_uobject_put(uobj);
 		uobj = ERR_CAST(filp);
-		goto err_uobj;
+		goto err_fd;
 	}
 	uobj->object = filp;
 
 	uobj->id = new_fd;
 	return uobj;
 
-err_uobj:
-	uverbs_uobject_put(uobj);
 err_fd:
 	put_unused_fd(new_fd);
 	return uobj;

From: Leon Romanovsky <[email protected]>

commit f0abc761bbb9418876cc4d1ebc473e4ea6352e42 upstream.

The call to ->lookup_put() was too early and it caused an unlock of the read/write protection of the uobject after the FD was put. This allows a race:

CPU1 CPU2
rdma_lookup_put_uobject()
lookup_put_fd_uobject()

     fput()
				   fput()
				     uverbs_uobject_fd_release()
				       WARN_ON(uverbs_try_lock_object(uobj,
					       UVERBS_LOOKUP_WRITE));

atomic_dec(usecnt)

Fix the code by changing the order, first unlock and call to ->lookup_put() after that.

---
 drivers/infiniband/core/rdma_core.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -678,7 +678,6 @@ void rdma_lookup_put_uobject(struct ib_u
 			     enum rdma_lookup_mode mode)
 {
 	assert_uverbs_usecnt(uobj, mode);
-	uobj->uapi_object->type_class->lookup_put(uobj, mode);
 	/*
 	 * In order to unlock an object, either decrease its usecnt for
 	 * read access or zero it in case of exclusive access. See

@@ -695,6 +694,7 @@ void rdma_lookup_put_uobject(struct ib_u
 		break;
 	}
 
+	uobj->uapi_object->type_class->lookup_put(uobj, mode);
 	/* Pairs with the kref obtained by type->lookup_get */
 	uverbs_uobject_put(uobj);
 }

From: Jason Gunthorpe <[email protected]>

commit e8dc4e885c459343970b25acd9320fe9ee5492e7 upstream.

xa_alloc_cyclic() is a SMP release to be paired with some later acquire during xa_load() as part of cm_acquire_id().

As such, xa_alloc_cyclic() must be done after the cm_id is fully initialized, in particular, it absolutely must be after the refcount_set(), otherwise the refcount_inc() in cm_acquire_id() may not see the set.

As there are several cases where a reader will be able to use the id.local_id after cm_acquire_id in the IB_CM_IDLE state there needs to be an unfortunate split into a NULL allocate and a finalizing xa_store.

---
 drivers/infiniband/core/cm.c |   27 +++++++++++----------------
 1 file changed, 11 insertions(+), 16 deletions(-)

@@ -572,18 +572,6 @@ static int cm_init_av_by_path(struct sa_
 	return 0;
 }
 
-static int cm_alloc_id(struct cm_id_private *cm_id_priv)
-{
-	int err;
-	u32 id;
-
-	err = xa_alloc_cyclic_irq(&cm.local_id_table, &id, cm_id_priv,
-			xa_limit_32b, &cm.local_id_next, GFP_KERNEL);
-
-	cm_id_priv->id.local_id = (__force __be32)id ^ cm.random_id_operand;
-	return err;
-}
-
 static u32 cm_local_id(__be32 local_id)
 {
 	return (__force u32) (local_id ^ cm.random_id_operand);

@@ -825,6 +813,7 @@ struct ib_cm_id *ib_create_cm_id(struct
 				 void *context)
 {
 	struct cm_id_private *cm_id_priv;
+	u32 id;
 	int ret;
 
 	cm_id_priv = kzalloc(sizeof *cm_id_priv, GFP_KERNEL);

@@ -836,9 +825,6 @@ struct ib_cm_id *ib_create_cm_id(struct
 	cm_id_priv->id.cm_handler = cm_handler;
 	cm_id_priv->id.context = context;
 	cm_id_priv->id.remote_cm_qpn = 1;
-	ret = cm_alloc_id(cm_id_priv);
-	if (ret)
-		goto error;
 
 	spin_lock_init(&cm_id_priv->lock);
 	init_completion(&cm_id_priv->comp);

@@ -847,11 +833,20 @@ struct ib_cm_id *ib_create_cm_id(struct
 	INIT_LIST_HEAD(&cm_id_priv->altr_list);
 	atomic_set(&cm_id_priv->work_count, -1);
 	refcount_set(&cm_id_priv->refcount, 1);
+
+	ret = xa_alloc_cyclic_irq(&cm.local_id_table, &id, NULL, xa_limit_32b,
+				  &cm.local_id_next, GFP_KERNEL);
+	if (ret)
+		goto error;
+	cm_id_priv->id.local_id = (__force __be32)id ^ cm.random_id_operand;
+	xa_store_irq(&cm.local_id_table, cm_local_id(cm_id_priv->id.local_id),
+		     cm_id_priv, GFP_KERNEL);
+
 	return &cm_id_priv->id;
 
 error:
 	kfree(cm_id_priv);
-	return ERR_PTR(-ENOMEM);
+	return ERR_PTR(ret);
 }
 EXPORT_SYMBOL(ib_create_cm_id);
 

From: Dan Carpenter <[email protected]>

commit 983653515849fb56b78ce55d349bb384d43030f6 upstream.

The xa_alloc_cyclic_irq() function returns either 0 or 1 on success and negatives on error. This code treats 1 as an error and returns ERR_PTR(1) which will cause an Oops in the caller.

---
 drivers/infiniband/core/cm.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -836,7 +836,7 @@ struct ib_cm_id *ib_create_cm_id(struct
 
 	ret = xa_alloc_cyclic_irq(&cm.local_id_table, &id, NULL, xa_limit_32b,
 				  &cm.local_id_next, GFP_KERNEL);
-	if (ret)
+	if (ret < 0)
 		goto error;
 	cm_id_priv->id.local_id = (__force __be32)id ^ cm.random_id_operand;
 	xa_store_irq(&cm.local_id_table, cm_local_id(cm_id_priv->id.local_id),

From: Rayagonda Kokatanur <[email protected]>

commit 068143a8195fb0fdeea1f3ca430b3db0f6d04a53 upstream.

When slave status is I2C_SLAVE_RX_END, generate I2C_SLAVE_STOP event to i2c_client.

---
 drivers/i2c/busses/i2c-bcm-iproc.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -360,6 +360,9 @@ static bool bcm_iproc_i2c_slave_isr(stru
 			value = (u8)((val >> S_RX_DATA_SHIFT) & S_RX_DATA_MASK);
 			i2c_slave_event(iproc_i2c->slave,
 					I2C_SLAVE_WRITE_RECEIVED, &value);
+			if (rx_status == I2C_SLAVE_RX_END)
+				i2c_slave_event(iproc_i2c->slave,
+						I2C_SLAVE_STOP, &value);
 		}
 	} else if (status & BIT(IS_S_TX_UNDERRUN_SHIFT)) {
 		/* Master read other than start */

From: YueHaibing <[email protected]>

commit ae148b43516d90756ff8255925fb7df142b0c76e upstream.

If PCI_MSI is not set, building fais:

drivers/dma/hisi_dma.c: In function ‘hisi_dma_free_irq_vectors’:
drivers/dma/hisi_dma.c:138:2: error: implicit declaration of function ‘pci_free_irq_vectors’;
did you mean ‘pci_alloc_irq_vectors’? [-Werror=implicit-function-declaration]
pci_free_irq_vectors(data);
^~~~~~~~~~~~~~~~~~~~

Make HISI_DMA depends on PCI_MSI to fix this.

---
 drivers/dma/Kconfig |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -241,7 +241,8 @@ config FSL_RAID
 
 config HISI_DMA
 	tristate "HiSilicon DMA Engine support"
-	depends on ARM64 || (COMPILE_TEST && PCI_MSI)
+	depends on ARM64 || COMPILE_TEST
+	depends on PCI_MSI
 	select DMA_ENGINE
 	select DMA_VIRTUAL_CHANNELS
 	help

From: Yan Zhao <[email protected]>

commit 0ea971f8dcd6dee78a9a30ea70227cf305f11ff7 upstream.

add parentheses to avoid possible vaddr overflow.

---
 drivers/vfio/vfio_iommu_type1.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -554,7 +554,7 @@ static int vfio_iommu_type1_pin_pages(vo
 			continue;
 		}
 
-		remote_vaddr = dma->vaddr + iova - dma->iova;
+		remote_vaddr = dma->vaddr + (iova - dma->iova);
 		ret = vfio_pin_page_external(dma, remote_vaddr, &phys_pfn[i],
 					     do_accounting);
 		if (ret)

From: Sean Christopherson <[email protected]>

commit 5cbf3264bc715e9eb384e2b68601f8c02bb9a61d upstream.

Use follow_pfn() to get the PFN of a PFNMAP VMA instead of assuming that vma->vm_pgoff holds the base PFN of the VMA. This fixes a bug where attempting to do VFIO_IOMMU_MAP_DMA on an arbitrary PFNMAP'd region of memory calculates garbage for the PFN.

Hilariously, this only got detected because the first "PFN" calculated by vaddr_get_pfn() is PFN 0 (vma->vm_pgoff==0), and iommu_iova_to_phys() uses PA==0 as an error, which triggers a WARN in vfio_unmap_unpin() because the translation "failed". PFN 0 is now unconditionally reserved on x86 in order to mitigate L1TF, which causes is_invalid_reserved_pfn() to return true and in turns results in vaddr_get_pfn() returning success for PFN 0. Eventually the bogus calculation runs into PFNs that aren't reserved and leads to failure in vfio_pin_map_dma(). The subsequent call to vfio_remove_dma() attempts to unmap PFN 0 and WARNs.

  WARNING: CPU: 8 PID: 5130 at drivers/vfio/vfio_iommu_type1.c:750 vfio_unmap_unpin+0x2e1/0x310 [vfio_iommu_type1]
  Modules linked in: vfio_pci vfio_virqfd vfio_iommu_type1 vfio ...
  CPU: 8 PID: 5130 Comm: sgx Tainted: G        W         5.6.0-rc5-705d787c7fee-vfio+ #3
  Hardware name: Intel Corporation Mehlow UP Server Platform/Moss Beach Server, BIOS CNLSE2R1.D00.X119.B49.1803010910 03/01/2018
  RIP: 0010:vfio_unmap_unpin+0x2e1/0x310 [vfio_iommu_type1]
  Code: <0f> 0b 49 81 c5 00 10 00 00 e9 c5 fe ff ff bb 00 10 00 00 e9 3d fe
  RSP: 0018:ffffbeb5039ebda8 EFLAGS: 00010246
  RAX: 0000000000000000 RBX: ffff9a55cbf8d480 RCX: 0000000000000000
  RDX: 0000000000000000 RSI: 0000000000000001 RDI: ffff9a52b771c200
  RBP: 0000000000000000 R08: 0000000000000040 R09: 00000000fffffff2
  R10: 0000000000000001 R11: ffff9a51fa896000 R12: 0000000184010000
  R13: 0000000184000000 R14: 0000000000010000 R15: ffff9a55cb66ea08
  FS:  00007f15d3830b40(0000) GS:ffff9a55d5600000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 0000561cf39429e0 CR3: 000000084f75f005 CR4: 00000000003626e0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  Call Trace:
   vfio_remove_dma+0x17/0x70 [vfio_iommu_type1]
   vfio_iommu_type1_ioctl+0x9e3/0xa7b [vfio_iommu_type1]
   ksys_ioctl+0x92/0xb0
   __x64_sys_ioctl+0x16/0x20
   do_syscall_64+0x4c/0x180
   entry_SYSCALL_64_after_hwframe+0x44/0xa9
  RIP: 0033:0x7f15d04c75d7
  Code: <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 81 48 2d 00 f7 d8 64 89 01 48

---
 drivers/vfio/vfio_iommu_type1.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -341,8 +341,8 @@ static int vaddr_get_pfn(struct mm_struc
 	vma = find_vma_intersection(mm, vaddr, vaddr + 1);
 
 	if (vma && vma->vm_flags & VM_PFNMAP) {
-		*pfn = ((vaddr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
-		if (is_invalid_reserved_pfn(*pfn))
+		if (!follow_pfn(vma, vaddr, pfn) &&
+		    is_invalid_reserved_pfn(*pfn))
 			ret = 0;
 	}
 done:

From: Tang Bin <[email protected]>

commit b52649aee6243ea661905bdc5fbe28cc5f6dec76 upstream.

The function qcom_iommu_device_probe() does not perform sufficient error checking after executing devm_ioremap_resource(), which can result in crashes if a critical error path is encountered.

---
 drivers/iommu/qcom_iommu.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -813,8 +813,11 @@ static int qcom_iommu_device_probe(struc
 	qcom_iommu->dev = dev;
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (res)
+	if (res) {
 		qcom_iommu->local_base = devm_ioremap_resource(dev, res);
+		if (IS_ERR(qcom_iommu->local_base))
+			return PTR_ERR(qcom_iommu->local_base);
+	}
 
 	qcom_iommu->iface_clk = devm_clk_get(dev, "iface");
 	if (IS_ERR(qcom_iommu->iface_clk)) {

From: Grygorii Strashko <[email protected]>

commit 172d59ecd61b89f535ad99a7e531c0f111453b9a upstream.

The mutex_unlock() is missed on error path of psil_get_ep_config() which causes deadlock, so add missed mutex_unlock().

---
 drivers/dma/ti/k3-psil.c |    1 +
 1 file changed, 1 insertion(+)

@@ -27,6 +27,7 @@ struct psil_endpoint_config *psil_get_ep
 			soc_ep_map = &j721e_ep_map;
 		} else {
 			pr_err("PSIL: No compatible machine found for map\n");
+			mutex_unlock(&ep_map_mutex);
 			return ERR_PTR(-ENOTSUPP);
 		}
 		pr_debug("%s: Using map for %s\n", __func__, soc_ep_map->name);

From: Dave Jiang <[email protected]>

commit 0821009445a8261ac4d32a6df4b83938e007c765 upstream.

When the channel register code was changed to allow hotplug operations, dynamic indexing wasn't taken into account. When channels are randomly plugged and unplugged out of order, the serial indexing breaks. Convert channel indexing to using IDA tracking in order to allow dynamic assignment. The previous code does not cause any regression bug for existing channel allocation besides idxd driver since the hotplug usage case is only used by idxd at this point.

With this change, the chan->idr_ref is also not needed any longer. We can have a device with no channels registered due to hot plug. The channel device release code no longer should attempt to free the dma device id on the last channel release.

---
 drivers/dma/dmaengine.c   |   60 +++++++++++++++++++---------------------------
 include/linux/dmaengine.h |    4 +--
 2 files changed, 28 insertions(+), 36 deletions(-)

@@ -151,10 +151,6 @@ static void chan_dev_release(struct devi
 	struct dma_chan_dev *chan_dev;
 
 	chan_dev = container_of(dev, typeof(*chan_dev), device);
-	if (atomic_dec_and_test(chan_dev->idr_ref)) {
-		ida_free(&dma_ida, chan_dev->dev_id);
-		kfree(chan_dev->idr_ref);
-	}
 	kfree(chan_dev);
 }
 

@@ -952,27 +948,9 @@ static int get_dma_id(struct dma_device
 }
 
 static int __dma_async_device_channel_register(struct dma_device *device,
-					       struct dma_chan *chan,
-					       int chan_id)
+					       struct dma_chan *chan)
 {
 	int rc = 0;
-	int chancnt = device->chancnt;
-	atomic_t *idr_ref;
-	struct dma_chan *tchan;
-
-	tchan = list_first_entry_or_null(&device->channels,
-					 struct dma_chan, device_node);
-	if (!tchan)
-		return -ENODEV;
-
-	if (tchan->dev) {
-		idr_ref = tchan->dev->idr_ref;
-	} else {
-		idr_ref = kmalloc(sizeof(*idr_ref), GFP_KERNEL);
-		if (!idr_ref)
-			return -ENOMEM;
-		atomic_set(idr_ref, 0);
-	}
 
 	chan->local = alloc_percpu(typeof(*chan->local));
 	if (!chan->local)

@@ -988,29 +966,36 @@ static int __dma_async_device_channel_re
 	 * When the chan_id is a negative value, we are dynamically adding
 	 * the channel. Otherwise we are static enumerating.
 	 */
-	chan->chan_id = chan_id < 0 ? chancnt : chan_id;
+	mutex_lock(&device->chan_mutex);
+	chan->chan_id = ida_alloc(&device->chan_ida, GFP_KERNEL);
+	mutex_unlock(&device->chan_mutex);
+	if (chan->chan_id < 0) {
+		pr_err("%s: unable to alloc ida for chan: %d\n",
+		       __func__, chan->chan_id);
+		goto err_out;
+	}
+
 	chan->dev->device.class = &dma_devclass;
 	chan->dev->device.parent = device->dev;
 	chan->dev->chan = chan;
-	chan->dev->idr_ref = idr_ref;
 	chan->dev->dev_id = device->dev_id;
-	atomic_inc(idr_ref);
 	dev_set_name(&chan->dev->device, "dma%dchan%d",
 		     device->dev_id, chan->chan_id);
-
 	rc = device_register(&chan->dev->device);
 	if (rc)
-		goto err_out;
+		goto err_out_ida;
 	chan->client_count = 0;
-	device->chancnt = chan->chan_id + 1;
+	device->chancnt++;
 
 	return 0;
 
+ err_out_ida:
+	mutex_lock(&device->chan_mutex);
+	ida_free(&device->chan_ida, chan->chan_id);
+	mutex_unlock(&device->chan_mutex);
  err_out:
 	free_percpu(chan->local);
 	kfree(chan->dev);
-	if (atomic_dec_return(idr_ref) == 0)
-		kfree(idr_ref);
 	return rc;
 }
 

@@ -1019,7 +1004,7 @@ int dma_async_device_channel_register(st
 {
 	int rc;
 
-	rc = __dma_async_device_channel_register(device, chan, -1);
+	rc = __dma_async_device_channel_register(device, chan);
 	if (rc < 0)
 		return rc;
 

@@ -1039,6 +1024,9 @@ static void __dma_async_device_channel_u
 	device->chancnt--;
 	chan->dev->chan = NULL;
 	mutex_unlock(&dma_list_mutex);
+	mutex_lock(&device->chan_mutex);
+	ida_free(&device->chan_ida, chan->chan_id);
+	mutex_unlock(&device->chan_mutex);
 	device_unregister(&chan->dev->device);
 	free_percpu(chan->local);
 }

@@ -1061,7 +1049,7 @@ EXPORT_SYMBOL_GPL(dma_async_device_chann
  */
 int dma_async_device_register(struct dma_device *device)
 {
-	int rc, i = 0;
+	int rc;
 	struct dma_chan* chan;
 
 	if (!device)

@@ -1166,9 +1154,12 @@ int dma_async_device_register(struct dma
 	if (rc != 0)
 		return rc;
 
+	mutex_init(&device->chan_mutex);
+	ida_init(&device->chan_ida);
+
 	/* represent channels in sysfs. Probably want devs too */
 	list_for_each_entry(chan, &device->channels, device_node) {
-		rc = __dma_async_device_channel_register(device, chan, i++);
+		rc = __dma_async_device_channel_register(device, chan);
 		if (rc < 0)
 			goto err_out;
 	}

@@ -1239,6 +1230,7 @@ void dma_async_device_unregister(struct
 	 */
 	dma_cap_set(DMA_PRIVATE, device->cap_mask);
 	dma_channel_rebalance();
+	ida_free(&dma_ida, device->dev_id);
 	dma_device_put(device);
 	mutex_unlock(&dma_list_mutex);
 }

@@ -336,13 +336,11 @@ struct dma_chan {
  * @chan: driver channel device
  * @device: sysfs device
  * @dev_id: parent dma_device dev_id
- * @idr_ref: reference count to gate release of dma_device dev_id
  */
 struct dma_chan_dev {
 	struct dma_chan *chan;
 	struct device device;
 	int dev_id;
-	atomic_t *idr_ref;
 };
 
 /**

@@ -827,6 +825,8 @@ struct dma_device {
 	int dev_id;
 	struct device *dev;
 	struct module *owner;
+	struct ida chan_ida;
+	struct mutex chan_mutex;	/* to protect chan_ida */
 
 	u32 src_addr_widths;
 	u32 dst_addr_widths;