4.14.298-rc1 review

This is the start of the stable review cycle for the 4.14.298 release. There are 60 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri, 04 Nov 2022 02:20:38 +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/v4.x/stable-review/patch-4.14.298-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-4.14.y
and the diffstat can be found below.

thanks,

greg k-h

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

Greg Kroah-Hartman <[email protected]>

    Linux 4.14.298-rc1

Biju Das <[email protected]>
can: rcar_canfd: rcar_canfd_handle_global_receive(): fix IRQ storm on global FIFO receive

Yang Yingliang <[email protected]>
net: ehea: fix possible memory leak in ehea_register_port()

Aaron Conole <[email protected]>
openvswitch: switch from WARN to pr_warn

Takashi Iwai <[email protected]>
ALSA: aoa: Fix I2S device accounting

Yang Yingliang <[email protected]>
ALSA: aoa: i2sbus: fix possible memory leak in i2sbus_add_dev()

Sudeep Holla <[email protected]>
PM: domains: Fix handling of unavailable/disabled idle states

Yang Yingliang <[email protected]>
net: ksz884x: fix missing pci_disable_device() on error in pcidev_init()

Slawomir Laba <[email protected]>
i40e: Fix flow-type by setting GL_HASH_INSET registers

Slawomir Laba <[email protected]>
i40e: Fix ethtool rx-flow-hash setting for X722

Hans Verkuil <[email protected]>
media: videodev2.h: V4L2_DV_BT_BLANKING_HEIGHT should check 'interlaced'

Hans Verkuil <[email protected]>
media: v4l2-dv-timings: add sanity checks for blanking values

Hans Verkuil <[email protected]>

    media: vivid: dev->bitmap_cap wasn't freed in all cases

Hans Verkuil <[email protected]>
media: vivid: s_fbuf: add more sanity checks

Mario Limonciello <[email protected]>
PM: hibernate: Allow hybrid sleep to work with s2idle

Dongliang Mu <[email protected]>
can: mscan: mpc5xxx: mpc5xxx_can_probe(): add missing put_clock() in error path

Neal Cardwell <[email protected]>
tcp: fix indefinite deferral of RTO with SACK reneging

Zhang Changzhong <[email protected]>
net: lantiq_etop: don't free skb when returning NETDEV_TX_BUSY

Eric Dumazet <[email protected]>

    kcm: annotate data-races around kcm->rx_wait

Eric Dumazet <[email protected]>

    kcm: annotate data-races around kcm->rx_psock

Raju Rangoju <[email protected]>
amd-xgbe: add the bit rate quirk for Molex cables

Raju Rangoju <[email protected]>
amd-xgbe: fix the SFP compliance codes check for DAC cables

Chen Zhongjin <[email protected]>
x86/unwind/orc: Fix unreliable stack dump with gcov

Yang Yingliang <[email protected]>
ALSA: ac97: fix possible memory leak in snd_ac97_dev_register()

Randy Dunlap <[email protected]>
arc: iounmap() arg is volatile

Nathan Huckleberry <[email protected]>
drm/msm: Fix return type of mdp4_lvds_connector_mode_valid

Wei Yongjun <[email protected]>
net: ieee802154: fix error return code in dgram_bind()

Rik van Riel <[email protected]>

    mm,hugetlb: take hugetlb_lock before decrementing h->resv_huge_pages

M. Vefa Bicakci <[email protected]>
xen/gntdev: Prevent leaking grants

Jan Beulich <[email protected]>
Xen/gntdev: don't ignore kernel unmapping error

Heiko Carstens <[email protected]>
s390/futex: add missing EX_TABLE entry to __futex_atomic_op()

Christian A. Ehrhardt <[email protected]>
kernfs: fix use-after-free in __kernfs_remove

Matthew Ma <[email protected]>
mmc: core: Fix kernel panic when remove non-standard SDIO card

Johan Hovold <[email protected]>
drm/msm/hdmi: fix memory corruption with too many bridges

Miquel Raynal <[email protected]>
mac802154: Fix LQI recording

Hyunwoo Kim <[email protected]>
fbdev: smscufx: Fix several use-after-free bugs

Shreeya Patel <[email protected]>
iio: light: tsl2583: Fix module unloading

Matti Vaittinen <[email protected]>
tools: iio: iio_utils: fix digit calculation

Mathias Nyman <[email protected]>
xhci: Remove device endpoints from bandwidth list when freeing the device

Jens Glathe <[email protected]>
usb: xhci: add XHCI_SPURIOUS_SUCCESS to ASM1042 despite being a V0.96 controller

Justin Chen <[email protected]>
usb: bdc: change state when port disconnected

Thinh Nguyen <[email protected]>
usb: dwc3: gadget: Don't set IMI for no_interrupt

Hannu Hartikainen <[email protected]>
USB: add RESET_RESUME quirk for NVIDIA Jetson devices in RCM

Jason A. Donenfeld <[email protected]>
ALSA: au88x0: use explicitly signed char

Steven Rostedt (Google) <[email protected]>
ALSA: Use del_timer_sync() before freeing timer

Werner Sembach <[email protected]>
ACPI: video: Force backlight native for more TongFang devices

Chen-Yu Tsai <[email protected]>
media: v4l2-mem2mem: Apply DST_QUEUE_OFF_BASE on MMAP buffers across ioctls

Jerry Snitselaar <[email protected]>
iommu/vt-d: Clean up si_domain in the init_dmars() error path

Yang Yingliang <[email protected]>
net: hns: fix possible memory leak in hnae_ae_register()

Xiaobo Liu <[email protected]>
net/atm: fix proc_mpc_write incorrect return value

José Expósito <[email protected]>
HID: magicmouse: Do not set BTN_MOUSE on double report

Tony Luck <[email protected]>
ACPI: extlog: Handle multiple records

Filipe Manana <[email protected]>
btrfs: fix processing of delayed data refs during backref walking

Jean-Francois Le Fillatre <[email protected]>
r8152: add PID for the Lenovo OneLink+ Dock

James Morse <[email protected]>
arm64: errata: Remove AES hwcap for COMPAT tasks

Eric Ren <[email protected]>
KVM: arm64: vgic: Fix exit condition in scan_its_table()

Kai-Heng Feng <[email protected]>
ata: ahci: Match EM_MAX_SLOTS with SATA_PMP_MAX_PORTS

Alexander Stein <[email protected]>
ata: ahci-imx: Fix MODULE_ALIAS

Borislav Petkov <[email protected]>
x86/microcode/AMD: Apply the patch early on every logical thread

Joseph Qi <[email protected]>
ocfs2: fix BUG when iput after ocfs2_mknod fails

Joseph Qi <[email protected]>
ocfs2: clear dinode links count in case of error

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

Diffstat:

 Documentation/arm64/silicon-errata.txt             |   2 +
 Makefile                                           |   4 +-
 arch/arc/include/asm/io.h                          |   2 +-
 arch/arc/mm/ioremap.c                              |   2 +-
 arch/arm64/Kconfig                                 |  16 ++++
 arch/arm64/include/asm/cpucaps.h                   |   3 +-
 arch/arm64/kernel/cpu_errata.c                     |  16 ++++
 arch/arm64/kernel/cpufeature.c                     |  13 ++-
 arch/s390/include/asm/futex.h                      |   3 +-
 arch/x86/kernel/cpu/microcode/amd.c                |  16 +++-
 arch/x86/kernel/unwind_orc.c                       |   2 +-
 drivers/acpi/acpi_extlog.c                         |  33 ++++---
 drivers/acpi/video_detect.c                        |  64 +++++++++++++
 drivers/ata/ahci.h                                 |   2 +-
 drivers/ata/ahci_imx.c                             |   2 +-
 drivers/base/power/domain.c                        |   4 +
 drivers/gpu/drm/msm/hdmi/hdmi.c                    |   5 ++
 drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_connector.c |   5 +-
 drivers/hid/hid-magicmouse.c                       |   2 +-
 drivers/iio/light/tsl2583.c                        |   2 +-
 drivers/iommu/intel-iommu.c                        |   5 ++
 drivers/media/platform/vivid/vivid-core.c          |  22 +++++
 drivers/media/platform/vivid/vivid-core.h          |   2 +
 drivers/media/platform/vivid/vivid-vid-cap.c       |  27 ++++--
 drivers/media/v4l2-core/v4l2-dv-timings.c          |  14 +++
 drivers/media/v4l2-core/v4l2-mem2mem.c             |  62 +++++++++----
 drivers/mmc/core/sdio_bus.c                        |   3 +-
 drivers/net/can/mscan/mpc5xxx_can.c                |   8 +-
 drivers/net/can/rcar/rcar_canfd.c                  |   6 +-
 drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c        |  17 ++--
 drivers/net/ethernet/hisilicon/hns/hnae.c          |   4 +-
 drivers/net/ethernet/ibm/ehea/ehea_main.c          |   1 +
 drivers/net/ethernet/intel/i40e/i40e_ethtool.c     | 100 ++++++++++++---------
 drivers/net/ethernet/intel/i40e/i40e_type.h        |   4 +
 drivers/net/ethernet/lantiq_etop.c                 |   1 -
 drivers/net/ethernet/micrel/ksz884x.c              |   2 +-
 drivers/net/usb/cdc_ether.c                        |   7 ++
 drivers/net/usb/r8152.c                            |   1 +
 drivers/usb/core/quirks.c                          |   9 ++
 drivers/usb/dwc3/gadget.c                          |   4 +-
 drivers/usb/gadget/udc/bdc/bdc_udc.c               |   1 +
 drivers/usb/host/xhci-mem.c                        |  20 +++--
 drivers/usb/host/xhci-pci.c                        |   8 +-
 drivers/video/fbdev/smscufx.c                      |  55 ++++++------
 drivers/xen/gntdev.c                               |  30 +++++--
 fs/btrfs/backref.c                                 |  33 +++++--
 fs/kernfs/dir.c                                    |   5 +-
 fs/ocfs2/namei.c                                   |  23 +++--
 include/uapi/linux/videodev2.h                     |   3 +-
 kernel/power/hibernate.c                           |   2 +-
 mm/hugetlb.c                                       |   2 +-
 net/atm/mpoa_proc.c                                |   3 +-
 net/ieee802154/socket.c                            |   4 +-
 net/ipv4/tcp_input.c                               |   3 +-
 net/kcm/kcmsock.c                                  |  23 +++--
 net/mac802154/rx.c                                 |   5 +-
 net/openvswitch/datapath.c                         |   3 +-
 sound/aoa/soundbus/i2sbus/core.c                   |   7 +-
 sound/pci/ac97/ac97_codec.c                        |   1 +
 sound/pci/au88x0/au88x0.h                          |   6 +-
 sound/pci/au88x0/au88x0_core.c                     |   2 +-
 sound/synth/emux/emux.c                            |   7 +-
 tools/iio/iio_utils.c                              |   4 +
 virt/kvm/arm/vgic/vgic-its.c                       |   5 +-
 64 files changed, 553 insertions(+), 199 deletions(-)

From: Joseph Qi <[email protected]>

commit 28f4821b1b53e0649706912e810c6c232fc506f9 upstream.

In ocfs2_mknod(), if error occurs after dinode successfully allocated, ocfs2 i_links_count will not be 0.

So even though we clear inode i_nlink before iput in error handling, it still won't wipe inode since we'll refresh inode from dinode during inode lock. So just like clear inode i_nlink, we clear ocfs2 i_links_count as well. Also do the same change for ocfs2_symlink().

---
 fs/ocfs2/namei.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

@@ -244,6 +244,7 @@ static int ocfs2_mknod(struct inode *dir
 	handle_t *handle = NULL;
 	struct ocfs2_super *osb;
 	struct ocfs2_dinode *dirfe;
+	struct ocfs2_dinode *fe = NULL;
 	struct buffer_head *new_fe_bh = NULL;
 	struct inode *inode = NULL;
 	struct ocfs2_alloc_context *inode_ac = NULL;

@@ -394,6 +395,7 @@ static int ocfs2_mknod(struct inode *dir
 		goto leave;
 	}
 
+	fe = (struct ocfs2_dinode *) new_fe_bh->b_data;
 	if (S_ISDIR(mode)) {
 		status = ocfs2_fill_new_dir(osb, handle, dir, inode,
 					    new_fe_bh, data_ac, meta_ac);

@@ -459,8 +461,11 @@ static int ocfs2_mknod(struct inode *dir
 leave:
 	if (status < 0 && did_quota_inode)
 		dquot_free_inode(inode);
-	if (handle)
+	if (handle) {
+		if (status < 0 && fe)
+			ocfs2_set_links_count(fe, 0);
 		ocfs2_commit_trans(osb, handle);
+	}
 
 	ocfs2_inode_unlock(dir, 1);
 	if (did_block_signals)

@@ -2030,8 +2035,11 @@ bail:
 					ocfs2_clusters_to_bytes(osb->sb, 1));
 	if (status < 0 && did_quota_inode)
 		dquot_free_inode(inode);
-	if (handle)
+	if (handle) {
+		if (status < 0 && fe)
+			ocfs2_set_links_count(fe, 0);
 		ocfs2_commit_trans(osb, handle);
+	}
 
 	ocfs2_inode_unlock(dir, 1);
 	if (did_block_signals)

From: Joseph Qi <[email protected]>

commit 759a7c6126eef5635506453e9b9d55a6a3ac2084 upstream.

Commit b1529a41f777 "ocfs2: should reclaim the inode if '__ocfs2_mknod_locked' returns an error" tried to reclaim the claimed inode if __ocfs2_mknod_locked() fails later. But this introduce a race, the freed bit may be reused immediately by another thread, which will update dinode, e.g. i_generation. Then iput this inode will lead to BUG:
inode->i_generation != le32_to_cpu(fe->i_generation)

We could make this inode as bad, but we did want to do operations like wipe in some cases. Since the claimed inode bit can only affect that an dinode is missing and will return back after fsck, it seems not a big problem. So just leave it as is by revert the reclaim logic.

---
 fs/ocfs2/namei.c |   11 +----------
 1 file changed, 1 insertion(+), 10 deletions(-)

@@ -643,18 +643,9 @@ static int ocfs2_mknod_locked(struct ocf
 		return status;
 	}
 
-	status = __ocfs2_mknod_locked(dir, inode, dev, new_fe_bh,
+	return __ocfs2_mknod_locked(dir, inode, dev, new_fe_bh,
 				    parent_fe_bh, handle, inode_ac,
 				    fe_blkno, suballoc_loc, suballoc_bit);
-	if (status < 0) {
-		u64 bg_blkno = ocfs2_which_suballoc_group(fe_blkno, suballoc_bit);
-		int tmp = ocfs2_free_suballoc_bits(handle, inode_ac->ac_inode,
-				inode_ac->ac_bh, suballoc_bit, bg_blkno, 1);
-		if (tmp)
-			mlog_errno(tmp);
-	}
-
-	return status;
 }
 
 static int ocfs2_mkdir(struct inode *dir,

From: Borislav Petkov <[email protected]>

commit e7ad18d1169c62e6c78c01ff693fd362d9d65278 upstream.

Currently, the patch application logic checks whether the revision needs to be applied on each logical CPU (SMT thread). Therefore, on SMT designs where the microcode engine is shared between the two threads, the application happens only on one of them as that is enough to update the shared microcode engine.

However, there are microcode patches which do per-thread modification, see Link tag below.

Therefore, drop the revision check and try applying on each thread. This is what the BIOS does too so this method is very much tested.

Btw, change only the early paths. On the late loading paths, there's no point in doing per-thread modification because if is it some case like in the bugzilla below - removing a CPUID flag - the kernel cannot go and un-use features it has detected are there early. For that, one should use early loading anyway.

[ bp: Fixes does not contain the oldest commit which did check for
equality but that is good enough. ]

---
 arch/x86/kernel/cpu/microcode/amd.c |   16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

@@ -222,7 +222,13 @@ apply_microcode_early_amd(u32 cpuid_1_ea
 		return ret;
 
 	native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy);
-	if (rev >= mc->hdr.patch_id)
+
+	/*
+	 * Allow application of the same revision to pick up SMT-specific
+	 * changes even if the revision of the other SMT thread is already
+	 * up-to-date.
+	 */
+	if (rev > mc->hdr.patch_id)
 		return ret;
 
 	if (!__apply_microcode_amd(mc)) {

@@ -304,8 +310,12 @@ void load_ucode_amd_ap(unsigned int cpui
 
 	native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy);
 
-	/* Check whether we have saved a new patch already: */
-	if (*new_rev && rev < mc->hdr.patch_id) {
+	/*
+	 * Check whether a new patch has been saved already. Also, allow application of
+	 * the same revision in order to pick up SMT-thread-specific configuration even
+	 * if the sibling SMT thread already has an up-to-date revision.
+	 */
+	if (*new_rev && rev <= mc->hdr.patch_id) {
 		if (!__apply_microcode_amd(mc)) {
 			*new_rev = mc->hdr.patch_id;
 			return;

From: Alexander Stein <[email protected]>

commit 979556f1521a835a059de3b117b9c6c6642c7d58 upstream.

'ahci:' is an invalid prefix, preventing the module from autoloading. Fix this by using the 'platform:' prefix and DRV_NAME.

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

@@ -887,4 +887,4 @@ module_platform_driver(imx_ahci_driver);
 MODULE_DESCRIPTION("Freescale i.MX AHCI SATA platform driver");
 MODULE_AUTHOR("Richard Zhu <[email protected]>");
 MODULE_LICENSE("GPL");
-MODULE_ALIAS("ahci:imx");
+MODULE_ALIAS("platform:" DRV_NAME);

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

commit 1e41e693f458eef2d5728207dbd327cd3b16580a upstream.

UBSAN complains about array-index-out-of-bounds:

[ 1.980703] kernel: UBSAN: array-index-out-of-bounds in /build/linux-9H675w/linux-5.15.0/drivers/ata/libahci.c:968:41
[ 1.980709] kernel: index 15 is out of range for type 'ahci_em_priv [8]'
[ 1.980713] kernel: CPU: 0 PID: 209 Comm: scsi_eh_8 Not tainted 5.15.0-25-generic #25-Ubuntu
[ 1.980716] kernel: Hardware name: System manufacturer System Product Name/P5Q3, BIOS 1102 06/11/2010
[ 1.980718] kernel: Call Trace:
[ 1.980721] kernel: <TASK>
[ 1.980723] kernel: show_stack+0x52/0x58
[ 1.980729] kernel: dump_stack_lvl+0x4a/0x5f
[ 1.980734] kernel: dump_stack+0x10/0x12
[ 1.980736] kernel: ubsan_epilogue+0x9/0x45
[ 1.980739] kernel: __ubsan_handle_out_of_bounds.cold+0x44/0x49
[ 1.980742] kernel: ahci_qc_issue+0x166/0x170 [libahci]
[ 1.980748] kernel: ata_qc_issue+0x135/0x240
[ 1.980752] kernel: ata_exec_internal_sg+0x2c4/0x580
[ 1.980754] kernel: ? vprintk_default+0x1d/0x20
[ 1.980759] kernel: ata_exec_internal+0x67/0xa0
[ 1.980762] kernel: sata_pmp_read+0x8d/0xc0
[ 1.980765] kernel: sata_pmp_read_gscr+0x3c/0x90
[ 1.980768] kernel: sata_pmp_attach+0x8b/0x310
[ 1.980771] kernel: ata_eh_revalidate_and_attach+0x28c/0x4b0
[ 1.980775] kernel: ata_eh_recover+0x6b6/0xb30
[ 1.980778] kernel: ? ahci_do_hardreset+0x180/0x180 [libahci]
[ 1.980783] kernel: ? ahci_stop_engine+0xb0/0xb0 [libahci]
[ 1.980787] kernel: ? ahci_do_softreset+0x290/0x290 [libahci]
[ 1.980792] kernel: ? trace_event_raw_event_ata_eh_link_autopsy_qc+0xe0/0xe0
[ 1.980795] kernel: sata_pmp_eh_recover.isra.0+0x214/0x560
[ 1.980799] kernel: sata_pmp_error_handler+0x23/0x40
[ 1.980802] kernel: ahci_error_handler+0x43/0x80 [libahci]
[ 1.980806] kernel: ata_scsi_port_error_handler+0x2b1/0x600
[ 1.980810] kernel: ata_scsi_error+0x9c/0xd0
[ 1.980813] kernel: scsi_error_handler+0xa1/0x180
[ 1.980817] kernel: ? scsi_unjam_host+0x1c0/0x1c0
[ 1.980820] kernel: kthread+0x12a/0x150
[ 1.980823] kernel: ? set_kthread_struct+0x50/0x50
[ 1.980826] kernel: ret_from_fork+0x22/0x30
[ 1.980831] kernel: </TASK>

This happens because sata_pmp_init_links() initialize link->pmp up to SATA_PMP_MAX_PORTS while em_priv is declared as 8 elements array.

I can't find the maximum Enclosure Management ports specified in AHCI spec v1.3.1, but "12.2.1 LED message type" states that "Port Multiplier Information" can utilize 4 bits, which implies it can support up to 16 ports. Hence, use SATA_PMP_MAX_PORTS as EM_MAX_SLOTS to resolve the issue.

---
 drivers/ata/ahci.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -260,7 +260,7 @@ enum {
 	ICH_MAP				= 0x90, /* ICH MAP register */
 
 	/* em constants */
-	EM_MAX_SLOTS			= 8,
+	EM_MAX_SLOTS			= SATA_PMP_MAX_PORTS,
 	EM_MAX_RETRY			= 5,
 
 	/* em_ctl bits */

From: Eric Ren <[email protected]>

commit c000a2607145d28b06c697f968491372ea56c23a upstream.

With some PCIe topologies, restoring a guest fails while parsing the ITS device tables.

Reproducer hints:
1. Create ARM virt VM with pxb-pcie bus which adds extra host bridges, with qemu command like:

```
  -device pxb-pcie,bus_nr=8,id=pci.x,numa_node=0,bus=pcie.0 \
  -device pcie-root-port,..,bus=pci.x \
  ...
  -device pxb-pcie,bus_nr=37,id=pci.y,numa_node=1,bus=pcie.0 \
  -device pcie-root-port,..,bus=pci.y \
  ...

```

2. Ensure the guest uses 2-level device table
3. Perform VM migration which calls save/restore device tables

In that setup, we get a big "offset" between 2 device_ids, which makes unsigned "len" round up a big positive number, causing the scan loop to continue with a bad GPA. For example:

1. L1 table has 2 entries;
2. and we are now scanning at L2 table entry index 2075 (pointed to by L1 first entry)
3. if next device id is 9472, we will get a big offset: 7397;
4. with unsigned 'len', 'len -= offset * esz', len will underflow to a positive number, mistakenly into next iteration with a bad GPA; (It should break out of the current L2 table scanning, and jump into the next L1 table entry)
5. that bad GPA fails the guest read.

Fix it by stopping the L2 table scan when the next device id is outside of the current table, allowing the scan to continue from the next L1 table entry.

Thanks to Eric Auger for the fix suggestion.

---
 virt/kvm/arm/vgic/vgic-its.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -1835,7 +1835,7 @@ static int scan_its_table(struct vgic_it
 
 	memset(entry, 0, esz);
 
-	while (len > 0) {
+	while (true) {
 		int next_offset;
 		size_t byte_offset;
 

@@ -1848,6 +1848,9 @@ static int scan_its_table(struct vgic_it
 			return next_offset;
 
 		byte_offset = next_offset * esz;
+		if (byte_offset >= len)
+			break;
+
 		id += next_offset;
 		gpa += byte_offset;
 		len -= byte_offset;

From: James Morse <[email protected]>

commit 44b3834b2eed595af07021b1c64e6f9bc396398b upstream.

Cortex-A57 and Cortex-A72 have an erratum where an interrupt that occurs between a pair of AES instructions in aarch32 mode may corrupt the ELR. The task will subsequently produce the wrong AES result.

The AES instructions are part of the cryptographic extensions, which are optional. User-space software will detect the support for these instructions from the hwcaps. If the platform doesn't support these instructions a software implementation should be used.

Remove the hwcap bits on affected parts to indicate user-space should not use the AES instructions.

---
 Documentation/arm64/silicon-errata.txt |    2 ++
 arch/arm64/Kconfig                     |   16 ++++++++++++++++
 arch/arm64/include/asm/cpucaps.h       |    3 ++-
 arch/arm64/kernel/cpu_errata.c         |   16 ++++++++++++++++
 arch/arm64/kernel/cpufeature.c         |   13 ++++++++++++-
 5 files changed, 48 insertions(+), 2 deletions(-)

@@ -53,7 +53,9 @@ stable kernels.
 | ARM            | Cortex-A57      | #832075         | ARM64_ERRATUM_832075        |
 | ARM            | Cortex-A57      | #852523         | N/A                         |
 | ARM            | Cortex-A57      | #834220         | ARM64_ERRATUM_834220        |
+| ARM            | Cortex-A57      | #1742098        | ARM64_ERRATUM_1742098       |
 | ARM            | Cortex-A72      | #853709         | N/A                         |
+| ARM            | Cortex-A72      | #1655431        | ARM64_ERRATUM_1742098       |
 | ARM            | Cortex-A73      | #858921         | ARM64_ERRATUM_858921        |
 | ARM            | Cortex-A55      | #1024718        | ARM64_ERRATUM_1024718       |
 | ARM            | Cortex-A76      | #1188873        | ARM64_ERRATUM_1188873       |

@@ -472,6 +472,22 @@ config ARM64_ERRATUM_1188873
 
 	  If unsure, say Y.
 
+config ARM64_ERRATUM_1742098
+	bool "Cortex-A57/A72: 1742098: ELR recorded incorrectly on interrupt taken between cryptographic instructions in a sequence"
+	depends on COMPAT
+	default y
+	help
+	  This option removes the AES hwcap for aarch32 user-space to
+	  workaround erratum 1742098 on Cortex-A57 and Cortex-A72.
+
+	  Affected parts may corrupt the AES state if an interrupt is
+	  taken between a pair of AES instructions. These instructions
+	  are only present if the cryptography extensions are present.
+	  All software should have a fallback implementation for CPUs
+	  that don't implement the cryptography extensions.
+
+	  If unsure, say Y.
+
 config CAVIUM_ERRATUM_22375
 	bool "Cavium erratum 22375, 24313"
 	default y

@@ -47,7 +47,8 @@
 #define ARM64_SSBS				27
 #define ARM64_WORKAROUND_1188873		28
 #define ARM64_SPECTRE_BHB			29
+#define ARM64_WORKAROUND_1742098		30
 
-#define ARM64_NCAPS				30
+#define ARM64_NCAPS				31
 
 #endif /* __ASM_CPUCAPS_H */

@@ -576,6 +576,14 @@ check_branch_predictor(const struct arm6
 	return (need_wa > 0);
 }
 
+#ifdef CONFIG_ARM64_ERRATUM_1742098
+static struct midr_range broken_aarch32_aes[] = {
+	MIDR_RANGE(MIDR_CORTEX_A57, 0, 1, 0xf, 0xf),
+	MIDR_ALL_VERSIONS(MIDR_CORTEX_A72),
+	{},
+};
+#endif
+
 const struct arm64_cpu_capabilities arm64_errata[] = {
 #if	defined(CONFIG_ARM64_ERRATUM_826319) || \
 	defined(CONFIG_ARM64_ERRATUM_827319) || \

@@ -741,6 +749,14 @@ const struct arm64_cpu_capabilities arm6
 		.matches = is_spectre_bhb_affected,
 		.cpu_enable = spectre_bhb_enable_mitigation,
 	},
+#ifdef CONFIG_ARM64_ERRATUM_1742098
+	{
+		.desc = "ARM erratum 1742098",
+		.capability = ARM64_WORKAROUND_1742098,
+		CAP_MIDR_RANGE_LIST(broken_aarch32_aes),
+		.type = ARM64_CPUCAP_LOCAL_CPU_ERRATUM,
+	},
+#endif
 	{
 	}
 };

@@ -30,6 +30,7 @@
 #include <asm/cpu.h>
 #include <asm/cpufeature.h>
 #include <asm/cpu_ops.h>
+#include <asm/hwcap.h>
 #include <asm/mmu_context.h>
 #include <asm/processor.h>
 #include <asm/sysreg.h>

@@ -1010,6 +1011,14 @@ static void cpu_enable_ssbs(const struct
 }
 #endif /* CONFIG_ARM64_SSBD */
 
+static void elf_hwcap_fixup(void)
+{
+#ifdef CONFIG_ARM64_ERRATUM_1742098
+	if (cpus_have_const_cap(ARM64_WORKAROUND_1742098))
+		compat_elf_hwcap2 &= ~COMPAT_HWCAP2_AES;
+#endif /* ARM64_ERRATUM_1742098 */
+}
+
 static const struct arm64_cpu_capabilities arm64_features[] = {
 	{
 		.desc = "GIC system register CPU interface",

@@ -1588,8 +1597,10 @@ void __init setup_cpu_features(void)
 	mark_const_caps_ready();
 	setup_elf_hwcaps(arm64_elf_hwcaps);
 
-	if (system_supports_32bit_el0())
+	if (system_supports_32bit_el0()) {
 		setup_elf_hwcaps(compat_elf_hwcaps);
+		elf_hwcap_fixup();
+	}
 
 	/* Advertise that we have computed the system capabilities */
 	set_sys_caps_initialised();

From: Werner Sembach <[email protected]>

commit 3dbc80a3e4c55c4a5b89ef207bed7b7de36157b4 upstream.

This commit is very different from the upstream commit! It fixes the same issue by adding more quirks, rather then the general fix from the 6.1 kernel, because the general fix from the 6.1 kernel is part of a larger refactoring of the backlight code which is not suitable for the stable series.

As described in "ACPI: video: Drop NL5x?U, PF4NU1F and PF5?U?? acpi_backlight=native quirks" (10212754a0d2) the upstream commit "ACPI:
video: Make backlight class device registration a separate step (v2)" (3dbc80a3e4c5) makes these quirks unnecessary. However as mentioned in this bugtracker ticket https://bugzilla.kernel.org/show_bug.cgi?id=215683#c17
the upstream fix is part of a larger patchset that is overall too complex for stable.

The TongFang GKxNRxx, GMxNGxx, GMxZGxx, and GMxRGxx / TUXEDO Stellaris/Polaris Gen 1-4, have the same problem as the Clevo NL5xRU and NL5xNU / TUXEDO Aura 15 Gen1 and Gen2:
They have a working native and video interface for screen backlight. However the default detection mechanism first registers the video interface before unregistering it again and switching to the native interface during boot. This results in a dangling SBIOS request for backlight change for some reason, causing the backlight to switch to ~2% once per boot on the first power cord connect or disconnect event. Setting the native interface explicitly circumvents this buggy behaviour by avoiding the unregistering process.

---
 drivers/acpi/video_detect.c |   64 ++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 64 insertions(+)

@@ -448,6 +448,70 @@ static const struct dmi_system_id video_
 		},
 	},
 	/*
+	 * More Tongfang devices with the same issue as the Clevo NL5xRU and
+	 * NL5xNU/TUXEDO Aura 15 Gen1 and Gen2. See the description above.
+	 */
+	{
+	.callback = video_detect_force_native,
+	.ident = "TongFang GKxNRxx",
+	.matches = {
+		DMI_MATCH(DMI_BOARD_NAME, "GKxNRxx"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "TongFang GKxNRxx",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "TUXEDO"),
+		DMI_MATCH(DMI_BOARD_NAME, "POLARIS1501A1650TI"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "TongFang GKxNRxx",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "TUXEDO"),
+		DMI_MATCH(DMI_BOARD_NAME, "POLARIS1501A2060"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "TongFang GKxNRxx",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "TUXEDO"),
+		DMI_MATCH(DMI_BOARD_NAME, "POLARIS1701A1650TI"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "TongFang GKxNRxx",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "TUXEDO"),
+		DMI_MATCH(DMI_BOARD_NAME, "POLARIS1701A2060"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "TongFang GMxNGxx",
+	.matches = {
+		DMI_MATCH(DMI_BOARD_NAME, "GMxNGxx"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "TongFang GMxZGxx",
+	.matches = {
+		DMI_MATCH(DMI_BOARD_NAME, "GMxZGxx"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "TongFang GMxRGxx",
+	.matches = {
+		DMI_MATCH(DMI_BOARD_NAME, "GMxRGxx"),
+		},
+	},
+	/*
 	 * Desktops which falsely report a backlight and which our heuristics
 	 * for this do not catch.
 	 */

From: Jean-Francois Le Fillatre <[email protected]>

commit 1bd3a383075c64d638e65d263c9267b08ee7733c upstream.

The Lenovo OneLink+ Dock contains an RTL8153 controller that behaves as a broken CDC device by default. Add the custom Lenovo PID to the r8152 driver to support it properly.

Also, systems compatible with this dock provide a BIOS option to enable MAC address passthrough (as per Lenovo document "ThinkPad Docking Solutions 2017"). Add the custom PID to the MAC passthrough list too.

Tested on a ThinkPad 13 1st gen with the expected results:

passthrough disabled: Invalid header when reading pass-thru MAC addr
passthrough enabled:  Using pass-thru MAC addr XX:XX:XX:XX:XX:XX

---
 drivers/net/usb/cdc_ether.c |    7 +++++++
 drivers/net/usb/r8152.c     |    1 +
 2 files changed, 8 insertions(+)

@@ -777,6 +777,13 @@ static const struct usb_device_id	produc
 },
 #endif
 
+/* Lenovo ThinkPad OneLink+ Dock (based on Realtek RTL8153) */
+{
+	USB_DEVICE_AND_INTERFACE_INFO(LENOVO_VENDOR_ID, 0x3054, USB_CLASS_COMM,
+			USB_CDC_SUBCLASS_ETHERNET, USB_CDC_PROTO_NONE),
+	.driver_info = 0,
+},
+
 /* ThinkPad USB-C Dock (based on Realtek RTL8153) */
 {
 	USB_DEVICE_AND_INTERFACE_INFO(LENOVO_VENDOR_ID, 0x3062, USB_CLASS_COMM,

@@ -5312,6 +5312,7 @@ static const struct usb_device_id rtl815
 	{REALTEK_USB_DEVICE(VENDOR_ID_MICROSOFT, 0x0927)},
 	{REALTEK_USB_DEVICE(VENDOR_ID_SAMSUNG, 0xa101)},
 	{REALTEK_USB_DEVICE(VENDOR_ID_LENOVO,  0x304f)},
+	{REALTEK_USB_DEVICE(VENDOR_ID_LENOVO,  0x3054)},
 	{REALTEK_USB_DEVICE(VENDOR_ID_LENOVO,  0x3062)},
 	{REALTEK_USB_DEVICE(VENDOR_ID_LENOVO,  0x3069)},
 	{REALTEK_USB_DEVICE(VENDOR_ID_LENOVO,  0x7205)},

From: Filipe Manana <[email protected]>

[ Upstream commit 4fc7b57228243d09c0d878873bf24fa64a90fa01 ]

When processing delayed data references during backref walking and we are using a share context (we are being called through fiemap), whenever we find a delayed data reference for an inode different from the one we are interested in, then we immediately exit and consider the data extent as shared. This is wrong, because:

1) This might be a DROP reference that will cancel out a reference in the extent tree;

2) Even if it's an ADD reference, it may be followed by a DROP reference that cancels it out.

In either case we should not exit immediately.

Fix this by never exiting when we find a delayed data reference for another inode - instead add the reference and if it does not cancel out other delayed reference, we will exit early when we call extent_is_shared() after processing all delayed references. If we find a drop reference, then signal the code that processes references from the extent tree (add_inline_refs() and add_keyed_refs()) to not exit immediately if it finds there a reference for another inode, since we have delayed drop references that may cancel it out. In this later case we exit once we don't have references in the rb trees that cancel out each other and have two references for different inodes.

Example reproducer for case 1):

   $ cat test-1.sh
   #!/bin/bash

DEV=/dev/sdj
MNT=/mnt/sdj

mkfs.btrfs -f $DEV
mount $DEV $MNT

   xfs_io -f -c "pwrite 0 64K" $MNT/foo
   cp --reflink=always $MNT/foo $MNT/bar

echo
echo "fiemap after cloning:"
xfs_io -c "fiemap -v" $MNT/foo

rm -f $MNT/bar
echo
echo "fiemap after removing file bar:"
xfs_io -c "fiemap -v" $MNT/foo

umount $MNT

Running it before this patch, the extent is still listed as shared, it has the flag 0x2000 (FIEMAP_EXTENT_SHARED) set:

   $ ./test-1.sh
   fiemap after cloning:
   /mnt/sdj/foo:
    EXT: FILE-OFFSET      BLOCK-RANGE      TOTAL FLAGS
      0: [0..127]:        26624..26751       128 0x2001

fiemap after removing file bar:
/mnt/sdj/foo:
EXT: FILE-OFFSET BLOCK-RANGE TOTAL FLAGS
0: [0..127]: 26624..26751 128 0x2001

Example reproducer for case 2):

   $ cat test-2.sh
   #!/bin/bash

DEV=/dev/sdj
MNT=/mnt/sdj

mkfs.btrfs -f $DEV
mount $DEV $MNT

   xfs_io -f -c "pwrite 0 64K" $MNT/foo
   cp --reflink=always $MNT/foo $MNT/bar

# Flush delayed references to the extent tree and commit current

   # transaction.
   sync

echo
echo "fiemap after cloning:"
xfs_io -c "fiemap -v" $MNT/foo

rm -f $MNT/bar
echo
echo "fiemap after removing file bar:"
xfs_io -c "fiemap -v" $MNT/foo

umount $MNT

Running it before this patch, the extent is still listed as shared, it has the flag 0x2000 (FIEMAP_EXTENT_SHARED) set:

   $ ./test-2.sh
   fiemap after cloning:
   /mnt/sdj/foo:
    EXT: FILE-OFFSET      BLOCK-RANGE      TOTAL FLAGS
      0: [0..127]:        26624..26751       128 0x2001

fiemap after removing file bar:
/mnt/sdj/foo:
EXT: FILE-OFFSET BLOCK-RANGE TOTAL FLAGS
0: [0..127]: 26624..26751 128 0x2001

After this patch, after deleting bar in both tests, the extent is not reported with the 0x2000 flag anymore, it gets only the flag 0x1 (which is FIEMAP_EXTENT_LAST):

   $ ./test-1.sh
   fiemap after cloning:
   /mnt/sdj/foo:
    EXT: FILE-OFFSET      BLOCK-RANGE      TOTAL FLAGS
      0: [0..127]:        26624..26751       128 0x2001

fiemap after removing file bar:
/mnt/sdj/foo:
EXT: FILE-OFFSET BLOCK-RANGE TOTAL FLAGS
0: [0..127]: 26624..26751 128 0x1

   $ ./test-2.sh
   fiemap after cloning:
   /mnt/sdj/foo:
    EXT: FILE-OFFSET      BLOCK-RANGE      TOTAL FLAGS
      0: [0..127]:        26624..26751       128 0x2001

fiemap after removing file bar:
/mnt/sdj/foo:
EXT: FILE-OFFSET BLOCK-RANGE TOTAL FLAGS
0: [0..127]: 26624..26751 128 0x1

These tests will later be converted to a test case for fstests.

---
 fs/btrfs/backref.c | 33 ++++++++++++++++++++++++---------
 1 file changed, 24 insertions(+), 9 deletions(-)

index 58dc96d7ecaf..93cfbdada40f 100644
@@ -146,6 +146,7 @@ struct share_check {
 	u64 root_objectid;
 	u64 inum;
 	int share_count;
+	bool have_delayed_delete_refs;
 };
 
 static inline int extent_is_shared(struct share_check *sc)

@@ -832,13 +833,22 @@ static int add_delayed_refs(const struct btrfs_fs_info *fs_info,
 			key.offset = ref->offset;
 
 			/*
-			 * Found a inum that doesn't match our known inum, we
-			 * know it's shared.
+			 * If we have a share check context and a reference for
+			 * another inode, we can't exit immediately. This is
+			 * because even if this is a BTRFS_ADD_DELAYED_REF
+			 * reference we may find next a BTRFS_DROP_DELAYED_REF
+			 * which cancels out this ADD reference.
+			 *
+			 * If this is a DROP reference and there was no previous
+			 * ADD reference, then we need to signal that when we
+			 * process references from the extent tree (through
+			 * add_inline_refs() and add_keyed_refs()), we should
+			 * not exit early if we find a reference for another
+			 * inode, because one of the delayed DROP references
+			 * may cancel that reference in the extent tree.
 			 */
-			if (sc && sc->inum && ref->objectid != sc->inum) {
-				ret = BACKREF_FOUND_SHARED;
-				goto out;
-			}
+			if (sc && count < 0)
+				sc->have_delayed_delete_refs = true;
 
 			ret = add_indirect_ref(fs_info, preftrees, ref->root,
 					       &key, 0, node->bytenr, count, sc,

@@ -868,7 +878,7 @@ static int add_delayed_refs(const struct btrfs_fs_info *fs_info,
 	}
 	if (!ret)
 		ret = extent_is_shared(sc);
-out:
+
 	spin_unlock(&head->lock);
 	return ret;
 }

@@ -972,7 +982,8 @@ static int add_inline_refs(const struct btrfs_fs_info *fs_info,
 			key.type = BTRFS_EXTENT_DATA_KEY;
 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
 
-			if (sc && sc->inum && key.objectid != sc->inum) {
+			if (sc && sc->inum && key.objectid != sc->inum &&
+			    !sc->have_delayed_delete_refs) {
 				ret = BACKREF_FOUND_SHARED;
 				break;
 			}

@@ -982,6 +993,7 @@ static int add_inline_refs(const struct btrfs_fs_info *fs_info,
 			ret = add_indirect_ref(fs_info, preftrees, root,
 					       &key, 0, bytenr, count,
 					       sc, GFP_NOFS);
+
 			break;
 		}
 		default:

@@ -1071,7 +1083,8 @@ static int add_keyed_refs(struct btrfs_fs_info *fs_info,
 			key.type = BTRFS_EXTENT_DATA_KEY;
 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
 
-			if (sc && sc->inum && key.objectid != sc->inum) {
+			if (sc && sc->inum && key.objectid != sc->inum &&
+			    !sc->have_delayed_delete_refs) {
 				ret = BACKREF_FOUND_SHARED;
 				break;
 			}

@@ -1490,6 +1503,7 @@ int btrfs_check_shared(struct btrfs_root *root, u64 inum, u64 bytenr)
 		.root_objectid = root->objectid,
 		.inum = inum,
 		.share_count = 0,
+		.have_delayed_delete_refs = false,
 	};
 
 	tmp = ulist_alloc(GFP_NOFS);

@@ -1528,6 +1542,7 @@ int btrfs_check_shared(struct btrfs_root *root, u64 inum, u64 bytenr)
 			break;
 		bytenr = node->val;
 		shared.share_count = 0;
+		shared.have_delayed_delete_refs = false;
 		cond_resched();
 	}
 

2.35.1

From: Tony Luck <[email protected]>

[ Upstream commit f6ec01da40e4139b41179f046044ee7c4f6370dc ]

If there is no user space consumer of extlog_mem trace records, then Linux properly handles multiple error records in an ELOG block

	extlog_print()
	  print_extlog_rcd()
	    __print_extlog_rcd()
	      cper_estatus_print()
		apei_estatus_for_each_section()

But the other code path hard codes looking for a single record to output a trace record.

Fix by using the same apei_estatus_for_each_section() iterator to step over all records.

---
 drivers/acpi/acpi_extlog.c | 33 ++++++++++++++++++++-------------
 1 file changed, 20 insertions(+), 13 deletions(-)

index 943b1dc2d0b3..e05309bc41cc 100644
@@ -13,6 +13,7 @@
 #include <linux/ratelimit.h>
 #include <linux/edac.h>
 #include <linux/ras.h>
+#include <acpi/ghes.h>
 #include <asm/cpu.h>
 #include <asm/mce.h>
 

@@ -141,8 +142,8 @@ static int extlog_print(struct notifier_block *nb, unsigned long val,
 	int	cpu = mce->extcpu;
 	struct acpi_hest_generic_status *estatus, *tmp;
 	struct acpi_hest_generic_data *gdata;
-	const guid_t *fru_id = &guid_null;
-	char *fru_text = "";
+	const guid_t *fru_id;
+	char *fru_text;
 	guid_t *sec_type;
 	static u32 err_seq;
 

@@ -163,17 +164,23 @@ static int extlog_print(struct notifier_block *nb, unsigned long val,
 
 	/* log event via trace */
 	err_seq++;
-	gdata = (struct acpi_hest_generic_data *)(tmp + 1);
-	if (gdata->validation_bits & CPER_SEC_VALID_FRU_ID)
-		fru_id = (guid_t *)gdata->fru_id;
-	if (gdata->validation_bits & CPER_SEC_VALID_FRU_TEXT)
-		fru_text = gdata->fru_text;
-	sec_type = (guid_t *)gdata->section_type;
-	if (guid_equal(sec_type, &CPER_SEC_PLATFORM_MEM)) {
-		struct cper_sec_mem_err *mem = (void *)(gdata + 1);
-		if (gdata->error_data_length >= sizeof(*mem))
-			trace_extlog_mem_event(mem, err_seq, fru_id, fru_text,
-					       (u8)gdata->error_severity);
+	apei_estatus_for_each_section(tmp, gdata) {
+		if (gdata->validation_bits & CPER_SEC_VALID_FRU_ID)
+			fru_id = (guid_t *)gdata->fru_id;
+		else
+			fru_id = &guid_null;
+		if (gdata->validation_bits & CPER_SEC_VALID_FRU_TEXT)
+			fru_text = gdata->fru_text;
+		else
+			fru_text = "";
+		sec_type = (guid_t *)gdata->section_type;
+		if (guid_equal(sec_type, &CPER_SEC_PLATFORM_MEM)) {
+			struct cper_sec_mem_err *mem = (void *)(gdata + 1);
+
+			if (gdata->error_data_length >= sizeof(*mem))
+				trace_extlog_mem_event(mem, err_seq, fru_id, fru_text,
+						       (u8)gdata->error_severity);
+		}
 	}
 
 out:

2.35.1

From: José Expósito <[email protected]>

[ Upstream commit bb5f0c855dcfc893ae5ed90e4c646bde9e4498bf ]

Under certain conditions the Magic Trackpad can group 2 reports in a single packet. The packet is split and the raw event function is invoked recursively for each part.

However, after processing each part, the BTN_MOUSE status is updated, sending multiple click events. [1]

Return after processing double reports to avoid this issue.

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

index 8c993f95e3ba..396a3c720b51 100644
@@ -342,7 +342,7 @@ static int magicmouse_raw_event(struct hid_device *hdev,
 		magicmouse_raw_event(hdev, report, data + 2, data[1]);
 		magicmouse_raw_event(hdev, report, data + 2 + data[1],
 			size - 2 - data[1]);
-		break;
+		return 0;
 	default:
 		return 0;
 	}

2.35.1

From: Xiaobo Liu <[email protected]>

[ Upstream commit d8bde3bf7f82dac5fc68a62c2816793a12cafa2a ]

Then the input contains '\0' or '\n', proc_mpc_write has read them, so the return value needs +1.

---
 net/atm/mpoa_proc.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index 8a0c17e1c203..4d5f8690e914 100644
@@ -220,11 +220,12 @@ static ssize_t proc_mpc_write(struct file *file, const char __user *buff,
 	if (!page)
 		return -ENOMEM;
 
-	for (p = page, len = 0; len < nbytes; p++, len++) {
+	for (p = page, len = 0; len < nbytes; p++) {
 		if (get_user(*p, buff++)) {
 			free_page((unsigned long)page);
 			return -EFAULT;
 		}
+		len += 1;
 		if (*p == '\0' || *p == '\n')
 			break;
 	}

2.35.1

From: Yang Yingliang <[email protected]>

[ Upstream commit ff2f5ec5d009844ec28f171123f9e58750cef4bf ]

Inject fault while probing module, if device_register() fails, but the refcount of kobject is not decreased to 0, the name allocated in dev_set_name() is leaked. Fix this by calling put_device(), so that name can be freed in callback function kobject_cleanup().

unreferenced object 0xffff00c01aba2100 (size 128):
comm "systemd-udevd", pid 1259, jiffies 4294903284 (age 294.152s) hex dump (first 32 bytes):

    68 6e 61 65 30 00 00 00 18 21 ba 1a c0 00 ff ff  hnae0....!......
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<0000000034783f26>] slab_post_alloc_hook+0xa0/0x3e0
    [<00000000748188f2>] __kmem_cache_alloc_node+0x164/0x2b0
    [<00000000ab0743e8>] __kmalloc_node_track_caller+0x6c/0x390
    [<000000006c0ffb13>] kvasprintf+0x8c/0x118
    [<00000000fa27bfe1>] kvasprintf_const+0x60/0xc8
    [<0000000083e10ed7>] kobject_set_name_vargs+0x3c/0xc0
    [<000000000b87affc>] dev_set_name+0x7c/0xa0
    [<000000003fd8fe26>] hnae_ae_register+0xcc/0x190 [hnae]
    [<00000000fe97edc9>] hns_dsaf_ae_init+0x9c/0x108 [hns_dsaf]
    [<00000000c36ff1eb>] hns_dsaf_probe+0x548/0x748 [hns_dsaf]

---
 drivers/net/ethernet/hisilicon/hns/hnae.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

index c7fa97a7e1f4..b591b05b956b 100644
@@ -424,8 +424,10 @@ int hnae_ae_register(struct hnae_ae_dev *hdev, struct module *owner)
 	hdev->cls_dev.release = hnae_release;
 	(void)dev_set_name(&hdev->cls_dev, "hnae%d", hdev->id);
 	ret = device_register(&hdev->cls_dev);
-	if (ret)
+	if (ret) {
+		put_device(&hdev->cls_dev);
 		return ret;
+	}
 
 	__module_get(THIS_MODULE);
 

2.35.1

From: Jerry Snitselaar <[email protected]>

[ Upstream commit 620bf9f981365c18cc2766c53d92bf8131c63f32 ]

A splat from kmem_cache_destroy() was seen with a kernel prior to commit ee2653bbe89d ("iommu/vt-d: Remove domain and devinfo mempool") when there was a failure in init_dmars(), because the iommu_domain cache still had objects. While the mempool code is now gone, there still is a leak of the si_domain memory if init_dmars() fails. So clean up si_domain in the init_dmars() error path.

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

index 74bfd7d29338..a04d4664edb4 100644
@@ -2785,6 +2785,7 @@ static int __init si_domain_init(int hw)
 
 	if (md_domain_init(si_domain, DEFAULT_DOMAIN_ADDRESS_WIDTH)) {
 		domain_exit(si_domain);
+		si_domain = NULL;
 		return -EFAULT;
 	}
 

@@ -3475,6 +3476,10 @@ static int __init init_dmars(void)
 		disable_dmar_iommu(iommu);
 		free_dmar_iommu(iommu);
 	}
+	if (si_domain) {
+		domain_exit(si_domain);
+		si_domain = NULL;
+	}
 
 	kfree(g_iommus);
 

2.35.1

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

commit 8310ca94075e784bbb06593cd6c068ee6b6e4ca6 upstream.

DST_QUEUE_OFF_BASE is applied to offset/mem_offset on MMAP capture buffers only for the VIDIOC_QUERYBUF ioctl, while the userspace fields (including offset/mem_offset) are filled in for VIDIOC_{QUERY,PREPARE,Q,DQ}BUF ioctls. This leads to differences in the values presented to userspace. If userspace attempts to mmap the capture buffer directly using values from DQBUF, it will fail.

Move the code that applies the magic offset into a helper, and call that helper from all four ioctl entry points.

[hverkuil: drop unnecessary '= 0' in v4l2_m2m_querybuf() for ret]

---
 drivers/media/v4l2-core/v4l2-mem2mem.c |   62 +++++++++++++++++++++++----------
 1 file changed, 45 insertions(+), 17 deletions(-)

@@ -358,19 +358,14 @@ int v4l2_m2m_reqbufs(struct file *file,
 }
 EXPORT_SYMBOL_GPL(v4l2_m2m_reqbufs);
 
-int v4l2_m2m_querybuf(struct file *file, struct v4l2_m2m_ctx *m2m_ctx,
-		      struct v4l2_buffer *buf)
+static void v4l2_m2m_adjust_mem_offset(struct vb2_queue *vq,
+				       struct v4l2_buffer *buf)
 {
-	struct vb2_queue *vq;
-	int ret = 0;
-	unsigned int i;
-
-	vq = v4l2_m2m_get_vq(m2m_ctx, buf->type);
-	ret = vb2_querybuf(vq, buf);
-
 	/* Adjust MMAP memory offsets for the CAPTURE queue */
 	if (buf->memory == V4L2_MEMORY_MMAP && !V4L2_TYPE_IS_OUTPUT(vq->type)) {
 		if (V4L2_TYPE_IS_MULTIPLANAR(vq->type)) {
+			unsigned int i;
+
 			for (i = 0; i < buf->length; ++i)
 				buf->m.planes[i].m.mem_offset
 					+= DST_QUEUE_OFF_BASE;

@@ -378,8 +373,23 @@ int v4l2_m2m_querybuf(struct file *file,
 			buf->m.offset += DST_QUEUE_OFF_BASE;
 		}
 	}
+}
+
+int v4l2_m2m_querybuf(struct file *file, struct v4l2_m2m_ctx *m2m_ctx,
+		      struct v4l2_buffer *buf)
+{
+	struct vb2_queue *vq;
+	int ret;
 
-	return ret;
+	vq = v4l2_m2m_get_vq(m2m_ctx, buf->type);
+	ret = vb2_querybuf(vq, buf);
+	if (ret)
+		return ret;
+
+	/* Adjust MMAP memory offsets for the CAPTURE queue */
+	v4l2_m2m_adjust_mem_offset(vq, buf);
+
+	return 0;
 }
 EXPORT_SYMBOL_GPL(v4l2_m2m_querybuf);
 

@@ -391,10 +401,15 @@ int v4l2_m2m_qbuf(struct file *file, str
 
 	vq = v4l2_m2m_get_vq(m2m_ctx, buf->type);
 	ret = vb2_qbuf(vq, buf);
-	if (!ret)
-		v4l2_m2m_try_schedule(m2m_ctx);
+	if (ret)
+		return ret;
+
+	/* Adjust MMAP memory offsets for the CAPTURE queue */
+	v4l2_m2m_adjust_mem_offset(vq, buf);
+
+	v4l2_m2m_try_schedule(m2m_ctx);
 
-	return ret;
+	return 0;
 }
 EXPORT_SYMBOL_GPL(v4l2_m2m_qbuf);
 

@@ -402,9 +417,17 @@ int v4l2_m2m_dqbuf(struct file *file, st
 		   struct v4l2_buffer *buf)
 {
 	struct vb2_queue *vq;
+	int ret;
 
 	vq = v4l2_m2m_get_vq(m2m_ctx, buf->type);
-	return vb2_dqbuf(vq, buf, file->f_flags & O_NONBLOCK);
+	ret = vb2_dqbuf(vq, buf, file->f_flags & O_NONBLOCK);
+	if (ret)
+		return ret;
+
+	/* Adjust MMAP memory offsets for the CAPTURE queue */
+	v4l2_m2m_adjust_mem_offset(vq, buf);
+
+	return 0;
 }
 EXPORT_SYMBOL_GPL(v4l2_m2m_dqbuf);
 

@@ -416,10 +439,15 @@ int v4l2_m2m_prepare_buf(struct file *fi
 
 	vq = v4l2_m2m_get_vq(m2m_ctx, buf->type);
 	ret = vb2_prepare_buf(vq, buf);
-	if (!ret)
-		v4l2_m2m_try_schedule(m2m_ctx);
+	if (ret)
+		return ret;
+
+	/* Adjust MMAP memory offsets for the CAPTURE queue */
+	v4l2_m2m_adjust_mem_offset(vq, buf);
+
+	v4l2_m2m_try_schedule(m2m_ctx);
 
-	return ret;
+	return 0;
 }
 EXPORT_SYMBOL_GPL(v4l2_m2m_prepare_buf);
 

From: Steven Rostedt (Google) <[email protected]>

commit f0a868788fcbf63cdab51f5adcf73b271ede8164 upstream.

The current code for freeing the emux timer is extremely dangerous:

CPU0 CPU1
---- ----
snd_emux_timer_callback()

			    snd_emux_free()
			      spin_lock(&emu->voice_lock)
			      del_timer(&emu->tlist); <-- returns immediately
			      spin_unlock(&emu->voice_lock);
			      [..]
			      kfree(emu);

  spin_lock(&emu->voice_lock);

[BOOM!]

Instead just use del_timer_sync() which will wait for the timer to finish before continuing. No need to check if the timer is active or not when doing so.

This doesn't fix the race of a possible re-arming of the timer, but at least it won't use the data that has just been freed.

[ Fixed unused variable warning by tiwai ]

---
 sound/synth/emux/emux.c |    7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

@@ -138,15 +138,10 @@ EXPORT_SYMBOL(snd_emux_register);
  */
 int snd_emux_free(struct snd_emux *emu)
 {
-	unsigned long flags;
-
 	if (! emu)
 		return -EINVAL;
 
-	spin_lock_irqsave(&emu->voice_lock, flags);
-	if (emu->timer_active)
-		del_timer(&emu->tlist);
-	spin_unlock_irqrestore(&emu->voice_lock, flags);
+	del_timer_sync(&emu->tlist);
 
 	snd_emux_proc_free(emu);
 	snd_emux_delete_virmidi(emu);

From: Jason A. Donenfeld <[email protected]>

commit ee03c0f200eb0d9f22dd8732d9fb7956d91019c2 upstream.

With char becoming unsigned by default, and with char alone being ambiguous and based on architecture, signed chars need to be marked explicitly as such. This fixes warnings like:

sound/pci/au88x0/au88x0_core.c:2029 vortex_adb_checkinout() warn: signedness bug returning '(-22)' sound/pci/au88x0/au88x0_core.c:2046 vortex_adb_checkinout() warn: signedness bug returning '(-12)' sound/pci/au88x0/au88x0_core.c:2125 vortex_adb_allocroute() warn: 'vortex_adb_checkinout(vortex, (0), en, 0)' is unsigned sound/pci/au88x0/au88x0_core.c:2170 vortex_adb_allocroute() warn: 'vortex_adb_checkinout(vortex, stream->resources, en, 4)' is unsigned

As well, since one function returns errnos, return an int rather than a signed char.

---
 sound/pci/au88x0/au88x0.h      |    6 +++---
 sound/pci/au88x0/au88x0_core.c |    2 +-
 2 files changed, 4 insertions(+), 4 deletions(-)

@@ -153,7 +153,7 @@ struct snd_vortex {
 #ifndef CHIP_AU8810
 	stream_t dma_wt[NR_WT];
 	wt_voice_t wt_voice[NR_WT];	/* WT register cache. */
-	char mixwt[(NR_WT / NR_WTPB) * 6];	/* WT mixin objects */
+	s8 mixwt[(NR_WT / NR_WTPB) * 6];	/* WT mixin objects */
 #endif
 
 	/* Global resources */

@@ -247,8 +247,8 @@ static int vortex_alsafmt_aspfmt(int als
 static void vortex_connect_default(vortex_t * vortex, int en);
 static int vortex_adb_allocroute(vortex_t * vortex, int dma, int nr_ch,
 				 int dir, int type, int subdev);
-static char vortex_adb_checkinout(vortex_t * vortex, int resmap[], int out,
-				  int restype);
+static int vortex_adb_checkinout(vortex_t * vortex, int resmap[], int out,
+				 int restype);
 #ifndef CHIP_AU8810
 static int vortex_wt_allocroute(vortex_t * vortex, int dma, int nr_ch);
 static void vortex_wt_connect(vortex_t * vortex, int en);

@@ -2004,7 +2004,7 @@ static int resnum[VORTEX_RESOURCE_LAST]
  out: Mean checkout if != 0. Else mean Checkin resource.
  restype: Indicates type of resource to be checked in or out.
 */
-static char
+static int
 vortex_adb_checkinout(vortex_t * vortex, int resmap[], int out, int restype)
 {
 	int i, qty = resnum[restype], resinuse = 0;

From: Hannu Hartikainen <[email protected]>

commit fc4ade55c617dc73c7e9756b57f3230b4ff24540 upstream.

NVIDIA Jetson devices in Force Recovery mode (RCM) do not support suspending, ie. flashing fails if the device has been suspended. The devices are still visible in lsusb and seem to work otherwise, making the issue hard to debug. This has been discovered in various forum posts, eg. [1].

The patch has been tested on NVIDIA Jetson AGX Xavier, but I'm adding all the Jetson models listed in [2] on the assumption that they all behave similarly.

[1]: https://forums.developer.nvidia.com/t/flashing-not-working/72365
[2]: https://docs.nvidia.com/jetson/archives/l4t-archived/l4t-3271/index.html#page/Tegra%20Linux%20Driver%20Package%20Development%20Guide/quick_start.html

---
 drivers/usb/core/quirks.c |    9 +++++++++
 1 file changed, 9 insertions(+)

@@ -227,6 +227,15 @@ static const struct usb_device_id usb_qu
 	/* Kingston DataTraveler 3.0 */
 	{ USB_DEVICE(0x0951, 0x1666), .driver_info = USB_QUIRK_NO_LPM },
 
+	/* NVIDIA Jetson devices in Force Recovery mode */
+	{ USB_DEVICE(0x0955, 0x7018), .driver_info = USB_QUIRK_RESET_RESUME },
+	{ USB_DEVICE(0x0955, 0x7019), .driver_info = USB_QUIRK_RESET_RESUME },
+	{ USB_DEVICE(0x0955, 0x7418), .driver_info = USB_QUIRK_RESET_RESUME },
+	{ USB_DEVICE(0x0955, 0x7721), .driver_info = USB_QUIRK_RESET_RESUME },
+	{ USB_DEVICE(0x0955, 0x7c18), .driver_info = USB_QUIRK_RESET_RESUME },
+	{ USB_DEVICE(0x0955, 0x7e19), .driver_info = USB_QUIRK_RESET_RESUME },
+	{ USB_DEVICE(0x0955, 0x7f21), .driver_info = USB_QUIRK_RESET_RESUME },
+
 	/* X-Rite/Gretag-Macbeth Eye-One Pro display colorimeter */
 	{ USB_DEVICE(0x0971, 0x2000), .driver_info = USB_QUIRK_NO_SET_INTF },
 

From: Thinh Nguyen <[email protected]>

commit 308c316d16cbad99bb834767382baa693ac42169 upstream.

The gadget driver may have a certain expectation of how the request completion flow should be from to its configuration. Make sure the controller driver respect that. That is, don't set IMI (Interrupt on Missed Isoc) when usb_request->no_interrupt is set. Also, the driver should only set IMI to the last TRB of a chain.

---
 drivers/usb/dwc3/gadget.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -965,8 +965,8 @@ static void __dwc3_prepare_one_trb(struc
 			trb->ctrl = DWC3_TRBCTL_ISOCHRONOUS;
 		}
 
-		/* always enable Interrupt on Missed ISOC */
-		trb->ctrl |= DWC3_TRB_CTRL_ISP_IMI;
+		if (!no_interrupt && !chain)
+			trb->ctrl |= DWC3_TRB_CTRL_ISP_IMI;
 		break;
 
 	case USB_ENDPOINT_XFER_BULK:

From: Justin Chen <[email protected]>

commit fb8f60dd1b67520e0e0d7978ef17d015690acfc1 upstream.

When port is connected and then disconnected, the state stays as configured. Which is incorrect as the port is no longer configured, but in a not attached state.

---
 drivers/usb/gadget/udc/bdc/bdc_udc.c |    1 +
 1 file changed, 1 insertion(+)

@@ -156,6 +156,7 @@ static void bdc_uspc_disconnected(struct
 	bdc->delayed_status = false;
 	bdc->reinit = reinit;
 	bdc->test_mode = false;
+	usb_gadget_set_state(&bdc->gadget, USB_STATE_NOTATTACHED);
 }
 
 /* TNotify wkaeup timer */

From: Jens Glathe <[email protected]>

commit 4f547472380136718b56064ea5689a61e135f904 upstream.

This appears to fix the error:
"xhci_hcd <address>; ERROR Transfer event TRB DMA ptr not part of current TD ep_index 2 comp_code 13" that appear spuriously (or pretty often) when using a r8152 USB3 ethernet adapter with integrated hub.

ASM1042 reports as a 0.96 controller, but appears to behave more like 1.0

Inspired by this email thread: https://markmail.org/thread/7vzqbe7t6du6qsw3

---
 drivers/usb/host/xhci-pci.c |    8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

@@ -239,8 +239,14 @@ static void xhci_pci_quirks(struct devic
 		xhci->quirks |= XHCI_BROKEN_STREAMS;
 
 	if (pdev->vendor == PCI_VENDOR_ID_ASMEDIA &&
-		pdev->device == PCI_DEVICE_ID_ASMEDIA_1042_XHCI)
+		pdev->device == PCI_DEVICE_ID_ASMEDIA_1042_XHCI) {
+		/*
+		 * try to tame the ASMedia 1042 controller which reports 0.96
+		 * but appears to behave more like 1.0
+		 */
+		xhci->quirks |= XHCI_SPURIOUS_SUCCESS;
 		xhci->quirks |= XHCI_BROKEN_STREAMS;
+	}
 	if (pdev->vendor == PCI_VENDOR_ID_ASMEDIA &&
 		pdev->device == PCI_DEVICE_ID_ASMEDIA_1042A_XHCI) {
 		xhci->quirks |= XHCI_TRUST_TX_LENGTH;

From: Mathias Nyman <[email protected]>

commit 5aed5b7c2430ce318a8e62f752f181e66f0d1053 upstream.

Endpoints are normally deleted from the bandwidth list when they are dropped, before the virt device is freed.

If xHC host is dying or being removed then the endpoints aren't dropped cleanly due to functions returning early to avoid interacting with a non-accessible host controller.

So check and delete endpoints that are still on the bandwidth list when freeing the virt device.

Solves a list_del corruption kernel crash when unbinding xhci-pci, caused by xhci_mem_cleanup() when it later tried to delete already freed endpoints from the bandwidth list.

This only affects hosts that use software bandwidth checking, which currenty is only the xHC in intel Panther Point PCH (Ivy Bridge)

---
 drivers/usb/host/xhci-mem.c |   20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)

@@ -911,15 +911,19 @@ void xhci_free_virt_device(struct xhci_h
 		if (dev->eps[i].stream_info)
 			xhci_free_stream_info(xhci,
 					dev->eps[i].stream_info);
-		/* Endpoints on the TT/root port lists should have been removed
-		 * when usb_disable_device() was called for the device.
-		 * We can't drop them anyway, because the udev might have gone
-		 * away by this point, and we can't tell what speed it was.
+		/*
+		 * Endpoints are normally deleted from the bandwidth list when
+		 * endpoints are dropped, before device is freed.
+		 * If host is dying or being removed then endpoints aren't
+		 * dropped cleanly, so delete the endpoint from list here.
+		 * Only applicable for hosts with software bandwidth checking.
 		 */
-		if (!list_empty(&dev->eps[i].bw_endpoint_list))
-			xhci_warn(xhci, "Slot %u endpoint %u "
-					"not removed from BW list!\n",
-					slot_id, i);
+
+		if (!list_empty(&dev->eps[i].bw_endpoint_list)) {
+			list_del_init(&dev->eps[i].bw_endpoint_list);
+			xhci_dbg(xhci, "Slot %u endpoint %u not removed from BW list!\n",
+				 slot_id, i);
+		}
 	}
 	/* If this is a hub, free the TT(s) from the TT list */
 	xhci_free_tt_info(xhci, dev, slot_id);

From: Matti Vaittinen <[email protected]>

commit 72b2aa38191bcba28389b0e20bf6b4f15017ff2b upstream.

The iio_utils uses a digit calculation in order to know length of the file name containing a buffer number. The digit calculation does not work for number 0.

This leads to allocation of one character too small buffer for the file-name when file name contains value '0'. (Eg. buffer0).

Fix digit calculation by returning one digit to be present for number '0'.

---
 tools/iio/iio_utils.c |    4 ++++
 1 file changed, 4 insertions(+)

@@ -546,6 +546,10 @@ static int calc_digits(int num)
 {
 	int count = 0;
 
+	/* It takes a digit to represent zero */
+	if (!num)
+		return 1;
+
 	while (num != 0) {
 		num /= 10;
 		count++;

From: Shreeya Patel <[email protected]>

commit 0dec4d2f2636b9e54d9d29f17afc7687c5407f78 upstream.

tsl2583 probe() uses devm_iio_device_register() and calling iio_device_unregister() causes the unregister to occur twice. s Switch to iio_device_register() instead of devm_iio_device_register() in probe to avoid the device managed cleanup.

---
 drivers/iio/light/tsl2583.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -867,7 +867,7 @@ static int tsl2583_probe(struct i2c_clie
 					 TSL2583_POWER_OFF_DELAY_MS);
 	pm_runtime_use_autosuspend(&clientp->dev);
 
-	ret = devm_iio_device_register(indio_dev->dev.parent, indio_dev);
+	ret = iio_device_register(indio_dev);
 	if (ret) {
 		dev_err(&clientp->dev, "%s: iio registration failed\n",
 			__func__);

From: Hyunwoo Kim <[email protected]>

commit cc67482c9e5f2c80d62f623bcc347c29f9f648e1 upstream.

Several types of UAFs can occur when physically removing a USB device.

Adds ufx_ops_destroy() function to .fb_destroy of fb_ops, and in this function, there is kref_put() that finally calls ufx_free().

This fix prevents multiple UAFs.

---
 drivers/video/fbdev/smscufx.c |   55 ++++++++++++++++++++++--------------------
 1 file changed, 30 insertions(+), 25 deletions(-)

@@ -100,7 +100,6 @@ struct ufx_data {
 	struct kref kref;
 	int fb_count;
 	bool virtualized; /* true when physical usb device not present */
-	struct delayed_work free_framebuffer_work;
 	atomic_t usb_active; /* 0 = update virtual buffer, but no usb traffic */
 	atomic_t lost_pixels; /* 1 = a render op failed. Need screen refresh */
 	u8 *edid; /* null until we read edid from hw or get from sysfs */

@@ -1120,15 +1119,24 @@ static void ufx_free(struct kref *kref)
 {
 	struct ufx_data *dev = container_of(kref, struct ufx_data, kref);
 
-	/* this function will wait for all in-flight urbs to complete */
-	if (dev->urbs.count > 0)
-		ufx_free_urb_list(dev);
+	kfree(dev);
+}
 
-	pr_debug("freeing ufx_data %p", dev);
+static void ufx_ops_destory(struct fb_info *info)
+{
+	struct ufx_data *dev = info->par;
+	int node = info->node;
 
-	kfree(dev);
+	/* Assume info structure is freed after this point */
+	framebuffer_release(info);
+
+	pr_debug("fb_info for /dev/fb%d has been freed", node);
+
+	/* release reference taken by kref_init in probe() */
+	kref_put(&dev->kref, ufx_free);
 }
 
+
 static void ufx_release_urb_work(struct work_struct *work)
 {
 	struct urb_node *unode = container_of(work, struct urb_node,

@@ -1137,14 +1145,9 @@ static void ufx_release_urb_work(struct
 	up(&unode->dev->urbs.limit_sem);
 }
 
-static void ufx_free_framebuffer_work(struct work_struct *work)
+static void ufx_free_framebuffer(struct ufx_data *dev)
 {
-	struct ufx_data *dev = container_of(work, struct ufx_data,
-					    free_framebuffer_work.work);
 	struct fb_info *info = dev->info;
-	int node = info->node;
-
-	unregister_framebuffer(info);
 
 	if (info->cmap.len != 0)
 		fb_dealloc_cmap(&info->cmap);

@@ -1156,11 +1159,6 @@ static void ufx_free_framebuffer_work(st
 
 	dev->info = NULL;
 
-	/* Assume info structure is freed after this point */
-	framebuffer_release(info);
-
-	pr_debug("fb_info for /dev/fb%d has been freed", node);
-
 	/* ref taken in probe() as part of registering framebfufer */
 	kref_put(&dev->kref, ufx_free);
 }

@@ -1172,11 +1170,13 @@ static int ufx_ops_release(struct fb_inf
 {
 	struct ufx_data *dev = info->par;
 
+	mutex_lock(&disconnect_mutex);
+
 	dev->fb_count--;
 
 	/* We can't free fb_info here - fbmem will touch it when we return */
 	if (dev->virtualized && (dev->fb_count == 0))
-		schedule_delayed_work(&dev->free_framebuffer_work, HZ);
+		ufx_free_framebuffer(dev);
 
 	if ((dev->fb_count == 0) && (info->fbdefio)) {
 		fb_deferred_io_cleanup(info);

@@ -1190,6 +1190,8 @@ static int ufx_ops_release(struct fb_inf
 
 	kref_put(&dev->kref, ufx_free);
 
+	mutex_unlock(&disconnect_mutex);
+
 	return 0;
 }
 

@@ -1296,6 +1298,7 @@ static struct fb_ops ufx_ops = {
 	.fb_blank = ufx_ops_blank,
 	.fb_check_var = ufx_ops_check_var,
 	.fb_set_par = ufx_ops_set_par,
+	.fb_destroy = ufx_ops_destory,
 };
 
 /* Assumes &info->lock held by caller

@@ -1688,9 +1691,6 @@ static int ufx_usb_probe(struct usb_inte
 		goto error;
 	}
 
-	INIT_DELAYED_WORK(&dev->free_framebuffer_work,
-			  ufx_free_framebuffer_work);
-
 	retval = ufx_reg_read(dev, 0x3000, &id_rev);
 	check_warn_goto_error(retval, "error %d reading 0x3000 register from device", retval);
 	dev_dbg(dev->gdev, "ID_REV register value 0x%08x", id_rev);

@@ -1769,10 +1769,12 @@ error:
 static void ufx_usb_disconnect(struct usb_interface *interface)
 {
 	struct ufx_data *dev;
+	struct fb_info *info;
 
 	mutex_lock(&disconnect_mutex);
 
 	dev = usb_get_intfdata(interface);
+	info = dev->info;
 
 	pr_debug("USB disconnect starting\n");
 

@@ -1786,12 +1788,15 @@ static void ufx_usb_disconnect(struct us
 
 	/* if clients still have us open, will be freed on last close */
 	if (dev->fb_count == 0)
-		schedule_delayed_work(&dev->free_framebuffer_work, 0);
+		ufx_free_framebuffer(dev);
 
-	/* release reference taken by kref_init in probe() */
-	kref_put(&dev->kref, ufx_free);
+	/* this function will wait for all in-flight urbs to complete */
+	if (dev->urbs.count > 0)
+		ufx_free_urb_list(dev);
+
+	pr_debug("freeing ufx_data %p", dev);
 
-	/* consider ufx_data freed */
+	unregister_framebuffer(info);
 
 	mutex_unlock(&disconnect_mutex);
 }

From: Miquel Raynal <[email protected]>

commit 5a5c4e06fd03b595542d5590f2bc05a6b7fc5c2b upstream.

Back in 2014, the LQI was saved in the skb control buffer (skb->cb, or mac_cb(skb)) without any actual reset of this area prior to its use.

As part of a useful rework of the use of this region, 32edc40ae65c ("ieee802154: change _cb handling slightly") introduced mac_cb_init() to basically memset the cb field to 0. In particular, this new function got called at the beginning of mac802154_parse_frame_start(), right before the location where the buffer got actually filled.

What went through unnoticed however, is the fact that the very first helper called by device drivers in the receive path already used this area to save the LQI value for later extraction. Resetting the cb field "so late" led to systematically zeroing the LQI.

If we consider the reset of the cb field needed, we can make it as soon as we get an skb from a device driver, right before storing the LQI, as is the very first time we need to write something there.

---
 net/mac802154/rx.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

@@ -140,7 +140,7 @@ static int
 ieee802154_parse_frame_start(struct sk_buff *skb, struct ieee802154_hdr *hdr)
 {
 	int hlen;
-	struct ieee802154_mac_cb *cb = mac_cb_init(skb);
+	struct ieee802154_mac_cb *cb = mac_cb(skb);
 
 	skb_reset_mac_header(skb);
 

@@ -302,8 +302,9 @@ void
 ieee802154_rx_irqsafe(struct ieee802154_hw *hw, struct sk_buff *skb, u8 lqi)
 {
 	struct ieee802154_local *local = hw_to_local(hw);
+	struct ieee802154_mac_cb *cb = mac_cb_init(skb);
 
-	mac_cb(skb)->lqi = lqi;
+	cb->lqi = lqi;
 	skb->pkt_type = IEEE802154_RX_MSG;
 	skb_queue_tail(&local->skb_queue, skb);
 	tasklet_schedule(&local->tasklet);

From: Johan Hovold <[email protected]>

commit 4c1294da6aed1f16d47a417dcfe6602833c3c95c upstream.

Add the missing sanity check on the bridge counter to avoid corrupting data beyond the fixed-sized bridge array in case there are ever more than eight bridges.

Patchwork: https://patchwork.freedesktop.org/patch/502670/

---
 drivers/gpu/drm/msm/hdmi/hdmi.c |    5 +++++
 1 file changed, 5 insertions(+)

@@ -291,6 +291,11 @@ int msm_hdmi_modeset_init(struct hdmi *h
 	struct platform_device *pdev = hdmi->pdev;
 	int ret;
 
+	if (priv->num_bridges == ARRAY_SIZE(priv->bridges)) {
+		DRM_DEV_ERROR(dev->dev, "too many bridges\n");
+		return -ENOSPC;
+	}
+
 	hdmi->dev = dev;
 	hdmi->encoder = encoder;
 

From: Matthew Ma <[email protected]>

commit 9972e6b404884adae9eec7463e30d9b3c9a70b18 upstream.

SDIO tuple is only allocated for standard SDIO card, especially it causes memory corruption issues when the non-standard SDIO card has removed, which is because the card device's reference counter does not increase for it at sdio_init_func(), but all SDIO card device reference counter gets decreased at sdio_release_func().

---
 drivers/mmc/core/sdio_bus.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -264,7 +264,8 @@ static void sdio_release_func(struct dev
 {
 	struct sdio_func *func = dev_to_sdio_func(dev);
 
-	sdio_free_func_cis(func);
+	if (!(func->card->quirks & MMC_QUIRK_NONSTD_SDIO))
+		sdio_free_func_cis(func);
 
 	kfree(func->info);
 	kfree(func->tmpbuf);

From: Christian A. Ehrhardt <[email protected]>

commit 4abc99652812a2ddf932f137515d5c5a04723538 upstream.

Syzkaller managed to trigger concurrent calls to kernfs_remove_by_name_ns() for the same file resulting in a KASAN detected use-after-free. The race occurs when the root node is freed during kernfs_drain().

To prevent this acquire an additional reference for the root of the tree that is removed before calling __kernfs_remove().

Found by syzkaller with the following reproducer (slab_nomerge is required):

syz_mount_image$ext4(0x0, &(0x7f0000000100)='./file0\x00', 0x100000, 0x0, 0x0, 0x0, 0x0)
r0 = openat(0xffffffffffffff9c, &(0x7f0000000080)='/proc/self/exe\x00', 0x0, 0x0)
close(r0)
pipe2(&(0x7f0000000140)={0xffffffffffffffff, <r1=>0xffffffffffffffff}, 0x800)
mount$9p_fd(0x0, &(0x7f0000000040)='./file0\x00', &(0x7f00000000c0), 0x408, &(0x7f0000000280)={'trans=fd,', {'rfdno', 0x3d, r0}, 0x2c, {'wfdno', 0x3d, r1}, 0x2c, {[{@cache_loose}, {@mmap}, {@loose}, {@loose}, {@mmap}], [{@mask={'mask', 0x3d, '^MAY_EXEC'}}, {@fsmagic={'fsmagic', 0x3d, 0x10001}}, {@dont_hash}]}})

Sample report:

==================================================================

BUG: KASAN: use-after-free in kernfs_type include/linux/kernfs.h:335 [inline]
BUG: KASAN: use-after-free in kernfs_leftmost_descendant fs/kernfs/dir.c:1261 [inline]
BUG: KASAN: use-after-free in __kernfs_remove.part.0+0x843/0x960 fs/kernfs/dir.c:1369
Read of size 2 at addr ffff8880088807f0 by task syz-executor.2/857

CPU: 0 PID: 857 Comm: syz-executor.2 Not tainted 6.0.0-rc3-00363-g7726d4c3e60b #5
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.15.0-1 04/01/2014
Call Trace:
 <TASK>
 __dump_stack lib/dump_stack.c:88 [inline]
 dump_stack_lvl+0x6e/0x91 lib/dump_stack.c:106
 print_address_description mm/kasan/report.c:317 [inline]
 print_report.cold+0x5e/0x5e5 mm/kasan/report.c:433
 kasan_report+0xa3/0x130 mm/kasan/report.c:495
 kernfs_type include/linux/kernfs.h:335 [inline]
 kernfs_leftmost_descendant fs/kernfs/dir.c:1261 [inline]
 __kernfs_remove.part.0+0x843/0x960 fs/kernfs/dir.c:1369
 __kernfs_remove fs/kernfs/dir.c:1356 [inline]
 kernfs_remove_by_name_ns+0x108/0x190 fs/kernfs/dir.c:1589
 sysfs_slab_add+0x133/0x1e0 mm/slub.c:5943
 __kmem_cache_create+0x3e0/0x550 mm/slub.c:4899
 create_cache mm/slab_common.c:229 [inline]
 kmem_cache_create_usercopy+0x167/0x2a0 mm/slab_common.c:335
 p9_client_create+0xd4d/0x1190 net/9p/client.c:993
 v9fs_session_init+0x1e6/0x13c0 fs/9p/v9fs.c:408
 v9fs_mount+0xb9/0xbd0 fs/9p/vfs_super.c:126
 legacy_get_tree+0xf1/0x200 fs/fs_context.c:610
 vfs_get_tree+0x85/0x2e0 fs/super.c:1530
 do_new_mount fs/namespace.c:3040 [inline]
 path_mount+0x675/0x1d00 fs/namespace.c:3370
 do_mount fs/namespace.c:3383 [inline]
 __do_sys_mount fs/namespace.c:3591 [inline]
 __se_sys_mount fs/namespace.c:3568 [inline]
 __x64_sys_mount+0x282/0x300 fs/namespace.c:3568
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x38/0x90 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x7f725f983aed
Code: 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 f3 0f 1e fa 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 b0 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f725f0f7028 EFLAGS: 00000246 ORIG_RAX: 00000000000000a5
RAX: ffffffffffffffda RBX: 00007f725faa3f80 RCX: 00007f725f983aed
RDX: 00000000200000c0 RSI: 0000000020000040 RDI: 0000000000000000
RBP: 00007f725f9f419c R08: 0000000020000280 R09: 0000000000000000
R10: 0000000000000408 R11: 0000000000000246 R12: 0000000000000000
R13: 0000000000000006 R14: 00007f725faa3f80 R15: 00007f725f0d7000
 </TASK>

Allocated by task 855:
 kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
 kasan_set_track mm/kasan/common.c:45 [inline]
 set_alloc_info mm/kasan/common.c:437 [inline]
 __kasan_slab_alloc+0x66/0x80 mm/kasan/common.c:470
 kasan_slab_alloc include/linux/kasan.h:224 [inline]
 slab_post_alloc_hook mm/slab.h:727 [inline]
 slab_alloc_node mm/slub.c:3243 [inline]
 slab_alloc mm/slub.c:3251 [inline]
 __kmem_cache_alloc_lru mm/slub.c:3258 [inline]
 kmem_cache_alloc+0xbf/0x200 mm/slub.c:3268
 kmem_cache_zalloc include/linux/slab.h:723 [inline]
 __kernfs_new_node+0xd4/0x680 fs/kernfs/dir.c:593
 kernfs_new_node fs/kernfs/dir.c:655 [inline]
 kernfs_create_dir_ns+0x9c/0x220 fs/kernfs/dir.c:1010
 sysfs_create_dir_ns+0x127/0x290 fs/sysfs/dir.c:59
 create_dir lib/kobject.c:63 [inline]
 kobject_add_internal+0x24a/0x8d0 lib/kobject.c:223
 kobject_add_varg lib/kobject.c:358 [inline]
 kobject_init_and_add+0x101/0x160 lib/kobject.c:441
 sysfs_slab_add+0x156/0x1e0 mm/slub.c:5954
 __kmem_cache_create+0x3e0/0x550 mm/slub.c:4899
 create_cache mm/slab_common.c:229 [inline]
 kmem_cache_create_usercopy+0x167/0x2a0 mm/slab_common.c:335
 p9_client_create+0xd4d/0x1190 net/9p/client.c:993
 v9fs_session_init+0x1e6/0x13c0 fs/9p/v9fs.c:408
 v9fs_mount+0xb9/0xbd0 fs/9p/vfs_super.c:126
 legacy_get_tree+0xf1/0x200 fs/fs_context.c:610
 vfs_get_tree+0x85/0x2e0 fs/super.c:1530
 do_new_mount fs/namespace.c:3040 [inline]
 path_mount+0x675/0x1d00 fs/namespace.c:3370
 do_mount fs/namespace.c:3383 [inline]
 __do_sys_mount fs/namespace.c:3591 [inline]
 __se_sys_mount fs/namespace.c:3568 [inline]
 __x64_sys_mount+0x282/0x300 fs/namespace.c:3568
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x38/0x90 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x63/0xcd

Freed by task 857:
 kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
 kasan_set_track+0x21/0x30 mm/kasan/common.c:45
 kasan_set_free_info+0x20/0x40 mm/kasan/generic.c:370
 ____kasan_slab_free mm/kasan/common.c:367 [inline]
 ____kasan_slab_free mm/kasan/common.c:329 [inline]
 __kasan_slab_free+0x108/0x190 mm/kasan/common.c:375
 kasan_slab_free include/linux/kasan.h:200 [inline]
 slab_free_hook mm/slub.c:1754 [inline]
 slab_free_freelist_hook mm/slub.c:1780 [inline]
 slab_free mm/slub.c:3534 [inline]
 kmem_cache_free+0x9c/0x340 mm/slub.c:3551
 kernfs_put.part.0+0x2b2/0x520 fs/kernfs/dir.c:547
 kernfs_put+0x42/0x50 fs/kernfs/dir.c:521
 __kernfs_remove.part.0+0x72d/0x960 fs/kernfs/dir.c:1407
 __kernfs_remove fs/kernfs/dir.c:1356 [inline]
 kernfs_remove_by_name_ns+0x108/0x190 fs/kernfs/dir.c:1589
 sysfs_slab_add+0x133/0x1e0 mm/slub.c:5943
 __kmem_cache_create+0x3e0/0x550 mm/slub.c:4899
 create_cache mm/slab_common.c:229 [inline]
 kmem_cache_create_usercopy+0x167/0x2a0 mm/slab_common.c:335
 p9_client_create+0xd4d/0x1190 net/9p/client.c:993
 v9fs_session_init+0x1e6/0x13c0 fs/9p/v9fs.c:408
 v9fs_mount+0xb9/0xbd0 fs/9p/vfs_super.c:126
 legacy_get_tree+0xf1/0x200 fs/fs_context.c:610
 vfs_get_tree+0x85/0x2e0 fs/super.c:1530
 do_new_mount fs/namespace.c:3040 [inline]
 path_mount+0x675/0x1d00 fs/namespace.c:3370
 do_mount fs/namespace.c:3383 [inline]
 __do_sys_mount fs/namespace.c:3591 [inline]
 __se_sys_mount fs/namespace.c:3568 [inline]
 __x64_sys_mount+0x282/0x300 fs/namespace.c:3568
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x38/0x90 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x63/0xcd

The buggy address belongs to the object at ffff888008880780
 which belongs to the cache kernfs_node_cache of size 128

The buggy address is located 112 bytes inside of
128-byte region [ffff888008880780, ffff888008880800)

The buggy address belongs to the physical page:
page:00000000732833f8 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x8880
flags: 0x100000000000200(slab|node=0|zone=1)
raw: 0100000000000200 0000000000000000 dead000000000122 ffff888001147280
raw: 0000000000000000 0000000000150015 00000001ffffffff 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
 ffff888008880680: fc fc fc fc fc fc fc fc fa fb fb fb fb fb fb fb
 ffff888008880700: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
>ffff888008880780: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
                                                             ^
 ffff888008880800: fc fc fc fc fc fc fc fc fa fb fb fb fb fb fb fb
 ffff888008880880: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
==================================================================

---
 fs/kernfs/dir.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -1490,8 +1490,11 @@ int kernfs_remove_by_name_ns(struct kern
 	mutex_lock(&kernfs_mutex);
 
 	kn = kernfs_find_ns(parent, name, ns);
-	if (kn)
+	if (kn) {
+		kernfs_get(kn);
 		__kernfs_remove(kn);
+		kernfs_put(kn);
+	}
 
 	mutex_unlock(&kernfs_mutex);
 

From: Heiko Carstens <[email protected]>

commit a262d3ad6a433e4080cecd0a8841104a5906355e upstream.

For some exception types the instruction address points behind the instruction that caused the exception. Take that into account and add the missing exception table entry.

---
 arch/s390/include/asm/futex.h |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -16,7 +16,8 @@
 		"3: jl    1b\n"						\
 		"   lhi   %0,0\n"					\
 		"4: sacf  768\n"					\
-		EX_TABLE(0b,4b) EX_TABLE(2b,4b) EX_TABLE(3b,4b)		\
+		EX_TABLE(0b,4b) EX_TABLE(1b,4b)				\
+		EX_TABLE(2b,4b) EX_TABLE(3b,4b)				\
 		: "=d" (ret), "=&d" (oldval), "=&d" (newval),		\
 		  "=m" (*uaddr)						\
 		: "0" (-EFAULT), "d" (oparg), "a" (uaddr),		\

From: Jan Beulich <[email protected]>

commit f28347cc66395e96712f5c2db0a302ee75bafce6 upstream.

While working on XSA-361 and its follow-ups, I failed to spot another place where the kernel mapping part of an operation was not treated the same as the user space part. Detect and propagate errors and add a 2nd pr_debug().

---
 drivers/xen/gntdev.c |    8 ++++++++
 1 file changed, 8 insertions(+)

@@ -399,6 +399,14 @@ static void __unmap_grant_pages_done(int
 			map->unmap_ops[offset+i].handle,
 			map->unmap_ops[offset+i].status);
 		map->unmap_ops[offset+i].handle = -1;
+		if (use_ptemod) {
+			WARN_ON(map->kunmap_ops[offset+i].status &&
+				map->kunmap_ops[offset+i].handle != -1);
+			pr_debug("kunmap handle=%u st=%d\n",
+				 map->kunmap_ops[offset+i].handle,
+				 map->kunmap_ops[offset+i].status);
+			map->kunmap_ops[offset+i].handle = -1;
+		}
 	}
 	/*
 	 * Decrease the live-grant counter.  This must happen after the loop to

From: M. Vefa Bicakci <[email protected]>

commit 0991028cd49567d7016d1b224fe0117c35059f86 upstream.

Prior to this commit, if a grant mapping operation failed partially, some of the entries in the map_ops array would be invalid, whereas all of the entries in the kmap_ops array would be valid. This in turn would cause the following logic in gntdev_map_grant_pages to become invalid:

  for (i = 0; i < map->count; i++) {
    if (map->map_ops[i].status == GNTST_okay) {
      map->unmap_ops[i].handle = map->map_ops[i].handle;
      if (!use_ptemod)
        alloced++;
    }
    if (use_ptemod) {
      if (map->kmap_ops[i].status == GNTST_okay) {
        if (map->map_ops[i].status == GNTST_okay)
          alloced++;
        map->kunmap_ops[i].handle = map->kmap_ops[i].handle;
      }
    }
  }
  ...
  atomic_add(alloced, &map->live_grants);

Assume that use_ptemod is true (i.e., the domain mapping the granted pages is a paravirtualized domain). In the code excerpt above, note that the "alloced" variable is only incremented when both kmap_ops[i].status and map_ops[i].status are set to GNTST_okay (i.e., both mapping operations are successful). However, as also noted above, there are cases where a grant mapping operation fails partially, breaking the assumption of the code excerpt above.

The aforementioned causes map->live_grants to be incorrectly set. In some cases, all of the map_ops mappings fail, but all of the kmap_ops mappings succeed, meaning that live_grants may remain zero. This in turn makes it impossible to unmap the successfully grant-mapped pages pointed to by kmap_ops, because unmap_grant_pages has the following snippet of code at its beginning:

if (atomic_read(&map->live_grants) == 0)
return; /* Nothing to do */

In other cases where only some of the map_ops mappings fail but all kmap_ops mappings succeed, live_grants is made positive, but when the user requests unmapping the grant-mapped pages, __unmap_grant_pages_done will then make map->live_grants negative, because the latter function does not check if all of the pages that were requested to be unmapped were actually unmapped, and the same function unconditionally subtracts "data->count" (i.e., a value that can be greater than map->live_grants) from map->live_grants. The side effects of a negative live_grants value have not been studied.

The net effect of all of this is that grant references are leaked in one of the above conditions. In Qubes OS v4.1 (which uses Xen's grant mechanism extensively for X11 GUI isolation), this issue manifests itself with warning messages like the following to be printed out by the Linux kernel in the VM that had granted pages (that contain X11 GUI window data) to dom0: "g.e. 0x1234 still pending", especially after the user rapidly resizes GUI VM windows (causing some grant-mapping operations to partially or completely fail, due to the fact that the VM unshares some of the pages as part of the window resizing, making the pages impossible to grant-map from dom0).

The fix for this issue involves counting all successful map_ops and kmap_ops mappings separately, and then adding the sum to live_grants. During unmapping, only the number of successfully unmapped grants is subtracted from live_grants. The code is also modified to check for negative live_grants values after the subtraction and warn the user.

---
 drivers/xen/gntdev.c |   22 +++++++++++++++++-----
 1 file changed, 17 insertions(+), 5 deletions(-)

@@ -364,8 +364,7 @@ static int map_grant_pages(struct grant_
 	for (i = 0; i < map->count; i++) {
 		if (map->map_ops[i].status == GNTST_okay) {
 			map->unmap_ops[i].handle = map->map_ops[i].handle;
-			if (!use_ptemod)
-				alloced++;
+			alloced++;
 		} else if (!err)
 			err = -EINVAL;
 

@@ -374,8 +373,7 @@ static int map_grant_pages(struct grant_
 
 		if (use_ptemod) {
 			if (map->kmap_ops[i].status == GNTST_okay) {
-				if (map->map_ops[i].status == GNTST_okay)
-					alloced++;
+				alloced++;
 				map->kunmap_ops[i].handle = map->kmap_ops[i].handle;
 			} else if (!err)
 				err = -EINVAL;

@@ -391,8 +389,14 @@ static void __unmap_grant_pages_done(int
 	unsigned int i;
 	struct grant_map *map = data->data;
 	unsigned int offset = data->unmap_ops - map->unmap_ops;
+	int successful_unmaps = 0;
+	int live_grants;
 
 	for (i = 0; i < data->count; i++) {
+		if (map->unmap_ops[offset + i].status == GNTST_okay &&
+		    map->unmap_ops[offset + i].handle != -1)
+			successful_unmaps++;
+
 		WARN_ON(map->unmap_ops[offset+i].status &&
 			map->unmap_ops[offset+i].handle != -1);
 		pr_debug("unmap handle=%d st=%d\n",

@@ -400,6 +404,10 @@ static void __unmap_grant_pages_done(int
 			map->unmap_ops[offset+i].status);
 		map->unmap_ops[offset+i].handle = -1;
 		if (use_ptemod) {
+			if (map->kunmap_ops[offset + i].status == GNTST_okay &&
+			    map->kunmap_ops[offset + i].handle != -1)
+				successful_unmaps++;
+
 			WARN_ON(map->kunmap_ops[offset+i].status &&
 				map->kunmap_ops[offset+i].handle != -1);
 			pr_debug("kunmap handle=%u st=%d\n",

@@ -408,11 +416,15 @@ static void __unmap_grant_pages_done(int
 			map->kunmap_ops[offset+i].handle = -1;
 		}
 	}
+
 	/*
 	 * Decrease the live-grant counter.  This must happen after the loop to
 	 * prevent premature reuse of the grants by gnttab_mmap().
 	 */
-	atomic_sub(data->count, &map->live_grants);
+	live_grants = atomic_sub_return(successful_unmaps, &map->live_grants);
+	if (WARN_ON(live_grants < 0))
+		pr_err("%s: live_grants became negative (%d) after unmapping %d pages!\n",
+		       __func__, live_grants, successful_unmaps);
 
 	/* Release reference taken by unmap_grant_pages */
 	gntdev_put_map(NULL, map);

From: Rik van Riel <[email protected]>

commit 12df140f0bdfae5dcfc81800970dd7f6f632e00c upstream.

The h->*_huge_pages counters are protected by the hugetlb_lock, but alloc_huge_page has a corner case where it can decrement the counter outside of the lock.

This could lead to a corrupted value of h->resv_huge_pages, which we have observed on our systems.

Take the hugetlb_lock before decrementing h->resv_huge_pages to avoid a potential race.

---
 mm/hugetlb.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -2089,11 +2089,11 @@ struct page *alloc_huge_page(struct vm_a
 		page = __alloc_buddy_huge_page_with_mpol(h, vma, addr);
 		if (!page)
 			goto out_uncharge_cgroup;
+		spin_lock(&hugetlb_lock);
 		if (!avoid_reserve && vma_has_reserves(vma, gbl_chg)) {
 			SetPagePrivate(page);
 			h->resv_huge_pages--;
 		}
-		spin_lock(&hugetlb_lock);
 		list_move(&page->lru, &h->hugepage_activelist);
 		/* Fall through */
 	}

From: Wei Yongjun <[email protected]>

commit 444d8ad4916edec8a9fc684e841287db9b1e999f upstream.

Fix to return error code -EINVAL from the error handling case instead of 0, as done elsewhere in this function.

---
 net/ieee802154/socket.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -515,8 +515,10 @@ static int dgram_bind(struct sock *sk, s
 	if (err < 0)
 		goto out;
 
-	if (addr->family != AF_IEEE802154)
+	if (addr->family != AF_IEEE802154) {
+		err = -EINVAL;
 		goto out;
+	}
 
 	ieee802154_addr_from_sa(&haddr, &addr->addr);
 	dev = ieee802154_get_dev(sock_net(sk), &haddr);

From: Nathan Huckleberry <[email protected]>

[ Upstream commit 0b33a33bd15d5bab73b87152b220a8d0153a4587 ]

The mode_valid field in drm_connector_helper_funcs is expected to be of type:
enum drm_mode_status (* mode_valid) (struct drm_connector *connector,

                                     struct drm_display_mode *mode);

The mismatched return type breaks forward edge kCFI since the underlying function definition does not match the function hook definition.

The return type of mdp4_lvds_connector_mode_valid should be changed from int to enum drm_mode_status.

Patchwork: https://patchwork.freedesktop.org/patch/502878/

---
 drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_connector.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

index e3b1c86b7aae..5c932b3fb831 100644
@@ -71,8 +71,9 @@ static int mdp4_lvds_connector_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int mdp4_lvds_connector_mode_valid(struct drm_connector *connector,
-				 struct drm_display_mode *mode)
+static enum drm_mode_status
+mdp4_lvds_connector_mode_valid(struct drm_connector *connector,
+			       struct drm_display_mode *mode)
 {
 	struct mdp4_lvds_connector *mdp4_lvds_connector =
 			to_mdp4_lvds_connector(connector);

2.35.1

From: Randy Dunlap <[email protected]>

[ Upstream commit c44f15c1c09481d50fd33478ebb5b8284f8f5edb ]

Add 'volatile' to iounmap()'s argument to prevent build warnings. This make it the same as other major architectures.

Placates these warnings: (12 such warnings)

../drivers/video/fbdev/riva/fbdev.c: In function 'rivafb_probe':
../drivers/video/fbdev/riva/fbdev.c:2067:42: error: passing argument 1 of 'iounmap' discards 'volatile' qualifier from pointer target type [-Werror=discarded-qualifiers]
2067 | iounmap(default_par->riva.PRAMIN);

---
 arch/arc/include/asm/io.h | 2 +-
 arch/arc/mm/ioremap.c     | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

index 2f39d9b3886e..19d0cab60a39 100644
@@ -35,7 +35,7 @@ static inline void ioport_unmap(void __iomem *addr)
 {
 }
 
-extern void iounmap(const void __iomem *addr);
+extern void iounmap(const volatile void __iomem *addr);
 
 #define ioremap_nocache(phy, sz)	ioremap(phy, sz)
 #define ioremap_wc(phy, sz)		ioremap(phy, sz)

index 9881bd740ccc..0719b1280ef8 100644
@@ -95,7 +95,7 @@ void __iomem *ioremap_prot(phys_addr_t paddr, unsigned long size,
 EXPORT_SYMBOL(ioremap_prot);
 
 
-void iounmap(const void __iomem *addr)
+void iounmap(const volatile void __iomem *addr)
 {
 	/* weird double cast to handle phys_addr_t > 32 bits */
 	if (arc_uncached_addr_space((phys_addr_t)(u32)addr))

2.35.1

From: Yang Yingliang <[email protected]>

[ Upstream commit 4881bda5ea05c8c240fc8afeaa928e2bc43f61fa ]

If device_register() fails in snd_ac97_dev_register(), it should call put_device() to give up reference, or the name allocated in dev_set_name() is leaked.

---
 sound/pci/ac97/ac97_codec.c | 1 +
 1 file changed, 1 insertion(+)

index 1cb2a1ecf3cf..d5dfc7349e70 100644
@@ -1965,6 +1965,7 @@ static int snd_ac97_dev_register(struct snd_device *device)
 		     snd_ac97_get_short_name(ac97));
 	if ((err = device_register(&ac97->dev)) < 0) {
 		ac97_err(ac97, "Can't register ac97 bus\n");
+		put_device(&ac97->dev);
 		ac97->dev.bus = NULL;
 		return err;
 	}

2.35.1

From: Chen Zhongjin <[email protected]>

[ Upstream commit 230db82413c091bc16acee72650f48d419cebe49 ]

When a console stack dump is initiated with CONFIG_GCOV_PROFILE_ALL enabled, show_trace_log_lvl() gets out of sync with the ORC unwinder, causing the stack trace to show all text addresses as unreliable:

  # echo l > /proc/sysrq-trigger
  [  477.521031] sysrq: Show backtrace of all active CPUs
  [  477.523813] NMI backtrace for cpu 0
  [  477.524492] CPU: 0 PID: 1021 Comm: bash Not tainted 6.0.0 #65
  [  477.525295] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.0-1.fc36 04/01/2014
  [  477.526439] Call Trace:
  [  477.526854]  <TASK>
  [  477.527216]  ? dump_stack_lvl+0xc7/0x114
  [  477.527801]  ? dump_stack+0x13/0x1f
  [  477.528331]  ? nmi_cpu_backtrace.cold+0xb5/0x10d
  [  477.528998]  ? lapic_can_unplug_cpu+0xa0/0xa0
  [  477.529641]  ? nmi_trigger_cpumask_backtrace+0x16a/0x1f0
  [  477.530393]  ? arch_trigger_cpumask_backtrace+0x1d/0x30
  [  477.531136]  ? sysrq_handle_showallcpus+0x1b/0x30
  [  477.531818]  ? __handle_sysrq.cold+0x4e/0x1ae
  [  477.532451]  ? write_sysrq_trigger+0x63/0x80
  [  477.533080]  ? proc_reg_write+0x92/0x110
  [  477.533663]  ? vfs_write+0x174/0x530
  [  477.534265]  ? handle_mm_fault+0x16f/0x500
  [  477.534940]  ? ksys_write+0x7b/0x170
  [  477.535543]  ? __x64_sys_write+0x1d/0x30
  [  477.536191]  ? do_syscall_64+0x6b/0x100
  [  477.536809]  ? entry_SYSCALL_64_after_hwframe+0x63/0xcd
  [  477.537609]  </TASK>

This happens when the compiled code for show_stack() has a single word on the stack, and doesn't use a tail call to show_stack_log_lvl(). (CONFIG_GCOV_PROFILE_ALL=y is the only known case of this.) Then the __unwind_start() skip logic hits an off-by-one bug and fails to unwind all the way to the intended starting frame.

Fix it by reverting the following commit:

f1d9a2abff66 ("x86/unwind/orc: Don't skip the first frame for inactive tasks")

The original justification for that commit no longer exists. That original issue was later fixed in a different way, with the following commit:

f2ac57a4c49d ("x86/unwind/orc: Fix inactive tasks with stack pointer in %sp on GCC 10 compiled kernels")

---
 arch/x86/kernel/unwind_orc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index e64c5b78fbfd..350f40f9a0bf 100644
@@ -579,7 +579,7 @@ void __unwind_start(struct unwind_state *state, struct task_struct *task,
 	/* Otherwise, skip ahead to the user-specified starting frame: */
 	while (!unwind_done(state) &&
 	       (!on_stack(&state->stack_info, first_frame, sizeof(long)) ||
-			state->sp < (unsigned long)first_frame))
+			state->sp <= (unsigned long)first_frame))
 		unwind_next_frame(state);
 
 	return;

2.35.1

From: Raju Rangoju <[email protected]>

[ Upstream commit 09c5f6bf11ac98874339e55f4f5f79a9dbc9b375 ]

The current XGBE code assumes that offset 6 of EEPROM SFP DAC (passive) cables is NULL. However, some cables (the 5 meter and 7 meter Molex passive cables) have non-zero data at offset 6. Fix the logic by moving the passive cable check above the active checks, so as not to be improperly identified as an active cable. This will fix the issue for any passive cable that advertises 1000Base-CX in offset 6.

---
 drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

index 4a4370a470fd..d432489c093a 100644
@@ -1095,7 +1095,10 @@ static void xgbe_phy_sfp_parse_eeprom(struct xgbe_prv_data *pdata)
 	}
 
 	/* Determine the type of SFP */
-	if (sfp_base[XGBE_SFP_BASE_10GBE_CC] & XGBE_SFP_BASE_10GBE_CC_SR)
+	if (phy_data->sfp_cable == XGBE_SFP_CABLE_PASSIVE &&
+	    xgbe_phy_sfp_bit_rate(sfp_eeprom, XGBE_SFP_SPEED_10000))
+		phy_data->sfp_base = XGBE_SFP_BASE_10000_CR;
+	else if (sfp_base[XGBE_SFP_BASE_10GBE_CC] & XGBE_SFP_BASE_10GBE_CC_SR)
 		phy_data->sfp_base = XGBE_SFP_BASE_10000_SR;
 	else if (sfp_base[XGBE_SFP_BASE_10GBE_CC] & XGBE_SFP_BASE_10GBE_CC_LR)
 		phy_data->sfp_base = XGBE_SFP_BASE_10000_LR;

@@ -1111,9 +1114,6 @@ static void xgbe_phy_sfp_parse_eeprom(struct xgbe_prv_data *pdata)
 		phy_data->sfp_base = XGBE_SFP_BASE_1000_CX;
 	else if (sfp_base[XGBE_SFP_BASE_1GBE_CC] & XGBE_SFP_BASE_1GBE_CC_T)
 		phy_data->sfp_base = XGBE_SFP_BASE_1000_T;
-	else if ((phy_data->sfp_cable == XGBE_SFP_CABLE_PASSIVE) &&
-		 xgbe_phy_sfp_bit_rate(sfp_eeprom, XGBE_SFP_SPEED_10000))
-		phy_data->sfp_base = XGBE_SFP_BASE_10000_CR;
 
 	switch (phy_data->sfp_base) {
 	case XGBE_SFP_BASE_1000_T:

2.35.1

From: Raju Rangoju <[email protected]>

[ Upstream commit 170a9e341a3b02c0b2ea0df16ef14a33a4f41de8 ]

The offset 12 (bit-rate) of EEPROM SFP DAC (passive) cables is expected to be in the range 0x64 to 0x68. However, the 5 meter and 7 meter Molex passive cables have the rate ceiling 0x78 at offset 12.

Add a quirk for Molex passive cables to extend the rate ceiling to 0x78.

---
 drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

index d432489c093a..3ccdac464cf5 100644
@@ -237,6 +237,7 @@ enum xgbe_sfp_speed {
 #define XGBE_SFP_BASE_BR_1GBE_MAX		0x0d
 #define XGBE_SFP_BASE_BR_10GBE_MIN		0x64
 #define XGBE_SFP_BASE_BR_10GBE_MAX		0x68
+#define XGBE_MOLEX_SFP_BASE_BR_10GBE_MAX	0x78
 
 #define XGBE_SFP_BASE_CU_CABLE_LEN		18
 

@@ -273,6 +274,8 @@ struct xgbe_sfp_eeprom {
 #define XGBE_BEL_FUSE_VENDOR	"BEL-FUSE        "
 #define XGBE_BEL_FUSE_PARTNO	"1GBT-SFP06      "
 
+#define XGBE_MOLEX_VENDOR	"Molex Inc.      "
+
 struct xgbe_sfp_ascii {
 	union {
 		char vendor[XGBE_SFP_BASE_VENDOR_NAME_LEN + 1];

@@ -840,7 +843,11 @@ static bool xgbe_phy_sfp_bit_rate(struct xgbe_sfp_eeprom *sfp_eeprom,
 		break;
 	case XGBE_SFP_SPEED_10000:
 		min = XGBE_SFP_BASE_BR_10GBE_MIN;
-		max = XGBE_SFP_BASE_BR_10GBE_MAX;
+		if (memcmp(&sfp_eeprom->base[XGBE_SFP_BASE_VENDOR_NAME],
+			   XGBE_MOLEX_VENDOR, XGBE_SFP_BASE_VENDOR_NAME_LEN) == 0)
+			max = XGBE_MOLEX_SFP_BASE_BR_10GBE_MAX;
+		else
+			max = XGBE_SFP_BASE_BR_10GBE_MAX;
 		break;
 	default:
 		return false;

2.35.1

From: Eric Dumazet <[email protected]>

[ Upstream commit 15e4dabda11b0fa31d510a915d1a580f47dfc92e ]

kcm->rx_psock can be read locklessly in kcm_rfree(). Annotate the read and writes accordingly.

We do the same for kcm->rx_wait in the following patch.

syzbot reported:

BUG: KCSAN: data-race in kcm_rfree / unreserve_rx_kcm

write to 0xffff888123d827b8 of 8 bytes by task 2758 on cpu 1:

unreserve_rx_kcm+0x72/0x1f0 net/kcm/kcmsock.c:313
kcm_rcv_strparser+0x2b5/0x3a0 net/kcm/kcmsock.c:373
__strp_recv+0x64c/0xd20 net/strparser/strparser.c:301
strp_recv+0x6d/0x80 net/strparser/strparser.c:335
tcp_read_sock+0x13e/0x5a0 net/ipv4/tcp.c:1703
strp_read_sock net/strparser/strparser.c:358 [inline]
do_strp_work net/strparser/strparser.c:406 [inline]
strp_work+0xe8/0x180 net/strparser/strparser.c:415
process_one_work+0x3d3/0x720 kernel/workqueue.c:2289
worker_thread+0x618/0xa70 kernel/workqueue.c:2436
kthread+0x1a9/0x1e0 kernel/kthread.c:376
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:306

read to 0xffff888123d827b8 of 8 bytes by task 5859 on cpu 0:

kcm_rfree+0x14c/0x220 net/kcm/kcmsock.c:181
skb_release_head_state+0x8e/0x160 net/core/skbuff.c:841
skb_release_all net/core/skbuff.c:852 [inline]
__kfree_skb net/core/skbuff.c:868 [inline]
kfree_skb_reason+0x5c/0x260 net/core/skbuff.c:891
kfree_skb include/linux/skbuff.h:1216 [inline]
kcm_recvmsg+0x226/0x2b0 net/kcm/kcmsock.c:1161
____sys_recvmsg+0x16c/0x2e0
___sys_recvmsg net/socket.c:2743 [inline]
do_recvmmsg+0x2f1/0x710 net/socket.c:2837
__sys_recvmmsg net/socket.c:2916 [inline]
__do_sys_recvmmsg net/socket.c:2939 [inline]
__se_sys_recvmmsg net/socket.c:2932 [inline]
__x64_sys_recvmmsg+0xde/0x160 net/socket.c:2932
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x2b/0x70 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd

value changed: 0xffff88812971ce00 -> 0x0000000000000000

Reported by Kernel Concurrency Sanitizer on:

CPU: 0 PID: 5859 Comm: syz-executor.3 Not tainted 6.0.0-syzkaller-12189-g19d17ab7c68b-dirty #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/22/2022

---
 net/kcm/kcmsock.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

index c364d849e7c3..ce022c2e320e 100644
@@ -180,7 +180,7 @@ static void kcm_rfree(struct sk_buff *skb)
 	/* For reading rx_wait and rx_psock without holding lock */
 	smp_mb__after_atomic();
 
-	if (!kcm->rx_wait && !kcm->rx_psock &&
+	if (!kcm->rx_wait && !READ_ONCE(kcm->rx_psock) &&
 	    sk_rmem_alloc_get(sk) < sk->sk_rcvlowat) {
 		spin_lock_bh(&mux->rx_lock);
 		kcm_rcv_ready(kcm);

@@ -285,7 +285,8 @@ static struct kcm_sock *reserve_rx_kcm(struct kcm_psock *psock,
 	kcm->rx_wait = false;
 
 	psock->rx_kcm = kcm;
-	kcm->rx_psock = psock;
+	/* paired with lockless reads in kcm_rfree() */
+	WRITE_ONCE(kcm->rx_psock, psock);
 
 	spin_unlock_bh(&mux->rx_lock);
 

@@ -312,7 +313,8 @@ static void unreserve_rx_kcm(struct kcm_psock *psock,
 	spin_lock_bh(&mux->rx_lock);
 
 	psock->rx_kcm = NULL;
-	kcm->rx_psock = NULL;
+	/* paired with lockless reads in kcm_rfree() */
+	WRITE_ONCE(kcm->rx_psock, NULL);
 
 	/* Commit kcm->rx_psock before sk_rmem_alloc_get to sync with
 	 * kcm_rfree

2.35.1

From: Eric Dumazet <[email protected]>

[ Upstream commit 0c745b5141a45a076f1cb9772a399f7ebcb0948a ]

kcm->rx_psock can be read locklessly in kcm_rfree(). Annotate the read and writes accordingly.

syzbot reported:

BUG: KCSAN: data-race in kcm_rcv_strparser / kcm_rfree

write to 0xffff88810784e3d0 of 1 bytes by task 1823 on cpu 1:

reserve_rx_kcm net/kcm/kcmsock.c:283 [inline]
kcm_rcv_strparser+0x250/0x3a0 net/kcm/kcmsock.c:363
__strp_recv+0x64c/0xd20 net/strparser/strparser.c:301
strp_recv+0x6d/0x80 net/strparser/strparser.c:335
tcp_read_sock+0x13e/0x5a0 net/ipv4/tcp.c:1703
strp_read_sock net/strparser/strparser.c:358 [inline]
do_strp_work net/strparser/strparser.c:406 [inline]
strp_work+0xe8/0x180 net/strparser/strparser.c:415
process_one_work+0x3d3/0x720 kernel/workqueue.c:2289
worker_thread+0x618/0xa70 kernel/workqueue.c:2436
kthread+0x1a9/0x1e0 kernel/kthread.c:376
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:306

read to 0xffff88810784e3d0 of 1 bytes by task 17869 on cpu 0:

kcm_rfree+0x121/0x220 net/kcm/kcmsock.c:181
skb_release_head_state+0x8e/0x160 net/core/skbuff.c:841
skb_release_all net/core/skbuff.c:852 [inline]
__kfree_skb net/core/skbuff.c:868 [inline]
kfree_skb_reason+0x5c/0x260 net/core/skbuff.c:891
kfree_skb include/linux/skbuff.h:1216 [inline]
kcm_recvmsg+0x226/0x2b0 net/kcm/kcmsock.c:1161
____sys_recvmsg+0x16c/0x2e0
___sys_recvmsg net/socket.c:2743 [inline]
do_recvmmsg+0x2f1/0x710 net/socket.c:2837
__sys_recvmmsg net/socket.c:2916 [inline]
__do_sys_recvmmsg net/socket.c:2939 [inline]
__se_sys_recvmmsg net/socket.c:2932 [inline]
__x64_sys_recvmmsg+0xde/0x160 net/socket.c:2932
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x2b/0x70 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd

value changed: 0x01 -> 0x00

Reported by Kernel Concurrency Sanitizer on:

CPU: 0 PID: 17869 Comm: syz-executor.2 Not tainted 6.1.0-rc1-syzkaller-00010-gbb1a1146467a-dirty #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/22/2022

---
 net/kcm/kcmsock.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

index ce022c2e320e..629103c7337b 100644
@@ -164,7 +164,8 @@ static void kcm_rcv_ready(struct kcm_sock *kcm)
 	/* Buffer limit is okay now, add to ready list */
 	list_add_tail(&kcm->wait_rx_list,
 		      &kcm->mux->kcm_rx_waiters);
-	kcm->rx_wait = true;
+	/* paired with lockless reads in kcm_rfree() */
+	WRITE_ONCE(kcm->rx_wait, true);
 }
 
 static void kcm_rfree(struct sk_buff *skb)

@@ -180,7 +181,7 @@ static void kcm_rfree(struct sk_buff *skb)
 	/* For reading rx_wait and rx_psock without holding lock */
 	smp_mb__after_atomic();
 
-	if (!kcm->rx_wait && !READ_ONCE(kcm->rx_psock) &&
+	if (!READ_ONCE(kcm->rx_wait) && !READ_ONCE(kcm->rx_psock) &&
 	    sk_rmem_alloc_get(sk) < sk->sk_rcvlowat) {
 		spin_lock_bh(&mux->rx_lock);
 		kcm_rcv_ready(kcm);

@@ -239,7 +240,8 @@ static void requeue_rx_msgs(struct kcm_mux *mux, struct sk_buff_head *head)
 		if (kcm_queue_rcv_skb(&kcm->sk, skb)) {
 			/* Should mean socket buffer full */
 			list_del(&kcm->wait_rx_list);
-			kcm->rx_wait = false;
+			/* paired with lockless reads in kcm_rfree() */
+			WRITE_ONCE(kcm->rx_wait, false);
 
 			/* Commit rx_wait to read in kcm_free */
 			smp_wmb();

@@ -282,7 +284,8 @@ static struct kcm_sock *reserve_rx_kcm(struct kcm_psock *psock,
 	kcm = list_first_entry(&mux->kcm_rx_waiters,
 			       struct kcm_sock, wait_rx_list);
 	list_del(&kcm->wait_rx_list);
-	kcm->rx_wait = false;
+	/* paired with lockless reads in kcm_rfree() */
+	WRITE_ONCE(kcm->rx_wait, false);
 
 	psock->rx_kcm = kcm;
 	/* paired with lockless reads in kcm_rfree() */

@@ -1242,7 +1245,8 @@ static void kcm_recv_disable(struct kcm_sock *kcm)
 	if (!kcm->rx_psock) {
 		if (kcm->rx_wait) {
 			list_del(&kcm->wait_rx_list);
-			kcm->rx_wait = false;
+			/* paired with lockless reads in kcm_rfree() */
+			WRITE_ONCE(kcm->rx_wait, false);
 		}
 
 		requeue_rx_msgs(mux, &kcm->sk.sk_receive_queue);

@@ -1800,7 +1804,8 @@ static void kcm_done(struct kcm_sock *kcm)
 
 	if (kcm->rx_wait) {
 		list_del(&kcm->wait_rx_list);
-		kcm->rx_wait = false;
+		/* paired with lockless reads in kcm_rfree() */
+		WRITE_ONCE(kcm->rx_wait, false);
 	}
 	/* Move any pending receive messages to other kcm sockets */
 	requeue_rx_msgs(mux, &sk->sk_receive_queue);

2.35.1

From: Zhang Changzhong <[email protected]>

[ Upstream commit 9c1eaa27ec599fcc25ed4970c0b73c247d147a2b ]

The ndo_start_xmit() method must not free skb when returning NETDEV_TX_BUSY, since caller is going to requeue freed skb.

---
 drivers/net/ethernet/lantiq_etop.c | 1 -
 1 file changed, 1 deletion(-)

index afc810069440..2a14520e4798 100644
@@ -479,7 +479,6 @@ ltq_etop_tx(struct sk_buff *skb, struct net_device *dev)
 	len = skb->len < ETH_ZLEN ? ETH_ZLEN : skb->len;
 
 	if ((desc->ctl & (LTQ_DMA_OWN | LTQ_DMA_C)) || ch->skb[ch->dma.desc]) {
-		dev_kfree_skb_any(skb);
 		netdev_err(dev, "tx ring full\n");
 		netif_tx_stop_queue(txq);
 		return NETDEV_TX_BUSY;

2.35.1

From: Neal Cardwell <[email protected]>

[ Upstream commit 3d2af9cce3133b3bc596a9d065c6f9d93419ccfb ]

This commit fixes a bug that can cause a TCP data sender to repeatedly defer RTOs when encountering SACK reneging.

The bug is that when we're in fast recovery in a scenario with SACK reneging, every time we get an ACK we call tcp_check_sack_reneging() and it can note the apparent SACK reneging and rearm the RTO timer for srtt/2 into the future. In some SACK reneging scenarios that can happen repeatedly until the receive window fills up, at which point the sender can't send any more, the ACKs stop arriving, and the RTO fires at srtt/2 after the last ACK. But that can take far too long (O(10 secs)), since the connection is stuck in fast recovery with a low cwnd that cannot grow beyond ssthresh, even if more bandwidth is available.

This fix changes the logic in tcp_check_sack_reneging() to only rearm the RTO timer if data is cumulatively ACKed, indicating forward progress. This avoids this kind of nearly infinite loop of RTO timer re-arming. In addition, this meets the goals of tcp_check_sack_reneging() in handling Windows TCP behavior that looks temporarily like SACK reneging but is not really.

Many thanks to Jakub Kicinski and Neil Spring, who reported this issue and provided critical packet traces that enabled root-causing this issue. Also, many thanks to Jakub Kicinski for testing this fix.

---
 net/ipv4/tcp_input.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index eeed79ce8f9f..87095d5ecf95 100644
@@ -2058,7 +2058,8 @@ void tcp_enter_loss(struct sock *sk)
  */
 static bool tcp_check_sack_reneging(struct sock *sk, int flag)
 {
-	if (flag & FLAG_SACK_RENEGING) {
+	if (flag & FLAG_SACK_RENEGING &&
+	    flag & FLAG_SND_UNA_ADVANCED) {
 		struct tcp_sock *tp = tcp_sk(sk);
 		unsigned long delay = max(usecs_to_jiffies(tp->srtt_us >> 4),
 					  msecs_to_jiffies(10));

2.35.1

From: Dongliang Mu <[email protected]>

[ Upstream commit 3e5b3418827cefb5e1cc658806f02965791b8f07 ]

The commit 1149108e2fbf ("can: mscan: improve clock API use") only adds put_clock() in mpc5xxx_can_remove() function, forgetting to add put_clock() in the error handling code.

Fix this bug by adding put_clock() in the error handling code.

---
 drivers/net/can/mscan/mpc5xxx_can.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

index 2949a381a94d..21993ba7ae2a 100644
@@ -336,14 +336,14 @@ static int mpc5xxx_can_probe(struct platform_device *ofdev)
 					       &mscan_clksrc);
 	if (!priv->can.clock.freq) {
 		dev_err(&ofdev->dev, "couldn't get MSCAN clock properties\n");
-		goto exit_free_mscan;
+		goto exit_put_clock;
 	}
 
 	err = register_mscandev(dev, mscan_clksrc);
 	if (err) {
 		dev_err(&ofdev->dev, "registering %s failed (err=%d)\n",
 			DRV_NAME, err);
-		goto exit_free_mscan;
+		goto exit_put_clock;
 	}
 
 	dev_info(&ofdev->dev, "MSCAN at 0x%p, irq %d, clock %d Hz\n",

@@ -351,7 +351,9 @@ static int mpc5xxx_can_probe(struct platform_device *ofdev)
 
 	return 0;
 
-exit_free_mscan:
+exit_put_clock:
+	if (data->put_clock)
+		data->put_clock(ofdev);
 	free_candev(dev);
 exit_dispose_irq:
 	irq_dispose_mapping(irq);

2.35.1

From: Mario Limonciello <[email protected]>

[ Upstream commit 85850af4fc47132f3f2f0dd698b90f67906600b4 ]

Hybrid sleep is currently hardcoded to only operate with S3 even on systems that might not support it.

Instead of assuming this mode is what the user wants to use, for hybrid sleep follow the setting of mem_sleep_current which will respect mem_sleep_default kernel command line and policy decisions made by the presence of the FADT low power idle bit.

---
 kernel/power/hibernate.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index f4ecb23c9194..e97f2395e15d 100644
@@ -618,7 +618,7 @@ static void power_down(void)
 	int error;
 
 	if (hibernation_mode == HIBERNATION_SUSPEND) {
-		error = suspend_devices_and_enter(PM_SUSPEND_MEM);
+		error = suspend_devices_and_enter(mem_sleep_current);
 		if (error) {
 			hibernation_mode = hibernation_ops ?
 						HIBERNATION_PLATFORM :

2.35.1

From: Hans Verkuil <[email protected]>

[ Upstream commit f8bcaf714abfc94818dff8c0db84d750433984f4 ]

VIDIOC_S_FBUF is by definition a scary ioctl, which is why only root can use it. But at least check if the framebuffer parameters match that of one of the framebuffer created by vivid, and reject anything else.

---
 drivers/media/platform/vivid/vivid-core.c    | 22 ++++++++++++++++++++
 drivers/media/platform/vivid/vivid-core.h    |  2 ++
 drivers/media/platform/vivid/vivid-vid-cap.c |  9 +++++++-
 3 files changed, 32 insertions(+), 1 deletion(-)

index 6754e5fcc4c4..c4bbdd5196bf 100644
@@ -309,6 +309,28 @@ static int vidioc_g_fbuf(struct file *file, void *fh, struct v4l2_framebuffer *a
 	return vivid_vid_out_g_fbuf(file, fh, a);
 }
 
+/*
+ * Only support the framebuffer of one of the vivid instances.
+ * Anything else is rejected.
+ */
+bool vivid_validate_fb(const struct v4l2_framebuffer *a)
+{
+	struct vivid_dev *dev;
+	int i;
+
+	for (i = 0; i < n_devs; i++) {
+		dev = vivid_devs[i];
+		if (!dev || !dev->video_pbase)
+			continue;
+		if ((unsigned long)a->base == dev->video_pbase &&
+		    a->fmt.width <= dev->display_width &&
+		    a->fmt.height <= dev->display_height &&
+		    a->fmt.bytesperline <= dev->display_byte_stride)
+			return true;
+	}
+	return false;
+}
+
 static int vidioc_s_fbuf(struct file *file, void *fh, const struct v4l2_framebuffer *a)
 {
 	struct video_device *vdev = video_devdata(file);

index 5cdf95bdc4d1..6de8eb8d2cef 100644
@@ -562,4 +562,6 @@ static inline bool vivid_is_hdmi_out(const struct vivid_dev *dev)
 	return dev->output_type[dev->output] == HDMI;
 }
 
+bool vivid_validate_fb(const struct v4l2_framebuffer *a);
+
 #endif

index c66568e8f388..182d8a2b3c98 100644
@@ -1255,7 +1255,14 @@ int vivid_vid_cap_s_fbuf(struct file *file, void *fh,
 		return -EINVAL;
 	if (a->fmt.bytesperline < (a->fmt.width * fmt->bit_depth[0]) / 8)
 		return -EINVAL;
-	if (a->fmt.height * a->fmt.bytesperline < a->fmt.sizeimage)
+	if (a->fmt.bytesperline > a->fmt.sizeimage / a->fmt.height)
+		return -EINVAL;
+
+	/*
+	 * Only support the framebuffer of one of the vivid instances.
+	 * Anything else is rejected.
+	 */
+	if (!vivid_validate_fb(a))
 		return -EINVAL;
 
 	dev->fb_vbase_cap = phys_to_virt((unsigned long)a->base);

2.35.1

From: Hans Verkuil <[email protected]>

[ Upstream commit 1f65ea411cc7b6ff128d82a3493d7b5648054e6f ]

Whenever the compose width/height values change, the dev->bitmap_cap vmalloc'ed array must be freed and dev->bitmap_cap set to NULL.

This was done in some places, but not all. This is only an issue if overlay support is enabled and the bitmap clipping is used.

---
 drivers/media/platform/vivid/vivid-vid-cap.c | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

index 182d8a2b3c98..459cff1626a6 100644
@@ -458,6 +458,12 @@ void vivid_update_format_cap(struct vivid_dev *dev, bool keep_controls)
 	tpg_reset_source(&dev->tpg, dev->src_rect.width, dev->src_rect.height, dev->field_cap);
 	dev->crop_cap = dev->src_rect;
 	dev->crop_bounds_cap = dev->src_rect;
+	if (dev->bitmap_cap &&
+	    (dev->compose_cap.width != dev->crop_cap.width ||
+	     dev->compose_cap.height != dev->crop_cap.height)) {
+		vfree(dev->bitmap_cap);
+		dev->bitmap_cap = NULL;
+	}
 	dev->compose_cap = dev->crop_cap;
 	if (V4L2_FIELD_HAS_T_OR_B(dev->field_cap))
 		dev->compose_cap.height /= 2;

@@ -886,6 +892,8 @@ int vivid_vid_cap_s_selection(struct file *file, void *fh, struct v4l2_selection
 	struct vivid_dev *dev = video_drvdata(file);
 	struct v4l2_rect *crop = &dev->crop_cap;
 	struct v4l2_rect *compose = &dev->compose_cap;
+	unsigned orig_compose_w = compose->width;
+	unsigned orig_compose_h = compose->height;
 	unsigned factor = V4L2_FIELD_HAS_T_OR_B(dev->field_cap) ? 2 : 1;
 	int ret;
 

@@ -1002,17 +1010,17 @@ int vivid_vid_cap_s_selection(struct file *file, void *fh, struct v4l2_selection
 			s->r.height /= factor;
 		}
 		v4l2_rect_map_inside(&s->r, &dev->fmt_cap_rect);
-		if (dev->bitmap_cap && (compose->width != s->r.width ||
-					compose->height != s->r.height)) {
-			vfree(dev->bitmap_cap);
-			dev->bitmap_cap = NULL;
-		}
 		*compose = s->r;
 		break;
 	default:
 		return -EINVAL;
 	}
 
+	if (dev->bitmap_cap && (compose->width != orig_compose_w ||
+				compose->height != orig_compose_h)) {
+		vfree(dev->bitmap_cap);
+		dev->bitmap_cap = NULL;
+	}
 	tpg_s_crop_compose(&dev->tpg, crop, compose);
 	return 0;
 }

2.35.1

From: Hans Verkuil <[email protected]>

[ Upstream commit 4b6d66a45ed34a15721cb9e11492fa1a24bc83df ]

Add sanity checks to v4l2_valid_dv_timings() to ensure that the provided blanking values are reasonable.

---
 drivers/media/v4l2-core/v4l2-dv-timings.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

index bed6b7db43f5..4353e624a585 100644
@@ -172,6 +172,20 @@ bool v4l2_valid_dv_timings(const struct v4l2_dv_timings *t,
 	    (bt->interlaced && !(caps & V4L2_DV_BT_CAP_INTERLACED)) ||
 	    (!bt->interlaced && !(caps & V4L2_DV_BT_CAP_PROGRESSIVE)))
 		return false;
+
+	/* sanity checks for the blanking timings */
+	if (!bt->interlaced &&
+	    (bt->il_vbackporch || bt->il_vsync || bt->il_vfrontporch))
+		return false;
+	if (bt->hfrontporch > 2 * bt->width ||
+	    bt->hsync > 1024 || bt->hbackporch > 1024)
+		return false;
+	if (bt->vfrontporch > 4096 ||
+	    bt->vsync > 128 || bt->vbackporch > 4096)
+		return false;
+	if (bt->interlaced && (bt->il_vfrontporch > 4096 ||
+	    bt->il_vsync > 128 || bt->il_vbackporch > 4096))
+		return false;
 	return fnc == NULL || fnc(t, fnc_handle);
 }
 EXPORT_SYMBOL_GPL(v4l2_valid_dv_timings);

2.35.1

From: Hans Verkuil <[email protected]>

[ Upstream commit 8da7f0976b9071b528c545008de9d10cc81883b1 ]

If it is a progressive (non-interlaced) format, then ignore the interlaced timing values.

---
 include/uapi/linux/videodev2.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index b773e96b4a28..b8fd2c303ed0 100644
@@ -1392,7 +1392,8 @@ struct v4l2_bt_timings {
 	((bt)->width + V4L2_DV_BT_BLANKING_WIDTH(bt))
 #define V4L2_DV_BT_BLANKING_HEIGHT(bt) \
 	((bt)->vfrontporch + (bt)->vsync + (bt)->vbackporch + \
-	 (bt)->il_vfrontporch + (bt)->il_vsync + (bt)->il_vbackporch)
+	 ((bt)->interlaced ? \
+	  ((bt)->il_vfrontporch + (bt)->il_vsync + (bt)->il_vbackporch) : 0))
 #define V4L2_DV_BT_FRAME_HEIGHT(bt) \
 	((bt)->height + V4L2_DV_BT_BLANKING_HEIGHT(bt))
 

2.35.1

From: Slawomir Laba <[email protected]>

[ Upstream commit 54b5af5a438076082d482cab105b1bd484ab5074 ]

When enabling flow type for RSS hash via ethtool:

ethtool -N $pf rx-flow-hash tcp4|tcp6|udp4|udp6 s|d

the driver would fail to setup this setting on X722 device since it was using the mask on the register dedicated for X710 devices.

Apply a different mask on the register when setting the RSS hash for the X722 device.

When displaying the flow types enabled via ethtool:

ethtool -n $pf rx-flow-hash tcp4|tcp6|udp4|udp6

the driver would print wrong values for X722 device.

Fix this issue by testing masks for X722 device in i40e_get_rss_hash_opts function.

---
 .../net/ethernet/intel/i40e/i40e_ethtool.c    | 31 ++++++++++++++-----
 drivers/net/ethernet/intel/i40e/i40e_type.h   |  4 +++
 2 files changed, 27 insertions(+), 8 deletions(-)

index 21648dab13e0..615558bb545a 100644
@@ -2375,10 +2375,17 @@ static int i40e_get_rss_hash_opts(struct i40e_pf *pf, struct ethtool_rxnfc *cmd)
 
 		if (cmd->flow_type == TCP_V4_FLOW ||
 		    cmd->flow_type == UDP_V4_FLOW) {
-			if (i_set & I40E_L3_SRC_MASK)
-				cmd->data |= RXH_IP_SRC;
-			if (i_set & I40E_L3_DST_MASK)
-				cmd->data |= RXH_IP_DST;
+			if (hw->mac.type == I40E_MAC_X722) {
+				if (i_set & I40E_X722_L3_SRC_MASK)
+					cmd->data |= RXH_IP_SRC;
+				if (i_set & I40E_X722_L3_DST_MASK)
+					cmd->data |= RXH_IP_DST;
+			} else {
+				if (i_set & I40E_L3_SRC_MASK)
+					cmd->data |= RXH_IP_SRC;
+				if (i_set & I40E_L3_DST_MASK)
+					cmd->data |= RXH_IP_DST;
+			}
 		} else if (cmd->flow_type == TCP_V6_FLOW ||
 			  cmd->flow_type == UDP_V6_FLOW) {
 			if (i_set & I40E_L3_V6_SRC_MASK)

@@ -2683,12 +2690,15 @@ static int i40e_get_rxnfc(struct net_device *netdev, struct ethtool_rxnfc *cmd,
 
 /**
  * i40e_get_rss_hash_bits - Read RSS Hash bits from register
+ * @hw: hw structure
  * @nfc: pointer to user request
  * @i_setc bits currently set
  *
  * Returns value of bits to be set per user request
  **/
-static u64 i40e_get_rss_hash_bits(struct ethtool_rxnfc *nfc, u64 i_setc)
+static u64 i40e_get_rss_hash_bits(struct i40e_hw *hw,
+				  struct ethtool_rxnfc *nfc,
+				  u64 i_setc)
 {
 	u64 i_set = i_setc;
 	u64 src_l3 = 0, dst_l3 = 0;

@@ -2707,8 +2717,13 @@ static u64 i40e_get_rss_hash_bits(struct ethtool_rxnfc *nfc, u64 i_setc)
 		dst_l3 = I40E_L3_V6_DST_MASK;
 	} else if (nfc->flow_type == TCP_V4_FLOW ||
 		  nfc->flow_type == UDP_V4_FLOW) {
-		src_l3 = I40E_L3_SRC_MASK;
-		dst_l3 = I40E_L3_DST_MASK;
+		if (hw->mac.type == I40E_MAC_X722) {
+			src_l3 = I40E_X722_L3_SRC_MASK;
+			dst_l3 = I40E_X722_L3_DST_MASK;
+		} else {
+			src_l3 = I40E_L3_SRC_MASK;
+			dst_l3 = I40E_L3_DST_MASK;
+		}
 	} else {
 		/* Any other flow type are not supported here */
 		return i_set;

@@ -2823,7 +2838,7 @@ static int i40e_set_rss_hash_opt(struct i40e_pf *pf, struct ethtool_rxnfc *nfc)
 					       flow_pctype)) |
 			((u64)i40e_read_rx_ctl(hw, I40E_GLQF_HASH_INSET(1,
 					       flow_pctype)) << 32);
-		i_set = i40e_get_rss_hash_bits(nfc, i_setc);
+		i_set = i40e_get_rss_hash_bits(&pf->hw, nfc, i_setc);
 		i40e_write_rx_ctl(hw, I40E_GLQF_HASH_INSET(0, flow_pctype),
 				  (u32)i_set);
 		i40e_write_rx_ctl(hw, I40E_GLQF_HASH_INSET(1, flow_pctype),

index fd4bbdd88b57..8338bd348d26 100644
@@ -1433,6 +1433,10 @@ struct i40e_lldp_variables {
 #define I40E_PFQF_CTL_0_HASHLUTSIZE_512	0x00010000
 
 /* INPUT SET MASK for RSS, flow director, and flexible payload */
+#define I40E_X722_L3_SRC_SHIFT		49
+#define I40E_X722_L3_SRC_MASK		(0x3ULL << I40E_X722_L3_SRC_SHIFT)
+#define I40E_X722_L3_DST_SHIFT		41
+#define I40E_X722_L3_DST_MASK		(0x3ULL << I40E_X722_L3_DST_SHIFT)
 #define I40E_L3_SRC_SHIFT		47
 #define I40E_L3_SRC_MASK		(0x3ULL << I40E_L3_SRC_SHIFT)
 #define I40E_L3_V6_SRC_SHIFT		43

2.35.1

From: Slawomir Laba <[email protected]>

[ Upstream commit 3b32c9932853e11d71f9db012d69e92e4669ba23 ]

Fix setting bits for specific flow_type for GLQF_HASH_INSET register. In previous version all of the bits were set only in hena register, while in inset only one bit was set. In order for this working correctly on all types of cards these bits needs to be set correctly for both hena and inset registers.

---
 .../net/ethernet/intel/i40e/i40e_ethtool.c    | 71 ++++++++++---------
 1 file changed, 38 insertions(+), 33 deletions(-)

index 615558bb545a..0691027e9ce1 100644
@@ -2741,6 +2741,7 @@ static u64 i40e_get_rss_hash_bits(struct i40e_hw *hw,
 	return i_set;
 }
 
+#define FLOW_PCTYPES_SIZE 64
 /**
  * i40e_set_rss_hash_opt - Enable/Disable flow types for RSS hash
  * @pf: pointer to the physical function struct

@@ -2753,9 +2754,11 @@ static int i40e_set_rss_hash_opt(struct i40e_pf *pf, struct ethtool_rxnfc *nfc)
 	struct i40e_hw *hw = &pf->hw;
 	u64 hena = (u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(0)) |
 		   ((u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(1)) << 32);
-	u8 flow_pctype = 0;
+	DECLARE_BITMAP(flow_pctypes, FLOW_PCTYPES_SIZE);
 	u64 i_set, i_setc;
 
+	bitmap_zero(flow_pctypes, FLOW_PCTYPES_SIZE);
+
 	if (pf->flags & I40E_FLAG_MFP_ENABLED) {
 		dev_err(&pf->pdev->dev,
 			"Change of RSS hash input set is not supported when MFP mode is enabled\n");

@@ -2771,36 +2774,35 @@ static int i40e_set_rss_hash_opt(struct i40e_pf *pf, struct ethtool_rxnfc *nfc)
 
 	switch (nfc->flow_type) {
 	case TCP_V4_FLOW:
-		flow_pctype = I40E_FILTER_PCTYPE_NONF_IPV4_TCP;
+		set_bit(I40E_FILTER_PCTYPE_NONF_IPV4_TCP, flow_pctypes);
 		if (pf->hw_features & I40E_HW_MULTIPLE_TCP_UDP_RSS_PCTYPE)
-			hena |=
-			  BIT_ULL(I40E_FILTER_PCTYPE_NONF_IPV4_TCP_SYN_NO_ACK);
+			set_bit(I40E_FILTER_PCTYPE_NONF_IPV4_TCP_SYN_NO_ACK,
+				flow_pctypes);
 		break;
 	case TCP_V6_FLOW:
-		flow_pctype = I40E_FILTER_PCTYPE_NONF_IPV6_TCP;
-		if (pf->hw_features & I40E_HW_MULTIPLE_TCP_UDP_RSS_PCTYPE)
-			hena |=
-			  BIT_ULL(I40E_FILTER_PCTYPE_NONF_IPV4_TCP_SYN_NO_ACK);
+		set_bit(I40E_FILTER_PCTYPE_NONF_IPV6_TCP, flow_pctypes);
 		if (pf->hw_features & I40E_HW_MULTIPLE_TCP_UDP_RSS_PCTYPE)
-			hena |=
-			  BIT_ULL(I40E_FILTER_PCTYPE_NONF_IPV6_TCP_SYN_NO_ACK);
+			set_bit(I40E_FILTER_PCTYPE_NONF_IPV6_TCP_SYN_NO_ACK,
+				flow_pctypes);
 		break;
 	case UDP_V4_FLOW:
-		flow_pctype = I40E_FILTER_PCTYPE_NONF_IPV4_UDP;
-		if (pf->hw_features & I40E_HW_MULTIPLE_TCP_UDP_RSS_PCTYPE)
-			hena |=
-			  BIT_ULL(I40E_FILTER_PCTYPE_NONF_UNICAST_IPV4_UDP) |
-			  BIT_ULL(I40E_FILTER_PCTYPE_NONF_MULTICAST_IPV4_UDP);
-
+		set_bit(I40E_FILTER_PCTYPE_NONF_IPV4_UDP, flow_pctypes);
+		if (pf->hw_features & I40E_HW_MULTIPLE_TCP_UDP_RSS_PCTYPE) {
+			set_bit(I40E_FILTER_PCTYPE_NONF_UNICAST_IPV4_UDP,
+				flow_pctypes);
+			set_bit(I40E_FILTER_PCTYPE_NONF_MULTICAST_IPV4_UDP,
+				flow_pctypes);
+		}
 		hena |= BIT_ULL(I40E_FILTER_PCTYPE_FRAG_IPV4);
 		break;
 	case UDP_V6_FLOW:
-		flow_pctype = I40E_FILTER_PCTYPE_NONF_IPV6_UDP;
-		if (pf->hw_features & I40E_HW_MULTIPLE_TCP_UDP_RSS_PCTYPE)
-			hena |=
-			  BIT_ULL(I40E_FILTER_PCTYPE_NONF_UNICAST_IPV6_UDP) |
-			  BIT_ULL(I40E_FILTER_PCTYPE_NONF_MULTICAST_IPV6_UDP);
-
+		set_bit(I40E_FILTER_PCTYPE_NONF_IPV6_UDP, flow_pctypes);
+		if (pf->hw_features & I40E_HW_MULTIPLE_TCP_UDP_RSS_PCTYPE) {
+			set_bit(I40E_FILTER_PCTYPE_NONF_UNICAST_IPV6_UDP,
+				flow_pctypes);
+			set_bit(I40E_FILTER_PCTYPE_NONF_MULTICAST_IPV6_UDP,
+				flow_pctypes);
+		}
 		hena |= BIT_ULL(I40E_FILTER_PCTYPE_FRAG_IPV6);
 		break;
 	case AH_ESP_V4_FLOW:

@@ -2833,17 +2835,20 @@ static int i40e_set_rss_hash_opt(struct i40e_pf *pf, struct ethtool_rxnfc *nfc)
 		return -EINVAL;
 	}
 
-	if (flow_pctype) {
-		i_setc = (u64)i40e_read_rx_ctl(hw, I40E_GLQF_HASH_INSET(0,
-					       flow_pctype)) |
-			((u64)i40e_read_rx_ctl(hw, I40E_GLQF_HASH_INSET(1,
-					       flow_pctype)) << 32);
-		i_set = i40e_get_rss_hash_bits(&pf->hw, nfc, i_setc);
-		i40e_write_rx_ctl(hw, I40E_GLQF_HASH_INSET(0, flow_pctype),
-				  (u32)i_set);
-		i40e_write_rx_ctl(hw, I40E_GLQF_HASH_INSET(1, flow_pctype),
-				  (u32)(i_set >> 32));
-		hena |= BIT_ULL(flow_pctype);
+	if (bitmap_weight(flow_pctypes, FLOW_PCTYPES_SIZE)) {
+		u8 flow_id;
+
+		for_each_set_bit(flow_id, flow_pctypes, FLOW_PCTYPES_SIZE) {
+			i_setc = (u64)i40e_read_rx_ctl(hw, I40E_GLQF_HASH_INSET(0, flow_id)) |
+				 ((u64)i40e_read_rx_ctl(hw, I40E_GLQF_HASH_INSET(1, flow_id)) << 32);
+			i_set = i40e_get_rss_hash_bits(&pf->hw, nfc, i_setc);
+
+			i40e_write_rx_ctl(hw, I40E_GLQF_HASH_INSET(0, flow_id),
+					  (u32)i_set);
+			i40e_write_rx_ctl(hw, I40E_GLQF_HASH_INSET(1, flow_id),
+					  (u32)(i_set >> 32));
+			hena |= BIT_ULL(flow_id);
+		}
 	}
 
 	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(0), (u32)hena);

2.35.1

From: Yang Yingliang <[email protected]>

[ Upstream commit 5da6d65590a0698199df44d095e54b0ed1708178 ]

pci_disable_device() need be called while module exiting, switch to use pcim_enable(), pci_disable_device() will be called in pcim_release() while unbinding device.

---
 drivers/net/ethernet/micrel/ksz884x.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index b178e59baa2e..8212f24711bb 100644
@@ -6939,7 +6939,7 @@ static int pcidev_init(struct pci_dev *pdev, const struct pci_device_id *id)
 	char banner[sizeof(version)];
 	struct ksz_switch *sw = NULL;
 
-	result = pci_enable_device(pdev);
+	result = pcim_enable_device(pdev);
 	if (result)
 		return result;
 

2.35.1

From: Sudeep Holla <[email protected]>

[ Upstream commit e0c57a5c70c13317238cb19a7ded0eab4a5f7de5 ]

Platforms can provide the information about the availability of each idle states via status flag. Platforms may have to disable one or more idle states for various reasons like broken firmware or other unmet dependencies.

Fix handling of such unavailable/disabled idle states by ignoring them while parsing the states.

---
 drivers/base/power/domain.c | 4 ++++
 1 file changed, 4 insertions(+)

index e811f2414889..a64b093a88cf 100644
@@ -2232,6 +2232,10 @@ static int genpd_iterate_idle_states(struct device_node *dn,
 		np = it.node;
 		if (!of_match_node(idle_state_match, np))
 			continue;
+
+		if (!of_device_is_available(np))
+			continue;
+
 		if (states) {
 			ret = genpd_parse_state(&states[i], np);
 			if (ret) {

2.35.1

From: Yang Yingliang <[email protected]>

[ Upstream commit 4a4c8482e370d697738a78dcd7bf2780832cb712 ]

dev_set_name() in soundbus_add_one() allocates memory for name, it need be freed when of_device_register() fails, call soundbus_dev_put() to give up the reference that hold in device_initialize(), so that it can be freed in kobject_cleanup() when the refcount hit to 0. And other resources are also freed in i2sbus_release_dev(), so it can return 0 directly.

---
 sound/aoa/soundbus/i2sbus/core.c | 4 ++++
 1 file changed, 4 insertions(+)

index 000b58522106..c016df586992 100644
@@ -302,6 +302,10 @@ static int i2sbus_add_dev(struct macio_dev *macio,
 
 	if (soundbus_add_one(&dev->sound)) {
 		printk(KERN_DEBUG "i2sbus: device registration error!\n");
+		if (dev->sound.ofdev.dev.kobj.state_initialized) {
+			soundbus_dev_put(&dev->sound);
+			return 0;
+		}
 		goto err;
 	}
 

2.35.1

From: Takashi Iwai <[email protected]>

[ Upstream commit f1fae475f10a26b7e34da4ff2e2f19b7feb3548e ]

i2sbus_add_dev() is supposed to return the number of probed devices, i.e. either 1 or 0. However, i2sbus_add_dev() has one error handling that returns -ENODEV; this will screw up the accumulation number counted in the caller, i2sbus_probe().

Fix the return value to 0 and add the comment for better understanding for readers.

---
 sound/aoa/soundbus/i2sbus/core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index c016df586992..2811e1f1e2fa 100644
@@ -148,6 +148,7 @@ static int i2sbus_get_and_fixup_rsrc(struct device_node *np, int index,
 	return rc;
 }
 
+/* Returns 1 if added, 0 for otherwise; don't return a negative value! */
 /* FIXME: look at device node refcounting */
 static int i2sbus_add_dev(struct macio_dev *macio,
 			  struct i2sbus_control *control,

@@ -213,7 +214,7 @@ static int i2sbus_add_dev(struct macio_dev *macio,
 	 * either as the second one in that case is just a modem. */
 	if (!ok) {
 		kfree(dev);
-		return -ENODEV;
+		return 0;
 	}
 
 	mutex_init(&dev->lock);

2.35.1

From: Aaron Conole <[email protected]>

[ Upstream commit fd954cc1919e35cb92f78671cab6e42d661945a3 ]

As noted by Paolo Abeni, pr_warn doesn't generate any splat and can still preserve the warning to the user that feature downgrade occurred. We likely cannot introduce other kinds of checks / enforcement here because syzbot can generate different genl versions to the datapath.

---
 net/openvswitch/datapath.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index 8319628ab428..a57a3755611d 100644
@@ -1578,7 +1578,8 @@ static void ovs_dp_reset_user_features(struct sk_buff *skb, struct genl_info *in
 	if (IS_ERR(dp))
 		return;
 
-	WARN(dp->user_features, "Dropping previously announced user features\n");
+	pr_warn("%s: Dropping previously announced user features\n",
+		ovs_dp_name(dp));
 	dp->user_features = 0;
 }
 

2.35.1

From: Yang Yingliang <[email protected]>

[ Upstream commit 0e7ce23a917a9cc83ca3c779fbba836bca3bcf1e ]

If of_device_register() returns error, the of node and the name allocated in dev_set_name() is leaked, call put_device() to give up the reference that was set in device_initialize(), so that of node is put in logical_port_release() and the name is freed in kobject_cleanup().