Kconfig.debug: disable CONFIG_FRAME_WARN for KMSAN by default
KMSAN adds a lot of instrumentation to the code, which results in
increased stack usage (up to 2048 bytes and more in some cases). It's
hard to predict how big the stack frames can be, so we disable the
warnings for KMSAN instead.
Baolin Wang [Mon, 24 Oct 2022 08:34:21 +0000 (16:34 +0800)]
mm: migrate: fix return value if all subpages of THPs are migrated successfully
During THP migration, if THPs are not migrated but they are split and all
subpages are migrated successfully, migrate_pages() will still return the
number of THP pages that were not migrated. This will confuse the callers
of migrate_pages(). For example, the longterm pinning will failed though
all pages are migrated successfully.
Thus we should return 0 to indicate that all pages are migrated in this
case
I just got time to revisit this and found that the root cause is we simply
messed up with the vma check, so that for !PTE_MARKER_UFFD_WP system, we
will allow UFFDIO_REGISTER of MINOR & WP upon shmem as the check was
wrong:
if (vm_flags & VM_UFFD_MINOR)
return is_vm_hugetlb_page(vma) || vma_is_shmem(vma);
Where we'll allow anything to pass on shmem as long as minor mode is
requested.
Axel did it right when introducing minor mode but I messed it up in b1f9e876862d when moving code around. Fix it.
Hugh Dickins [Sat, 22 Oct 2022 07:51:06 +0000 (00:51 -0700)]
mm: prep_compound_tail() clear page->private
Although page allocation always clears page->private in the first page or
head page of an allocation, it has never made a point of clearing
page->private in the tails (though 0 is often what is already there).
But now commit 71e2d666ef85 ("mm/huge_memory: do not clobber swp_entry_t
during THP split") issues a warning when page_tail->private is found to be
non-0 (unless it's swapcache).
We could just delete the warning, but today's consensus appears to want
page->private to be 0, unless there's a good reason for it to be set: so
now clear it in prep_compound_tail() (more general than just for THP; but
not for high order allocation, which makes no pass down the tails).
Rik van Riel [Fri, 21 Oct 2022 23:28:05 +0000 (19:28 -0400)]
mm,madvise,hugetlb: fix unexpected data loss with MADV_DONTNEED on hugetlbfs
A common use case for hugetlbfs is for the application to create
memory pools backed by huge pages, which then get handed over to
some malloc library (eg. jemalloc) for further management.
That malloc library may be doing MADV_DONTNEED calls on memory
that is no longer needed, expecting those calls to happen on
PAGE_SIZE boundaries.
However, currently the MADV_DONTNEED code rounds up any such
requests to HPAGE_PMD_SIZE boundaries. This leads to undesired
outcomes when jemalloc expects a 4kB MADV_DONTNEED, but 2MB of
memory get zeroed out, instead.
Use of pre-built shared libraries means that user code does not
always know the page size of every memory arena in use.
Avoid unexpected data loss with MADV_DONTNEED by rounding up
only to PAGE_SIZE (in do_madvise), and rounding down to huge
page granularity.
That way programs will only get as much memory zeroed out as
they requested.
Waiman Long [Thu, 20 Oct 2022 17:56:19 +0000 (13:56 -0400)]
mm/kmemleak: prevent soft lockup in kmemleak_scan()'s object iteration loops
Commit 6edda04ccc7c ("mm/kmemleak: prevent soft lockup in first object
iteration loop of kmemleak_scan()") adds cond_resched() in the first
object iteration loop of kmemleak_scan(). However, it turns that the 2nd
objection iteration loop can still cause soft lockup to happen in some
cases. So add a cond_resched() call in the 2nd and 3rd loops as well to
prevent that and for completeness.
Phillip Lougher [Thu, 20 Oct 2022 22:36:15 +0000 (23:36 +0100)]
squashfs: fix extending readahead beyond end of file
The readahead code will try to extend readahead to the entire size of the
Squashfs data block.
But, it didn't take into account that the last block at the end of the
file may not be a whole block. In this case, the code would extend
readahead to beyond the end of the file, leaving trailing pages.
Fix this by only requesting the expected number of pages.
Phillip Lougher [Thu, 20 Oct 2022 22:36:14 +0000 (23:36 +0100)]
squashfs: fix read regression introduced in readahead code
Patch series "squashfs: fix some regressions introduced in the readahead
code".
This patchset fixes 3 regressions introduced by the recent readahead code
changes. The first regression is causing "snaps" to randomly fail after a
couple of hours or days, which how the regression came to light.
This patch (of 3):
If a file isn't a whole multiple of the page size, the last page will have
trailing bytes unfilled.
There was a mistake in the readahead code which did this. In particular
it incorrectly assumed that the last page in the readahead page array
(page[nr_pages - 1]) will always contain the last page in the block, which
if we're at file end, will be the page that needs to be zero filled.
But the readahead code may not return the last page in the block, which
means it is unmapped and will be skipped by the decompressors (a temporary
buffer used).
In this case the zero filling code will zero out the wrong page, leading
to data corruption.
Fix this by by extending the "page actor" to return the last page if
present, or NULL if a temporary buffer was used.
Linus Torvalds [Fri, 28 Oct 2022 20:24:57 +0000 (13:24 -0700)]
Merge tag 'rtc-6.1-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux
Pull RTC fixes from Alexandre Belloni:
"Fix wakeup support that broke on multiple platforms"
* tag 'rtc-6.1-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux:
rtc: cmos: fix build on non-ACPI platforms
rtc: cmos: Fix wake alarm breakage
Linus Torvalds [Fri, 28 Oct 2022 20:00:54 +0000 (13:00 -0700)]
Merge tag 'mmc-v6.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc
Pull MMC fixes from Ulf Hansson:
"MMC core:
- Cancel recovery work on cleanup to avoid NULL pointer dereference
- Fix error path in the read/write error recovery path
- Fix kernel panic when remove non-standard SDIO card
- Fix WRITE_ZEROES handling for CQE
MMC host:
- sdhci_am654: Fixup Kconfig dependency for REGMAP_MMIO
- sdhci-esdhc-imx: Avoid warning of misconfigured bus-width
- sdhci-pci: Disable broken HS400 ES mode for ASUS BIOS on Jasper
Lake"
* tag 'mmc-v6.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
mmc: sdhci_am654: 'select', not 'depends' REGMAP_MMIO
mmc: core: Fix WRITE_ZEROES CQE handling
mmc: core: Fix kernel panic when remove non-standard SDIO card
mmc: sdhci-pci-core: Disable ES for ASUS BIOS on Jasper Lake
mmc: sdhci-esdhc-imx: Propagate ESDHC_FLAG_HS400* only on 8bit bus
mmc: queue: Cancel recovery work on cleanup
mmc: block: Remove error check of hw_reset on reset
Linus Torvalds [Fri, 28 Oct 2022 19:20:31 +0000 (12:20 -0700)]
Merge tag 'sound-6.1-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"A collection of small fixes:
- fixes for regressions by the recent ALSA control hash usages
- fixes for UAF with del_timer() at removals in a few drivers
- char signedness fixes
- a few memory leak fixes in error paths
- device-specific fixes / quirks for Intel SOF, AMD, HD-audio,
USB-audio, and various ASoC codecs"
* tag 'sound-6.1-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (50 commits)
ALSA: aoa: Fix I2S device accounting
ALSA: Use del_timer_sync() before freeing timer
ALSA: aoa: i2sbus: fix possible memory leak in i2sbus_add_dev()
ALSA: rme9652: use explicitly signed char
ALSA: au88x0: use explicitly signed char
ALSA: hda/realtek: Add another HP ZBook G9 model quirks
ALSA: usb-audio: Add quirks for M-Audio Fast Track C400/600
ASoC: SOF: Intel: hda-codec: fix possible memory leak in hda_codec_device_init()
ASoC: amd: yc: Add Lenovo Thinkbook 14+ 2022 21D0 to quirks table
ASoC: Intel: Skylake: fix possible memory leak in skl_codec_device_init()
ALSA: ac97: Use snd_ctl_rename() to rename a control
ALSA: ca0106: Use snd_ctl_rename() to rename a control
ALSA: emu10k1: Use snd_ctl_rename() to rename a control
ALSA: hda/realtek: Use snd_ctl_rename() to rename a control
ALSA: usb-audio: Use snd_ctl_rename() to rename a control
ALSA: control: add snd_ctl_rename()
ALSA: ac97: fix possible memory leak in snd_ac97_dev_register()
ASoC: SOF: Intel: pci-tgl: fix ADL-N descriptor
ASoC: qcom: lpass-cpu: Mark HDMI TX parity register as volatile
ASoC: amd: yc: Adding Lenovo ThinkBook 14 Gen 4+ ARA and Lenovo ThinkBook 16 Gen 4+ ARA to the Quirks List
...
Linus Torvalds [Fri, 28 Oct 2022 19:10:43 +0000 (12:10 -0700)]
Merge tag 'drm-fixes-2022-10-28' of git://anongit.freedesktop.org/drm/drm
Pull drm fixes from Dave Airlie:
"Regularly scheduled fixes for drm, live from a Red Hat office for the
first time in a while.
The core has two fixes, one for scheduler leak and one for aperture
uninit read.
Otherwise a single bridge fix, and msm, amdgpu/kfd and i915 have a set
of fixes each.
sched:
- Stop leaking fences when killing a sched entity.
aperture:
- Avoid uninitialized read in aperture_remove_conflicting_pci_device()
bridge:
- Fix HPD on bridge/ps8640.
msm:
- Fix shrinker deadlock
- Fix crash during suspend after unbind
- Fix IRQ lifetime issues
- Fix potential memory corruption with too many bridges
- Fix memory corruption on GPU state capture
amdkfd:
- Fix possible memory leak
- Fix GC 10.x cache info reporting
i915:
- Extend Wa_1607297627 to Alderlake-P
- Keep PCI autosuspend control 'on' by default on all dGPU
- Reset frl trained flag before restarting FRL training"
* tag 'drm-fixes-2022-10-28' of git://anongit.freedesktop.org/drm/drm: (39 commits)
fbdev/core: Avoid uninitialized read in aperture_remove_conflicting_pci_device()
drm/amdgpu: disallow gfxoff until GC IP blocks complete s2idle resume
drm/scheduler: fix fence ref counting
drm/amd/display: Revert logic for plane modifiers
drm/amdkfd: correct the cache info for gfx1036
drm/amdkfd: update gfx1037 Lx cache setting
drm/amdgpu: skip mes self test for gc 11.0.3 in recover
drm/amd: Add IMU fw version to fw version queries
drm/amd/display: Don't return false if no stream
drm/amd/display: Remove wrong pipe control lock
drm/amd/pm: allow gfxoff on gc_11_0_3
drm/amdkfd: Fix memory leak in kfd_mem_dmamap_userptr()
drm/amdgpu: Remove ATC L2 access for MMHUB 2.1.x
drm/i915/dp: Reset frl trained flag before restarting FRL training
drm/i915/dgfx: Keep PCI autosuspend control 'on' by default on all dGPU
drm/i915: Extend Wa_1607297627 to Alderlake-P
drm/amdgpu: Adjust MES polling timeout for sriov
drm/amd/pm: update driver-if header for smu_v13_0_10
drm/amdgpu: fix pstate setting issue
drm/bridge: ps8640: Add back the 50 ms mystery delay after HPD
...
Linus Torvalds [Fri, 28 Oct 2022 16:53:30 +0000 (09:53 -0700)]
Merge tag 'v6.1-p3' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Pull crypto fix from Herbert Xu:
"Fix an alignment crash in x86/polyval"
* tag 'v6.1-p3' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
crypto: x86/polyval - Fix crashes when keys are not 16-byte aligned
Ard Biesheuvel [Thu, 27 Oct 2022 13:52:31 +0000 (15:52 +0200)]
efi: efivars: Fix variable writes with unsupported query_variable_store()
Commit 8a254d90a775 ("efi: efivars: Fix variable writes without
query_variable_store()") addressed an issue that was introduced during
the EFI variable store refactor, where alternative implementations of
the efivars layer that lacked query_variable_store() would no longer
work.
Unfortunately, there is another case to consider here, which was missed:
if the efivars layer is backed by the EFI runtime services as usual, but
the EFI implementation predates the introduction of QueryVariableInfo(),
we will return EFI_UNSUPPORTED, and this is no longer being dealt with
correctly.
So let's fix this, and while at it, clean up the code a bit, by merging
the check_var_size() routines as well as their callers.
ACPI: x86: Add another system to quirk list for forcing StorageD3Enable
commit 018d6711c26e4 ("ACPI: x86: Add a quirk for Dell Inspiron 14 2-in-1
for StorageD3Enable") introduced a quirk to allow a system with ambiguous
use of _ADR 0 to force StorageD3Enable.
Julius Brockmann reports that Inspiron 16 5625 suffers that same symptoms.
Add this other system to the list as well.
Matti Vaittinen [Fri, 28 Oct 2022 05:15:45 +0000 (08:15 +0300)]
MAINTAINERS: Change myself to a maintainer
After some off-list discussion with Marek Vasut and Geert Uytterhoeven
and finally a kx022a driver related discussion with Joe Perches
https://lore.kernel.org/lkml/92c3f72e60bc99bf4a21da259b4d78c1bdca447d[email protected]/
it seems that my status as a reviewer has been wrong. I do look after
the ROHM/Kionix drivers I've authored and currently I am also paid to do
so as is reflected by the 'S: Supported'. According to Joe, the reviewer
entry in MAINTAINERS do not indicate such level of support and having a
reviewer supporting an IC is a contradiction.
Switch undersigned from a reviewer to a maintainer for IC drivers I am
taking care of.
John Garry [Wed, 26 Oct 2022 10:35:13 +0000 (18:35 +0800)]
blk-mq: Properly init requests from blk_mq_alloc_request_hctx()
Function blk_mq_alloc_request_hctx() is missing zeroing/init of rq->bio,
biotail, __sector, and __data_len members, which blk_mq_alloc_request()
has, so duplicate what we do in blk_mq_alloc_request().
Gaosheng Cui [Tue, 25 Oct 2022 13:33:57 +0000 (21:33 +0800)]
capabilities: fix potential memleak on error path from vfs_getxattr_alloc()
In cap_inode_getsecurity(), we will use vfs_getxattr_alloc() to
complete the memory allocation of tmpbuf, if we have completed
the memory allocation of tmpbuf, but failed to call handler->get(...),
there will be a memleak in below logic:
|-- ret = (int)vfs_getxattr_alloc(mnt_userns, ...)
| /* ^^^ alloc for tmpbuf */
|-- value = krealloc(*xattr_value, error + 1, flags)
| /* ^^^ alloc memory */
|-- error = handler->get(handler, ...)
| /* error! */
|-- *xattr_value = value
| /* xattr_value is &tmpbuf (memory leak!) */
So we will try to free(tmpbuf) after vfs_getxattr_alloc() fails to fix it.
Maxim Levitsky [Tue, 25 Oct 2022 12:47:31 +0000 (15:47 +0300)]
KVM: x86: emulator: update the emulation mode after CR0 write
Update the emulation mode when handling writes to CR0, because
toggling CR0.PE switches between Real and Protected Mode, and toggling
CR0.PG when EFER.LME=1 switches between Long and Protected Mode.
This is likely a benign bug because there is no writeback of state,
other than the RIP increment, and when toggling CR0.PE, the CPU has
to execute code from a very low memory address.
Maxim Levitsky [Tue, 25 Oct 2022 12:47:30 +0000 (15:47 +0300)]
KVM: x86: emulator: update the emulation mode after rsm
Update the emulation mode after RSM so that RIP will be correctly
written back, because the RSM instruction can switch the CPU mode from
32 bit (or less) to 64 bit.
This fixes a guest crash in case the #SMI is received while the guest
runs a code from an address > 32 bit.
Some instructions update the cpu execution mode, which needs to update the
emulation mode.
Extract this code, and make assign_eip_far use it.
assign_eip_far now reads CS, instead of getting it via a parameter,
which is ok, because callers always assign CS to the same value
before calling this function.
Maxim Levitsky [Tue, 25 Oct 2022 12:47:28 +0000 (15:47 +0300)]
KVM: x86: emulator: em_sysexit should update ctxt->mode
SYSEXIT is one of the instructions that can change the
processor mode, thus ctxt->mode should be updated after it.
Note that this is likely a benign bug, because the only problematic
mode change is from 32 bit to 64 bit which can lead to truncation of RIP,
and it is not possible to do with sysexit,
since sysexit running in 32 bit mode will be limited to 32 bit version.
KVM: selftests: Mark "guest_saw_irq" as volatile in xen_shinfo_test
Tag "guest_saw_irq" as "volatile" to ensure that the compiler will never
optimize away lookups. Relying on the compiler thinking that the flag
is global and thus might change also works, but it's subtle, less robust,
and looks like a bug at first glance, e.g. risks being "fixed" and
breaking the test.
Make the flag "static" as well since convincing the compiler it's global
is no longer necessary.
Alternatively, the flag could be accessed with {READ,WRITE}_ONCE(), but
literally every access would need the wrappers, and eking out performance
isn't exactly top priority for selftests.
Michal Luczaj [Thu, 13 Oct 2022 21:12:33 +0000 (21:12 +0000)]
KVM: selftests: Add tests in xen_shinfo_test to detect lock races
Tests for races between shinfo_cache (de)activation and hypercall+ioctl()
processing. KVM has had bugs where activating the shared info cache
multiple times and/or with concurrent users results in lock corruption,
NULL pointer dereferences, and other fun.
For the timer injection testcase (#22), re-arm the timer until the IRQ
is successfully injected. If the timer expires while the shared info
is deactivated (invalid), KVM will drop the event.
There are two reasons for memleak in dsa_loop_init().
First, fixed_phy_register() create and register phy_device:
fixed_phy_register()
get_phy_device()
phy_device_create() # freed by phy_device_free()
phy_device_register() # freed by phy_device_remove()
But fixed_phy_unregister() only calls phy_device_remove().
So the memory allocated in phy_device_create() is leaked.
Second, when mdio_driver_register() fail in dsa_loop_init(),
it just returns and there is no cleanup for phydevs.
Fix the problems by catching the error of mdio_driver_register()
in dsa_loop_init(), then calling both fixed_phy_unregister() and
phy_device_free() to release phydevs.
Also add a function for phydevs cleanup to avoid duplacate.
Fixes: 98cd1552ea27 ("net: dsa: Mock-up driver") Signed-off-by: Chen Zhongjin <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Zeng Heng [Thu, 27 Oct 2022 12:45:28 +0000 (20:45 +0800)]
cifs: fix use-after-free caused by invalid pointer `hostname`
`hostname` needs to be set as null-pointer after free in
`cifs_put_tcp_session` function, or when `cifsd` thread attempts
to resolve hostname and reconnect the host, the thread would deref
the invalid pointer.
Here is one of practical backtrace examples as reference:
cifsd
---------------------------
kthread
cifs_demultiplex_thread
cifs_reconnect
reconn_set_ipaddr_from_hostname
--> if (!server->hostname)
--> if (server->hostname[0] == '\0') // !! UAF fault here
CIFS: VFS: cifs_mount failed w/return code = -112
mount error(112): Host is down
BUG: KASAN: use-after-free in reconn_set_ipaddr_from_hostname+0x2ba/0x310
Read of size 1 at addr ffff888108f35380 by task cifsd/480
CPU: 2 PID: 480 Comm: cifsd Not tainted 6.1.0-rc2-00106-gf705792f89dd-dirty #25
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1.1 04/01/2014
Call Trace:
<TASK>
dump_stack_lvl+0x68/0x85
print_report+0x16c/0x4a3
kasan_report+0x95/0x190
reconn_set_ipaddr_from_hostname+0x2ba/0x310
__cifs_reconnect.part.0+0x241/0x800
cifs_reconnect+0x65f/0xb60
cifs_demultiplex_thread+0x1570/0x2570
kthread+0x2c5/0x380
ret_from_fork+0x22/0x30
</TASK>
Allocated by task 477:
kasan_save_stack+0x1e/0x40
kasan_set_track+0x21/0x30
__kasan_kmalloc+0x7e/0x90
__kmalloc_node_track_caller+0x52/0x1b0
kstrdup+0x3b/0x70
cifs_get_tcp_session+0xbc/0x19b0
mount_get_conns+0xa9/0x10c0
cifs_mount+0xdf/0x1970
cifs_smb3_do_mount+0x295/0x1660
smb3_get_tree+0x352/0x5e0
vfs_get_tree+0x8e/0x2e0
path_mount+0xf8c/0x1990
do_mount+0xee/0x110
__x64_sys_mount+0x14b/0x1f0
do_syscall_64+0x3b/0x90
entry_SYSCALL_64_after_hwframe+0x63/0xcd
Freed by task 477:
kasan_save_stack+0x1e/0x40
kasan_set_track+0x21/0x30
kasan_save_free_info+0x2a/0x50
__kasan_slab_free+0x10a/0x190
__kmem_cache_free+0xca/0x3f0
cifs_put_tcp_session+0x30c/0x450
cifs_mount+0xf95/0x1970
cifs_smb3_do_mount+0x295/0x1660
smb3_get_tree+0x352/0x5e0
vfs_get_tree+0x8e/0x2e0
path_mount+0xf8c/0x1990
do_mount+0xee/0x110
__x64_sys_mount+0x14b/0x1f0
do_syscall_64+0x3b/0x90
entry_SYSCALL_64_after_hwframe+0x63/0xcd
The buggy address belongs to the object at ffff888108f35380
which belongs to the cache kmalloc-16 of size 16
The buggy address is located 0 bytes inside of
16-byte region [ffff888108f35380, ffff888108f35390)
The buggy address belongs to the physical page:
page:00000000333f8e58 refcount:1 mapcount:0 mapping:0000000000000000 index:0xffff888108f350e0 pfn:0x108f35
flags: 0x200000000000200(slab|node=0|zone=2)
raw: 02000000000002000000000000000000dead000000000122ffff8881000423c0
raw: ffff888108f350e0000000008080007a00000001ffffffff0000000000000000
page dumped because: kasan: bad access detected
Memory state around the buggy address: ffff888108f35280: fa fb fc fc fa fb fc fc fa fb fc fc fa fb fc fc ffff888108f35300: fa fb fc fc fa fb fc fc fa fb fc fc fa fb fc fc
>ffff888108f35380: fa fb fc fc fa fb fc fc fa fb fc fc fa fb fc fc
^ ffff888108f35400: fa fb fc fc fc fc fc fc fc fc fc fc fc fc fc fc ffff888108f35480: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
Fixes: 7be3248f3139 ("cifs: To match file servers, make sure the server hostname matches") Signed-off-by: Zeng Heng <[email protected]> Reviewed-by: Paulo Alcantara (SUSE) <[email protected]> Signed-off-by: Steve French <[email protected]>
Theodore Ts'o [Thu, 27 Oct 2022 20:04:36 +0000 (16:04 -0400)]
ext4: update the backup superblock's at the end of the online resize
When expanding a file system using online resize, various fields in
the superblock (e.g., s_blocks_count, s_inodes_count, etc.) change.
To update the backup superblocks, the online resize uses the function
update_backups() in fs/ext4/resize.c. This function was not updating
the checksum field in the backup superblocks. This wasn't a big deal
previously, because e2fsck didn't care about the checksum field in the
backup superblock. (And indeed, update_backups() goes all the way
back to the ext3 days, well before we had support for metadata
checksums.)
However, there is an alternate, more general way of updating
superblock fields, ext4_update_primary_sb() in fs/ext4/ioctl.c. This
function does check the checksum of the backup superblock, and if it
doesn't match will mark the file system as corrupted. That was
clearly not the intent, so avoid to aborting the resize when a bad
superblock is found.
In addition, teach update_backups() to properly update the checksum in
the backup superblocks. We will eventually want to unify
updapte_backups() with the infrasture in ext4_update_primary_sb(), but
that's for another day.
Note: The problem has been around for a while; it just didn't really
matter until ext4_update_primary_sb() was added by commit bbc605cdb1e1
("ext4: implement support for get/set fs label"). And it became
trivially easy to reproduce after commit 827891a38acc ("ext4: update
the s_overhead_clusters in the backup sb's when resizing") in v6.0.
Cc: [email protected] # 5.17+ Fixes: bbc605cdb1e1 ("ext4: implement support for get/set fs label") Signed-off-by: Theodore Ts'o <[email protected]>
Dave Airlie [Fri, 28 Oct 2022 02:58:33 +0000 (12:58 +1000)]
Merge tag 'drm-misc-fixes-2022-10-27' of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes
drm-misc-fixes for v6.1-rc3:
- Fix HPD on bridge/ps8640.
- Stop leaking fences when killing a sched entity.
- Avoid uninitialized read in aperture_remove_conflicting_pci_device()
Dave Airlie [Fri, 28 Oct 2022 02:57:18 +0000 (12:57 +1000)]
Merge tag 'drm-intel-fixes-2022-10-27-1' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes
- Extend Wa_1607297627 to Alderlake-P (José Roberto de Souza)
- Keep PCI autosuspend control 'on' by default on all dGPU (Anshuman Gupta)
- Reset frl trained flag before restarting FRL training (Ankit Nautiyal)
Taniya Das [Wed, 19 Oct 2022 06:05:35 +0000 (11:35 +0530)]
clk: qcom: Update the force mem core bit for GPU clocks
There are few GPU clocks which are powering up the memories
and thus enable the FORCE_MEM_PERIPH always for these clocks
to force the periph_on signal to remain active during halt
state of the clock.
Maxime Ripard [Tue, 18 Oct 2022 13:52:58 +0000 (15:52 +0200)]
clk: Initialize max_rate in struct clk_rate_request
Since commit b46fd8dbe8ad ("clk: Zero the clk_rate_request structure"),
the clk_core_init_rate_req() function clears the struct clk_rate_request
passed as argument.
However, the default value for max_rate isn't 0 but ULONG_MAX, and we
end up creating a clk_rate_request instance where the maximum rate is 0.
Maxime Ripard [Tue, 18 Oct 2022 13:52:57 +0000 (15:52 +0200)]
clk: Initialize the clk_rate_request even if clk_core is NULL
Since commit c35e84b09776 ("clk: Introduce clk_hw_init_rate_request()"),
users that used to initialize their clk_rate_request by initializing
their local structure now rely on clk_hw_init_rate_request().
This function is backed by clk_core_init_rate_req(), which will skip the
initialization if either the pointer to struct clk_core or to struct
clk_rate_request are NULL.
However, the core->parent pointer might be NULL because the clock is
orphan, and we will thus end up with our local struct clk_rate_request
left untouched.
And since clk_hw_init_rate_request() doesn't return an error, we will
then call a determine_rate variant with that unitialized structure.
In order to avoid this, let's clear our clk_rate_request if the pointer
to it is valid but the pointer to struct clk_core isn't.
Maxime Ripard [Tue, 18 Oct 2022 13:52:56 +0000 (15:52 +0200)]
clk: Remove WARN_ON NULL parent in clk_core_init_rate_req()
If a clock has CLK_SET_RATE_PARENT, but core->parent is NULL (most
likely because it's orphan), callers of clk_core_init_rate_req() will
blindly call this function leading to a very verbose warning.
Since it's a fairly common situation, let's just remove the WARN_ON but
keep the check that prevents us from dereferencing the pointer.
Interestingly, it fixes a regression on the Mediatek MT8195 where the
GPU would stall during a clk_set_rate for its main clock. We couldn't
come up with a proper explanation since the condition is essentially the
same.
It was then assumed that it could be timing related since printing the
warning stacktrace takes a while, but we couldn't replicate the failure
by using fairly large (10ms) mdelays.
Andrew Jones [Fri, 14 Oct 2022 15:58:44 +0000 (17:58 +0200)]
RISC-V: Fix /proc/cpuinfo cpumask warning
Commit 78e5a3399421 ("cpumask: fix checking valid cpu range") has
started issuing warnings[*] when cpu indices equal to nr_cpu_ids - 1
are passed to cpumask_next* functions. seq_read_iter() and cpuinfo's
start and next seq operations implement a pattern like
n = cpumask_next(n - 1, mask);
show(n);
while (1) {
++n;
n = cpumask_next(n - 1, mask);
if (n >= nr_cpu_ids)
break;
show(n);
}
which will issue the warning when reading /proc/cpuinfo. Ensure no
warning is generated by validating the cpu index before calling
cpumask_next().
[*] Warnings will only appear with DEBUG_PER_CPU_MAPS enabled.
This came up due to a report from Kevin @ kernel-ci, who had been
running a mixed configuration of GNU binutils and clang. Their compiler
was relatively recent & supports Zicbom but binutils @ 2.35.2 did not.
Our current checks for extension support only cover the compiler, but it
appears to me that we need to check both the compiler & linker support
in case of "pot-luck" configurations that mix different versions of
LD,AS,CC etc.
Linker support does not seem possible to actually check, since the ISA
string is emitted into the object files - so I put in version checks for
that. The checks have gotten a bit ugly since 32 & 64 bit support need
to be checked independently but ahh well.
As I was going, I fell into the trap of there being duplicated checks
for CC support in both the Makefile and Kconfig, so as part of renaming
the Kconfig symbol to TOOLCHAIN_HAS_FOO, I dropped the extra checks in
the Makefile. This has the added advantage of the TOOLCHAIN_HAS_FOO
symbol for Zihintpause appearing in .config.
I pushed out a version of this that specificly checked for assember
support for LKP to test & it looked /okay/ - but I did some more testing
today and realised that this is redudant & have since dropped the as
check.
I tested locally with a fair few different combinations, to try and
cover each of AS, LD, CC missing support for the extension.
* b4-shazam-merge:
riscv: fix detection of toolchain Zihintpause support
riscv: fix detection of toolchain Zicbom support
Conor Dooley [Thu, 6 Oct 2022 17:35:21 +0000 (18:35 +0100)]
riscv: fix detection of toolchain Zihintpause support
It is not sufficient to check if a toolchain supports a particular
extension without checking if the linker supports that extension
too. For example, Clang 15 supports Zihintpause but GNU bintutils
2.35.2 does not, leading build errors like so:
riscv64-linux-gnu-ld: -march=rv64i2p0_m2p0_a2p0_c2p0_zihintpause2p0: Invalid or unknown z ISA extension: 'zihintpause'
Add a TOOLCHAIN_HAS_ZIHINTPAUSE which checks if each of the compiler,
assembler and linker support the extension. Replace the ifdef in the
vdso with one depending on this new symbol.
Conor Dooley [Thu, 6 Oct 2022 17:35:20 +0000 (18:35 +0100)]
riscv: fix detection of toolchain Zicbom support
It is not sufficient to check if a toolchain supports a particular
extension without checking if the linker supports that extension too.
For example, Clang 15 supports Zicbom but GNU bintutils 2.35.2 does
not, leading build errors like so:
riscv64-linux-gnu-ld: -march=rv64i2p0_m2p0_a2p0_c2p0_zicbom1p0_zihintpause2p0: Invalid or unknown z ISA extension: 'zicbom'
Convert CC_HAS_ZICBOM to TOOLCHAIN_HAS_ZICBOM & check if the linker
also supports Zicbom.
Qinglin Pan [Sun, 9 Oct 2022 08:30:50 +0000 (16:30 +0800)]
riscv: mm: add missing memcpy in kasan_init
Hi Atish,
It seems that the panic is due to the missing memcpy during kasan_init.
Could you please check whether this patch is helpful?
When doing kasan_populate, the new allocated base_pud/base_p4d should
contain kasan_early_shadow_{pud, p4d}'s content. Add the missing memcpy
to avoid page fault when read/write kasan shadow region.
Tested on:
- qemu with sv57 and CONFIG_KASAN on.
- qemu with sv48 and CONFIG_KASAN on.
Linus Torvalds [Thu, 27 Oct 2022 20:36:59 +0000 (13:36 -0700)]
Merge tag 'net-6.1-rc3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Jakub Kicinski:
"Including fixes from 802.15.4 (Zigbee et al).
Current release - regressions:
- ipa: fix bugs in the register conversion for IPA v3.1 and v3.5.1
Current release - new code bugs:
- mptcp: fix abba deadlock on fastopen
- eth: stmmac: rk3588: allow multiple gmac controllers in one system
Previous releases - regressions:
- ip: rework the fix for dflt addr selection for connected nexthop
- net: couple more fixes for misinterpreting bits in struct page
after the signature was added
Previous releases - always broken:
- ipv6: ensure sane device mtu in tunnels
- openvswitch: switch from WARN to pr_warn on a user-triggerable path
- ethtool: eeprom: fix null-deref on genl_info in dump
- ieee802154: more return code fixes for corner cases in
dgram_sendmsg
- mac802154: fix link-quality-indicator recording
- eth: mlx5: fixes for IPsec, PTP timestamps, OvS and conntrack
offload
- eth: fec: limit register access on i.MX6UL
- eth: bcm4908_enet: update TX stats after actual transmission
- can: rcar_canfd: improve IRQ handling for RZ/G2L
Misc:
- genetlink: piggy back on the newly added resv_op_start to enforce
more sanity checks on new commands"
* tag 'net-6.1-rc3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (57 commits)
net: enetc: survive memory pressure without crashing
kcm: do not sense pfmemalloc status in kcm_sendpage()
net: do not sense pfmemalloc status in skb_append_pagefrags()
net/mlx5e: Fix macsec sci endianness at rx sa update
net/mlx5e: Fix wrong bitwise comparison usage in macsec_fs_rx_add_rule function
net/mlx5e: Fix macsec rx security association (SA) update/delete
net/mlx5e: Fix macsec coverity issue at rx sa update
net/mlx5: Fix crash during sync firmware reset
net/mlx5: Update fw fatal reporter state on PCI handlers successful recover
net/mlx5e: TC, Fix cloned flow attr instance dests are not zeroed
net/mlx5e: TC, Reject forwarding from internal port to internal port
net/mlx5: Fix possible use-after-free in async command interface
net/mlx5: ASO, Create the ASO SQ with the correct timestamp format
net/mlx5e: Update restore chain id for slow path packets
net/mlx5e: Extend SKB room check to include PTP-SQ
net/mlx5: DR, Fix matcher disconnect error flow
net/mlx5: Wait for firmware to enable CRS before pci_restore_state
net/mlx5e: Do not increment ESN when updating IPsec ESN state
netdevsim: remove dir in nsim_dev_debugfs_init() when creating ports dir failed
netdevsim: fix memory leak in nsim_drv_probe() when nsim_dev_resources_register() failed
...
Linus Torvalds [Thu, 27 Oct 2022 20:16:36 +0000 (13:16 -0700)]
Merge tag 'execve-v6.1-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux
Pull execve fixes from Kees Cook:
- Fix an ancient signal action copy race (Bernd Edlinger)
- Fix a memory leak in ELF loader, when under memory pressure (Li
Zetao)
* tag 'execve-v6.1-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
fs/binfmt_elf: Fix memory leak in load_elf_binary()
exec: Copy oldsighand->action under spin-lock
Fixes: abf79bb341bf ("NFS: Add a slot table to struct nfs_client for NFSv4.0 transport blocking") Signed-off-by: Zhang Xiaoxu <[email protected]> Signed-off-by: Anna Schumaker <[email protected]>
NFSv4.2: Fixup CLONE dest file size for zero-length count
When holding a delegation, the NFS client optimizes away setting the
attributes of a file from the GETATTR in the compound after CLONE, and for
a zero-length CLONE we will end up setting the inode's size to zero in
nfs42_copy_dest_done(). Handle this case by computing the resulting count
from the server's reported size after CLONE's GETATTR.
Suggested-by: Trond Myklebust <[email protected]> Signed-off-by: Benjamin Coddington <[email protected]> Fixes: 94d202d5ca39 ("NFSv42: Copy offload should update the file size when appropriate") Signed-off-by: Anna Schumaker <[email protected]>
Chuck Lever [Sat, 8 Oct 2022 18:58:29 +0000 (14:58 -0400)]
SUNRPC: Fix crasher in gss_unwrap_resp_integ()
If a zero length is passed to kmalloc() it returns 0x10, which is
not a valid address. gss_unwrap_resp_integ() subsequently crashes
when it attempts to dereference that pointer.
NFSv4: Retry LOCK on OLD_STATEID during delegation return
There's a small window where a LOCK sent during a delegation return can
race with another OPEN on client, but the open stateid has not yet been
updated. In this case, the client doesn't handle the OLD_STATEID error
from the server and will lose this lock, emitting:
"NFS: nfs4_handle_delegation_recall_error: unhandled error -10024".
Fix this by sending the task through the nfs4 error handling in
nfs4_lock_done() when we may have to reconcile our stateid with what the
server believes it to be. For this case, the result is a retry of the
LOCK operation with the updated stateid.
Zhang Xiaoxu [Thu, 20 Oct 2022 03:42:17 +0000 (11:42 +0800)]
SUNRPC: Fix null-ptr-deref when xps sysfs alloc failed
There is a null-ptr-deref when xps sysfs alloc failed:
BUG: KASAN: null-ptr-deref in sysfs_do_create_link_sd+0x40/0xd0
Read of size 8 at addr 0000000000000030 by task gssproxy/457
When the xprt_switch sysfs alloc failed, should not add xprt and
switch sysfs to it, otherwise, maybe null-ptr-deref; also initialize
the 'xps_sysfs' to NULL to avoid oops when destroy it.
Fixes: 2a338a543163 ("sunrpc: add a symlink from rpc-client directory to the xprt_switch") Fixes: d408ebe04ac5 ("sunrpc: add add sysfs directory per xprt under each xprt_switch") Fixes: baea99445dd4 ("sunrpc: add xprt_switch direcotry to sunrpc's sysfs") Signed-off-by: Zhang Xiaoxu <[email protected]> Signed-off-by: Anna Schumaker <[email protected]>
Trond Myklebust [Sun, 16 Oct 2022 18:44:33 +0000 (14:44 -0400)]
NFSv4.1: We must always send RECLAIM_COMPLETE after a reboot
Currently, we are only guaranteed to send RECLAIM_COMPLETE if we have
open state to recover. Fix the client to always send RECLAIM_COMPLETE
after setting up the lease.
Trond Myklebust [Sun, 16 Oct 2022 18:44:31 +0000 (14:44 -0400)]
NFSv4: Fix a potential state reclaim deadlock
If the server reboots while we are engaged in a delegation return, and
there is a pNFS layout with return-on-close set, then the current code
can end up deadlocking in pnfs_roc() when nfs_inode_set_delegation()
tries to return the old delegation.
Now that delegreturn actually uses its own copy of the stateid, it
should be safe to just always update the delegation stateid in place.
Fixes: 078000d02d57 ("pNFS: We want return-on-close to complete when evicting the inode") Signed-off-by: Trond Myklebust <[email protected]> Signed-off-by: Anna Schumaker <[email protected]>
Kees Cook [Mon, 17 Oct 2022 04:36:50 +0000 (21:36 -0700)]
NFS: Avoid memcpy() run-time warning for struct sockaddr overflows
The 'nfs_server' and 'mount_server' structures include a union of
'struct sockaddr' (with the older 16 bytes max address size) and
'struct sockaddr_storage' which is large enough to hold all the
supported sa_family types (128 bytes max size). The runtime memcpy()
buffer overflow checker is seeing attempts to write beyond the 16
bytes as an overflow, but the actual expected size is that of 'struct
sockaddr_storage'. Plumb the use of 'struct sockaddr_storage' more
completely through-out NFS, which results in adjusting the memcpy()
buffers to the correct union members. Avoids this false positive run-time
warning under CONFIG_FORTIFY_SOURCE:
memcpy: detected field-spanning write (size 28) of single field "&ctx->nfs_server.address" at fs/nfs/namespace.c:178 (size 16)
Linus Torvalds [Thu, 27 Oct 2022 19:31:57 +0000 (12:31 -0700)]
Merge tag 'hardening-v6.1-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux
Pull hardening fixes from Kees Cook:
- Fix older Clang vs recent overflow KUnit test additions (Nick
Desaulniers, Kees Cook)
- Fix kern-doc visibility for overflow helpers (Kees Cook)
* tag 'hardening-v6.1-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
overflow: Refactor test skips for Clang-specific issues
overflow: disable failing tests for older clang versions
overflow: Fix kern-doc markup for functions
Linus Torvalds [Thu, 27 Oct 2022 19:21:57 +0000 (12:21 -0700)]
Merge tag 'media/v6.1-3' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media fixes from Mauro Carvalho Chehab:
"A bunch of patches addressing issues in the vivid driver and adding
new checks in V4L2 to validate the input parameters from some ioctls"
* tag 'media/v6.1-3' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
media: vivid.rst: loop_video is set on the capture devnode
media: vivid: set num_in/outputs to 0 if not supported
media: vivid: drop GFP_DMA32
media: vivid: fix control handler mutex deadlock
media: videodev2.h: V4L2_DV_BT_BLANKING_HEIGHT should check 'interlaced'
media: v4l2-dv-timings: add sanity checks for blanking values
media: vivid: dev->bitmap_cap wasn't freed in all cases
media: vivid: s_fbuf: add more sanity checks
Ryan Roberts [Thu, 27 Oct 2022 12:09:45 +0000 (13:09 +0100)]
KVM: arm64: Fix bad dereference on MTE-enabled systems
enter_exception64() performs an MTE check, which involves dereferencing
vcpu->kvm. While vcpu has already been fixed up to be a HYP VA pointer,
kvm is still a pointer in the kernel VA space.
This only affects nVHE configurations with MTE enabled, as in other
cases, the pointer is either valid (VHE) or not dereferenced (!MTE).
Fix this by first converting kvm to a HYP VA pointer.
Vladimir Oltean [Thu, 27 Oct 2022 18:29:25 +0000 (21:29 +0300)]
net: enetc: survive memory pressure without crashing
Under memory pressure, enetc_refill_rx_ring() may fail, and when called
during the enetc_open() -> enetc_setup_rxbdr() procedure, this is not
checked for.
An extreme case of memory pressure will result in exactly zero buffers
being allocated for the RX ring, and in such a case it is expected that
hardware drops all RX packets due to lack of buffers.
This does not happen, because the reset-default value of the consumer
and produces index is 0, and this makes the ENETC think that all buffers
have been initialized and that it owns them (when in reality none were).
The hardware guide explains this best:
| Configure the receive ring producer index register RBaPIR with a value
| of 0. The producer index is initially configured by software but owned
| by hardware after the ring has been enabled. Hardware increments the
| index when a frame is received which may consume one or more BDs.
| Hardware is not allowed to increment the producer index to match the
| consumer index since it is used to indicate an empty condition. The ring
| can hold at most RBLENR[LENGTH]-1 received BDs.
|
| Configure the receive ring consumer index register RBaCIR. The
| consumer index is owned by software and updated during operation of the
| of the BD ring by software, to indicate that any receive data occupied
| in the BD has been processed and it has been prepared for new data.
| - If consumer index and producer index are initialized to the same
| value, it indicates that all BDs in the ring have been prepared and
| hardware owns all of the entries.
| - If consumer index is initialized to producer index plus N, it would
| indicate N BDs have been prepared. Note that hardware cannot start if
| only a single buffer is prepared due to the restrictions described in
| (2).
| - Software may write consumer index to match producer index anytime
| while the ring is operational to indicate all received BDs prior have
| been processed and new BDs prepared for hardware.
Normally, the value of rx_ring->rcir (consumer index) is brought in sync
with the rx_ring->next_to_use software index, but this only happens if
page allocation ever succeeded.
When PI==CI==0, the hardware appears to receive frames and write them to
DMA address 0x0 (?!), then set the READY bit in the BD.
The enetc_clean_rx_ring() function (and its XDP derivative) is naturally
not prepared to handle such a condition. It will attempt to process
those frames using the rx_swbd structure associated with index i of the
RX ring, but that structure is not fully initialized (enetc_new_page()
does all of that). So what happens next is undefined behavior.
To operate using no buffer, we must initialize the CI to PI + 1, which
will block the hardware from advancing the CI any further, and drop
everything.
The issue was seen while adding support for zero-copy AF_XDP sockets,
where buffer memory comes from user space, which can even decide to
supply no buffers at all (example: "xdpsock --txonly"). However, the bug
is present also with the network stack code, even though it would take a
very determined person to trigger a page allocation failure at the
perfect time (a series of ifup/ifdown under memory pressure should
eventually reproduce it given enough retries).
Eric Dumazet [Thu, 27 Oct 2022 04:03:46 +0000 (04:03 +0000)]
net: do not sense pfmemalloc status in skb_append_pagefrags()
skb_append_pagefrags() is used by af_unix and udp sendpage()
implementation so far.
In commit 326140063946 ("tcp: TX zerocopy should not sense
pfmemalloc status") we explained why we should not sense
pfmemalloc status for pages owned by user space.
We should also use skb_fill_page_desc_noacc()
in skb_append_pagefrags() to avoid following KCSAN report:
BUG: KCSAN: data-race in lru_add_fn / skb_append_pagefrags
value changed: 0x0000000000000000 -> 0xffffea00058fc188
Reported by Kernel Concurrency Sanitizer on:
CPU: 1 PID: 17325 Comm: syz-executor.0 Not tainted 6.1.0-rc1-syzkaller-00158-g440b7895c990-dirty #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/11/2022
Raed Salem [Wed, 26 Oct 2022 13:51:53 +0000 (14:51 +0100)]
net/mlx5e: Fix macsec sci endianness at rx sa update
The cited commit at rx sa update operation passes the sci object
attribute, in the wrong endianness and not as expected by the HW
effectively create malformed hw sa context in case of update rx sa
consequently, HW produces unexpected MACsec packets which uses this
sa.
Fix by passing sci to create macsec object with the correct endianness,
while at it add __force u64 to prevent sparse check error of type
"sparse: error: incorrect type in assignment".
Raed Salem [Wed, 26 Oct 2022 13:51:52 +0000 (14:51 +0100)]
net/mlx5e: Fix wrong bitwise comparison usage in macsec_fs_rx_add_rule function
The cited commit produces a sparse check error of type
"sparse: error: restricted __be64 degrades to integer". The
offending line wrongly did a bitwise operation between two different
storage types one of 64 bit when the other smaller side is 16 bit
which caused the above sparse error, furthermore bitwise operation
usage here is wrong in the first place as the constant MACSEC_PORT_ES
is not a bitwise field.
Fix by using the right mask to get the lower 16 bit if the sci number,
and use comparison operator '==' instead of bitwise '&' operator.
Raed Salem [Wed, 26 Oct 2022 13:51:51 +0000 (14:51 +0100)]
net/mlx5e: Fix macsec rx security association (SA) update/delete
The cited commit adds the support for update/delete MACsec Rx SA,
naturally, these operations need to check if the SA in question exists
to update/delete the SA and return error code otherwise, however they
do just the opposite i.e. return with error if the SA exists
Fix by change the check to return error in case the SA in question does
not exist, adjust error message and code accordingly.
Raed Salem [Wed, 26 Oct 2022 13:51:50 +0000 (14:51 +0100)]
net/mlx5e: Fix macsec coverity issue at rx sa update
The cited commit at update rx sa operation passes object attributes
to MACsec object create function without initializing/setting all
attributes fields leaving some of them with garbage values, therefore
violating the implicit assumption at create object function, which
assumes that all input object attributes fields are set.
Fix by initializing the object attributes struct to zero, thus leaving
unset fields with the legal zero value.
When setting Bluefield to DPU NIC mode using mlxconfig tool + sync
firmware reset flow, we run into scenario where the host was not
eswitch manager at the time of mlx5 driver load but becomes eswitch manager
after the sync firmware reset flow. This results in null pointer
access of mpfs structure during mac filter add. This change prevents null
pointer access but mpfs table entries will not be added.
Roy Novich [Wed, 26 Oct 2022 13:51:48 +0000 (14:51 +0100)]
net/mlx5: Update fw fatal reporter state on PCI handlers successful recover
Update devlink health fw fatal reporter state to "healthy" is needed by
strictly calling devlink_health_reporter_state_update() after recovery
was done by PCI error handler. This is needed when fw_fatal reporter was
triggered due to PCI error. Poll health is called and set reporter state
to error. Health recovery failed (since EEH didn't re-enable the PCI).
PCI handlers keep on recover flow and succeed later without devlink
acknowledgment. Fix this by adding devlink state update at the end of
the PCI handler recovery process.
Roi Dayan [Wed, 26 Oct 2022 13:51:47 +0000 (14:51 +0100)]
net/mlx5e: TC, Fix cloned flow attr instance dests are not zeroed
On multi table split the driver creates a new attr instance with
data being copied from prev attr instance zeroing action flags.
Also need to reset dests properties to avoid incorrect dests per attr.
Ariel Levkovich [Wed, 26 Oct 2022 13:51:46 +0000 (14:51 +0100)]
net/mlx5e: TC, Reject forwarding from internal port to internal port
Reject TC rules that forward from internal port to internal port
as it is not supported.
This include rules that are explicitly have internal port as
the filter device as well as rules that apply on tunnel interfaces
as the route device for the tunnel interface can be an internal
port.
Tariq Toukan [Wed, 26 Oct 2022 13:51:45 +0000 (14:51 +0100)]
net/mlx5: Fix possible use-after-free in async command interface
mlx5_cmd_cleanup_async_ctx should return only after all its callback
handlers were completed. Before this patch, the below race between
mlx5_cmd_cleanup_async_ctx and mlx5_cmd_exec_cb_handler was possible and
lead to a use-after-free:
1. mlx5_cmd_cleanup_async_ctx is called while num_inflight is 2 (i.e.
elevated by 1, a single inflight callback).
2. mlx5_cmd_cleanup_async_ctx decreases num_inflight to 1.
3. mlx5_cmd_exec_cb_handler is called, decreases num_inflight to 0 and
is about to call wake_up().
4. mlx5_cmd_cleanup_async_ctx calls wait_event, which returns
immediately as the condition (num_inflight == 0) holds.
5. mlx5_cmd_cleanup_async_ctx returns.
6. The caller of mlx5_cmd_cleanup_async_ctx frees the mlx5_async_ctx
object.
7. mlx5_cmd_exec_cb_handler goes on and calls wake_up() on the freed
object.
Fix it by syncing using a completion object. Mark it completed when
num_inflight reaches 0.
Trace:
BUG: KASAN: use-after-free in do_raw_spin_lock+0x23d/0x270
Read of size 4 at addr ffff888139cd12f4 by task swapper/5/0
Saeed Mahameed [Wed, 26 Oct 2022 13:51:44 +0000 (14:51 +0100)]
net/mlx5: ASO, Create the ASO SQ with the correct timestamp format
mlx5 SQs must select the timestamp format explicitly according to the
active clock mode, select the current active timestamp mode so ASO SQ create
will succeed.
This fixes the following error prints when trying to create ipsec ASO SQ
while the timestamp format is real time mode.
mlx5_cmd_out_err:778:(pid 34874): CREATE_SQ(0x904) op_mod(0x0) failed, status bad parameter(0x3), syndrome (0xd61c0b), err(-22)
mlx5_aso_create_sq:285:(pid 34874): Failed to open aso wq sq, err=-22
mlx5e_ipsec_init:436:(pid 34874): IPSec initialization failed, -22
Paul Blakey [Wed, 26 Oct 2022 13:51:43 +0000 (14:51 +0100)]
net/mlx5e: Update restore chain id for slow path packets
Currently encap slow path rules just forward to software without
setting the chain id miss register, so driver doesn't restore
the chain, and packets hitting this rule will restart from tc chain
0 instead of continuing to the chain the encap rule was on.
Fix this by setting the chain id miss register to the chain id mapping.
Aya Levin [Wed, 26 Oct 2022 13:51:42 +0000 (14:51 +0100)]
net/mlx5e: Extend SKB room check to include PTP-SQ
When tx_port_ts is set, the driver diverts all UPD traffic over PTP port
to a dedicated PTP-SQ. The SKBs are cached until the wire-CQE arrives.
When the packet size is greater then MTU, the firmware might drop it and
the packet won't be transmitted to the wire, hence the wire-CQE won't
reach the driver. In this case the SKBs are accumulated in the SKB fifo.
Add room check to consider the PTP-SQ SKB fifo, when the SKB fifo is
full, driver stops the queue resulting in a TX timeout. Devlink
TX-reporter can recover from it.
Rongwei Liu [Wed, 26 Oct 2022 13:51:41 +0000 (14:51 +0100)]
net/mlx5: DR, Fix matcher disconnect error flow
When 2nd flow rules arrives, it will merge together with the
1st one if matcher criteria is the same.
If merge fails, driver will rollback the merge contents, and
reject the 2nd rule. At rollback stage, matcher can't be
disconnected unconditionally, otherise the 1st rule can't be
hit anymore.
Add logic to check if the matcher should be disconnected or not.
Moshe Shemesh [Wed, 26 Oct 2022 13:51:40 +0000 (14:51 +0100)]
net/mlx5: Wait for firmware to enable CRS before pci_restore_state
After firmware reset driver should verify firmware already enabled CRS
and became responsive to pci config cycles before restoring pci state.
Fix that by waiting till device_id is readable through PCI again.
Hyong Youb Kim [Wed, 26 Oct 2022 13:51:39 +0000 (14:51 +0100)]
net/mlx5e: Do not increment ESN when updating IPsec ESN state
An offloaded SA stops receiving after about 2^32 + replay_window
packets. For example, when SA reaches <seq-hi 0x1, seq 0x2c>, all
subsequent packets get dropped with SA-icv-failure (integrity_failed).
To reproduce the bug:
- ConnectX-6 Dx with crypto enabled (FW 22.30.1004)
- ipsec.conf:
nic-offload = yes
replay-window = 32
esn = yes
salifetime=24h
- Run netperf for a long time to send more than 2^32 packets
netperf -H <device-under-test> -t TCP_STREAM -l 20000
When 2^32 + replay_window packets are received, the replay window
moves from the 2nd half of subspace (overlap=1) to the 1st half
(overlap=0). The driver then updates the 'esn' value in NIC
(i.e. seq_hi) as follows.
seq_hi = xfrm_replay_seqhi(seq_bottom)
new esn in NIC = seq_hi + 1
The +1 increment is wrong, as seq_hi already contains the correct
seq_hi. For example, when seq_hi=1, the driver actually tells NIC to
use seq_hi=2 (esn). This incorrect esn value causes all subsequent
packets to fail integrity checks (SA-icv-failure). So, do not
increment.
Zhengchao Shao [Wed, 26 Oct 2022 01:46:42 +0000 (09:46 +0800)]
netdevsim: remove dir in nsim_dev_debugfs_init() when creating ports dir failed
Remove dir in nsim_dev_debugfs_init() when creating ports dir failed.
Otherwise, the netdevsim device will not be created next time. Kernel
reports an error: debugfs: Directory 'netdevsim1' with parent 'netdevsim'
already present!
Fixes: ab1d0cc004d7 ("netdevsim: change debugfs tree topology") Signed-off-by: Zhengchao Shao <[email protected]> Signed-off-by: Jakub Kicinski <[email protected]>
Zhengchao Shao [Wed, 26 Oct 2022 01:54:05 +0000 (09:54 +0800)]
netdevsim: fix memory leak in nsim_bus_dev_new()
If device_register() failed in nsim_bus_dev_new(), the value of reference
in nsim_bus_dev->dev is 1. obj->name in nsim_bus_dev->dev will not be
released.
Jakub Kicinski [Thu, 27 Oct 2022 17:30:41 +0000 (10:30 -0700)]
Merge tag 'linux-can-fixes-for-6.1-20221027' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can
Marc Kleine-Budde says:
====================
pull-request: can 2022-10-27
Anssi Hannula fixes the use of the completions in the kvaser_usb
driver.
Biju Das contributes 2 patches for the rcar_canfd driver. A IRQ storm
that can be triggered by high CAN bus load and channel specific IRQ
handlers are fixed.
Yang Yingliang fixes the j1939 transport protocol by moving a
kfree_skb() out of a spin_lock_irqsave protected section.
* tag 'linux-can-fixes-for-6.1-20221027' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
can: j1939: transport: j1939_session_skb_drop_old(): spin_unlock_irqrestore() before kfree_skb()
can: rcar_canfd: fix channel specific IRQ handling for RZ/G2L
can: rcar_canfd: rcar_canfd_handle_global_receive(): fix IRQ storm on global FIFO receive
can: kvaser_usb: Fix possible completions during init_completion
====================
Rafał Miłecki [Thu, 27 Oct 2022 11:24:30 +0000 (13:24 +0200)]
net: broadcom: bcm4908_enet: update TX stats after actual transmission
Queueing packets doesn't guarantee their transmission. Update TX stats
after hardware confirms consuming submitted data.
This also fixes a possible race and NULL dereference.
bcm4908_enet_start_xmit() could try to access skb after freeing it in
the bcm4908_enet_poll_tx().
====================
ip: rework the fix for dflt addr selection for connected nexthop"
This series reworks the fix that is reverted in the second commit.
As Julian explained, nhc_scope is related to nhc_gw, it's not the scope of
the route.
====================
Nicolas Dichtel [Thu, 20 Oct 2022 10:09:52 +0000 (12:09 +0200)]
nh: fix scope used to find saddr when adding non gw nh
As explained by Julian, fib_nh_scope is related to fib_nh_gw4, but
fib_info_update_nhc_saddr() needs the scope of the route, which is
the scope "before" fib_nh_scope, ie fib_nh_scope - 1.
This patch fixes the problem described in commit 747c14307214 ("ip: fix
dflt addr selection for connected nexthop").
projects / linux.git / log