Yunseong Kim [Mon, 24 Jun 2024 17:33:23 +0000 (02:33 +0900)]
tracing/net_sched: NULL pointer dereference in perf_trace_qdisc_reset()
In the TRACE_EVENT(qdisc_reset) NULL dereference occurred from
qdisc->dev_queue->dev <NULL> ->name
This situation simulated from bunch of veths and Bluetooth disconnection
and reconnection.
During qdisc initialization, qdisc was being set to noop_queue.
In veth_init_queue, the initial tx_num was reduced back to one,
causing the qdisc reset to be called with noop, which led to the kernel
panic.
I've attached the GitHub gist link that C converted syz-execprogram
source code and 3 log of reproduced vmcore-dmesg.
However, without our patch applied, I tested upstream 6.10.0-rc3 kernel
using the qdisc_reset event and the ip command on my qemu virtual machine.
This 2 commands makes always kernel panic.
Linux version: 6.10.0-rc3
[ 0.000000] Linux version 6.10.0-rc3-00164-g44ef20baed8e-dirty
(paran@fedora) (gcc (GCC) 14.1.1 20240522 (Red Hat 14.1.1-4), GNU ld
version 2.41-34.fc40) #20 SMP PREEMPT Sat Jun 15 16:51:25 KST 2024
Yeoreum and I don't know if the patch we wrote will fix the underlying
cause, but we think that priority is to prevent kernel panic happening.
So, we're sending this patch.
Dave Airlie [Thu, 27 Jun 2024 07:22:13 +0000 (17:22 +1000)]
Merge tag 'drm-misc-fixes-2024-06-26' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-fixes
drm-misc-fixes for v6.10-rc6:
- nouveau tv mode fixes.
- Add KOE TX26D202VM0BWA timings.
- Fix fb_info when vmalloc is used, regression from
CONFIG_DRM_FBDEV_LEAK_PHYS_SMEM.
Linus Torvalds [Thu, 27 Jun 2024 00:51:39 +0000 (17:51 -0700)]
Merge tag 'mm-hotfixes-stable-2024-06-26-17-28' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Pull misc fixes from Andrew Morton:
"13 hotfixes, 7 are cc:stable.
All are MM related apart from a MAINTAINERS update. There is no
identifiable theme here - just singleton patches in various places"
* tag 'mm-hotfixes-stable-2024-06-26-17-28' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
mm/memory: don't require head page for do_set_pmd()
mm/page_alloc: Separate THP PCP into movable and non-movable categories
nfs: drop the incorrect assertion in nfs_swap_rw()
mm/migrate: make migrate_pages_batch() stats consistent
MAINTAINERS: TPM DEVICE DRIVER: update the W-tag
selftests/mm:fix test_prctl_fork_exec return failure
mm: convert page type macros to enum
ocfs2: fix DIO failure due to insufficient transaction credits
kasan: fix bad call to unpoison_slab_object
mm: handle profiling for fake memory allocations during compaction
mm/slab: fix 'variable obj_exts set but not used' warning
/proc/pid/smaps: add mseal info for vma
mm: fix incorrect vbq reference in purge_fragmented_block
netfilter: nf_tables: fully validate NFT_DATA_VALUE on store to data registers
register store validation for NFT_DATA_VALUE is conditional, however,
the datatype is always either NFT_DATA_VALUE or NFT_DATA_VERDICT. This
only requires a new helper function to infer the register type from the
set datatype so this conditional check can be removed. Otherwise,
pointer to chain object can be leaked through the registers.
Linus Torvalds [Wed, 26 Jun 2024 22:01:33 +0000 (15:01 -0700)]
Merge tag 'wq-for-6.10-rc5-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq
Pull workqueue fixes from Tejun Heo:
"Two patches to fix kworker name formatting"
* tag 'wq-for-6.10-rc5-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq:
workqueue: Increase worker desc's length to 32
workqueue: Refactor worker ID formatting and make wq_worker_comm() use full ID string
Takashi Iwai [Wed, 26 Jun 2024 20:02:55 +0000 (22:02 +0200)]
Merge tag 'asoc-fix-v6.10-rc5' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus
ASoC: Fixes for v6.10
A relatively large batch of updates, largely due to the long interval
since I last sent fixes due to various travel and holidays. There's a
lot of driver specific fixes and quirks in here, none of them too major,
and also some fixes for recently introduced memory safety issues in the
topology code.
Joel Granados [Wed, 26 Jun 2024 12:06:16 +0000 (14:06 +0200)]
kbuild: scripts/gdb: bring the "abspath" back
Use the "abspath" call when symlinking the gdb python scripts in
scripts/gdb/linux. This call is needed to avoid broken links when
running the scripts_gdb target on a build directory located directly
under the source tree (e.g., O=builddir).
Fixes: 659bbf7e1b08 ("kbuild: scripts/gdb: Replace missed $(srctree)/$(src) w/ $(src)") Signed-off-by: Joel Granados <[email protected]> Reviewed-by: Douglas Anderson <[email protected]> Tested-by: Douglas Anderson <[email protected]> Signed-off-by: Masahiro Yamada <[email protected]>
nvmet-fc: Remove __counted_by from nvmet_fc_tgt_queue.fod[]
Work for __counted_by on generic pointers in structures (not just
flexible array members) has started landing in Clang 19 (current tip of
tree). During the development of this feature, a restriction was added
to __counted_by to prevent the flexible array member's element type from
including a flexible array member itself such as:
struct foo {
int count;
char buf[];
};
struct bar {
int count;
struct foo data[] __counted_by(count);
};
because the size of data cannot be calculated with the standard array
size formula:
sizeof(struct foo) * count
This restriction was downgraded to a warning but due to CONFIG_WERROR,
it can still break the build. The application of __counted_by on the fod
member of 'struct nvmet_fc_tgt_queue' triggers this restriction,
resulting in:
drivers/nvme/target/fc.c:151:2: error: 'counted_by' should not be applied to an array with element of unknown size because 'struct nvmet_fc_fcp_iod' is a struct type with a flexible array member. This will be an error in a future compiler version [-Werror,-Wbounds-safety-counted-by-elt-type-unknown-size]
151 | struct nvmet_fc_fcp_iod fod[] __counted_by(sqsize);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1 error generated.
Remove this use of __counted_by to fix the warning/error. However,
rather than remove it altogether, leave it commented, as it may be
possible to support this in future compiler releases.
Takashi Iwai [Wed, 26 Jun 2024 14:51:13 +0000 (16:51 +0200)]
ALSA: seq: Fix missing MSB in MIDI2 SPP conversion
The conversion of SPP to MIDI2 UMP called a wrong function, and the
secondary argument wasn't taken. As a result, MSB of SPP was always
zero. Fix to call the right function.
Andy Chiu [Thu, 13 Jun 2024 07:11:06 +0000 (15:11 +0800)]
riscv: stacktrace: convert arch_stack_walk() to noinstr
arch_stack_walk() is called intensively in function_graph when the
kernel is compiled with CONFIG_TRACE_IRQFLAGS. As a result, the kernel
logs a lot of arch_stack_walk and its sub-functions into the ftrace
buffer. However, these functions should not appear on the trace log
because they are part of the ftrace itself. This patch references what
arm64 does for the smae function. So it further prevent the re-enter
kprobe issue, which is also possible on riscv.
Alexandre Ghiti [Mon, 24 Jun 2024 08:21:41 +0000 (10:21 +0200)]
riscv: patch: Flush the icache right after patching to avoid illegal insns
We cannot delay the icache flush after patching some functions as we may
have patched a function that will get called before the icache flush.
The only way to completely avoid such scenario is by flushing the icache
as soon as we patch a function. This will probably be costly as we don't
batch the icache maintenance anymore.
Arnd Bergmann [Tue, 28 May 2024 12:06:30 +0000 (14:06 +0200)]
i2c: viai2c: turn common code into a proper module
The i2c-viai2c-common.c file is used by two drivers, but is not a proper
abstraction and can get linked into both modules in the same configuration,
which results in a warning:
scripts/Makefile.build:236: drivers/i2c/busses/Makefile: i2c-viai2c-common.o is added to multiple modules: i2c-wmt i2c-zhaoxin
The other problems with this include the incorrect use of a __weak function
when both are built-in, and the fact that the "common" module is sprinked
with 'if (i2c->plat == ...)' checks that have knowledge about the differences
between the drivers using it.
Avoid the link time warning by making the common driver a proper module
with MODULE_LICENCE()/MODULE_AUTHOR() tags, and remove the __weak function
by slightly rearranging the code.
This adds a little more duplication between the two main drivers, but
those versions get more readable in the process.
Darrick J. Wong [Wed, 19 Jun 2024 17:32:46 +0000 (10:32 -0700)]
xfs: honor init_xattrs in xfs_init_new_inode for !ATTR fs
xfs_init_new_inode ignores the init_xattrs parameter for filesystems
that do not have ATTR enabled. As a result, the first init_xattrs file
to be created by the kernel will not have an attr fork created to store
acls. Storing that first acl will add ATTR to the superblock flags, so
subsequent files will be created with attr forks. The overhead of this
is so small that chances are that nobody has noticed this behavior.
However, this is disastrous on a filesystem with parent pointers because
it requires that a new linkable file /must/ have a pre-existing attr
fork, and the parent pointers code uses init_xattrs to create that fork.
The preproduction version of mkfs.xfs used to set this, but the V5 sb
verifier only requires ATTR2, not ATTR. There is no guard for
filesystems with (PARENT && !ATTR).
It turns out that I misunderstood the two flags -- ATTR means that we at
some point created an attr fork to store xattrs in a file; ATTR2
apparently means only that inodes have dynamic fork offsets or that the
filesystem was mounted with the "attr2" option.
Fixes: 2442ee15bb1e ("xfs: eager inode attr fork init needs attr feature awareness") Signed-off-by: Darrick J. Wong <[email protected]> Reviewed-by: Christoph Hellwig <[email protected]> Signed-off-by: Chandan Babu R <[email protected]>
Darrick J. Wong [Wed, 19 Jun 2024 17:32:45 +0000 (10:32 -0700)]
xfs: allow unlinked symlinks and dirs with zero size
For a very very long time, inode inactivation has set the inode size to
zero before unmapping the extents associated with the data fork.
Unfortunately, commit 3c6f46eacd876 changed the inode verifier to
prohibit zero-length symlinks and directories. If an inode happens to
get logged in this state and the system crashes before freeing the
inode, log recovery will also fail on the broken inode.
Therefore, allow zero-size symlinks and directories as long as the link
count is zero; nobody will be able to open these files by handle so
there isn't any risk of data exposure.
Fixes: 3c6f46eacd876 ("xfs: sanity check directory inode di_size") Signed-off-by: Darrick J. Wong <[email protected]> Reviewed-by: Christoph Hellwig <[email protected]> Signed-off-by: Chandan Babu R <[email protected]>
Darrick J. Wong [Wed, 19 Jun 2024 17:32:44 +0000 (10:32 -0700)]
xfs: restrict when we try to align cow fork delalloc to cowextsz hints
xfs/205 produces the following failure when always_cow is enabled:
--- a/tests/xfs/205.out 2024-02-28 16:20:24.437887970 -0800
+++ b/tests/xfs/205.out.bad 2024-06-03 21:13:40.584000000 -0700
@@ -1,4 +1,5 @@
QA output created by 205
*** one file
+ !!! disk full (expected)
*** one file, a few bytes at a time
*** done
This is the result of overly aggressive attempts to align cow fork
delalloc reservations to the CoW extent size hint. Looking at the trace
data, we're trying to append a single fsblock to the "fred" file.
Trying to create a speculative post-eof reservation fails because
there's not enough space.
We then set @prealloc_blocks to zero and try again, but the cowextsz
alignment code triggers, which expands our request for a 1-fsblock
reservation into a 39-block reservation. There's not enough space for
that, so the whole write fails with ENOSPC even though there's
sufficient space in the filesystem to allocate the single block that we
need to land the write.
There are two things wrong here -- first, we shouldn't be attempting
speculative preallocations beyond what was requested when we're low on
space. Second, if we've already computed a posteof preallocation, we
shouldn't bother trying to align that to the cowextsize hint.
Fix both of these problems by adding a flag that only enables the
expansion of the delalloc reservation to the cowextsize if we're doing a
non-extending write, and only if we're not doing an ENOSPC retry. This
requires us to move the ENOSPC retry logic to xfs_bmapi_reserve_delalloc.
I probably should have caught this six years ago when 6ca30729c206d was
being reviewed, but oh well. Update the comments to reflect what the
code does now.
xfs: fix freeing speculative preallocations for preallocated files
xfs_can_free_eofblocks returns false for files that have persistent
preallocations unless the force flag is passed and there are delayed
blocks. This means it won't free delalloc reservations for files
with persistent preallocations unless the force flag is set, and it
will also free the persistent preallocations if the force flag is
set and the file happens to have delayed allocations.
Both of these are bad, so do away with the force flag and always free
only post-EOF delayed allocations for files with the XFS_DIFLAG_PREALLOC
or APPEND flags set.
Takashi Iwai [Tue, 25 Jun 2024 15:52:12 +0000 (17:52 +0200)]
ALSA: hda/realtek: Fix conflicting quirk for PCI SSID 17aa:3820
The recent fix for Lenovo IdeaPad 330-17IKB replaced the quirk entry,
and this eventually breaks the existing quirk for Lenovo Yoga Duet 7
13ITL6 equipped with the same PCI SSID 17aa:3820.
For applying a proper quirk for each model, check the codec SSID
additionally. Fortunately Yoga Duet has a different codec SSID,
0x17aa3802.
(Interestingly, 17aa:3802 has another conflict of SSID between another
Yoga model vs 14IRP8 which we had to work around similarly.)
Neal Cardwell [Mon, 24 Jun 2024 14:43:23 +0000 (14:43 +0000)]
tcp: fix tcp_rcv_fastopen_synack() to enter TCP_CA_Loss for failed TFO
Testing determined that the recent commit 9e046bb111f1 ("tcp: clear
tp->retrans_stamp in tcp_rcv_fastopen_synack()") has a race, and does
not always ensure retrans_stamp is 0 after a TFO payload retransmit.
If transmit completion for the SYN+data skb happens after the client
TCP stack receives the SYNACK (which sometimes happens), then
retrans_stamp can erroneously remain non-zero for the lifetime of the
connection, causing a premature ETIMEDOUT later.
Testing and tracing showed that the buggy scenario is the following
somewhat tricky sequence:
+ Client attempts a TFO handshake. tcp_send_syn_data() sends SYN + TFO
cookie + data in a single packet in the syn_data skb. It hands the
syn_data skb to tcp_transmit_skb(), which makes a clone. Crucially,
it then reuses the same original (non-clone) syn_data skb,
transforming it by advancing the seq by one byte and removing the
FIN bit, and enques the resulting payload-only skb in the
sk->tcp_rtx_queue.
+ Client sets retrans_stamp to the start time of the three-way
handshake.
+ Cookie mismatches or server has TFO disabled, and server only ACKs
SYN.
+ tcp_ack() sees SYN is acked, tcp_clean_rtx_queue() clears
retrans_stamp.
+ Since the client SYN was acked but not the payload, the TFO failure
code path in tcp_rcv_fastopen_synack() tries to retransmit the
payload skb. However, in some cases the transmit completion for the
clone of the syn_data (which had SYN + TFO cookie + data) hasn't
happened. In those cases, skb_still_in_host_queue() returns true
for the retransmitted TFO payload, because the clone of the syn_data
skb has not had its tx completetion.
+ Because skb_still_in_host_queue() finds skb_fclone_busy() is true,
it sets the TSQ_THROTTLED bit and the retransmit does not happen in
the tcp_rcv_fastopen_synack() call chain.
+ The tcp_rcv_fastopen_synack() code next implicitly assumes the
retransmit process is finished, and sets retrans_stamp to 0 to clear
it, but this is later overwritten (see below).
+ Later, upon tx completion, tcp_tsq_write() calls
tcp_xmit_retransmit_queue(), which puts the retransmit in flight and
sets retrans_stamp to a non-zero value.
+ The client receives an ACK for the retransmitted TFO payload data.
+ Since we're in CA_Open and there are no dupacks/SACKs/DSACKs/ECN to
make tcp_ack_is_dubious() true and make us call
tcp_fastretrans_alert() and reach a code path that clears
retrans_stamp, retrans_stamp stays nonzero.
+ Later, if there is a TLP, RTO, RTO sequence, then the connection
will suffer an early ETIMEDOUT due to the erroneously ancient
retrans_stamp.
The fix: this commit refactors the code to have
tcp_rcv_fastopen_synack() retransmit by reusing the relevant parts of
tcp_simple_retransmit() that enter CA_Loss (without changing cwnd) and
call tcp_xmit_retransmit_queue(). We have tcp_simple_retransmit() and
tcp_rcv_fastopen_synack() share code in this way because in both cases
we get a packet indicating non-congestion loss (MTU reduction or TFO
failure) and thus in both cases we want to retransmit as many packets
as cwnd allows, without reducing cwnd. And given that retransmits will
set retrans_stamp to a non-zero value (and may do so in a later
calling context due to TSQ), we also want to enter CA_Loss so that we
track when all retransmitted packets are ACked and clear retrans_stamp
when that happens (to ensure later recurring RTOs are using the
correct retrans_stamp and don't declare ETIMEDOUT prematurely).
Shannon Nelson [Mon, 24 Jun 2024 17:50:15 +0000 (10:50 -0700)]
ionic: use dev_consume_skb_any outside of napi
If we're not in a NAPI softirq context, we need to be careful
about how we call napi_consume_skb(), specifically we need to
call it with budget==0 to signal to it that we're not in a
safe context.
This was found while running some configuration stress testing
of traffic and a change queue config loop running, and this
curious note popped out:
I found that ionic_tx_clean() calls napi_consume_skb() which calls
napi_skb_cache_put(), but before that last call is the note
/* Zero budget indicate non-NAPI context called us, like netpoll */
and
DEBUG_NET_WARN_ON_ONCE(!in_softirq());
Those are pretty big hints that we're doing it wrong. We can pass a
context hint down through the calls to let ionic_tx_clean() know what
we're doing so it can call napi_consume_skb() correctly.
Kent Overstreet [Sun, 23 Jun 2024 04:53:44 +0000 (00:53 -0400)]
bcachefs: Discard, invalidate workers are now per device
There's no reason for discards to be single threaded across all devices;
this will improve performance on multi device setups.
Additionally, making them per-device simplifies the refcounting on
bch_dev->io_ref; we now hold it for the duration that the discard path
is running, which fixes a race between the discard path and device
removal.
Ma Ke [Tue, 25 Jun 2024 08:18:28 +0000 (16:18 +0800)]
drm/nouveau/dispnv04: fix null pointer dereference in nv17_tv_get_ld_modes
In nv17_tv_get_ld_modes(), the return value of drm_mode_duplicate() is
assigned to mode, which will lead to a possible NULL pointer dereference
on failure of drm_mode_duplicate(). Add a check to avoid npd.
Ma Ke [Tue, 25 Jun 2024 08:10:29 +0000 (16:10 +0800)]
drm/nouveau/dispnv04: fix null pointer dereference in nv17_tv_get_hd_modes
In nv17_tv_get_hd_modes(), the return value of drm_mode_duplicate() is
assigned to mode, which will lead to a possible NULL pointer dereference
on failure of drm_mode_duplicate(). The same applies to drm_cvt_mode().
Add a check to avoid null pointer dereference.
Julia Zhang [Mon, 3 Jun 2024 11:31:09 +0000 (19:31 +0800)]
drm/amdgpu: avoid using null object of framebuffer
Instead of using state->fb->obj[0] directly, get object from framebuffer
by calling drm_gem_fb_get_obj() and return error code when object is
null to avoid using null object of framebuffer.
Alex Deucher [Fri, 14 Jun 2024 17:48:26 +0000 (13:48 -0400)]
drm/amdgpu/atomfirmware: fix parsing of vram_info
v3.x changed the how vram width was encoded. The previous
implementation actually worked correctly for most boards.
Fix the implementation to work correctly everywhere.
This fixes the vram width reported in the kernel log on
some boards.
Nicolas Schier [Mon, 24 Jun 2024 11:12:14 +0000 (13:12 +0200)]
kbuild: Use $(obj)/%.cc to fix host C++ module builds
Use $(obj)/ instead of $(src)/ prefix when building C++ modules for
host, as explained in commit b1992c3772e6 ("kbuild: use $(src) instead
of $(srctree)/$(src) for source directory"). This fixes build failures
of 'xconfig':
$ make O=build/ xconfig
make[1]: Entering directory '/data/linux/kbuild-review/build'
GEN Makefile
make[3]: *** No rule to make target '../scripts/kconfig/qconf-moc.cc', needed by 'scripts/kconfig/qconf-moc.o'. Stop.
Fixes: b1992c3772e6 ("kbuild: use $(src) instead of $(srctree)/$(src) for source directory") Reported-by: Rolf Eike Beer <[email protected]> Signed-off-by: Nicolas Schier <[email protected]> Tested-by: Rolf Eike Beer <[email protected]> Signed-off-by: Masahiro Yamada <[email protected]>
Thayne Harbaugh [Sun, 16 Jun 2024 05:34:54 +0000 (23:34 -0600)]
kbuild: Fix build target deb-pkg: ln: failed to create hard link
The make deb-pkg target calls debian-orig which attempts to either
hard link the source .tar to the build-output location or copy the
source .tar to the build-output location. The test to determine
whether to ln or cp is incorrectly expanded by Make and consequently
always attempts to ln the source .tar. This fix corrects the escaping
of '$' so that the test is expanded by the shell rather than by Make
and appropriately selects between ln and cp.
Fixes: b44aa8c96e9e ("kbuild: deb-pkg: make .orig tarball a hard link if possible") Signed-off-by: Thayne Harbaugh <[email protected]> Signed-off-by: Masahiro Yamada <[email protected]>
Mark-PK Tsai [Fri, 14 Jun 2024 07:15:02 +0000 (15:15 +0800)]
kbuild: doc: Update default INSTALL_MOD_DIR from extra to updates
The default INSTALL_MOD_DIR was changed from 'extra' to
'updates' in commit b74d7bb7ca24 ("kbuild: Modify default
INSTALL_MOD_DIR from extra to updates").
This commit updates the documentation to align with the
latest kernel.
Fixes: b74d7bb7ca24 ("kbuild: Modify default INSTALL_MOD_DIR from extra to updates") Signed-off-by: Mark-PK Tsai <[email protected]> Signed-off-by: Masahiro Yamada <[email protected]>
Dragan Simic [Mon, 10 Jun 2024 05:21:12 +0000 (07:21 +0200)]
kbuild: Install dtb files as 0644 in Makefile.dtbinst
The compiled dtb files aren't executable, so install them with 0644 as their
permission mode, instead of defaulting to 0755 for the permission mode and
installing them with the executable bits set.
Some Linux distributions, including Debian, [1][2][3] already include fixes
in their kernel package build recipes to change the dtb file permissions to
0644 in their kernel packages. These changes, when additionally propagated
into the long-term kernel versions, will allow such distributions to remove
their downstream fixes.
Phil Chang [Mon, 10 Jun 2024 13:31:36 +0000 (21:31 +0800)]
hrtimer: Prevent queuing of hrtimer without a function callback
The hrtimer function callback must not be NULL. It has to be specified by
the call side but it is not validated by the hrtimer code. When a hrtimer
is queued without a function callback, the kernel crashes with a null
pointer dereference when trying to execute the callback in __run_hrtimer().
Introduce a validation before queuing the hrtimer in
hrtimer_start_range_ns().
The reverted commit moves a test on a field protected by a mutex outside
of the protection of that mutex, and so is obviously racey.
Depending on how the race goes, si->serv might be NULL when dereferenced
in svc_pool_stats_start(), or svc_pool_stats_stop() might unlock a mutex
that hadn't been locked.
This bug that the commit tried to fix has been addressed by initialising
->mutex earlier.
Fixes: 8e948c365d9c ("nfsd: fix oops when reading pool_stats before server is started") Signed-off-by: NeilBrown <[email protected]> Reviewed-by: Jeff Layton <[email protected]> Signed-off-by: Chuck Lever <[email protected]>
Arnd Bergmann [Thu, 20 Jun 2024 12:47:27 +0000 (14:47 +0200)]
linux/syscalls.h: add missing __user annotations
A couple of declarations in linux/syscalls.h are missing __user
annotations on their pointers, which can lead to warnings from
sparse because these don't match the implementation that have
the correct address space annotations.
Arnd Bergmann [Tue, 4 Jun 2024 12:20:26 +0000 (14:20 +0200)]
syscalls: mmap(): use unsigned offset type consistently
Most architectures that implement the old-style mmap() with byte offset
use 'unsigned long' as the type for that offset, but microblaze and
riscv have the off_t type that is shared with userspace, matching the
prototype in include/asm-generic/syscalls.h.
Make this consistent by using an unsigned argument everywhere. This
changes the behavior slightly, as the argument is shifted to a page
number, and an user input with the top bit set would result in a
negative page offset rather than a large one as we use elsewhere.
For riscv, the 32-bit sys_mmap2() definition actually used a custom
type that is different from the global declaration, but this was
missed due to an incorrect type check.
Arnd Bergmann [Thu, 20 Jun 2024 13:24:11 +0000 (15:24 +0200)]
hexagon: fix fadvise64_64 calling conventions
fadvise64_64() has two 64-bit arguments at the wrong alignment
for hexagon, which turns them into a 7-argument syscall that is
not supported by Linux.
The downstream musl port for hexagon actually asks for a 6-argument
version the same way we do it on arm, csky, powerpc, so make the
kernel do it the same way to avoid having to change both.
Arnd Bergmann [Fri, 14 Jun 2024 07:54:20 +0000 (09:54 +0200)]
csky, hexagon: fix broken sys_sync_file_range
Both of these architectures require u64 function arguments to be
passed in even/odd pairs of registers or stack slots, which in case of
sync_file_range would result in a seven-argument system call that is
not currently possible. The system call is therefore incompatible with
all existing binaries.
While it would be possible to implement support for seven arguments
like on mips, it seems better to use a six-argument version, either
with the normal argument order but misaligned as on most architectures
or with the reordered sync_file_range2() calling conventions as on
arm and powerpc.
Arnd Bergmann [Tue, 11 Jun 2024 20:12:43 +0000 (22:12 +0200)]
sh: rework sync_file_range ABI
The unusual function calling conventions on SuperH ended up causing
sync_file_range to have the wrong argument order, with the 'flags'
argument getting sorted before 'nbytes' by the compiler.
In userspace, I found that musl, glibc, uclibc and strace all expect the
normal calling conventions with 'nbytes' last, so changing the kernel
to match them should make all of those work.
In order to be able to also fix libc implementations to work with existing
kernels, they need to be able to tell which ABI is used. An easy way
to do this is to add yet another system call using the sync_file_range2
ABI that works the same on all architectures.
Old user binaries can now work on new kernels, and new binaries can
try the new sync_file_range2() to work with new kernels or fall back
to the old sync_file_range() version if that doesn't exist.
Arnd Bergmann [Fri, 7 Jun 2024 11:40:45 +0000 (13:40 +0200)]
parisc: use generic sys_fanotify_mark implementation
The sys_fanotify_mark() syscall on parisc uses the reverse word order
for the two halves of the 64-bit argument compared to all syscalls on
all 32-bit architectures. As far as I can tell, the problem is that
the function arguments on parisc are sorted backwards (26, 25, 24, 23,
...) compared to everyone else, so the calling conventions of using an
even/odd register pair in native word order result in the lower word
coming first in function arguments, matching the expected behavior
on little-endian architectures. The system call conventions however
ended up matching what the other 32-bit architectures do.
A glibc cleanup in 2020 changed the userspace behavior in a way that
handles all architectures consistently, but this inadvertently broke
parisc32 by changing to the same method as everyone else.
The change made it into glibc-2.35 and subsequently into debian 12
(bookworm), which is the latest stable release. This means we
need to choose between reverting the glibc change or changing the
kernel to match it again, but either hange will leave some systems
broken.
Pick the option that is more likely to help current and future
users and change the kernel to match current glibc. This also
means the behavior is now consistent across architectures, but
it breaks running new kernels with old glibc builds before 2.35.
Arnd Bergmann [Wed, 19 Jun 2024 12:27:55 +0000 (14:27 +0200)]
parisc: use correct compat recv/recvfrom syscalls
Johannes missed parisc back when he introduced the compat version
of these syscalls, so receiving cmsg messages that require a compat
conversion is still broken.
Use the correct calls like the other architectures do.
Fixes: 1dacc76d0014 ("net/compat/wext: send different messages to compat tasks") Acked-by: Helge Deller <[email protected]> Signed-off-by: Arnd Bergmann <[email protected]>
Arnd Bergmann [Wed, 19 Jun 2024 10:49:39 +0000 (12:49 +0200)]
sparc: fix compat recv/recvfrom syscalls
sparc has the wrong compat version of recv() and recvfrom() for both the
direct syscalls and socketcall().
The direct syscalls just need to use the compat version. For socketcall,
the same thing could be done, but it seems better to completely remove
the custom assembler code for it and just use the same implementation that
everyone else has.
Fixes: 1dacc76d0014 ("net/compat/wext: send different messages to compat tasks") Signed-off-by: Arnd Bergmann <[email protected]>
Arnd Bergmann [Wed, 19 Jun 2024 12:07:30 +0000 (14:07 +0200)]
sparc: fix old compat_sys_select()
sparc has two identical select syscalls at numbers 93 and 230, respectively.
During the conversion to the modern syscall.tbl format, the older one of the
two broke in compat mode, and now refers to the native 64-bit syscall.
Restore the correct behavior. This has very little effect, as glibc has
been using the newer number anyway.
Using sys_io_pgetevents() as the entry point for compat mode tasks
works almost correctly, but misses the sign extension for the min_nr
and nr arguments.
This was addressed on parisc by switching to
compat_sys_io_pgetevents_time64() in commit 6431e92fc827 ("parisc:
io_pgetevents_time64() needs compat syscall in 32-bit compat mode"),
as well as by using more sophisticated system call wrappers on x86 and
s390. However, arm64, mips, powerpc, sparc and riscv still have the
same bug.
Change all of them over to use compat_sys_io_pgetevents_time64()
like parisc already does. This was clearly the intention when the
function was originally added, but it got hooked up incorrectly in
the tables.
Jens Remus [Thu, 20 Jun 2024 09:12:49 +0000 (11:12 +0200)]
s390/boot: Do not adjust GOT entries for undef weak sym
Since commit 778666df60f0 ("s390: compile relocatable kernel without
-fPIE") and commit 00cda11d3b2e ("s390: Compile kernel with -fPIC and
link with -no-pie") the kernel on s390x may have a Global Offset Table
(GOT) whose entries are adjusted for KASLR in kaslr_adjust_got().
The GOT may contain entries for undefined weak symbols that resolved to
zero. That is the resulting GOT entry value is zero. Adjusting those
entries unconditionally in kaslr_adjust_got() is wrong. Otherwise the
following sample code would erroneously assume foo to be defined, due to
the adjustment changing the zero-value to a non-zero one:
extern int foo __attribute__((weak));
if (*foo)
/* foo is defined [or undefined and erroneously adjusted] */
The vmlinux build at commit 00cda11d3b2e ("s390: Compile kernel with
-fPIC and link with -no-pie") with defconfig actually had two GOT
entries for the undefined weak symbols __start_BTF and __stop_BTF:
The s390-specific vmlinux linker script sets the section start to
__START_KERNEL, which is currently defined as 0x100000 on s390x. Access
to lowcore is performed via a pointer of 0 and not a symbol in a section
starting at 0. The first 64K are reserved for the loader on s390x. Thus
it is safe to assume that __START_KERNEL will never be 0. As a result
there cannot be any defined symbols resolving to zero in the kernel.
Note that the first three GOT entries are reserved for the dynamic
loader on s390x. [1] In the kernel they are zero. Therefore no extra
handling is required to skip these.
Skip adjusting GOT entries with a value of zero in kaslr_adjust_got().
While at it update the comment when a GOT exists on s390x. Since commit 00cda11d3b2e ("s390: Compile kernel with -fPIC and link with -no-pie")
it no longer only exists when compiling with Clang, but also with GCC.
thermal: gov_step_wise: Go straight to instance->lower when mitigation is over
Commit b6846826982b ("thermal: gov_step_wise: Restore passive polling
management") attempted to fix a Step-Wise thermal governor issue
introduced by commit 042a3d80f118 ("thermal: core: Move passive polling
management to the core"), which caused the governor to leave cooling
devices in high states, by partially reverting that commit.
However, this turns out to be insufficient on some systems due to
interactions between the governor code restored by commit b6846826982b
and the passive polling management in the thermal core.
For this reason, revert commit b6846826982b and make the governor set
the target cooling device state to the "lower" one as soon as the zone
temperature falls below the threshold of the trip point corresponding
to the given thermal instance, which means that thermal mitigation is
not necessary any more.
Before this change the "lower" cooling device state would be reached in
steps through the passive polling mechanism which was questionable for
three reasons: (1) cooling device were kept in high states when that was
not necessary (and it could adversely impact performance), (2) it only
worked for thermal zones with nonzero passive_delay_jiffies value, and
(3) passive polling belongs to the core and should not be hijacked by
governors for their internal purposes.
Tristram Ha [Fri, 21 Jun 2024 22:34:22 +0000 (15:34 -0700)]
net: dsa: microchip: fix wrong register write when masking interrupt
The switch global port interrupt mask, REG_SW_PORT_INT_MASK__4, is
defined as 0x001C in ksz9477_reg.h. The designers used 32-bit value in
anticipation for increase of port count in future product but currently
the maximum port count is 7 and the effective value is 0x7F in register
0x001F. Each port has its own interrupt mask and is defined as 0x#01F.
It uses only 4 bits for different interrupts.
The developer who implemented the current interrupt mechanism in the
switch driver noticed there are similarities between the mechanism to
mask port interrupts in global interrupt and individual interrupts in
each port and so used the same code to handle these interrupts. He
updated the code to use the new macro REG_SW_PORT_INT_MASK__1 which is
defined as 0x1F in ksz_common.h but he forgot to update the 32-bit write
to 8-bit as now the mask registers are 0x1F and 0x#01F.
In addition all KSZ switches other than the KSZ9897/KSZ9893 and LAN937X
families use only 8-bit access and so this common code will eventually
be changed to accommodate them.
Shengjiu Wang [Thu, 20 Jun 2024 02:40:18 +0000 (10:40 +0800)]
ALSA: dmaengine_pcm: terminate dmaengine before synchronize
When dmaengine supports pause function, in suspend state,
dmaengine_pause() is called instead of dmaengine_terminate_async(),
In end of playback stream, the runtime->state will go to
SNDRV_PCM_STATE_DRAINING, if system suspend & resume happen
at this time, application will not resume playback stream, the
stream will be closed directly, the dmaengine_terminate_async()
will not be called before the dmaengine_synchronize(), which
violates the call sequence for dmaengine_synchronize().
This behavior also happens for capture streams, but there is no
SNDRV_PCM_STATE_DRAINING state for capture. So use
dmaengine_tx_status() to check the DMA status if the status is
DMA_PAUSED, then call dmaengine_terminate_async() to terminate
dmaengine before dmaengine_synchronize().
luoxuanqiang [Fri, 21 Jun 2024 01:39:29 +0000 (09:39 +0800)]
Fix race for duplicate reqsk on identical SYN
When bonding is configured in BOND_MODE_BROADCAST mode, if two identical
SYN packets are received at the same time and processed on different CPUs,
it can potentially create the same sk (sock) but two different reqsk
(request_sock) in tcp_conn_request().
These two different reqsk will respond with two SYNACK packets, and since
the generation of the seq (ISN) incorporates a timestamp, the final two
SYNACK packets will have different seq values.
The consequence is that when the Client receives and replies with an ACK
to the earlier SYNACK packet, we will reset(RST) it.
This behavior is consistently reproducible in my local setup,
which comprises:
| NETA1 ------ NETB1 |
PC_A --- bond --- | | --- bond --- PC_B
| NETA2 ------ NETB2 |
- PC_A is the Server and has two network cards, NETA1 and NETA2. I have
bonded these two cards using BOND_MODE_BROADCAST mode and configured
them to be handled by different CPU.
- PC_B is the Client, also equipped with two network cards, NETB1 and
NETB2, which are also bonded and configured in BOND_MODE_BROADCAST mode.
If the client attempts a TCP connection to the server, it might encounter
a failure. Capturing packets from the server side reveals:
The attempted solution is as follows:
Add a return value to inet_csk_reqsk_queue_hash_add() to confirm if the
ehash insertion is successful (Up to now, the reason for unsuccessful
insertion is that a reqsk for the same connection has already been
inserted). If the insertion fails, release the reqsk.
Due to the refcnt, Kuniyuki suggests also adding a return value check
for the DCCP module; if ehash insertion fails, indicating a successful
insertion of the same connection, simply release the reqsk as well.
Simultaneously, In the reqsk_queue_hash_req(), the start of the
req->rsk_timer is adjusted to be after successful insertion.
Nick Child [Thu, 20 Jun 2024 15:23:11 +0000 (10:23 -0500)]
ibmvnic: Add tx check to prevent skb leak
Below is a summary of how the driver stores a reference to an skb during
transmit:
tx_buff[free_map[consumer_index]]->skb = new_skb;
free_map[consumer_index] = IBMVNIC_INVALID_MAP;
consumer_index ++;
Where variable data looks like this:
free_map == [4, IBMVNIC_INVALID_MAP, IBMVNIC_INVALID_MAP, 0, 3]
consumer_index^
tx_buff == [skb=null, skb=<ptr>, skb=<ptr>, skb=null, skb=null]
The driver has checks to ensure that free_map[consumer_index] pointed to
a valid index but there was no check to ensure that this index pointed
to an unused/null skb address. So, if, by some chance, our free_map and
tx_buff lists become out of sync then we were previously risking an
skb memory leak. This could then cause tcp congestion control to stop
sending packets, eventually leading to ETIMEDOUT.
Therefore, add a conditional to ensure that the skb address is null. If
not then warn the user (because this is still a bug that should be
patched) and free the old pointer to prevent memleak/tcp problems.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
Let's roll back all of the serial core and printk console changes that
went into 6.10-rc1 as there still are problems with them that need to be
sorted out.
mm/memory: don't require head page for do_set_pmd()
The requirement that the head page be passed to do_set_pmd() was added in
commit ef37b2ea08ac ("mm/memory: page_add_file_rmap() ->
folio_add_file_rmap_[pte|pmd]()") and prevents pmd-mapping in the
finish_fault() and filemap_map_pages() paths if the page to be inserted is
anything but the head page for an otherwise suitable vma and pmd-sized
page.
Matthew said:
: We're going to stop using PMDs to map large folios unless the fault is
: within the first 4KiB of the PMD. No idea how many workloads that
: affects, but it only needs to be backported as far as v6.8, so we may
: as well backport it.
yangge [Thu, 20 Jun 2024 00:59:50 +0000 (08:59 +0800)]
mm/page_alloc: Separate THP PCP into movable and non-movable categories
Since commit 5d0a661d808f ("mm/page_alloc: use only one PCP list for
THP-sized allocations") no longer differentiates the migration type of
pages in THP-sized PCP list, it's possible that non-movable allocation
requests may get a CMA page from the list, in some cases, it's not
acceptable.
If a large number of CMA memory are configured in system (for example, the
CMA memory accounts for 50% of the system memory), starting a virtual
machine with device passthrough will get stuck. During starting the
virtual machine, it will call pin_user_pages_remote(..., FOLL_LONGTERM,
...) to pin memory. Normally if a page is present and in CMA area,
pin_user_pages_remote() will migrate the page from CMA area to non-CMA
area because of FOLL_LONGTERM flag. But if non-movable allocation
requests return CMA memory, migrate_longterm_unpinnable_pages() will
migrate a CMA page to another CMA page, which will fail to pass the check
in check_and_migrate_movable_pages() and cause migration endless.
Call trace:
pin_user_pages_remote
--__gup_longterm_locked // endless loops in this function
----_get_user_pages_locked
----check_and_migrate_movable_pages
------migrate_longterm_unpinnable_pages
--------alloc_migration_target
This problem will also have a negative impact on CMA itself. For example,
when CMA is borrowed by THP, and we need to reclaim it through cma_alloc()
or dma_alloc_coherent(), we must move those pages out to ensure CMA's
users can retrieve that contigous memory. Currently, CMA's memory is
occupied by non-movable pages, meaning we can't relocate them. As a
result, cma_alloc() is more likely to fail.
To fix the problem above, we add one PCP list for THP, which will not
introduce a new cacheline for struct per_cpu_pages. THP will have 2 PCP
lists, one PCP list is used by MOVABLE allocation, and the other PCP list
is used by UNMOVABLE allocation. MOVABLE allocation contains GPF_MOVABLE,
and UNMOVABLE allocation contains GFP_UNMOVABLE and GFP_RECLAIMABLE.
nfs: drop the incorrect assertion in nfs_swap_rw()
Since commit 2282679fb20b ("mm: submit multipage write for SWP_FS_OPS
swap-space"), we can plug multiple pages then unplug them all together.
That means iov_iter_count(iter) could be way bigger than PAGE_SIZE, it
actually equals the size of iov_iter_npages(iter, INT_MAX).
Note this issue has nothing to do with large folios as we don't support
THP_SWPOUT to non-block devices.
Zi Yan [Tue, 18 Jun 2024 13:41:51 +0000 (09:41 -0400)]
mm/migrate: make migrate_pages_batch() stats consistent
As Ying pointed out in [1], stats->nr_thp_failed needs to be updated to
avoid stats inconsistency between MIGRATE_SYNC and MIGRATE_ASYNC when
calling migrate_pages_batch().
Because if not, when migrate_pages_batch() is called via
migrate_pages(MIGRATE_ASYNC), nr_thp_failed will not be increased and when
migrate_pages_batch() is called via migrate_pages(MIGRATE_SYNC*),
nr_thp_failed will be increase in migrate_pages_sync() by
stats->nr_thp_failed += astats.nr_thp_split.
After calling fork() in test_prctl_fork_exec(), the global variable
ksm_full_scans_fd is initialized to 0 in the child process upon entering
the main function of ./ksm_functional_tests.
In the function call chain test_child_ksm() -> __mmap_and_merge_range ->
ksm_merge-> ksm_get_full_scans, start_scans = ksm_get_full_scans() will
return an error. Therefore, the value of ksm_full_scans_fd needs to be
initialized before calling test_child_ksm in the child process.
Stephen Brennan [Fri, 7 Jun 2024 20:29:53 +0000 (13:29 -0700)]
mm: convert page type macros to enum
Changing PG_slab from a page flag to a page type in commit 46df8e73a4a3
("mm: free up PG_slab") in has the unintended consequence of removing the
PG_slab constant from kernel debuginfo. The commit does add the value to
the vmcoreinfo note, which allows debuggers to find the value without
hardcoding it. However it's most flexible to continue representing the
constant with an enum. To that end, convert the page type fields into an
enum. Debuggers will now be able to detect that PG_slab's type has
changed from enum pageflags to enum pagetype.
Jan Kara [Fri, 14 Jun 2024 14:52:43 +0000 (16:52 +0200)]
ocfs2: fix DIO failure due to insufficient transaction credits
The code in ocfs2_dio_end_io_write() estimates number of necessary
transaction credits using ocfs2_calc_extend_credits(). This however does
not take into account that the IO could be arbitrarily large and can
contain arbitrary number of extents.
Extent tree manipulations do often extend the current transaction but not
in all of the cases. For example if we have only single block extents in
the tree, ocfs2_mark_extent_written() will end up calling
ocfs2_replace_extent_rec() all the time and we will never extend the
current transaction and eventually exhaust all the transaction credits if
the IO contains many single block extents. Once that happens a
WARN_ON(jbd2_handle_buffer_credits(handle) <= 0) is triggered in
jbd2_journal_dirty_metadata() and subsequently OCFS2 aborts in response to
this error. This was actually triggered by one of our customers on a
heavily fragmented OCFS2 filesystem.
To fix the issue make sure the transaction always has enough credits for
one extent insert before each call of ocfs2_mark_extent_written().
Heming Zhao said:
------
PANIC: "Kernel panic - not syncing: OCFS2: (device dm-1): panic forced after error"
Andrey Konovalov [Fri, 14 Jun 2024 14:32:38 +0000 (16:32 +0200)]
kasan: fix bad call to unpoison_slab_object
Commit 29d7355a9d05 ("kasan: save alloc stack traces for mempool") messed
up one of the calls to unpoison_slab_object: the last two arguments are
supposed to be GFP flags and whether to init the object memory.
Fix the call.
Without this fix, __kasan_mempool_unpoison_object provides the object's
size as GFP flags to unpoison_slab_object, which can cause LOCKDEP reports
(and probably other issues).
mm: handle profiling for fake memory allocations during compaction
During compaction isolated free pages are marked allocated so that they
can be split and/or freed. For that, post_alloc_hook() is used inside
split_map_pages() and release_free_list(). split_map_pages() marks free
pages allocated, splits the pages and then lets
alloc_contig_range_noprof() free those pages. release_free_list() marks
free pages and immediately frees them. This usage of post_alloc_hook()
affect memory allocation profiling because these functions might not be
called from an instrumented allocator, therefore current->alloc_tag is
NULL and when debugging is enabled (CONFIG_MEM_ALLOC_PROFILING_DEBUG=y)
that causes warnings. To avoid that, wrap such post_alloc_hook() calls
into an instrumented function which acts as an allocator which will be
charged for these fake allocations. Note that these allocations are very
short lived until they are freed, therefore the associated counters should
usually read 0.
mm/slab: fix 'variable obj_exts set but not used' warning
slab_post_alloc_hook() uses prepare_slab_obj_exts_hook() to obtain
slabobj_ext object. Currently the only user of slabobj_ext object in this
path is memory allocation profiling, therefore when it's not enabled this
object is not needed. This also generates a warning when compiling with
CONFIG_MEM_ALLOC_PROFILING=n. Move the code under this configuration to
fix the warning. If more slabobj_ext users appear in the future, the code
will have to be changed back to call prepare_slab_obj_exts_hook().
Zhaoyang Huang [Fri, 7 Jun 2024 02:31:16 +0000 (10:31 +0800)]
mm: fix incorrect vbq reference in purge_fragmented_block
xa_for_each() in _vm_unmap_aliases() loops through all vbs. However,
since commit 062eacf57ad9 ("mm: vmalloc: remove a global vmap_blocks
xarray") the vb from xarray may not be on the corresponding CPU
vmap_block_queue. Consequently, purge_fragmented_block() might use the
wrong vbq->lock to protect the free list, leading to vbq->free breakage.
Incorrect lock protection can exhaust all vmalloc space as follows:
CPU0 CPU1
+--------------------------------------------+
| +--------------------+ +-----+ |
+--> | |---->| |------+
| CPU1:vbq free_list | | vb1 |
+--- | |<----| |<-----+
| +--------------------+ +-----+ |
+--------------------------------------------+
_vm_unmap_aliases() vb_alloc()
new_vmap_block()
xa_for_each(&vbq->vmap_blocks, idx, vb)
--> vb in CPU1:vbq->freelist
purge_fragmented_block(vb)
spin_lock(&vbq->lock) spin_lock(&vbq->lock)
--> use CPU0:vbq->lock --> use CPU1:vbq->lock
prev->next = next
+--------------------------------------------+
|----------------------------+ |
| +--------------------+ | +-----+ |
+--> | |--+ | |------+
| CPU1:vbq free_list | | vb2 |
+--- | |<----| |<-----+
| +--------------------+ +-----+ |
+--------------------------------------------+
Here’s a list breakdown. All vbs, which were to be added to
‘prev’, cannot be used by list_for_each_entry_rcu(vb, &vbq->free,
free_list) in vb_alloc(). Thus, vmalloc space is exhausted.
Jens Axboe [Tue, 25 Jun 2024 01:07:18 +0000 (19:07 -0600)]
io_uring: signal SQPOLL task_work with TWA_SIGNAL_NO_IPI
Before SQPOLL was transitioned to managing its own task_work, the core
used TWA_SIGNAL_NO_IPI to ensure that task_work was processed. If not,
we can't be sure that all task_work is processed at SQPOLL thread exit
time.
Jakub Kicinski [Tue, 25 Jun 2024 01:15:21 +0000 (18:15 -0700)]
Merge tag 'for-netdev' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Daniel Borkmann says:
====================
pull-request: bpf 2024-06-24
We've added 12 non-merge commits during the last 10 day(s) which contain
a total of 10 files changed, 412 insertions(+), 16 deletions(-).
The main changes are:
1) Fix a BPF verifier issue validating may_goto with a negative offset,
from Alexei Starovoitov.
2) Fix a BPF verifier validation bug with may_goto combined with jump to
the first instruction, also from Alexei Starovoitov.
3) Fix a bug with overrunning reservations in BPF ring buffer,
from Daniel Borkmann.
4) Fix a bug in BPF verifier due to missing proper var_off setting related
to movsx instruction, from Yonghong Song.
5) Silence unnecessary syzkaller-triggered warning in __xdp_reg_mem_model(),
from Daniil Dulov.
* tag 'for-netdev' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
xdp: Remove WARN() from __xdp_reg_mem_model()
selftests/bpf: Add tests for may_goto with negative offset.
bpf: Fix may_goto with negative offset.
selftests/bpf: Add more ring buffer test coverage
bpf: Fix overrunning reservations in ringbuf
selftests/bpf: Tests with may_goto and jumps to the 1st insn
bpf: Fix the corner case with may_goto and jump to the 1st insn.
bpf: Update BPF LSM maintainer list
bpf: Fix remap of arena.
selftests/bpf: Add a few tests to cover
bpf: Add missed var_off setting in coerce_subreg_to_size_sx()
bpf: Add missed var_off setting in set_sext32_default_val()
====================
Filipe Manana [Thu, 20 Jun 2024 11:32:00 +0000 (12:32 +0100)]
btrfs: qgroup: fix quota root leak after quota disable failure
If during the quota disable we fail when cleaning the quota tree or when
deleting the root from the root tree, we jump to the 'out' label without
ever dropping the reference on the quota root, resulting in a leak of the
root since fs_info->quota_root is no longer pointing to the root (we have
set it to NULL just before those steps).
Fix this by always doing a btrfs_put_root() call under the 'out' label.
This is a problem that exists since qgroups were first added in 2012 by
commit bed92eae26cc ("Btrfs: qgroup implementation and prototypes"), but
back then we missed a kfree on the quota root and free_extent_buffer()
calls on its root and commit root nodes, since back then roots were not
yet reference counted.
Qu Wenruo [Mon, 17 Jun 2024 05:48:44 +0000 (15:18 +0930)]
btrfs: scrub: handle RST lookup error correctly
[BUG]
When running btrfs/060 with forced RST feature, it would crash the
following ASSERT() inside scrub_read_endio():
ASSERT(sector_nr < stripe->nr_sectors);
Before that, we would have tree dump from
btrfs_get_raid_extent_offset(), as we failed to find the RST entry for
the range.
[CAUSE]
Inside scrub_submit_extent_sector_read() every time we allocated a new
bbio we immediately called btrfs_map_block() to make sure there was some
RST range covering the scrub target.
But if btrfs_map_block() fails, we immediately call endio for the bbio,
while the bbio is newly allocated, it's completely empty.
Then inside scrub_read_endio(), we go through the bvecs to find
the sector number (as bi_sector is no longer reliable if the bio is
submitted to lower layers).
And since the bio is empty, such bvecs iteration would not find any
sector matching the sector, and return sector_nr == stripe->nr_sectors,
triggering the ASSERT().
[FIX]
Instead of calling btrfs_map_block() after allocating a new bbio, call
btrfs_map_block() first.
Since our only objective of calling btrfs_map_block() is only to update
stripe_len, there is really no need to do that after btrfs_alloc_bio().
This new timing would avoid the problem of handling empty bbio
completely, and in fact fixes a possible race window for the old code,
where if the submission thread is the only owner of the pending_io, the
scrub would never finish (since we didn't decrease the pending_io
counter).
Although the root cause of RST lookup failure still needs to be
addressed.
Naohiro Aota [Tue, 11 Jun 2024 08:17:30 +0000 (17:17 +0900)]
btrfs: zoned: fix initial free space detection
When creating a new block group, it calls btrfs_add_new_free_space() to add
the entire block group range into the free space accounting.
__btrfs_add_free_space_zoned() checks if size == block_group->length to
detect the initial free space adding, and proceed that case properly.
However, if the zone_capacity == zone_size and the over-write speed is fast
enough, the entire zone can be over-written within one transaction. That
confuses __btrfs_add_free_space_zoned() to handle it as an initial free
space accounting. As a result, that block group becomes a strange state: 0
used bytes, 0 zone_unusable bytes, but alloc_offset == zone_capacity (no
allocation anymore).
The initial free space accounting can properly be checked by checking
alloc_offset too.
Filipe Manana [Thu, 13 Jun 2024 10:16:19 +0000 (11:16 +0100)]
btrfs: use NOFS context when getting inodes during logging and log replay
During inode logging (and log replay too), we are holding a transaction
handle and we often need to call btrfs_iget(), which will read an inode
from its subvolume btree if it's not loaded in memory and that results in
allocating an inode with GFP_KERNEL semantics at the btrfs_alloc_inode()
callback - and this may recurse into the filesystem in case we are under
memory pressure and attempt to commit the current transaction, resulting
in a deadlock since the logging (or log replay) task is holding a
transaction handle open.
Syzbot reported this with the following stack traces:
WARNING: possible circular locking dependency detected 6.10.0-rc2-syzkaller-00361-g061d1af7b030 #0 Not tainted
------------------------------------------------------
syz-executor.1/9919 is trying to acquire lock: ffffffff8dd3aac0 (fs_reclaim){+.+.}-{0:0}, at: might_alloc include/linux/sched/mm.h:334 [inline] ffffffff8dd3aac0 (fs_reclaim){+.+.}-{0:0}, at: slab_pre_alloc_hook mm/slub.c:3891 [inline] ffffffff8dd3aac0 (fs_reclaim){+.+.}-{0:0}, at: slab_alloc_node mm/slub.c:3981 [inline] ffffffff8dd3aac0 (fs_reclaim){+.+.}-{0:0}, at: kmem_cache_alloc_lru_noprof+0x58/0x2f0 mm/slub.c:4020
but task is already holding lock: ffff88804b569358 (&ei->log_mutex){+.+.}-{3:3}, at: btrfs_log_inode+0x39c/0x4660 fs/btrfs/tree-log.c:6481
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
Mostafa Saleh [Mon, 24 Jun 2024 20:37:28 +0000 (20:37 +0000)]
PCI/MSI: Fix UAF in msi_capability_init
KFENCE reports the following UAF:
BUG: KFENCE: use-after-free read in __pci_enable_msi_range+0x2c0/0x488
Use-after-free read at 0x0000000024629571 (in kfence-#12):
__pci_enable_msi_range+0x2c0/0x488
pci_alloc_irq_vectors_affinity+0xec/0x14c
pci_alloc_irq_vectors+0x18/0x28
allocated by task 81 on cpu 7 at 10.808142s:
__kmem_cache_alloc_node+0x1f0/0x2bc
kmalloc_trace+0x44/0x138
msi_alloc_desc+0x3c/0x9c
msi_domain_insert_msi_desc+0x30/0x78
msi_setup_msi_desc+0x13c/0x184
__pci_enable_msi_range+0x258/0x488
pci_alloc_irq_vectors_affinity+0xec/0x14c
pci_alloc_irq_vectors+0x18/0x28
freed by task 81 on cpu 7 at 10.811436s:
msi_domain_free_descs+0xd4/0x10c
msi_domain_free_locked.part.0+0xc0/0x1d8
msi_domain_alloc_irqs_all_locked+0xb4/0xbc
pci_msi_setup_msi_irqs+0x30/0x4c
__pci_enable_msi_range+0x2a8/0x488
pci_alloc_irq_vectors_affinity+0xec/0x14c
pci_alloc_irq_vectors+0x18/0x28
Freed in case of failure in __msi_domain_alloc_locked()
__pci_enable_msi_range
msi_capability_init
pci_msi_setup_msi_irqs
msi_domain_alloc_irqs_all_locked
msi_domain_alloc_locked
__msi_domain_alloc_locked => fails
msi_domain_free_locked
...
That failure propagates back to pci_msi_setup_msi_irqs() in
msi_capability_init() which accesses the descriptor for unmasking in the
error exit path.
Cure it by copying the descriptor and using the copy for the error exit path
unmask operation.
cpufreq: intel_pstate: Use HWP to initialize ITMT if CPPC is missing
It is reported that single-thread performance on some hybrid systems
dropped significantly after commit 7feec7430edd ("ACPI: CPPC: Only probe
for _CPC if CPPC v2 is acked") which prevented _CPC from being used if
the support for it had not been confirmed by the platform firmware.
The problem is that if the platform firmware does not confirm CPPC v2
support, cppc_get_perf_caps() returns an error which prevents the
intel_pstate driver from enabling ITMT. Consequently, the scheduler
does not get any hints on CPU performance differences, so in a hybrid
system some tasks may run on CPUs with lower capacity even though they
should be running on high-capacity CPUs.
To address this, modify intel_pstate to use the information from
MSR_HWP_CAPABILITIES to enable ITMT if CPPC is not available (which is
done already if the highest performance number coming from CPPC is not
realistic).
projects / linux.git / log