[ Upstream commit 52416a27aa ]
Don't account the free space in a data relocation space-info sub-group as
usable free space in statfs.
This is misleading as no user allocations can be made in this space-info
sub-group. It is only a target for relocation.
Fixes: f92ee31e03 ("btrfs: introduce btrfs_space_info sub-group")
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7f8581c70a ]
set_fan() parses user input with kstrtol() and passes the resulting
value to FAN16_TO_REG() on chips with 16-bit fan support.
Negative fan speeds are not meaningful and should be rejected before
conversion. Worst scenario, one may be able to abuse undefined
behaviour of signed overflow to possibly induce rpm * 2 == 0 in
FAN16_TO_REG(), thus causing a division by zero.
Instead, clamp val < 0 to zero and keep the conversion in its valid
input domain, avoiding unsafe arithmetic in the register conversion
path.
Found by Linux Verification Center (linuxtesting.org) with static
analysis tool SVACE.
Fixes: 17d648bf57 ("it87: Add support for the IT8716F")
Signed-off-by: Nikita Zhandarovich <n.zhandarovich@fintech.ru>
Link: https://lore.kernel.org/r/20260529141839.1639287-1-n.zhandarovich@fintech.ru
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d8421e0938 ]
The code that gets the frame buffer mode from OF has 'use after free',
'buffer overrun' and memory leaks.
info->edid_data isn't free if the probe functions fail or if
pd->def_mode is set.
If both the CRT and PANEL are enabled info->edid_data is used after
being freed and is freed twice.
The string returned by of_get_property(np, "mode", &len) is just
written over either the static "640x480-16@60" or the module parameter
string without any regard for the length (which is most likely longer).
Use kstrump() for the OF mode and free everything before freeing 'info.
Fixes: 4295f9bf74 ("video, sm501: add OF binding to support SM501")
Signed-off-by: David Laight <david.laight.linux@gmail.com>
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 486f8298b6 ]
In the beginning of the loop, we try to obtain a locked delayed ref head,
if 'locked_ref' is currently NULL, by calling btrfs_select_ref_head(),
which can return an error pointer. If the error pointer is -EAGAIN we do
a continue and go back to the beginning of the loop, which will not try
again to call btrfs_select_ref_head() since 'locked_ref' is no longer
NULL but it's ERR_PTR(-EAGAIN), and then we do:
spin_lock(&locked_ref->lock);
against a ERR_PTR(-EAGAIN) value, generating an invalid pointer
dereference.
Fix this by ensuring that 'locked_ref' is set to NULL when
btrfs_select_ref_head() returns ERR_PTR(-EAGAIN) and incrementing 'count'
as well, to prevent infinite looping. We do this by doing a goto to the
bottom of the loop that already sets 'locked_ref' to NULL and does a
cond_resched(), with an increment to 'count' right before the goto.
These measures were in place before the refactoring in commit 0110a4c434
("btrfs: refactor __btrfs_run_delayed_refs loop") but were unintentionally
lost afterwards.
Reported-by: Dan Carpenter <error27@gmail.com>
Link: https://lore.kernel.org/linux-btrfs/ag8ARRwykv8bpJ87@stanley.mountain/
Fixes: 0110a4c434 ("btrfs: refactor __btrfs_run_delayed_refs loop")
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3c5671ed81 ]
When FIELD_GET returns 0 for the retry count, subtracting 1 causes
an unsigned integer underflow, resulting in tx_retries becoming a
very large value (0xFFFFFFFF for u32).
Fix by checking if count is non-zero before subtracting 1.
Fixes: 9aecfa754c ("wifi: mt76: mt7921e: report tx retries/failed counts in tx free event")
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Link: https://patch.msgid.link/20260605113306.3485554-2-ryder.lee@mediatek.com
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 05e72b6167 ]
When FIELD_GET returns 0 for the retry count, subtracting 1 causes
an unsigned integer underflow, resulting in tx_retries becoming a
very large value (0xFFFFFFFF for u32).
Fix by checking if count is non-zero before subtracting 1.
Fixes: 943e4fb96e ("wifi: mt76: mt7915: report tx retries/failed counts for non-WED path")
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Link: https://patch.msgid.link/20260605113306.3485554-1-ryder.lee@mediatek.com
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 29922fdfc2 ]
If we take runnable_avg in max(runnable_avg, util_avg) in cpu_util(), we
should then add or subtract task runnable_avg, but the arithmetic below
is still with task util_avg. This mixes runnable_avg with util_avg which
is incorrect.
Fix by always doing arithmetic with runnable_avg and only take
max(runnable_avg, util_avg) at the last step.
Fixes: 7d0583cf9e ("sched/fair, cpufreq: Introduce 'runnable boosting'")
Signed-off-by: Hongyan Xia <hongyan.xia@transsion.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260605094318.37931-1-hongyan.xia@transsion.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7c08d43083 ]
tcm_loop_target_reset is supposed to handle all the LUNs on a target but
it's only doing a TMR_LUN_RESET so only that one LUN is handled. This
will cause us to return early while IOs to other LUNs are still hung in
lower layers. This just removes the target reset handler for the driver
because LIO doesn't support target resets and for the common case where
this is run from the scsi-ml error hamdler we have already tried an
abort and lun reset so waiting again is most likely useless.
Fixes: 1333eee56c ("scsi: target: tcm_loop: Drain commands in target_reset handler")
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Reviewed-by: Hannes Reinecke <hare@kernel.org>
Link: https://patch.msgid.link/20260530052349.5134-1-michael.christie@oracle.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7e161211f1 ]
A mutual CHAP handshake requires target processing of an initiator-sent
CHAP_I identifier. The RFC 3720 specification states:
11.1.4. Challenge Handshake Authentication Protocol (CHAP)
...
CHAP_A=<A> CHAP_I=<I> CHAP_C=<C>
...
Where N, (A,A1,A2), I, C, and R are (correspondingly) the Name,
Algorithm, Identifier, Challenge, and Response as defined in
[RFC1994], N is a text string, A,A1,A2, and I are numbers
CHAP_I parsing currently calls extract_param(), which returns the
@identifier string (stripped of any 0b/0B or 0x/0X prefix) and a @type
which indicates DECIMAL, HEX, or BASE64 encoding (based on any stripped
prefix).
Any HEX encoded CHAP_I string is further processed via:
ret = kstrtoul(&identifier[2], 0, &id);
This is incorrect for two reasons:
* The @identifier string has already been stripped of the 0x/0X prefix,
so skipping the first two bytes omits part of the number.
* The kstrtoul() call specifies a base of 0, which will see
&identifier[2] parsed as a decimal, unless a '0x' or (octal) '0' is
erroneously present at that offset.
Fix this by passing the (zero-offset) identifier string to kstrtoul()
along with a base=16 parameter. Also add an explicit error handler for
BASE64 encoding.
Hex-encoded CHAP_I handling can be testing using the libiscsi EncodedI
test linked below.
Reported-by: Sashiko (gemini/gemini-3.1-pro-preview)
Link: https://sashiko.dev/#/patchset/20260521151121.808477-1-hossu.alexandru%40gmail.com
Link: https://github.com/sahlberg/libiscsi/pull/473
Fixes: 85db739131 ("scsi: target: iscsi: Validate CHAP_R length before base64 decode")
Signed-off-by: David Disseldorp <ddiss@suse.de>
Reviewed-by: Lee Duncan <lduncan@suse.com>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Link: https://patch.msgid.link/20260605122019.24146-2-ddiss@suse.de
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a298c7302e ]
watchdog_register_device() registers wdd->pm_nb when
WDOG_NO_PING_ON_SUSPEND is set, but watchdog_unregister_device() does not
remove it. This leaves an embedded notifier block on the PM notifier chain
after the watchdog device has been unregistered.
A later suspend/resume notification can then call watchdog_pm_notifier()
with a stale watchdog_device pointer, or at minimum after wdd->wd_data has
been cleared by watchdog_dev_unregister().
Unregister the PM notifier before tearing down the watchdog device.
Fixes: 60bcd91aaf ("watchdog: introduce watchdog_dev_suspend/resume")
Signed-off-by: Yuho Choi <dbgh9129@gmail.com>
Link: https://lore.kernel.org/r/20260601192005.1970805-1-dbgh9129@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9b9e8bb81c ]
Having the parent directory locked protects entries from removal
by another thread, but it does *not* protect cursors from being
moved around by lseek() - or freed, for that matter.
Fixes: 6f61076406 ("configfs: Introduce configfs_dirent_lock")
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d3ec78f8f8 ]
A recursive locking deadlock can occur in the firmware loader's power
management notification handler.
During system suspend or hibernation preparation, fw_pm_notify() calls
device_cache_fw_images(). This function acquires fw_lock to set the
firmware cache state to FW_LOADER_START_CACHE and then iterates over all
devices using dpm_for_each_dev() while still holding the lock.
For each device, dev_cache_fw_image() schedules asynchronous work to cache
the firmware. If memory allocation for the async work entry fails (e.g., in
out-of-memory conditions), async_schedule_node_domain() falls back to
executing the work function synchronously in the current thread.
The synchronous execution path (__async_dev_cache_fw_image() ->
cache_firmware() -> request_firmware() -> assign_fw()) attempts to acquire
fw_lock again. Since the current thread already holds fw_lock, this results
in a recursive locking deadlock.
Fix this by releasing fw_lock immediately after updating the cache state
and before calling dpm_for_each_dev(). The lock is only needed to protect
the state update. Concurrent firmware requests will correctly see the
FW_LOADER_START_CACHE state and use the piggyback mechanism, which is
independently protected by its own fwc->name_lock.
Fixes: ac39b3ea73 ("firmware loader: let caching firmware piggyback on loading firmware")
Assisted-by: Gemini:gemini-3.1-pro-preview Gemini:gemini-3-flash-preview syzbot
Reported-by: syzbot+e70e4c6f6eee43357ba7@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=e70e4c6f6eee43357ba7
Link: https://syzkaller.appspot.com/ai_job?id=8b4af9fd-24af-423f-8acb-1159fd34c1a5
Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Link: https://patch.msgid.link/48b092a5-f49d-48a4-95f4-f65bebfc6bc3@mail.kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7886054b06 ]
If DMA setup fails after allocating the zeropage, the error path frees
the page but leaves espi->zeropage dangling. A subsequent call to
ep93xx_spi_release_dma() sees the non-NULL pointer and frees the page
again.
Clear the pointer after freeing in the error path of
ep93xx_spi_setup_dma().
Fixes: 626a96db11 ("spi/ep93xx: add DMA support")
Signed-off-by: Felix Gu <ustc.gu@gmail.com>
Link: https://patch.msgid.link/20260529-ep93xx-v1-1-9185070ca1fc@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 55d339d200 ]
Jumping to mlx5_ib_reg_user_mr() without IB_MR_REREG_TRANS set will use
garbage values for start, length, and iova. Recovering the original mr
parameters for ODP and DMABUF to properly recreate it is too hard in this
flow, so just fail it.
Fixes: ef3642c4f5 ("RDMA/mlx5: Fix error unwinds for rereg_mr")
Link: https://patch.msgid.link/r/1-v1-29ebd2c229b5+fd5-ib_mr_pd_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 03120e1425 ]
The conservative governor documentation incorrectly states that setting
freq_step to 0 will use the default 5% frequency step. In reality, since
at least commit 8e677ce83b ("[CPUFREQ] conservative: fixup governor to
function more like ondemand logic"), freq_step=0 has always caused the
governor to skip frequency updates entirely.
Correct the documentation to reflect the actual behavior: freq_step=0
disables frequency changes by the governor entirely.
Fixes: 2a0e492798 ("cpufreq: User/admin documentation update and consolidation")
Signed-off-by: Pengjie Zhang <zhangpengjie2@huawei.com>
Reviewed-by: Zhongqiu Han <zhongqiu.han@oss.qualcomm.com>
[ rjw: Subject adjustment ]
Link: https://patch.msgid.link/20260603055635.1549943-1-zhangpengjie2@huawei.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 63320db6a5 ]
acpi_ipmi_space_handler() takes an extra reference on tx_msg before
checking whether the selected IPMI device is dead. The reference
belongs to the tx_msg_list entry and is normally dropped by
ipmi_cancel_tx_msg() or ipmi_flush_tx_msg() after the message is removed
from the list.
On the dead-device path, the message has not been queued yet, but the
error path still calls ipmi_msg_release() directly. That bypasses
kref_put() and frees tx_msg while the queued-message reference is still
recorded in the kref count.
Take the queued-message reference only after the dead-device check
succeeds, immediately before adding tx_msg to the list.
Fixes: 7b98447722 ("ACPI / IPMI: Add reference counting for ACPI IPMI transfers")
Signed-off-by: Yuho Choi <dbgh9129@gmail.com>
Link: https://patch.msgid.link/20260603163108.2149359-1-dbgh9129@gmail.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 705dd6dcbc ]
snd_seq_read() copies a queued variable-length event header to userspace
before expanding the payload. Queued variable-length events use
SNDRV_SEQ_EXT_CHAINED internally, and data.ext.ptr points at the first
extension cell.
The read side strips SNDRV_SEQ_EXT_* bits from data.ext.len before the
copy, but it leaves data.ext.ptr untouched. A userspace sequencer client
can therefore write a direct variable event to itself and read back the
extension-cell kernel address from the returned header.
Clear the temporary header pointer before copy_to_user(). The original
queued event remains unchanged and is still passed to
snd_seq_expand_var_event(), so payload expansion keeps using the
internal chain.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kyle Zeng <kylebot@openai.com>
Link: https://patch.msgid.link/20260607004129.61345-1-kylebot@openai.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8ac35bac70 ]
In the non-frame-pointer version of walk_stackframe, each value read
from the stack is treated as a potential return address and has 0x4
subtracted before being used as the program counter. This was intended
to convert the return address (the instruction after a call) back to
the call site, but it is incorrect:
1. RISC-V has variable-length instructions due to the RVC (compressed
instruction) extension. A call instruction can be either 4 bytes
(regular) or 2 bytes (compressed, e.g. c.jal). Subtracting a fixed
0x4 assumes all call instructions are 4 bytes, which is wrong for
compressed instructions.
2. Stack traces conventionally report return addresses, not call sites.
Other architectures (ARM64, x86, ARM) do not subtract instruction
size from return addresses in their stack unwinding code.
3. The frame-pointer version of walk_stackframe already dropped the
-0x4 offset. Commit b785ec129b ("riscv/ftrace: Add
HAVE_FUNCTION_GRAPH_RET_ADDR_PTR support") replaced "pc =
frame->ra - 0x4" with ftrace_graph_ret_addr(), and the commit
message explicitly noted that "the original calculation, pc =
frame->ra - 4, is buggy when the instruction at the return address
happened to be a compressed inst." The non-FP version was simply
overlooked.
Remove the bogus -0x4 offset to match the FP version and the
conventions used by other architectures.
Fixes: 5d8544e2d0 ("RISC-V: Generic library routines and assembly")
Signed-off-by: Rui Qi <qirui.001@bytedance.com>
Link: https://patch.msgid.link/20260603115329.791603-2-qirui.001@bytedance.com
Signed-off-by: Paul Walmsley <pjw@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2b7bd6f548 ]
snd_seq_expand_var_event_at() clamps the number of bytes to copy to the
remaining variable-event length, but passes the original buffer size to
expand_var_event().
For SNDRV_SEQ_EXT_USRPTR events, expand_var_event() copies exactly the
size argument from userspace. On the final chunk, when the remaining
event data is shorter than the caller's buffer, this can read past the
declared event data and can spuriously fail with -EFAULT if the extra
bytes cross an unmapped page.
Pass the clamped length instead. The chained and kernel-backed paths
already reclamp in dump_var_event(), but the user-pointer path handles
the size directly.
Fixes: ea46f79709 ("ALSA: seq: Add snd_seq_expand_var_event_at() helper")
Signed-off-by: HyeongJun An <sammiee5311@gmail.com>
Link: https://patch.msgid.link/20260606040913.230213-1-sammiee5311@gmail.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 82d7d0adbc ]
Currently, bpf_lwt_push_ip_encap() does not update skb->transport_header.
When a driver, e.g. ice, reuses the stale skb->transport_header to
offload checksum computation to NIC hardware, VxLAN packets encapsulated
by bpf_lwt_push_encap() helper may be dropped due to incorrect checksum.
Update skb->transport_header in bpf_lwt_push_ip_encap() whenever the
encapsulated packet uses UDP, so checksum offload works correctly.
Fixes: 52f278774e ("bpf: implement BPF_LWT_ENCAP_IP mode in bpf_lwt_push_encap")
Cc: Leon Hwang <leon.huangfu@shopee.com>
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/r/20260602150931.49629-2-leon.hwang@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 748835fcb2 ]
With deferrable card binding the sound card driver tries to
get the mclk configuration before it is setup in sai3 node.
Fix this by setting the sai clock config for the audio codec as well.
Fixes: d8f9d81265 ("arm64: dts: imx8mp: Add analog audio output on i.MX8MP TQMa8MPxL/MBa8MPxL")
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1521d560b7 ]
Initialize iwmr->access during initial user mem registration so
that it contains a valid value during a subsequent rereg_mr.
Otherwise, a rereg_mr that doesn't set IB_MR_REREG_ACCESS (for
example, one that only changes the PD) ends up clearing the
access flags in HW since iwmr->access is zero-initialized, which
is not intended.
Fixes: 5ac388db27 ("RDMA/irdma: Add support to re-register a memory region")
Link: https://patch.msgid.link/r/20260604154104.4035581-1-jmoroni@google.com
Signed-off-by: Jacob Moroni <jmoroni@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4385ddd654 ]
Sashiko pointed out an unrelated bug during a previous patch:
https://sashiko.dev/#/patchset/20260512183852.614045-1-jmoroni%40google.com
This change fixes the bug by eliminating the cqmr->split field which
was not being set properly and instead just checks the CQ resize
feature flag directly.
The cqmr->split field essentially tracks whether IRDMA_FEATURE_CQ_RESIZE
is set, but it was not being set until CQ creation time, which is _after_
CQ memory registration (the only other place where it is referenced).
As a result, it would always be false during MR registration and would
therefore cause irdma_handle_q_mem to populate cqmr->shadow even for GEN_2
HW and beyond:
cqmr->shadow = (dma_addr_t)arr[req->cq_pages];
The issue is that for GEN_2 and beyond, req->cq_pages may be exactly equal
to iwmr->page_cnt and therefore equal to the size of arr, which would cause
an OOB read by one.
Fixes: b48c24c2d7 ("RDMA/irdma: Implement device supported verb APIs")
Link: https://patch.msgid.link/r/20260602214423.1315105-2-jmoroni@google.com
Signed-off-by: Jacob Moroni <jmoroni@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9b2207bc5c ]
Codex pointed out that cm_init_av_by_path() can call cm_set_av_port()
which takes a reference on the cm device, but then can immediately return
error if ib_init_ah_attr_from_path() fails.
Since callers like ib_send_cm_req() put the av on the stack this leaks
that cm device reference.
Re-order cm_init_av_by_path() so it doesn't touch the av until it has done
all its failable work, and then update the av in one shot so it is either
left alone or fully init'd.
Sashiko also pointed out that the cm_destroy_av() prior to
cm_init_av_by_path() is harmful as it leaves the AV broken in the error
case and thus the REJ won't send. Since cm_init_av_by_path() is now atomic
it is safe to delete the cm_destroy_av(). On succees the av from
cm_init_av_for_response() is cleaned up by cm_init_av_by_path(), on
failure the 'goto rejected' guarentees the av is destroyed during
ib_destroy_cm_id().
Fixes: 76039ac909 ("IB/cm: Protect cm_dev, cm_ports and mad_agent with kref and lock")
Link: https://patch.msgid.link/r/0-v1-38292501f539+14f-ib_cm_av_leak_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b0f02608fb ]
For GRE flows, validate that the ct master helper (if any) is pptp
before calling nf_ct_gre_keymap_destroy(), so the helper data area
can be accessed safely. Note that only the pptp helper provides a
.destroy callback.
Fixes: e56894356f ("netfilter: conntrack: remove l4proto destroy hook")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 35e21a4dcc ]
This infrastructure is not used anymore after moving ct timeout and
helper to use datapath refcount to track object use.
Revert commit c56716c69c ("netfilter: extensions: introduce extension
genid count") this patch disables all ct extensions (leading to NULL)
for unconfirmed conntracks, when this is only targeted at ct helper and
ct timeout. There is also codebase that dereferences the ct extension
without checking for NULL which could lead to crash.
Fixes: c56716c69c ("netfilter: extensions: introduce extension genid count")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9e37388b80 ]
nf_ct_seqadj_init() is called without holding the ct lock. This can race
with nf_ct_seq_adjust() when a connection is in CLOSE state due to an
RST or connection reopening. In addition for SYN_RECV state, concurrent
processing of packets can trigger nf_ct_seq_adjust() too. These
situations create a read/write data race.
As synproxy is the only user of nf_ct_seqadj_init() at the moment, fix
this by holding ct->lock inside nf_ct_seqadj_init() until all is done.
Fixes: 48b1de4c11 ("netfilter: add SYNPROXY core/target")
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a625c94144 ]
The MSS calculation in nf_osf_match_one() manually shifts bytes to
construct a 16-bit value before passing it to ntohs().
This works on little-endian hosts but it does not work on big-endian as
the bytes are being always shifted and set in the same way for all
architectures.
Use get_unaligned_be16() to fix this on big-endian systems. It also
simplifies the code.
Fixes: 11eeef41d5 ("netfilter: passive OS fingerprint xtables match")
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 03ad858ce8 ]
debug_lockres_open() copies inode->i_private into struct debug_lockres and
debug_lockres_release() later drops that pointer with dlm_put(). That
only works if open successfully pins the struct dlm_ctxt.
Today open calls dlm_grab(dlm) but ignores its return value. Once the
last domain unregister has removed the context from dlm_domains,
dlm_grab() returns NULL, yet open still stores the raw pointer and returns
success. The later release path is outside the debugfs removal barrier,
so it can call dlm_put() after dlm_free_ctxt_mem() has freed the context.
KASAN reports this as a slab-use-after-free in dlm_put() called from
debug_lockres_release().
Fail the open when dlm_grab() cannot acquire the reference and unwind the
seq_file private state before returning. That keeps locking_state from
handing out a file descriptor whose release path does not own the
dlm_ctxt.
The buggy scenario involves two paths, with each column showing the order
within that path:
locking_state debugfs open: last domain unregister:
1. debug_lockres_open() reads 1. dlm_unregister_domain() calls
inode->i_private. dlm_complete_dlm_shutdown().
2. debug_lockres_open() calls 2. shutdown removes the dlm_ctxt from
dlm_grab(dlm) and gets NULL. dlm_domains.
3. open still stores the raw dlm 3. final teardown reaches
pointer in dl->dl_ctxt and dlm_free_ctxt_mem() and frees it.
returns success.
4. debug_lockres_release() later
calls dlm_put(dl->dl_ctxt).
Validation reproduced this kernel report:
KASAN slab-use-after-free in dlm_put+0x82/0x200
RIP: 0033:0x7f4d349bc9e0
The buggy address belongs to the object at ffff888103a3c000 which belongs
to the cache kmalloc-2k of size 2048
The buggy address is located 816 bytes inside of freed 2048-byte region
[ffff888103a3c000, ffff888103a3c800)
Write of size 4
Call trace:
dump_stack_lvl+0x66/0xa0 (?:?)
print_report+0xd0/0x630 (?:?)
dlm_put+0x82/0x200 (?:?)
srso_alias_return_thunk+0x5/0xfbef5 (?:?)
__virt_addr_valid+0x188/0x2f0 (?:?)
kasan_report+0xe4/0x120 (?:?)
kasan_check_range+0x105/0x1b0 (?:?)
debug_lockres_release+0x53/0x80 (fs/ocfs2/dlm/dlmdebug.c:587)
dlm_put+0x9/0x200 (?:?)
debug_lockres_release+0x5c/0x80 (fs/ocfs2/dlm/dlmdebug.c:587)
full_proxy_release+0x67/0x90 (?:?)
__fput+0x1df/0x4b0 (?:?)
do_raw_spin_lock+0x10f/0x1b0 (?:?)
fput_close_sync+0xd2/0x170 (?:?)
__x64_sys_close+0x55/0x90 (?:?)
do_syscall_64+0x10c/0x640 (arch/x86/entry/syscall_64.c:87)
irqentry_exit+0xac/0x6e0 (?:?)
entry_SYSCALL_64_after_hwframe+0x77/0x7f (?:?)
Freed by task stack:
kasan_save_stack+0x33/0x60 (?:?)
kasan_save_track+0x14/0x30 (?:?)
kasan_save_free_info+0x3b/0x60 (?:?)
__kasan_slab_free+0x5f/0x80 (?:?)
kfree+0x30f/0x580 (?:?)
dlm_put+0x1ce/0x200 (?:?)
dlm_unregister_domain+0xf6/0xb30 (?:?)
o2cb_cluster_disconnect+0x6b/0x90 (?:?)
ocfs2_cluster_disconnect+0x41/0x70 (?:?)
ocfs2_dlm_shutdown+0x1c4/0x220 (?:?)
ocfs2_dismount_volume+0x38a/0x550 (?:?)
generic_shutdown_super+0xc3/0x220 (?:?)
kill_block_super+0x29/0x60 (?:?)
deactivate_locked_super+0x66/0xe0 (?:?)
cleanup_mnt+0x13d/0x210 (?:?)
task_work_run+0xfa/0x170 (?:?)
exit_to_user_mode_loop+0xd6/0x430 (?:?)
do_syscall_64+0x3cb/0x640 (arch/x86/entry/syscall_64.c:87)
entry_SYSCALL_64_after_hwframe+0x77/0x7f (?:?)
Link: https://lore.kernel.org/20260531044714.1640172-1-rollkingzzc@gmail.com
Fixes: 4e3d24ed1a ("ocfs2/dlm: Dumps the lockres' into a debugfs file")
Assisted-by: Codex:gpt-5.5
Signed-off-by: Zhang Cen <rollkingzzc@gmail.com>
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Jun Piao <piaojun@huawei.com>
Cc: Heming Zhao <heming.zhao@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ca1afd88f5 ]
ocfs2_trim_mainbm() trims the global bitmap in cluster units, but its
too-short range validation only checks sb->s_blocksize.
On filesystems with a cluster size larger than the block size, a FITRIM
range that is at least one block but shorter than one cluster is accepted
and shifted down to len == 0. The later start + len - 1 and len -= ...
arithmetic then underflows and can drive trimming past the requested
range.
Reject ranges shorter than s_clustersize instead. That preserves the
existing -EINVAL behavior for requests that cannot discard even one
allocation unit and keeps zero-cluster trims out of the group walk.
Link: https://lore.kernel.org/20260528151247.361854-1-rollkingzzc@gmail.com
Fixes: aa89762c54 ("ocfs2: return EINVAL if the given range to discard is less than block size")
Assisted-by: Codex:gpt-5.5
Signed-off-by: Zhang Cen <rollkingzzc@gmail.com>
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Jun Piao <piaojun@huawei.com>
Cc: Heming Zhao <heming.zhao@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 93612d48fa ]
The locking_state debugfs iterator snapshots struct ocfs2_lock_res by
value under ocfs2_dlm_tracking_lock and later formats that copy in
ocfs2_dlm_seq_show(). That is fine for the inline fields, but the
userspace fsdlm stack stores the LVB through lksb_fsdlm.sb_lvbptr. Once
the iterator drops the tracking lock, a copied non-NULL sb_lvbptr still
points into the original lockres owner, so teardown can free that
container before the debugfs dump walks the raw LVB bytes.
Rebase the copied sb_lvbptr to the copied l_lksb before dumping the raw
LVB. The seq snapshot already carries the inline LVB storage reserved in
struct ocfs2_dlm_lksb, so the debugfs reader can dump the copied bytes
without borrowing the original lockres lifetime.
The buggy scenario involves two paths, with each column showing the order
within that path:
locking_state reader: lockres teardown:
1. ocfs2_dlm_seq_start()/next() 1. file release or another owner
copies struct ocfs2_lock_res teardown reaches
2. ocfs2_dlm_seq_show() formats ocfs2_lock_res_free()
the copied row 2. the lockres is removed from the
3. ocfs2_dlm_lvb() follows the tracking list
copied sb_lvbptr 3. the owner frees the original
lockres container
Validation reproduced this kernel report:
KASAN slab-use-after-free in ocfs2_dlm_seq_show+0x1bd/0x430
RIP: 0033:0x7f8ec4b1e29d
The buggy address belongs to the object at ffff88810a1e0800 which belongs
to the cache kmalloc-1k of size 1024
The buggy address is located 368 bytes inside of freed 1024-byte region
[ffff88810a1e0800, ffff88810a1e0c00)
Read of size 1
Call trace:
dump_stack_lvl+0x66/0xa0
print_report+0xce/0x630
ocfs2_dlm_seq_show+0x1bd/0x430 (fs/ocfs2/dlmglue.c:3137)
srso_alias_return_thunk+0x5/0xfbef5
__virt_addr_valid+0x19f/0x330
kasan_report+0xe0/0x110
seq_read_iter+0x29d/0x790
seq_read+0x20a/0x280
find_held_lock+0x2b/0x80
rcu_read_unlock+0x18/0x70
full_proxy_read+0x9e/0xd0
vfs_read+0x12c/0x590
ksys_read+0xd2/0x170
do_user_addr_fault+0x65a/0x890
do_syscall_64+0x115/0x6a0 (arch/x86/entry/syscall_64.c:87)
entry_SYSCALL_64_after_hwframe+0x77/0x7f
Allocated by task stack:
kasan_save_stack+0x33/0x60
kasan_save_track+0x14/0x30
__kasan_kmalloc+0xaa/0xb0
ocfs2_file_open+0x13e/0x300
do_dentry_open+0x233/0x7f0
vfs_open+0x5a/0x1b0
path_openat+0x66d/0x1540
do_file_open+0x186/0x2b0
do_sys_openat2+0xce/0x150
__x64_sys_openat+0xd0/0x140
do_syscall_64+0x115/0x6a0 (arch/x86/entry/syscall_64.c:87)
entry_SYSCALL_64_after_hwframe+0x77/0x7f
Freed by task stack:
kasan_save_stack+0x33/0x60
kasan_save_track+0x14/0x30
kasan_save_free_info+0x3b/0x60
__kasan_slab_free+0x5f/0x80
kfree+0x313/0x590
ocfs2_file_release+0x138/0x260
__fput+0x1df/0x4b0
fput_close_sync+0xd2/0x170
__x64_sys_close+0x55/0x90
do_syscall_64+0x115/0x6a0 (arch/x86/entry/syscall_64.c:87)
entry_SYSCALL_64_after_hwframe+0x77/0x7f
Link: https://lore.kernel.org/20260525041726.4112882-1-rollkingzzc@gmail.com
Fixes: cf4d8d75d8 ("ocfs2: add fsdlm to stackglue")
Assisted-by: Codex:gpt-5.5
Signed-off-by: Zhang Cen <rollkingzzc@gmail.com>
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Jun Piao <piaojun@huawei.com>
Cc: Heming Zhao <heming.zhao@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 73a4c02f94 ]
Commit df4d29732f ("perf/x86/amd: Change/fix NMI latency mitigation
to use a timestamp") fixed handling of late-arriving NMIs but limited
the mitigation to processors having X86_FEATURE_PERFCTR_CORE. However,
it is unclear if processors without this feature are also affected.
When Mediated vPMU is enabled on affected hardware, it is also possible
to bypass the fix inside KVM guests if X86_FEATURE_PERFCTR_CORE is
removed from the guest CPUID (e.g. using "-cpu host,-perfctr-core" with
QEMU). Hence, use the mitigation at all times.
Fixes: df4d29732f ("perf/x86/amd: Change/fix NMI latency mitigation to use a timestamp")
Signed-off-by: Sandipan Das <sandipan.das@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/29a3c970da289ab8f24282933bdb36545c0403e8.1780325517.git.sandipan.das@amd.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8a7f2bff21 ]
bpf_prog_query() returns a negative errno on failure.
query_flow_dissector() currently closes the namespace fd and then reads
errno to decide whether -EINVAL means that the running kernel does not
support flow dissector queries.
That errno check controls behavior, not just diagnostics: -EINVAL is
handled as a non-fatal old-kernel case, while any other error makes bpftool
net fail.
The namespace fd is opened read-only, so close() is not expected to
commonly fail in normal use. Still, the BPF_PROG_QUERY error is already
available in err, and reading errno after an intervening close() is
fragile. If close() does change errno, the compatibility branch may be
based on close()'s error instead of the BPF_PROG_QUERY result.
This was reproduced with an LD_PRELOAD fault injector that forced
BPF_PROG_QUERY for BPF_FLOW_DISSECTOR to fail with EINVAL and then
forced close() on the netns fd to fail with EIO. The unpatched bpftool
reported "can't query prog: Input/output error". With this change, the
same injected failure is handled as the intended non-fatal EINVAL
compatibility case.
Use the libbpf-returned error code instead. Keep the existing errno reset
in the non-fatal path to preserve batch mode behavior. The success path
is unchanged.
Fixes: 7f0c57fec8 ("bpftool: show flow_dissector attachment status")
Signed-off-by: Woojin Ji <random6.xyz@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Leon Hwang <leon.hwang@linux.dev>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Acked-by: Quentin Monnet <qmo@kernel.org>
Link: https://lore.kernel.org/bpf/20260603003339.33791-1-random6.xyz@gmail.com
Assisted-by: ChatGPT:gpt-5.5
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit baa286df5f ]
The mca ras sub-blocks (mp0, mp1, mpio) all share the
AMDGPU_RAS_BLOCK__MCA block id and are distinguished only by
sub_block_index. The ras manager object for an mca block is selected
with:
con->objs[AMDGPU_RAS_BLOCK__LAST + head->sub_block_index]
Since the rework in commit 7f544c5488 ("drm/amdgpu: Rework mca ras
sw_init") moved the ras_comm setup into amdgpu_mca_mp*_ras_sw_init() but
left sub_block_index unset, mp0/mp1/mpio all default to index 0 and
collide on the same object slot. mp0 grabs the slot and creates its
sysfs node first; mp1 (and mpio) then find the slot already in use, so
amdgpu_ras_block_late_init() -> amdgpu_ras_sysfs_create() returns
-EINVAL:
amdgpu: mca.mp1 failed to execute ras_block_late_init_default! ret:-22
amdgpu: amdgpu_ras_late_init failed -22
amdgpu: amdgpu_device_ip_late_init failed
amdgpu: Fatal error during GPU init
The error is currently masked because amdgpu_ras_late_init() does not
check the return value of amdgpu_ras_block_late_init_default(), but it
already leaves mp1/mpio without their sysfs nodes and becomes a fatal
init failure as soon as that return value is honored.
Restore the per-sub-block sub_block_index assignment so each mca
sub-block maps to its own object slot.
Fixes: 7f544c5488 ("drm/amdgpu: Rework mca ras sw_init")
Signed-off-by: Yunxiang Li <Yunxiang.Li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 10da12d352 ]
Normally ->s_dentry is cleared when dentry it's pointing to becomes
negative (on eviction, realistically). However, that only happens
if dentry gets to be positive in the first place; in case of inode
allocation failure dentry never becomes positive, so ->d_iput()
is not called at all.
We do part of what normally would've been done by configfs_d_iput()
(dropping the reference to configfs_dirent) manually, but we do
not clear ->s_dentry there. Sloppy as it is, it does not matter in
case of configfs_create_{dir,link}() - there configfs_dirent does
not survive dropping the sole reference to it.
However, for configfs_lookup() it *does* survive, with a dangling
pointer to soon to be freed dentry sitting it its ->s_dentry.
Subsequent getdents(2) in that directory will end up dereferencing
that pointer in order to pick the inode number. Use after free...
This is the minimal fix; the right approach is to set the linkage
between dentry and configfs_dirent only after we know that we have
an inode, but that takes more surgery and the bug had been there
since 2006, so...
Fixes: 3d0f89bb16 ("configfs: Add permission and ownership to configfs objects") # 2.6.16-rc3
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 63c1a12bc0 ]
Commit ed4fb6d7ef ("hrtimer: Use and report correct timerslack values
for realtime tasks") sets timer_slack_ns to 0 for RT tasks in
__setscheduler_params(). However, when an RT task with SCHED_RESET_ON_FORK
creates child threads, the children inherit timer_slack_ns=0 from the
parent. sched_fork() resets the child's policy to SCHED_NORMAL but does
not restore timer_slack_ns, leaving the child permanently running with
zero slack.
Fix this by restoring timer_slack_ns from default_timer_slack_ns in
sched_fork() when resetting from RT/DL to NORMAL policy, matching the
existing behavior in __setscheduler_params().
Note: this fix alone requires a correct default_timer_slack_ns to be
effective. See the following patch for that fix.
Fixes: ed4fb6d7ef ("hrtimer: Use and report correct timerslack values for realtime tasks")
Reported-by: Qiaoting.Lin <linqiaoting@xiaomi.com>
Signed-off-by: Guanyou.Chen <chenguanyou@xiaomi.com>
Signed-off-by: Chunhui.Li <chunhui.li@mediatek.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260522131000.1664983-2-chenguanyou@xiaomi.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a4659be0bc ]
Fix ext2_dio_write_iter() to propagate the error returned by
generic_write_sync() instead of silently discarding it, which could
cause write(2) to return success to userspace on O_SYNC/O_DSYNC files
even when the sync failed.
The correct pattern, already used in ext2_dax_write_iter() in the same
file and in ext4, xfs, f2fs among others, is:
if (ret > 0)
ret = generic_write_sync(iocb, ret);
Found by Linux Verification Center (linuxtesting.org) with SVACE.
[JK: Reflect also filemap_write_and_wait() return value]
Fixes: fb5de4358e ("ext2: Move direct-io to use iomap")
Signed-off-by: Danila Chernetsov <listdansp@mail.ru>
Link: https://patch.msgid.link/20260530122311.136803-1-listdansp@mail.ru
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Johannes Thumshirn
Nikita Zhandarovich
David Laight
Filipe Manana
Ryder Lee
Bjoern A. Zeeb
Hongyan Xia
Wolfram Sang
Hui Zhu
Luca Leonardo Scorcia
Mike Christie
David Disseldorp
Yuho Choi
Al Viro
Dmitry Vyukov
Felix Gu
Jason Gunthorpe
Pengjie Zhang
Kyle Zeng
Rui Qi
HyeongJun An
Tristan Madani
Leon Hwang
Alexander Stein
Jacob Moroni
Pablo Neira Ayuso
Fernando Fernandez Mancera
Joseph Qi
Zhang Cen
Dmitry Antipov
Sandipan Das
Woojin Ji
Yunxiang Li
Alexander Potapenko
Guanyou.Chen
Danila Chernetsov