Have the NFSD encoders annotate the boundaries of every
direct-data-placement eligible result data payload. Then change
svcrdma to use that annotation instead of the xdr->page_len
when handling Write chunks.
For NFSv4 on RDMA, that enables the ability to recognize multiple
result payloads per compound. This is a pre-requisite for supporting
multiple Write chunks per RPC transaction.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
GFP_KERNEL may cause ida_alloc_range() to sleep, but the spinlock covering
this function is not allowed to sleep, so the spinlock needs to be changed
to mutex.
As there is a certain chance of memory allocation failure, GFP_ATOMIC is
not suitable for QP allocation scenarios.
Fixes: 71586dd20010 ("RDMA/hns: Create QP with selected QPN for bank load balance") Link: https://lore.kernel.org/r/1611048513-28663-1-git-send-email-liweihang@huawei.com Signed-off-by: Yangyang Li <liyangyang20@huawei.com> Signed-off-by: Weihang Li <liweihang@huawei.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Olga showed me a case where the client was sending multiple READ_PLUS
calls to the server in parallel, and the server replied
NFS4ERR_OPNOTSUPP to each. The client would fall back to READ for the
first reply, but fail to retry the other calls.
I fix this by removing the test for NFS_CAP_READ_PLUS in
nfs4_read_plus_not_supported(). This allows us to reschedule any
READ_PLUS call that has a NFS4ERR_OPNOTSUPP return value, even after the
capability has been cleared.
Reported-by: Olga Kornievskaia <kolga@netapp.com> Fixes: c567552612ec ("NFS: Add READ_PLUS data segment support") Cc: stable@vger.kernel.org # v5.10+ Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com> Reviewed-by: Benjamin Coddington <bcodding@redhat.com> Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
occured when /sys/bus/ap/a[pq]mask was updated with a relative mask value
(like +0x10-0x12,+60,-90) with one of the numeric values exceeding INT_MAX.
The fix is simple: use unsigned long values for the internal variables. The
correct checks are already in place in the function but a simple int for
the internal variables was used with the possibility to overflow.
Reported-by: Marc Hartmayer <mhartmay@linux.ibm.com> Signed-off-by: Harald Freudenberger <freude@linux.ibm.com> Tested-by: Marc Hartmayer <mhartmay@linux.ibm.com> Reviewed-by: Holger Dengler <dengler@linux.ibm.com> Cc: <stable@vger.kernel.org> Signed-off-by: Heiko Carstens <hca@linux.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is because when finding an entry in ext4_xattr_block_cache_find(), if
ext4_sb_bread() returns -ENOMEM, the ce's e_refcnt, which has already grown
in the __entry_find(), won't be put away, and eventually trigger the above
issue in mb_cache_destroy() due to reference count leakage.
So call mb_cache_entry_put() on the -ENOMEM error branch as a quick fix.
Reported-by: syzbot+dd43bd0f7474512edc47@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=dd43bd0f7474512edc47 Fixes: fb265c9cb49e ("ext4: add ext4_sb_bread() to disambiguate ENOMEM cases") Cc: stable@kernel.org Signed-off-by: Baokun Li <libaokun1@huawei.com> Reviewed-by: Jan Kara <jack@suse.cz> Link: https://lore.kernel.org/r/20240504075526.2254349-2-libaokun@huaweicloud.com Signed-off-by: Theodore Ts'o <tytso@mit.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
__dst_negative_advice() does not enforce proper RCU rules when
sk->dst_cache must be cleared, leading to possible UAF.
RCU rules are that we must first clear sk->sk_dst_cache,
then call dst_release(old_dst).
Note that sk_dst_reset(sk) is implementing this protocol correctly,
while __dst_negative_advice() uses the wrong order.
Given that ip6_negative_advice() has special logic
against RTF_CACHE, this means each of the three ->negative_advice()
existing methods must perform the sk_dst_reset() themselves.
Note the check against NULL dst is centralized in
__dst_negative_advice(), there is no need to duplicate
it in various callbacks.
Many thanks to Clement Lecigne for tracking this issue.
This old bug became visible after the blamed commit, using UDP sockets.
Fixes: a87cb3e48ee8 ("net: Facility to report route quality of connected sockets") Reported-by: Clement Lecigne <clecigne@google.com> Diagnosed-by: Clement Lecigne <clecigne@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Tom Herbert <tom@herbertland.com> Reviewed-by: David Ahern <dsahern@kernel.org> Link: https://lore.kernel.org/r/20240528114353.1794151-1-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
[Lee: Stable backport] Signed-off-by: Lee Jones <lee@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently when the current line should be removed from the display
kdb_read() uses memset() to fill a temporary buffer with spaces.
The problem is not that this could be trivially implemented using a
format string rather than open coding it. The real problem is that
it is possible, on systems with a long kdb_prompt_str, to write past
the end of the tmpbuffer.
Happily, as mentioned above, this can be trivially implemented using a
format string. Make it so!
Cc: stable@vger.kernel.org Reviewed-by: Douglas Anderson <dianders@chromium.org> Tested-by: Justin Stitt <justinstitt@google.com> Link: https://lore.kernel.org/r/20240424-kgdb_read_refactor-v3-5-f236dbe9828d@linaro.org Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The code that handles case 14 (down) and case 16 (up) has been copy and
pasted despite being byte-for-byte identical. Combine them.
Cc: stable@vger.kernel.org # Not a bug fix but it is needed for later bug fixes Reviewed-by: Douglas Anderson <dianders@chromium.org> Tested-by: Justin Stitt <justinstitt@google.com> Link: https://lore.kernel.org/r/20240424-kgdb_read_refactor-v3-4-f236dbe9828d@linaro.org Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, if the cursor position is not at the end of the command buffer
and the user uses the Tab-complete functions, then the console does not
leave the cursor in the correct position.
For example consider the following buffer with the cursor positioned
at the ^:
md kdb_pro 10
^
Pressing tab should result in:
md kdb_prompt_str 10
^
However this does not happen. Instead the cursor is placed at the end
(after then 10) and further cursor movement redraws incorrectly. The
same problem exists when we double-Tab but in a different part of the
code.
Fix this by sending a carriage return and then redisplaying the text to
the left of the cursor.
Cc: stable@vger.kernel.org Reviewed-by: Douglas Anderson <dianders@chromium.org> Tested-by: Justin Stitt <justinstitt@google.com> Link: https://lore.kernel.org/r/20240424-kgdb_read_refactor-v3-3-f236dbe9828d@linaro.org Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently when kdb_read() needs to reposition the cursor it uses copy and
paste code that works by injecting an '\0' at the cursor position before
delivering a carriage-return and reprinting the line (which stops at the
'\0').
Tidy up the code by hoisting the copy and paste code into an appropriately
named function. Additionally let's replace the '\0' injection with a
proper field width parameter so that the string will be abridged during
formatting instead.
Cc: stable@vger.kernel.org # Not a bug fix but it is needed for later bug fixes Tested-by: Justin Stitt <justinstitt@google.com> Reviewed-by: Douglas Anderson <dianders@chromium.org> Link: https://lore.kernel.org/r/20240424-kgdb_read_refactor-v3-2-f236dbe9828d@linaro.org Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, when the user attempts symbol completion with the Tab key, kdb
will use strncpy() to insert the completed symbol into the command buffer.
Unfortunately it passes the size of the source buffer rather than the
destination to strncpy() with predictably horrible results. Most obviously
if the command buffer is already full but cp, the cursor position, is in
the middle of the buffer, then we will write past the end of the supplied
buffer.
Fix this by replacing the dubious strncpy() calls with memmove()/memcpy()
calls plus explicit boundary checks to make sure we have enough space
before we start moving characters around.
Reported-by: Justin Stitt <justinstitt@google.com> Closes: https://lore.kernel.org/all/CAFhGd8qESuuifuHsNjFPR-Va3P80bxrw+LqvC8deA8GziUJLpw@mail.gmail.com/ Cc: stable@vger.kernel.org Reviewed-by: Douglas Anderson <dianders@chromium.org> Reviewed-by: Justin Stitt <justinstitt@google.com> Tested-by: Justin Stitt <justinstitt@google.com> Link: https://lore.kernel.org/r/20240424-kgdb_read_refactor-v3-1-f236dbe9828d@linaro.org Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On AM62x, the watchdog is pet before the valid window is open. Fix
min_hw_heartbeat and accommodate a 2% + static offset safety margin.
The static offset accounts for max hardware error.
Remove the hack in the driver which shifts the open window boundary,
since it is no longer necessary due to the fix mentioned above.
Nick Bowler reported:
When using newer kernels on my Ultra 60 with dual 450MHz UltraSPARC-II
CPUs, I noticed that only CPU 0 comes up, while older kernels (including
4.7) are working fine with both CPUs.
This is a small change that reverts very easily on top of 5.18: there is
just one trivial conflict. Once reverted, both CPUs work again.
Maybe this is related to the fact that the CPUs on this system are
numbered CPU0 and CPU2 (there is no CPU1)?
The current code that adjust cpu_possible based on nr_cpu_ids do not
take into account that CPU's may not come one after each other.
Move the chech to the function that setup the cpu_possible mask
so there is no need to adjust it later.
If p9_check_errors() fails early in p9_client_rpc(), req->rc.tag
will not be properly initialized. However, trace_9p_client_res()
ends up trying to print it out anyway before p9_client_rpc()
finishes.
Fix this issue by assigning default values to p9_fcall fields
such as 'tag' and (just in case KMSAN unearths something new) 'id'
during the tag allocation stage.
Problem
=========
After commit 67f695134703 ("ipv6: Move setting default metric for routes"),
we noticed that the logic of assigning the default value of fc_metirc
changed in the ioctl process. That is, when users use ioctl(fd, SIOCADDRT,
rt) with a non-zero metric to add a route, then they may fail to delete a
route with passing in a metric value of 0 to the kernel by ioctl(fd,
SIOCDELRT, rt). But iproute can succeed in deleting it.
As a reference, when using iproute tools by netlink to delete routes with
a metric parameter equals 0, like the command as follows:
ip -6 route del fe80::/64 via fe81::5054:ff:fe11:3451 dev eth0 metric 0
the user can still succeed in deleting the route entry with the smallest
metric.
Root Reason
===========
After commit 67f695134703 ("ipv6: Move setting default metric for routes"),
When ioctl() pass in SIOCDELRT with a zero metric, rtmsg_to_fib6_config()
will set a defalut value (1024) to cfg->fc_metric in kernel, and in
ip6_route_del() and the line 4074 at net/ipv3/route.c, it will check by
if (cfg->fc_metric && cfg->fc_metric != rt->fib6_metric)
continue;
and the condition is true and skip the later procedure (deleting route)
because cfg->fc_metric != rt->fib6_metric. But before that commit,
cfg->fc_metric is still zero there, so the condition is false and it
will do the following procedure (deleting).
Solution
========
In order to keep a consistent behaviour across netlink() and ioctl(), we
should allow to delete a route with a metric value of 0. So we only do
the default setting of fc_metric in route adding.
CC: stable@vger.kernel.org # 5.4+ Fixes: 67f695134703 ("ipv6: Move setting default metric for routes") Co-developed-by: Fan Yu <fan.yu9@zte.com.cn> Signed-off-by: Fan Yu <fan.yu9@zte.com.cn> Signed-off-by: xu xin <xu.xin16@zte.com.cn> Reviewed-by: David Ahern <dsahern@kernel.org> Link: https://lore.kernel.org/r/20240514201102055dD2Ba45qKbLlUMxu_DTHP@zte.com.cn Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Using completion_done to determine whether the caller has gone
away only works after a complete call. Furthermore it's still
possible that the caller has not yet called wait_for_completion,
resulting in another potential UAF.
Fix this by making the caller use cancel_work_sync and then freeing
the memory safely.
Fixes: 7d42e097607c ("crypto: qat - resolve race condition during AER recovery") Cc: <stable@vger.kernel.org> #6.8+ Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Reviewed-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add module alias with the algorithm cra_name similar to what we have for
RSA-related and other algorithms.
The kernel attempts to modprobe asymmetric algorithms using the names
"crypto-$cra_name" and "crypto-$cra_name-all." However, since these
aliases are currently missing, the modules are not loaded. For instance,
when using the `add_key` function, the hash algorithm is typically
loaded automatically, but the asymmetric algorithm is not.
Steps to test:
1. Cert is generated usings ima-evm-utils test suite with
`gen-keys.sh`, example cert is provided below:
The commit 04e5eac8f3ab("fbdev: savage: Error out if pixclock equals zero")
checks the value of pixclock to avoid divide-by-zero error. However
the function savagefb_probe doesn't handle the error return of
savagefb_check_var. When pixclock is 0, it will cause divide-by-zero error.
Fixes: 04e5eac8f3ab ("fbdev: savage: Error out if pixclock equals zero") Signed-off-by: Cai Xinchen <caixinchen1@huawei.com> Cc: stable@vger.kernel.org Signed-off-by: Helge Deller <deller@gmx.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Some mmc host drivers may need to fixup a card-detection GPIO's config
to e.g. enable the GPIO controllers builtin pull-up resistor on devices
where the firmware description of the GPIO is broken (e.g. GpioInt with
PullNone instead of PullUp in ACPI DSDT).
Since this is the exception rather then the rule adding a config
parameter to mmc_gpiod_request_cd() seems undesirable, so instead
add a new mmc_gpiod_set_cd_config() function. This is simply a wrapper
to call gpiod_set_config() on the card-detect GPIO acquired through
mmc_gpiod_request_cd().
Reviewed-by: Andy Shevchenko <andy@kernel.org> Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Adrian Hunter <adrian.hunter@intel.com> Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20240410191639.526324-2-hdegoede@redhat.com Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After the new V4L2 device node was registered, some additional
initialization was done before the device node was marked as
'registered'. During the time between creating the device node
and marking it as 'registered' it was possible to open the
device node, which would return -ENODEV since the 'registered'
flag was not yet set.
Hold the videodev_lock mutex from just before the device node
is registered until the 'registered' flag is set. Since v4l2_open
will take the same lock, it will wait until this registration
process is finished. This resolves this race condition.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Reviewed-by: Sakari Ailus <sakari.ailus@linux.intel.com> Cc: <stable@vger.kernel.org> # for vi4.18 and up Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is the result of the linked LLVM commit, which changes how the
arrays of structures in config_ts() get handled with
CONFIG_INIT_STACK_ZERO and CONFIG_INIT_STACK_PATTERN, which causes the
above warning in combination with inlining, as config_ts() gets inlined
into probe().
This warning can be easily fixed by moving the array of structures off
of the stackvia 'static const', which is a better location for these
variables anyways because they are static data that is only ever read
from, never modified, so allocating the stack space is wasteful.
This drops the stack usage from 2368 bytes to 256 bytes with the same
compiler and configuration.
First the media device node was created, and if successful it was
marked as 'registered'. This leaves a small race condition where
an application can open the device node and get an error back
because the 'registered' flag was not yet set.
Change the order: first set the 'registered' flag, then actually
register the media device node. If that fails, then clear the flag.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Fixes: cf4b9211b568 ("[media] media: Media device node support") Cc: stable@vger.kernel.org Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Structs rx_desc_92d and rx_fwinfo_92d will not work for big endian
systems.
Delete rx_desc_92d because it's big and barely used, and instead use
the get_rx_desc_rxmcs and get_rx_desc_rxht functions, which work on big
endian systems too.
Fix rx_fwinfo_92d by duplicating four of its members in the correct
order.
Tested only with RTL8192DU, which will use the same code.
Tested only on a little endian system.
Some (all?) management frames are incorrectly reported to mac80211 as
decrypted when actually the hardware did not decrypt them. This results
in speeds 3-5 times lower than expected, 20-30 Mbps instead of 100
Mbps.
Fix this by checking the encryption type field of the RX descriptor.
rtw88 does the same thing.
This fix was tested only with rtl8192du, which will use the same code.
Don't subtract 1 from the power index. This was added in commit 2fc0b8e5a17d ("rtl8xxxu: Add TX power base values for gen1 parts")
for unknown reasons. The vendor drivers don't do this.
Also correct the calculations of values written to
REG_OFDM0_X{C,D}_TX_IQ_IMBALANCE. According to the vendor driver,
these are used for TX power training.
With these changes rtl8xxxu sets the TX power of RTL8192CU the same
as the vendor driver.
None of this appears to have any effect on my RTL8192CU device.
Xiao reported that lvm2 test lvconvert-raid-takeover.sh can hang with
small possibility, the root cause is exactly the same as commit bed9e27baf52 ("Revert "md/raid5: Wait for MD_SB_CHANGE_PENDING in raid5d"")
However, Dan reported another hang after that, and junxiao investigated
the problem and found out that this is caused by plugged bio can't issue
from raid5d().
Current implementation in raid5d() has a weird dependence:
1) md_check_recovery() from raid5d() must hold 'reconfig_mutex' to clear
MD_SB_CHANGE_PENDING;
2) raid5d() handles IO in a deadloop, until all IO are issued;
3) IO from raid5d() must wait for MD_SB_CHANGE_PENDING to be cleared;
This behaviour is introduce before v2.6, and for consequence, if other
context hold 'reconfig_mutex', and md_check_recovery() can't update
super_block, then raid5d() will waste one cpu 100% by the deadloop, until
'reconfig_mutex' is released.
Refer to the implementation from raid1 and raid10, fix this problem by
skipping issue IO if MD_SB_CHANGE_PENDING is still set after
md_check_recovery(), daemon thread will be woken up when 'reconfig_mutex'
is released. Meanwhile, the hang problem will be fixed as well.
Fixes: 5e2cf333b7bd ("md/raid5: Wait for MD_SB_CHANGE_PENDING in raid5d") Cc: stable@vger.kernel.org # v5.19+ Reported-and-tested-by: Dan Moulding <dan@danm.net> Closes: https://lore.kernel.org/all/20240123005700.9302-1-dan@danm.net/ Investigated-by: Junxiao Bi <junxiao.bi@oracle.com> Signed-off-by: Yu Kuai <yukuai3@huawei.com> Link: https://lore.kernel.org/r/20240322081005.1112401-1-yukuai1@huaweicloud.com Signed-off-by: Song Liu <song@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The 'local-bd-address' property is used to pass a unique Bluetooth
device address from the boot firmware to the kernel and should otherwise
be left unset so that the OS can prevent the controller from being used
until a valid address has been provided through some other means (e.g.
using btmgmt).
There is no such device as "as3722@40", because its name is "pmic". Use
phandles for aliases to fix relying on full node path. This corrects
aliases for RTC devices and also fixes dtc W=1 warning:
tegra132-norrin.dts:12.3-36: Warning (alias_paths): /aliases:rtc0: aliases property is not a valid node (/i2c@7000d000/as3722@40)
Commit defc9cd826e4 ("pata_legacy: resychronize with upstream changes and
resubmit") missed to update legacy_exit(), so that it now fails to do any
cleanup -- the loop body there can never be entered. Fix that and finally
remove now useless nr_legacy_host variable...
Found by Linux Verification Center (linuxtesting.org) with the Svace static
analysis tool.
Fixes: defc9cd826e4 ("pata_legacy: resychronize with upstream changes and resubmit") Cc: stable@vger.kernel.org Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru> Reviewed-by: Niklas Cassel <cassel@kernel.org> Signed-off-by: Damien Le Moal <dlemoal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
if the sdma_v4_0_irq_id_to_seq return -EINVAL, the process should
be stop to avoid out-of-bounds read, so directly return -EINVAL.
Signed-off-by: Bob Zhou <bob.zhou@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Le Ma <le.ma@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
F2FS-fs (loop0): Mounted with checkpoint version = 48b305e4
==================================================================
BUG: KASAN: slab-out-of-bounds in f2fs_test_bit fs/f2fs/f2fs.h:2933 [inline]
BUG: KASAN: slab-out-of-bounds in current_nat_addr fs/f2fs/node.h:213 [inline]
BUG: KASAN: slab-out-of-bounds in f2fs_get_node_info+0xece/0x1200 fs/f2fs/node.c:600
Read of size 1 at addr ffff88807a58c76c by task syz-executor280/5076
The root cause is we missed to do sanity check on i_xattr_nid during
f2fs_iget(), so that in fiemap() path, current_nat_addr() will access
nat_bitmap w/ offset from invalid i_xattr_nid, result in triggering
kasan bug report, fix it.
nft_unregister_obj() can concurrent with __nft_obj_type_get(),
and there is not any protection when iterate over nf_tables_objects
list in __nft_obj_type_get(). Therefore, there is potential data-race
of nf_tables_objects list entry.
Use list_for_each_entry_rcu() to iterate over nf_tables_objects
list in __nft_obj_type_get(), and use rcu_read_lock() in the caller
nft_obj_type_get() to protect the entire type query process.
... and I think that's actually the important thing here:
- the first page fault is from user space, and triggers the vsyscall
emulation.
- the second page fault is from __do_sys_gettimeofday(), and that should
just have caused the exception that then sets the return value to
-EFAULT
- the third nested page fault is due to _raw_spin_unlock_irqrestore() ->
preempt_schedule() -> trace_sched_switch(), which then causes a BPF
trace program to run, which does that bpf_probe_read_compat(), which
causes that page fault under pagefault_disable().
It's quite the nasty backtrace, and there's a lot going on.
The problem is literally the vsyscall emulation, which sets
current->thread.sig_on_uaccess_err = 1;
and that causes the fixup_exception() code to send the signal *despite* the
exception being caught.
And I think that is in fact completely bogus. It's completely bogus
exactly because it sends that signal even when it *shouldn't* be sent -
like for the BPF user mode trace gathering.
In other words, I think the whole "sig_on_uaccess_err" thing is entirely
broken, because it makes any nested page-faults do all the wrong things.
Now, arguably, I don't think anybody should enable vsyscall emulation any
more, but this test case clearly does.
I think we should just make the "send SIGSEGV" be something that the
vsyscall emulation does on its own, not this broken per-thread state for
something that isn't actually per thread.
The x86 page fault code actually tried to deal with the "incorrect nesting"
by having that:
if (in_interrupt())
return;
which ignores the sig_on_uaccess_err case when it happens in interrupts,
but as shown by this example, these nested page faults do not need to be
about interrupts at all.
IOW, I think the only right thing is to remove that horrendously broken
code.
The attached patch looks like the ObviouslyCorrect(tm) thing to do.
NOTE! This broken code goes back to this commit in 2011:
4fc3490114bb ("x86-64: Set siginfo and context on vsyscall emulation faults")
... and back then the reason was to get all the siginfo details right.
Honestly, I do not for a moment believe that it's worth getting the siginfo
details right here, but part of the commit says:
This fixes issues with UML when vsyscall=emulate.
... and so my patch to remove this garbage will probably break UML in this
situation.
I do not believe that anybody should be running with vsyscall=emulate in
2024 in the first place, much less if you are doing things like UML. But
let's see if somebody screams.
Reported-and-tested-by: syzbot+83e7f982ca045ab4405c@syzkaller.appspotmail.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Tested-by: Jiri Olsa <jolsa@kernel.org> Acked-by: Andy Lutomirski <luto@kernel.org> Link: https://lore.kernel.org/r/CAHk-=wh9D6f7HUkDgZHKmDCHUQmp+Co89GP+b8+z+G56BKeyNg@mail.gmail.com Signed-off-by: Sasha Levin <sashal@kernel.org>
[gpiccoli: Backport the patch due to differences in the trees. The main change
between 5.10.y and 5.15.y is due to renaming the fixup function, by
commit 6456a2a69ee1 ("x86/fault: Rename no_context() to kernelmode_fixup_or_oops()").
Following 2 commits cause divergence in the diffs too (in the removed lines): cd072dab453a ("x86/fault: Add a helper function to sanitize error code") d4ffd5df9d18 ("x86/fault: Fix wrong signal when vsyscall fails with pkey")
Finally, there is context adjustment in the processor.h file.] Signed-off-by: Guilherme G. Piccoli <gpiccoli@igalia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit f58f45c1e5b9 ("vxlan: drop packets from invalid src-address")
has recently been added to vxlan mainly in the context of source
address snooping/learning so that when it is enabled, an entry in the
FDB is not being created for an invalid address for the corresponding
tunnel endpoint.
Before commit f58f45c1e5b9 vxlan was similarly behaving as geneve in
that it passed through whichever macs were set in the L2 header. It
turns out that this change in behavior breaks setups, for example,
Cilium with netkit in L3 mode for Pods as well as tunnel mode has been
passing before the change in f58f45c1e5b9 for both vxlan and geneve.
After mentioned change it is only passing for geneve as in case of
vxlan packets are dropped due to vxlan_set_mac() returning false as
source and destination macs are zero which for E/W traffic via tunnel
is totally fine.
Fix it by only opting into the is_valid_ether_addr() check in
vxlan_set_mac() when in fact source address snooping/learning is
actually enabled in vxlan. This is done by moving the check into
vxlan_snoop(). With this change, the Cilium connectivity test suite
passes again for both tunnel flavors.
Fixes: f58f45c1e5b9 ("vxlan: drop packets from invalid src-address") Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: David Bauer <mail@david-bauer.net> Cc: Ido Schimmel <idosch@nvidia.com> Cc: Nikolay Aleksandrov <razor@blackwall.org> Cc: Martin KaFai Lau <martin.lau@kernel.org> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org> Reviewed-by: David Bauer <mail@david-bauer.net> Signed-off-by: David S. Miller <davem@davemloft.net>
[ Backport note: vxlan snooping/learning not supported in 6.8 or older,
so commit is simply a revert. ] Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Patch series "nilfs2: fix log writer related issues".
This bug fix series covers three nilfs2 log writer-related issues,
including a timer use-after-free issue and potential deadlock issue on
unmount, and a potential freeze issue in event synchronization found
during their analysis. Details are described in each commit log.
This patch (of 3):
A use-after-free issue has been reported regarding the timer sc_timer on
the nilfs_sc_info structure.
The problem is that even though it is used to wake up a sleeping log
writer thread, sc_timer is not shut down until the nilfs_sc_info structure
is about to be freed, and is used regardless of the thread's lifetime.
Fix this issue by limiting the use of sc_timer only while the log writer
thread is alive.
Link: https://lkml.kernel.org/r/20240520132621.4054-1-konishi.ryusuke@gmail.com Link: https://lkml.kernel.org/r/20240520132621.4054-2-konishi.ryusuke@gmail.com Fixes: fdce895ea5dd ("nilfs2: change sc_timer from a pointer to an embedded one in struct nilfs_sc_info") Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com> Reported-by: "Bai, Shuangpeng" <sjb7183@psu.edu> Closes: https://groups.google.com/g/syzkaller/c/MK_LYqtt8ko/m/8rgdWeseAwAJ Tested-by: Ryusuke Konishi <konishi.ryusuke@gmail.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Don't cross a mountpoint that explicitly specifies a backup volume
(target is <vol>.backup) when starting from a backup volume.
It it not uncommon to mount a volume's backup directly in the volume
itself. This can cause tools that are not paying attention to get
into a loop mounting the volume onto itself as they attempt to
traverse the tree, leading to a variety of problems.
This doesn't prevent the general case of loops in a sequence of
mountpoints, but addresses a common special case in the same way
as other afs clients.
The NOP op flags should have been checked from beginning like any other
opcode, otherwise NOP may not be extended with the op flags.
Given both liburing and Rust io-uring crate always zeros SQE op flags, just
ignore users which play raw NOP uring interface without zeroing SQE, because
NOP is just for test purpose. Then we can save one NOP2 opcode.
Requesting a retune before switching to the RPMB partition has been
observed to cause CRC errors on the RPMB reads (-EILSEQ).
Since RPMB reads can not be retried, the clients would be directly
affected by the errors.
This commit disables the retune request prior to switching to the RPMB
partition: mmc_retune_pause() no longer triggers a retune before the
pause period begins.
This was verified with the sdhci-of-arasan driver (ZynqMP) configured
for HS200 using two separate eMMC cards (DG4064 and 064GB2). In both
cases, the error was easy to reproduce triggering every few tenths of
reads.
With this commit, systems that were utilizing OP-TEE to access RPMB
variables will experience an enhanced performance. Specifically, when
OP-TEE is configured to employ RPMB as a secure storage solution, it not
only writes the data but also the secure filesystem within the
partition. As a result, retrieving any variable involves multiple RPMB
reads, typically around five.
For context, on ZynqMP, each retune request consumed approximately
8ms. Consequently, reading any RPMB variable used to take at the very
minimum 40ms.
After droping the need to retune before switching to the RPMB partition,
this is no longer the case.
The type defined for the BINDER_SET_MAX_THREADS ioctl was changed from
size_t to __u32 in order to avoid incompatibility issues between 32 and
64-bit kernels. However, the internal types used to copy from user and
store the value were never updated. Use u32 to fix the inconsistency.
Fixes: a9350fc859ae ("staging: android: binder: fix BINDER_SET_MAX_THREADS declaration") Reported-by: Arve Hjønnevåg <arve@android.com> Cc: stable@vger.kernel.org Signed-off-by: Carlos Llamas <cmllamas@google.com> Reviewed-by: Alice Ryhl <aliceryhl@google.com> Link: https://lore.kernel.org/r/20240421173750.3117808-1-cmllamas@google.com
[cmllamas: resolve minor conflicts due to missing commit 421518a2740f] Signed-off-by: Carlos Llamas <cmllamas@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A potential deadlock was found by Zheng Zhang with a local syzkaller
instance.
The problem is that when a non-blocking CEC transmit is canceled by calling
cec_data_cancel, that in turn can call the high-level received() driver
callback, which can call cec_transmit_msg() to transmit a new message.
The cec_data_cancel() function is called with the adap->lock mutex held,
and cec_transmit_msg() tries to take that same lock.
The root cause is that the received() callback can either be used to pass
on a received message (and then adap->lock is not held), or to report a
canceled transmit (and then adap->lock is held).
This is confusing, so create a new low-level adap_nb_transmit_canceled
callback that reports back that a non-blocking transmit was canceled.
And the received() callback is only called when a message is received,
as was the case before commit f9d0ecbf56f4 ("media: cec: correctly pass
on reply results") complicated matters.
The absence of IRQD_MOVE_PCNTXT prevents immediate effectiveness of
interrupt affinity reconfiguration via procfs. Instead, the change is
deferred until the next instance of the interrupt being triggered on the
original CPU.
When the interrupt next triggers on the original CPU, the new affinity is
enforced within __irq_move_irq(). A vector is allocated from the new CPU,
but the old vector on the original CPU remains and is not immediately
reclaimed. Instead, apicd->move_in_progress is flagged, and the reclaiming
process is delayed until the next trigger of the interrupt on the new CPU.
Upon the subsequent triggering of the interrupt on the new CPU,
irq_complete_move() adds a task to the old CPU's vector_cleanup list if it
remains online. Subsequently, the timer on the old CPU iterates over its
vector_cleanup list, reclaiming old vectors.
However, a rare scenario arises if the old CPU is outgoing before the
interrupt triggers again on the new CPU.
In that case irq_force_complete_move() is not invoked on the outgoing CPU
to reclaim the old apicd->prev_vector because the interrupt isn't currently
affine to the outgoing CPU, and irq_needs_fixup() returns false. Even
though __vector_schedule_cleanup() is later called on the new CPU, it
doesn't reclaim apicd->prev_vector; instead, it simply resets both
apicd->move_in_progress and apicd->prev_vector to 0.
As a result, the vector remains unreclaimed in vector_matrix, leading to a
CPU vector leak.
To address this issue, move the invocation of irq_force_complete_move()
before the irq_needs_fixup() call to reclaim apicd->prev_vector, if the
interrupt is currently or used to be affine to the outgoing CPU.
Additionally, reclaim the vector in __vector_schedule_cleanup() as well,
following a warning message, although theoretically it should never see
apicd->move_in_progress with apicd->prev_cpu pointing to an offline CPU.
Fixes: f0383c24b485 ("genirq/cpuhotplug: Add support for cleaning up move in progress") Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20240522220218.162423-1-dongli.zhang@oracle.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently ALSA timer doesn't have the lower limit of the start tick
time, and it allows a very small size, e.g. 1 tick with 1ns resolution
for hrtimer. Such a situation may lead to an unexpected RCU stall,
where the callback repeatedly queuing the expire update, as reported
by fuzzer.
This patch introduces a sanity check of the timer start tick time, so
that the system returns an error when a too small start size is set.
As of this patch, the lower limit is hard-coded to 100us, which is
small enough but can still work somehow.
Reported-by: syzbot+43120c2af6ca2938cc38@syzkaller.appspotmail.com Closes: https://lore.kernel.org/r/000000000000fa00a1061740ab6d@google.com Cc: <stable@vger.kernel.org> Link: https://lore.kernel.org/r/20240514182745.4015-1-tiwai@suse.de Signed-off-by: Takashi Iwai <tiwai@suse.de>
[ backport note: the error handling is changed, as the original commit
is based on the recent cleanup with guard() in commit beb45974dd49
-- tiwai ] Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The warning triggers as this:
packet_sendmsg
packet_snd //skb->sk is packet sk
__dev_queue_xmit
__dev_xmit_skb //q->enqueue is not NULL
__qdisc_run
sch_direct_xmit
dev_hard_start_xmit
ipvlan_start_xmit
ipvlan_xmit_mode_l3 //l3 mode
ipvlan_process_outbound //vepa flag
ipvlan_process_v6_outbound
ip6_local_out
__ip6_finish_output
ip6_finish_output2 //multicast packet
sk_mc_loop //sk->sk_family is AF_PACKET
Call ip{6}_local_out() with NULL sk in ipvlan as other tunnels to fix this.
Fixes: 2ad7bf363841 ("ipvlan: Initial check-in of the IPVLAN driver.") Suggested-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Yue Haibing <yuehaibing@huawei.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/r/20240529095633.613103-1-yuehaibing@huawei.com Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The dev_warn to notify about a spurious interrupt was introduced with
the reasoning that these are unexpected. However spurious interrupts
tend to trigger continously and the error message on the serial console
prevents that the core's detection of spurious interrupts kicks in
(which disables the irq) and just floods the console.
Fixes: c64e7efe46b7 ("spi: stm32: make spurious and overrun interrupts visible") Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Link: https://msgid.link/r/20240521105241.62400-2-u.kleine-koenig@pengutronix.de Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Currently, comparisons to 'm' or 'n' result in incorrect output.
[Test Code]
config MODULES
def_bool y
modules
config A
def_tristate m
config B
def_bool A > n
CONFIG_B is unset, while CONFIG_B=y is expected.
The reason for the issue is because Kconfig compares the tristate values
as strings.
Currently, the .type fields in the constant symbol definitions,
symbol_{yes,mod,no} are unspecified, i.e., S_UNKNOWN.
When expr_calc_value() evaluates 'A > n', it checks the types of 'A' and
'n' to determine how to compare them.
The left-hand side, 'A', is a tristate symbol with a value of 'm', which
corresponds to a numeric value of 1. (Internally, 'y', 'm', and 'n' are
represented as 2, 1, and 0, respectively.)
The right-hand side, 'n', has an unknown type, so it is treated as the
string "n" during the comparison.
expr_calc_value() compares two values numerically only when both can
have numeric values. Otherwise, they are compared as strings.
symbol numeric value ASCII code
-------------------------------------
y 2 0x79
m 1 0x6d
n 0 0x6e
'm' is greater than 'n' if compared numerically (since 1 is greater
than 0), but smaller than 'n' if compared as strings (since the ASCII
code 0x6d is smaller than 0x6e).
Specifying .type=S_TRISTATE for symbol_{yes,mod,no} fixes the above
test code.
Doing so, however, would cause a regression to the following test code.
[Test Code 2]
config MODULES
def_bool n
modules
config A
def_tristate n
config B
def_bool A = m
You would get CONFIG_B=y, while CONFIG_B should not be set.
The reason is because sym_get_string_value() turns 'm' into 'n' when the
module feature is disabled. Consequently, expr_calc_value() evaluates
'A = n' instead of 'A = m'. This oddity has been hidden because the type
of 'm' was previously S_UNKNOWN instead of S_TRISTATE.
sym_get_string_value() should not tweak the string because the tristate
value has already been correctly calculated. There is no reason to
return the string "n" where its tristate value is mod.
Fixes: 31847b67bec0 ("kconfig: allow use of relations other than (in)equality") Signed-off-by: Masahiro Yamada <masahiroy@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
syzbot reports:
general protection fault, probably for non-canonical address 0xdffffc0000000003: 0000 [#1] PREEMPT SMP KASAN PTI
KASAN: null-ptr-deref in range [0x0000000000000018-0x000000000000001f]
[..]
RIP: 0010:nf_tproxy_laddr4+0xb7/0x340 net/ipv4/netfilter/nf_tproxy_ipv4.c:62
Call Trace:
nft_tproxy_eval_v4 net/netfilter/nft_tproxy.c:56 [inline]
nft_tproxy_eval+0xa9a/0x1a00 net/netfilter/nft_tproxy.c:168
__in_dev_get_rcu() can return NULL, so check for this.
Reported-and-tested-by: syzbot+b94a6818504ea90d7661@syzkaller.appspotmail.com Fixes: cc6eb4338569 ("tproxy: use the interface primary IP address as a default value for --on-ip") Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When fec_probe() fails or fec_drv_remove() needs to release the
fec queue and remove a NAPI context, therefore add a function
corresponding to fec_enet_init() and call fec_enet_deinit() which
does the opposite to release memory and remove a NAPI context.
Fixes: 59d0f7465644 ("net: fec: init multi queue date structure") Signed-off-by: Xiaolei Wang <xiaolei.wang@windriver.com> Reviewed-by: Wei Fang <wei.fang@nxp.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Link: https://lore.kernel.org/r/20240524050528.4115581-1-xiaolei.wang@windriver.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
We have seen an influx of syzkaller reports where a BPF program attached to
a tracepoint triggers a locking rule violation by performing a map_delete
on a sockmap/sockhash.
We don't intend to support this artificial use scenario. Extend the
existing verifier allowed-program-type check for updating sockmap/sockhash
to also cover deleting from a map.
From now on only BPF programs which were previously allowed to update
sockmap/sockhash can delete from these map types.
Fixes: ff9105993240 ("bpf, sockmap: Prevent lock inversion deadlock in map delete elem") Reported-by: Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp> Reported-by: syzbot+ec941d6e24f633a59172@syzkaller.appspotmail.com Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Tested-by: syzbot+ec941d6e24f633a59172@syzkaller.appspotmail.com Acked-by: John Fastabend <john.fastabend@gmail.com> Closes: https://syzkaller.appspot.com/bug?extid=ec941d6e24f633a59172 Link: https://lore.kernel.org/bpf/20240527-sockmap-verify-deletes-v1-1-944b372f2101@cloudflare.com Signed-off-by: Sasha Levin <sashal@kernel.org>
LED Select (LED_SEL) bit in the LED General Purpose IO Configuration
register is used to determine the functionality of external LED pins
(Speed Indicator, Link and Activity Indicator, Full Duplex Link
Indicator). The default value for this bit is 0 when no EEPROM is
present. If a EEPROM is present, the default value is the value of the
LED Select bit in the Configuration Flags of the EEPROM. A USB Reset or
Lite Reset (LRST) will cause this bit to be restored to the image value
last loaded from EEPROM, or to be set to 0 if no EEPROM is present.
While configuring the dual purpose GPIO/LED pins to LED outputs in the
LED General Purpose IO Configuration register, the LED_SEL bit is changed
as 0 and resulting the configured value from the EEPROM is cleared. The
issue is fixed by using read-modify-write approach.
Fixes: f293501c61c5 ("smsc95xx: configure LED outputs") Signed-off-by: Parthiban Veerasooran <Parthiban.Veerasooran@microchip.com> Reviewed-by: Simon Horman <horms@kernel.org> Reviewed-by: Woojung Huh <woojung.huh@microchip.com> Link: https://lore.kernel.org/r/20240523085314.167650-1-Parthiban.Veerasooran@microchip.com Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
enic_set_vf_port assumes that the nl attribute IFLA_PORT_PROFILE
is of length PORT_PROFILE_MAX and that the nl attributes
IFLA_PORT_INSTANCE_UUID, IFLA_PORT_HOST_UUID are of length PORT_UUID_MAX.
These attributes are validated (in the function do_setlink in rtnetlink.c)
using the nla_policy ifla_port_policy. The policy defines IFLA_PORT_PROFILE
as NLA_STRING, IFLA_PORT_INSTANCE_UUID as NLA_BINARY and
IFLA_PORT_HOST_UUID as NLA_STRING. That means that the length validation
using the policy is for the max size of the attributes and not on exact
size so the length of these attributes might be less than the sizes that
enic_set_vf_port expects. This might cause an out of bands
read access in the memcpys of the data of these
attributes in enic_set_vf_port.
Fixes: f8bd909183ac ("net: Add ndo_{set|get}_vf_port support for enic dynamic vnics") Signed-off-by: Roded Zats <rzats@paloaltonetworks.com> Link: https://lore.kernel.org/r/20240522073044.33519-1-rzats@paloaltonetworks.com Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
err is a 32-bit integer, but elf_update returns an off_t, which is 64-bit
at least on 64-bit platforms. If symbols_patch is called on a binary between
2-4GB in size, the result will be negative when cast to a 32-bit integer,
which the code assumes means an error occurred. This can wrongly trigger
build failures when building very large kernel images.
Fixes: fbbb68de80a4 ("bpf: Add resolve_btfids tool to resolve BTF IDs in ELF object") Signed-off-by: Friedrich Vock <friedrich.vock@gmx.de> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20240514070931.199694-1-friedrich.vock@gmx.de Signed-off-by: Sasha Levin <sashal@kernel.org>
Since commit a6aa8fca4d79 ("dma-buf/sw-sync: Reduce irqsave/irqrestore from
known context") by error replaced spin_unlock_irqrestore() with
spin_unlock_irq() for both sync_debugfs_show() and sync_print_obj() despite
sync_print_obj() is called from sync_debugfs_show(), lockdep complains
inconsistent lock state warning.
Use plain spin_{lock,unlock}() for sync_print_obj(), for
sync_debugfs_show() is already using spin_{lock,unlock}_irq().
Reported-by: syzbot <syzbot+a225ee3df7e7f9372dbe@syzkaller.appspotmail.com> Closes: https://syzkaller.appspot.com/bug?extid=a225ee3df7e7f9372dbe Fixes: a6aa8fca4d79 ("dma-buf/sw-sync: Reduce irqsave/irqrestore from known context") Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Reviewed-by: Christian König <christian.koenig@amd.com> Link: https://patchwork.freedesktop.org/patch/msgid/c2e46020-aaa6-4e06-bf73-f05823f913f0@I-love.SAKURA.ne.jp Signed-off-by: Christian König <christian.koenig@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Previously, the driver incorrectly used rx_dropped to report device
buffer exhaustion.
According to the documentation, rx_dropped should not be used to count
packets dropped due to buffer exhaustion, which is the purpose of
rx_missed_errors.
Use rx_missed_errors as intended for counting packets dropped due to
buffer exhaustion.
Fixes: 269e6b3af3bf ("net/mlx5e: Report additional error statistics in get stats ndo") Signed-off-by: Carolina Jubran <cjubran@nvidia.com> Signed-off-by: Tariq Toukan <tariqt@nvidia.com> Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
When disabling an nvmet namespace, there is a period where the
subsys->lock is released, as the ns disable waits for backend IO to
complete, and the ns percpu ref to be properly killed. The original
intent was to avoid taking the subsystem lock for a prolong period as
other processes may need to acquire it (for example new incoming
connections).
However, it opens up a window where another process may come in and
enable the ns, (re)intiailizing the ns percpu_ref, causing the disable
sequence to hang.
Solve this by taking the global nvmet_config_sem over the entire configfs
enable/disable sequence.
Fixes: a07b4970f464 ("nvmet: add a generic NVMe target") Signed-off-by: Sagi Grimberg <sagi@grimberg.me> Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com> Signed-off-by: Keith Busch <kbusch@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
There is no need to set the DMA mapped flag of the message if it has
no mapped transfers. Moreover, it may give the code a chance to take
the wrong paths, i.e. to exercise DMA related APIs on unmapped data.
Make __spi_map_msg() to bail earlier on the above mentioned cases.
Fixes: 99adef310f68 ("spi: Provide core support for DMA mapping transfers") Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://msgid.link/r/20240522171018.3362521-2-andriy.shevchenko@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
f41f72d09ee1 ("netfilter: nft_payload: simplify vlan header handling")
already allows to match on inner vlan tags by subtract the vlan header
size to the payload offset which has been popped and stored in skbuff
metadata fields.
When nci_rx_work() receives a zero-length payload packet, it should not
discard the packet and exit the loop. Instead, it should continue
processing subsequent packets.
Fixes: d24b03535e5e ("nfc: nci: Fix uninit-value in nci_dev_up and nci_ntf_packet") Signed-off-by: Ryosuke Yasuoka <ryasuoka@redhat.com> Reviewed-by: Simon Horman <horms@kernel.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/20240521153444.535399-1-ryasuoka@redhat.com Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit 7e8cdc97148c ("nfc: Add KCOV annotations") added
kcov_remote_start_common()/kcov_remote_stop() pair into nci_rx_work(),
with an assumption that kcov_remote_stop() is called upon continue of
the for loop. But commit d24b03535e5e ("nfc: nci: Fix uninit-value in
nci_dev_up and nci_ntf_packet") forgot to call kcov_remote_stop() before
break of the for loop.
Reported-by: syzbot <syzbot+0438378d6f157baae1a2@syzkaller.appspotmail.com> Closes: https://syzkaller.appspot.com/bug?extid=0438378d6f157baae1a2 Fixes: d24b03535e5e ("nfc: nci: Fix uninit-value in nci_dev_up and nci_ntf_packet") Suggested-by: Andrey Konovalov <andreyknvl@gmail.com> Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/6d10f829-5a0c-405a-b39a-d7266f3a1a0b@I-love.SAKURA.ne.jp Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Stable-dep-of: 6671e352497c ("nfc: nci: Fix handling of zero-length payload packets in nci_rx_work()") Signed-off-by: Sasha Levin <sashal@kernel.org>
In tls_init(), a write memory barrier is missing, and store-store
reordering may cause NULL dereference in tls_{setsockopt,getsockopt}.
CPU0 CPU1
----- -----
// In tls_init()
// In tls_ctx_create()
ctx = kzalloc()
ctx->sk_proto = READ_ONCE(sk->sk_prot) -(1)
// In update_sk_prot()
WRITE_ONCE(sk->sk_prot, tls_prots) -(2)
// In sock_common_setsockopt()
READ_ONCE(sk->sk_prot)->setsockopt()
// In tls_{setsockopt,getsockopt}()
ctx->sk_proto->setsockopt() -(3)
In the above scenario, when (1) and (2) are reordered, (3) can observe
the NULL value of ctx->sk_proto, causing NULL dereference.
To fix it, we rely on rcu_assign_pointer() which implies the release
barrier semantic. By moving rcu_assign_pointer() after ctx->sk_proto is
initialized, we can ensure that ctx->sk_proto are visible when
changing sk->sk_prot.
The assignment of pps_enable is protected by tmreg_lock, but the read
operation of pps_enable is not. So the Coverity tool reports a lock
evasion warning which may cause data race to occur when running in a
multithread environment. Although this issue is almost impossible to
occur, we'd better fix it, at least it seems more logically reasonable,
and it also prevents Coverity from continuing to issue warnings.
Fixes: 278d24047891 ("net: fec: ptp: Enable PPS output based on ptp clock") Signed-off-by: Wei Fang <wei.fang@nxp.com> Link: https://lore.kernel.org/r/20240521023800.17102-1-wei.fang@nxp.com Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
When request_irq() fails, error path calls vp_del_vqs(). There, as vq is
present in the list, free_irq() is called for the same vector. That
causes following splat:
When CONFIG_DEBUG_BUGVERBOSE=n, we fail to add necessary padding bytes
to bug_table entries, and as a result the last entry in a bug table will
be ignored, potentially leading to an unexpected panic(). All prior
entries in the table will be handled correctly.
The arm64 ABI requires that struct fields of up to 8 bytes are
naturally-aligned, with padding added within a struct such that struct
are suitably aligned within arrays.
When CONFIG_DEBUG_BUGVERPOSE=y, the layout of a bug_entry is:
struct bug_entry {
signed int bug_addr_disp; // 4 bytes
signed int file_disp; // 4 bytes
unsigned short line; // 2 bytes
unsigned short flags; // 2 bytes
}
... with 12 bytes total, requiring 4-byte alignment.
When CONFIG_DEBUG_BUGVERBOSE=n, the layout of a bug_entry is:
struct bug_entry {
signed int bug_addr_disp; // 4 bytes
unsigned short flags; // 2 bytes
< implicit padding > // 2 bytes
}
... with 8 bytes total, with 6 bytes of data and 2 bytes of trailing
padding, requiring 4-byte alginment.
When we create a bug_entry in assembly, we align the start of the entry
to 4 bytes, which implicitly handles padding for any prior entries.
However, we do not align the end of the entry, and so when
CONFIG_DEBUG_BUGVERBOSE=n, the final entry lacks the trailing padding
bytes.
For the main kernel image this is not a problem as find_bug() doesn't
depend on the trailing padding bytes when searching for entries:
for (bug = __start___bug_table; bug < __stop___bug_table; ++bug)
if (bugaddr == bug_addr(bug))
return bug;
However for modules, module_bug_finalize() depends on the trailing
bytes when calculating the number of entries:
Open vSwitch is originally intended to switch at layer 2, only dealing with
Ethernet frames. With the introduction of l3 tunnels support, it crossed
into the realm of needing to care a bit about some routing details when
making forwarding decisions. If an oversized packet would need to be
fragmented during this forwarding decision, there is a chance for pmtu
to get involved and generate a routing exception. This is gated by the
skbuff->pkt_type field.
When a flow is already loaded into the openvswitch module this field is
set up and transitioned properly as a packet moves from one port to
another. In the case that a packet execute is invoked after a flow is
newly installed this field is not properly initialized. This causes the
pmtud mechanism to omit sending the required exception messages across
the tunnel boundary and a second attempt needs to be made to make sure
that the routing exception is properly setup. To fix this, we set the
outgoing packet's pkt_type to PACKET_OUTGOING, since it can only get
to the openvswitch module via a port device or packet command.
Even for bridge ports as users, the pkt_type needs to be reset when
doing the transmit as the packet is truly outgoing and routing needs
to get involved post packet transformations, in the case of
VXLAN/GENEVE/udp-tunnel packets. In general, the pkt_type on output
gets ignored, since we go straight to the driver, but in the case of
tunnel ports they go through IP routing layer.
This issue is periodically encountered in complex setups, such as large
openshift deployments, where multiple sets of tunnel traversal occurs.
A way to recreate this is with the ovn-heater project that can setup
a networking environment which mimics such large deployments. We need
larger environments for this because we need to ensure that flow
misses occur. In these environment, without this patch, we can see:
./ovn_cluster.sh start
podman exec ovn-chassis-1 ip r a 170.168.0.5/32 dev eth1 mtu 1200
podman exec ovn-chassis-1 ip netns exec sw01p1 ip r flush cache
podman exec ovn-chassis-1 ip netns exec sw01p1 \
ping 21.0.0.3 -M do -s 1300 -c2
PING 21.0.0.3 (21.0.0.3) 1300(1328) bytes of data.
From 21.0.0.3 icmp_seq=2 Frag needed and DF set (mtu = 1142)
--- 21.0.0.3 ping statistics ---
...
Using tcpdump, we can also see the expected ICMP FRAG_NEEDED message is not
sent into the server.
With this patch, setting the pkt_type, we see the following:
podman exec ovn-chassis-1 ip netns exec sw01p1 \
ping 21.0.0.3 -M do -s 1300 -c2
PING 21.0.0.3 (21.0.0.3) 1300(1328) bytes of data.
From 21.0.0.3 icmp_seq=1 Frag needed and DF set (mtu = 1222)
ping: local error: message too long, mtu=1222
--- 21.0.0.3 ping statistics ---
...
In this case, the first ping request receives the FRAG_NEEDED message and
a local routing exception is created.
Under the scenario of IB device bonding, when bringing down one of the
ports, or all ports, we saw xprtrdma entering a non-recoverable state
where it is not even possible to complete the disconnect and shut it
down the mount, requiring a reboot. Following debug, we saw that
transport connect never ended after receiving the
RDMA_CM_EVENT_DEVICE_REMOVAL callback.
The DEVICE_REMOVAL callback is irrespective of whether the CM_ID is
connected, and ESTABLISHED may not have happened. So need to work with
each of these states accordingly.
Fixes: 2acc5cae2923 ('xprtrdma: Prevent dereferencing r_xprt->rx_ep after it is freed') Cc: Sagi Grimberg <sagi.grimberg@vastdata.com> Signed-off-by: Dan Aloni <dan.aloni@vastdata.com> Reviewed-by: Sagi Grimberg <sagi@grimberg.me> Reviewed-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
It used to be quite awhile ago since 1b63a75180c6 ('SUNRPC: Refactor
rpc_clone_client()'), in 2012, that `cl_timeout` was copied in so that
all mount parameters propagate to NFSACL clients. However since that
change, if mount options as follows are given:
These values lead to NFSACL operations not being retried under the
condition of transient network outages with soft mount. Instead, getacl
call fails after 60 seconds with EIO.
The simple fix is to pass the existing client's `cl_timeout` as the new
client timeout.
Cc: Chuck Lever <chuck.lever@oracle.com> Cc: Benjamin Coddington <bcodding@redhat.com> Link: https://lore.kernel.org/all/20231105154857.ryakhmgaptq3hb6b@gmail.com/T/ Fixes: 1b63a75180c6 ('SUNRPC: Refactor rpc_clone_client()') Signed-off-by: Dan Aloni <dan.aloni@vastdata.com> Reviewed-by: Benjamin Coddington <bcodding@redhat.com> Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
syzbot reported the following uninit-value access issue [1]
nci_rx_work() parses received packet from ndev->rx_q. It should be
validated header size, payload size and total packet size before
processing the packet. If an invalid packet is detected, it should be
silently discarded.
Fixes: d24b03535e5e ("nfc: nci: Fix uninit-value in nci_dev_up and nci_ntf_packet") Reported-and-tested-by: syzbot+d7b4dc6cd50410152534@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=d7b4dc6cd50410152534 [1] Signed-off-by: Ryosuke Yasuoka <ryasuoka@redhat.com> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
It took me some time to understand the purpose of the tricky code at
the end of arch/x86/Kconfig.debug.
Without it, the following would be shown:
WARNING: unmet direct dependencies detected for FRAME_POINTER
because
81d387190039 ("x86/kconfig: Consolidate unwinders into multiple choice selection")
removed 'select ARCH_WANT_FRAME_POINTERS'.
The correct and more straightforward approach should have been to move
it where 'select FRAME_POINTER' is located.
Several architectures properly handle the conditional selection of
ARCH_WANT_FRAME_POINTERS. For example, 'config UNWINDER_FRAME_POINTER'
in arch/arm/Kconfig.debug.
Some of the regulators on the BD71828 have common voltage setting for
RUN/SUSPEND/IDLE/LPSR states. The enable control can be set for each
state though.
The driver allows setting the voltage values for these states via
device-tree. As a side effect, setting the voltages for
SUSPEND/IDLE/LPSR will also change the RUN level voltage which is not
desired and can break the system.
The comment in code reflects this behaviour, but it is likely to not
make people any happier. The right thing to do is to allow setting the
enable/disable state at SUSPEND/IDLE/LPSR via device-tree, but to
disallow setting state specific voltages for those regulators.
BUCK1 is a bit different. It only shares the SUSPEND and LPSR state
voltages. The former behaviour of allowing to silently overwrite the
SUSPEND state voltage by LPSR state voltage is also changed here so that
the SUSPEND voltage is prioritized over LPSR voltage.
Prevent setting PMIC state specific voltages for regulators which do not
support it.
If, when waiting for a transmit to finish, the wait is interrupted,
then you might get a "transmit timed out" message, even though the
transmit was interrupted and did not actually time out.
Set transmit_in_progress_aborted to true if the
wait_for_completion_killable() call was interrupted and ensure
that the transmit is properly marked as ABORTED.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Reported-by: Yang, Chenyuan <cy54@illinois.edu> Closes: https://lore.kernel.org/linux-media/PH7PR11MB57688E64ADE4FE82E658D86DA09EA@PH7PR11MB5768.namprd11.prod.outlook.com/ Fixes: 590a8e564c6e ("media: cec: abort if the current transmit was canceled") Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Use call_(void_)op consistently in the CEC core framework. Ditto
for the cec pin ops. And check if !adap->devnode.unregistered before
calling each op. This avoids calls to ops when the device has been
unregistered and the underlying hardware may be gone.
The results of non-blocking transmits were not correctly communicated
to userspace.
Specifically:
1) if a non-blocking transmit was canceled, then rx_status wasn't set to 0
as it should.
2) if the non-blocking transmit succeeded, but the corresponding reply
never arrived (aborted or timed out), then tx_status wasn't set to 0
as it should, and rx_status was hardcoded to ABORTED instead of the
actual reason, such as TIMEOUT. In addition, adap->ops->received() was
never called, so drivers that want to do message processing themselves
would not be informed of the failed reply.
If a transmit-in-progress was canceled, then, once the transmit
is done, mark it as aborted and refrain from retrying the transmit.
To signal this situation the new transmit_in_progress_aborted field is
set to true.
The old implementation would just set adap->transmitting to NULL and
set adap->transmit_in_progress to false, but on the hardware level
the transmit was still ongoing. However, the framework would think
the transmit was aborted, and if a new transmit was issued, then
it could overwrite the HW buffer containing the old transmit with the
new transmit, leading to garbled data on the CEC bus.
Don't enable/disable the adapter if the first fh is opened or the
last fh is closed, instead do this when the adapter is configured
or unconfigured, and also when we enter Monitor All or Monitor Pin
mode for the first time or we exit the Monitor All/Pin mode for the
last time.
However, if needs_hpd is true, then do this when the physical
address is set or cleared: in that case the adapter typically is
powered by the HPD, so it really is disabled when the HPD is low.
This case (needs_hpd is true) was already handled in this way, so
this wasn't changed.
The problem with the old behavior was that if the HPD goes low when
no fh is open, and a transmit was in progress, then the adapter would
be disabled, typically stopping the transmit immediately which
leaves a partial message on the bus, which isn't nice and can confuse
some adapters.
It makes much more sense to disable it only when the adapter is
unconfigured and we're not monitoring the bus, since then you really
won't be using it anymore.
To keep track of this store a CEC activation count and call adap_enable
only when it goes from 0 to 1 or back to 0.
The cec_devnode struct has a lock meant to serialize access
to the fields of this struct. This lock is taken during
device node (un)registration and when opening or releasing a
filehandle to the device node. When the last open filehandle
is closed the cec adapter might be disabled by calling the
adap_enable driver callback with the devnode.lock held.
However, if during that callback a message or event arrives
then the driver will call one of the cec_queue_event()
variants in cec-adap.c, and those will take the same devnode.lock
to walk the open filehandle list.
This obviously causes a deadlock.
This is quite easy to reproduce with the cec-gpio driver since that
uses the cec-pin framework which generated lots of events and uses
a kernel thread for the processing, so when adap_enable is called
the thread is still running and can generate events.
But I suspect that it might also happen with other drivers if an
interrupt arrives signaling e.g. a received message before adap_enable
had a chance to disable the interrupts.
This patch adds a new mutex to serialize access to the fhs list.
When adap_enable() is called the devnode.lock mutex is held, but
not devnode.lock_fhs. The event functions in cec-adap.c will now
use devnode.lock_fhs instead of devnode.lock, ensuring that it is
safe to call those functions from the adap_enable callback.
This specific issue only happens if the last open filehandle is closed
and the physical address is invalid. This is not something that
happens during normal operation, but it does happen when monitoring
CEC traffic (e.g. cec-ctl --monitor) with an unconfigured CEC adapter.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Cc: <stable@vger.kernel.org> # for v5.13 and up Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Stable-dep-of: 47c82aac10a6 ("media: cec: core: avoid recursive cec_claim_log_addrs") Signed-off-by: Sasha Levin <sashal@kernel.org>
projects / thirdparty / kernel / stable.git / log
