103a4908ad4d ("x86/head/64: Disable stack protection for head$(BITS).o")
kernel/head{32,64}.c are compiled with -fno-stack-protector to allow
a call to set_bringup_idt_handler(), which would otherwise have stack
protection enabled with CONFIG_STACKPROTECTOR_STRONG.
While sufficient for that case, there may still be issues with calls to
any external functions that were compiled with stack protection enabled
that in-turn make stack-protected calls, or if the exception handlers
set up by set_bringup_idt_handler() make calls to stack-protected
functions.
Subsequent patches for SEV-SNP CPUID validation support will introduce
both such cases. Attempting to disable stack protection for everything
in scope to address that is prohibitive since much of the code, like the
SEV-ES #VC handler, is shared code that remains in use after boot and
could benefit from having stack protection enabled. Attempting to inline
calls is brittle and can quickly balloon out to library/helper code
where that's not really an option.
Instead, re-enable stack protection for head32.c/head64.c, and make the
appropriate changes to ensure the segment used for the stack canary is
initialized in advance of any stack-protected C calls.
For head64.c:
- The BSP will enter from startup_64() and call into C code
(startup_64_setup_env()) shortly after setting up the stack, which
may result in calls to stack-protected code. Set up %gs early to allow
for this safely.
- APs will enter from secondary_startup_64*(), and %gs will be set up
soon after. There is one call to C code prior to %gs being setup
(__startup_secondary_64()), but it is only to fetch 'sme_me_mask'
global, so just load 'sme_me_mask' directly instead, and remove the
now-unused __startup_secondary_64() function.
For head32.c:
- BSPs/APs will set %fs to __BOOT_DS prior to any C calls. In recent
kernels, the compiler is configured to access the stack canary at
%fs:__stack_chk_guard [1], which overlaps with the initial per-cpu
'__stack_chk_guard' variable in the initial/"master" .data..percpu
area. This is sufficient to allow access to the canary for use
during initial startup, so no changes are needed there.
[1] 3fb0fdb3bbe7 ("x86/stackprotector/32: Make the canary into a regular percpu variable")
[ bp: Massage commit message. ]
Suggested-by: Joerg Roedel <jroedel@suse.de> #for 64-bit %gs set up Signed-off-by: Michael Roth <michael.roth@amd.com> Signed-off-by: Brijesh Singh <brijesh.singh@amd.com> Signed-off-by: Borislav Petkov <bp@suse.de> Link: https://lore.kernel.org/r/20220307213356.2797205-24-brijesh.singh@amd.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
srso_alias_untrain_ret() is special code, even if it is a dummy
which is called in the !SRSO case, so annotate it like its real
counterpart, to address the following objtool splat:
vmlinux.o: warning: objtool: .export_symbol+0x2b290: data relocation to !ENDBR: srso_alias_untrain_ret+0x0
When the sd driver revalidates host-managed SMR disks, it calls
disk_set_zoned() which changes the zone_write_granularity attribute value
to the logical block size regardless of the device type. After that, the sd
driver overwrites the value in sd_zbc_read_zone() with the physical block
size, since ZBC/ZAC requires this for host-managed disks. Between the calls
to disk_set_zoned() and sd_zbc_read_zone(), there exists a window where the
attribute shows the logical block size as the zone_write_granularity value,
which is wrong for host-managed disks. The duration of the window is from
20ms to 200ms, depending on report zone command execution time.
To avoid the wrong zone_write_granularity value between disk_set_zoned()
and sd_zbc_read_zone(), modify the value not in sd_zbc_read_zone() but
just after disk_set_zoned() call.
Fixes: a805a4fa4fa3 ("block: introduce zone_write_granularity limit") Signed-off-by: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com> Link: https://lore.kernel.org/r/20230306063024.3376959-1-shinichiro.kawasaki@wdc.com Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 3fb0fdb3bbe7 ("x86/stackprotector/32: Make the canary into a regular
percpu variable") modified the stackprotector check on 32-bit x86 to check
if gcc supports using %fs as canary. Adjust dummy-tools gcc script to pass
this new test by returning "%fs" rather than "%gs" if it detects
-mstack-protector-guard-reg=fs on command line.
Fixes: 3fb0fdb3bbe7 ("x86/stackprotector/32: Make the canary into a regular percpu variable") Signed-off-by: Michal Kubecek <mkubecek@suse.cz> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The changes are similar to those given in the commit 19b070fefd0d
("VMCI: Fix memcpy() run-time warning in dg_dispatch_as_host()").
Fix filling of the msg and msg_payload in dg_info struct, which prevents a
possible "detected field-spanning write" of memcpy warning that is issued
by the tracking mechanism __fortify_memcpy_chk.
Originally, with strict in order execution, we could complete execution
only when the queue was empty. Preempt-to-busy allows replacement of an
active request that may complete before the preemption is processed by
HW. If that happens, the request is retired from the queue, but the
queue_priority_hint remains set, preventing direct submission until
after the next CS interrupt is processed.
This preempt-to-busy race can be triggered by the heartbeat, which will
also act as the power-management barrier and upon completion allow us to
idle the HW. We may process the completion of the heartbeat, and begin
parking the engine before the CS event that restores the
queue_priority_hint, causing us to fail the assertion that it is MIN.
It is necessary for the queue_priority_hint to be lower than the next
request submission upon waking up, as we rely on the hint to decide when
to kick the tasklet to submit that first request.
Fixes: 22b7a426bbe1 ("drm/i915/execlists: Preempt-to-busy") Closes: https://gitlab.freedesktop.org/drm/intel/issues/10154 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: <stable@vger.kernel.org> # v5.4+ Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240318135906.716055-2-janusz.krzysztofik@linux.intel.com
(cherry picked from commit 98850e96cf811dc2d0a7d0af491caff9f5d49c1e) Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
PAT handling won't do the right thing in COW mappings: the first PTE (or,
in fact, all PTEs) can be replaced during write faults to point at anon
folios. Reliably recovering the correct PFN and cachemode using
follow_phys() from PTEs will not work in COW mappings.
Using follow_phys(), we might just get the address+protection of the anon
folio (which is very wrong), or fail on swap/nonswap entries, failing
follow_phys() and triggering a WARN_ON_ONCE() in untrack_pfn() and
track_pfn_copy(), not properly calling free_pfn_range().
In free_pfn_range(), we either wouldn't call memtype_free() or would call
it with the wrong range, possibly leaking memory.
To fix that, let's update follow_phys() to refuse returning anon folios,
and fallback to using the stored PFN inside vma->vm_pgoff for COW mappings
if we run into that.
We will now properly handle untrack_pfn() with COW mappings, where we
don't need the cachemode. We'll have to fail fork()->track_pfn_copy() if
the first page was replaced by an anon folio, though: we'd have to store
the cachemode in the VMA to make this work, likely growing the VMA size.
For now, lets keep it simple and let track_pfn_copy() just fail in that
case: it would have failed in the past with swap/nonswap entries already,
and it would have done the wrong thing with anon folios.
Simple reproducer to trigger the WARN_ON_ONCE() in untrack_pfn():
Currently, we don't reenable the config if freezing the device failed.
For example, virtio-mem currently doesn't support suspend+resume, and
trying to freeze the device will always fail. Afterwards, the device
will no longer respond to resize requests, because it won't get notified
about config changes.
Let's fix this by re-enabling the config if freezing fails.
Fixes: 22b7050a024d ("virtio: defer config changed notifications") Cc: <stable@kernel.org> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Jason Wang <jasowang@redhat.com> Cc: Xuan Zhuo <xuanzhuo@linux.alibaba.com> Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20240213135425.795001-1-david@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Hook unregistration is deferred to the commit phase, same occurs with
hook updates triggered by the table dormant flag. When both commands are
combined, this results in deleting a basechain while leaving its hook
still registered in the core.
Fixes: 179d9ba5559a ("netfilter: nf_tables: fix table flag updates") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The commit mutex should not be released during the critical section
between nft_gc_seq_begin() and nft_gc_seq_end(), otherwise, async GC
worker could collect expired objects and get the released commit lock
within the same GC sequence.
nf_tables_module_autoload() temporarily releases the mutex to load
module dependencies, then it goes back to replay the transaction again.
Move it at the end of the abort phase after nft_gc_seq_end() is called.
Unlike early commit path stage which triggers a call to abort, an
explicit release of the batch is required on abort, otherwise mutex is
released and commit_list remains in place.
Add WARN_ON_ONCE to ensure commit_list is empty from the abort path
before releasing the mutex.
After this patch, commit_list is always assumed to be empty before
grabbing the mutex, therefore
03c1f1ef1584 ("netfilter: Cleanup nft_net->module_list from nf_tables_exit_net()")
only needs to release the pending modules for registration.
Cc: stable@vger.kernel.org Fixes: c0391b6ab810 ("netfilter: nf_tables: missing validation from the abort path") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
This commit adds NVME_QUIRK_NO_DEEPEST_PS and NVME_QUIRK_BOGUS_NID for
device [126f:2262], which appears to be a generic VID:PID pair used for
many SSDs based on the Silicon Motion SM2262/SM2262EN controller.
Two of my SSDs with this VID:PID pair exhibit the same behavior:
* They frequently have trouble exiting the deepest power state (5),
resulting in the entire disk unresponsive.
Verified by setting nvme_core.default_ps_max_latency_us=10000 and
observing them behaving normally.
* They produce all-zero nguid and eui64 with `nvme id-ns` command.
The offending products are:
* HP SSD EX950 1TB
* HIKVISION C2000Pro 2TB
Signed-off-by: Jiawei Fu <i@ibugone.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Sagi Grimberg <sagi@grimberg.me> Signed-off-by: Keith Busch <kbusch@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
There are some actions with value 'tmp' but 'dst_addr' is checked instead.
It is obvious that a copy-paste error was made here and the value
of variable 'tmp' should be checked here.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Replace seekdir() with rewinddir() in order to fix a localized glibc bug.
One of the glibc patches that stable Gentoo is using causes an improper
directory stream positioning bug on 32bit arm. That in turn ends up as a
floating point exception in iio_generic_buffer.
The attached patch provides a fix by using an equivalent function which
should not cause trouble for other distros and is easier to reason about
in general as it obviously always goes back to to the start.
In function ring_buffer_iter_empty(), cpu_buffer->commit_page is read
while other threads may change it. It may cause the time_stamp that read
in the next line come from a different page. Use READ_ONCE() to avoid
having to reason about compiler optimizations now and in future.
The test type "make_warnings_file" should have no mandatory configuration
parameters other than the ones required by the "build" test type, because
its purpose is to create a file with build warnings that may or may not be
used by other subsequent tests. Currently, the only way to use it as a
stand-alone test is by setting POWER_CYCLE, CONSOLE, SSH_USER,
BUILD_TARGET, TARGET_IMAGE, REBOOT_TYPE and GRUB_MENU.
Newer Lenovo Yogas and Legions with 60Hz/90Hz displays send a wmi event
when Fn + R is pressed. This is intended for use to switch between the
two refresh rates.
Allocate a new KEY_REFRESH_RATE_TOGGLE keycode for it.
Add timeout to cm_destroy_id, so that userspace can trigger any data
collection that would help in analyzing the cause of delay in destroying
the cm_id.
New noinline function helps dtrace/ebpf programs to hook on to it.
Existing functionality isn't changed except triggering a probe-able new
function at every timeout interval.
We have seen cases where CM messages stuck with MAD layer (either due to
software bug or faulty HCA), leading to cm_id getting stuck in the
following call stack. This patch helps in resolving such issues faster.
This reverts commit d52848620de00cde4a3a5df908e231b8c8868250, which was
originally put in place to work around a s2idle failure on this platform
where the NVMe device was inaccessible upon resume.
After extended testing, we found that the firmware's implementation of S3
is buggy and intermittently fails to wake up the system. We need to revert
to s2idle mode.
The NVMe issue has now been solved more precisely in the commit titled
"PCI: Disable D3cold on Asus B1400 PCI-NVMe bridge"
When the NFS client is under extreme load the rpc_wait_queue.qlen counter
can be overflowed. Here is an instant of the backlog queue overflow in a
real world environment shown by drgn helper:
There's issue as follows When do IO fault injection test:
Quota error (device dm-3): find_block_dqentry: Quota for id 101 referenced but not present
Quota error (device dm-3): qtree_read_dquot: Can't read quota structure for id 101
Quota error (device dm-3): do_check_range: Getting block 2021161007 out of range 1-186
Quota error (device dm-3): qtree_read_dquot: Can't read quota structure for id 661
Now, ext4_write_dquot()/ext4_acquire_dquot()/ext4_release_dquot() may commit
inconsistent quota data even if process failed. This may lead to filesystem
corruption.
To ensure filesystem consistent when errors=remount-ro there is need to call
ext4_handle_error() to abort journal.
If one group is marked as block bitmap corrupted, its free blocks cannot
be used and its free count is also deducted from the global
sbi->s_freeclusters_counter. User might be confused about the absent
free space because we can't query the information about corrupted block
groups except unreliable error messages in syslog. So add a hint to show
block bitmap corrupted groups in mb_groups.
I have a CD copy of the original Tom Clancy's Ghost Recon game from
2001. The disc mounts without error on Windows, but on Linux mounting
fails with the message "isofs_fill_super: get root inode failed". The
error originates in isofs_read_inode, which returns -EIO because de_len
is 0. The superblock on this disc appears to be intentionally corrupt as
a form of copy protection.
When the root inode is unusable, instead of giving up immediately, try
to continue with the Joliet file table. This fixes the Ghost Recon CD
and probably other copy-protected CDs too.
Signed-off-by: Alex Henrie <alexhenrie24@gmail.com> Signed-off-by: Jan Kara <jack@suse.cz>
Message-Id: <20240208022134.451490-1-alexhenrie24@gmail.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The call to lpfc_sli4_resume_rpi() in lpfc_rcv_padisc() may return an
unsuccessful status. In such cases, the elsiocb is not issued, the
completion is not called, and thus the elsiocb resource is leaked.
Check return value after calling lpfc_sli4_resume_rpi() and conditionally
release the elsiocb resource.
Signed-off-by: Justin Tee <justin.tee@broadcom.com> Link: https://lore.kernel.org/r/20240131185112.149731-3-justintee8345@gmail.com Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
syzbot is reporting sleep in atomic context in SysV filesystem [1], for
sb_bread() is called with rw_spinlock held.
A "write_lock(&pointers_lock) => read_lock(&pointers_lock) deadlock" bug
and a "sb_bread() with write_lock(&pointers_lock)" bug were introduced by
"Replace BKL for chain locking with sysvfs-private rwlock" in Linux 2.5.12.
Then, "[PATCH] err1-40: sysvfs locking fix" in Linux 2.6.8 fixed the
former bug by moving pointers_lock lock to the callers, but instead
introduced a "sb_bread() with read_lock(&pointers_lock)" bug (which made
this problem easier to hit).
Al Viro suggested that why not to do like get_branch()/get_block()/
find_shared() in Minix filesystem does. And doing like that is almost a
revert of "[PATCH] err1-40: sysvfs locking fix" except that get_branch()
from with find_shared() is called without write_lock(&pointers_lock).
If the number of provided enum IDs in a variable width config register
description does not match the expected number, the checker uses the
expected number for validating the individual enum IDs.
However, this may cause out-of-bounds accesses on the array holding the
enum IDs, leading to bogus enum_id conflict warnings. Worse, if the bug
is an incorrect bit field description (e.g. accidentally using "12"
instead of "-12" for a reserved field), thousands of warnings may be
printed, overflowing the kernel log buffer.
Fix this by limiting the enum ID check to the number of provided enum
IDs.
While input core can work with input->phys set to NULL userspace might
depend on it, so better fail probing if allocation fails. The system must
be in a pretty bad shape for it to happen anyway.
If hci_cmd_sync_complete() is triggered and skb is NULL, then
hdev->req_skb is NULL, which will cause this issue.
Reported-and-tested-by: syzbot+830d9e3fa61968246abd@syzkaller.appspotmail.com Signed-off-by: Edward Adam Davis <eadavis@qq.com> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Many syzbot reports show extreme rtnl pressure, and many of them hint
that smc acquires rtnl in netns creation for no good reason [1]
This patch returns early from smc_pnet_net_init()
if there is no netdevice yet.
I am not even sure why smc_pnet_create_pnetids_list() even exists,
because smc_pnet_netdev_event() is also calling
smc_pnet_add_base_pnetid() when handling NETDEV_UP event.
The get_parent handler looks up a parent of a given dentry, this can be
either a subvolume or a directory. The search is set up with offset -1
but it's never expected to find such item, as it would break allowed
range of inode number or a root id. This means it's a corruption (ext4
also returns this error code).
Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Anand Jain <anand.jain@oracle.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The unhandled case in btrfs_relocate_sys_chunks() loop is a corruption,
as it could be caused only by two impossible conditions:
- at first the search key is set up to look for a chunk tree item, with
offset -1, this is an inexact search and the key->offset will contain
the correct offset upon a successful search, a valid chunk tree item
cannot have an offset -1
- after first successful search, the found_key corresponds to a chunk
item, the offset is decremented by 1 before the next loop, it's
impossible to find a chunk item there due to alignment and size
constraints
Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Anand Jain <anand.jain@oracle.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
In function get_pkg_num() if fopen_or_die() succeeds it returns a file
pointer to be used. But fclose() is never called before returning from
the function.
Signed-off-by: Samasth Norway Ananda <samasth.norway.ananda@oracle.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
kasprintf() returns a pointer to dynamically allocated memory
which can be NULL upon failure. Ensure the allocation was successful
by checking the pointer validity.
We claim to have the AdminQ on our irq0 and thus cpu id 0,
but we need to be sure we set the affinity hint to try to
keep it there.
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com> Reviewed-by: Brett Creeley <brett.creeley@amd.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
If the kernel crashes in a context where printk() calls always
defer printing (such as in NMI or inside a printk_safe section)
then the final panic messages will be deferred to irq_work. But
if irq_work is not available, the messages will not get printed
unless explicitly flushed. The result is that the final
"end Kernel panic" banner does not get printed.
Add one final flush after the last printk() call to make sure
the final panic messages make it out as well.
In 'ath_ant_try_scan()', (most likely) the 2nd LNA's signal
strength should be used in comparison against RSSI when
selecting first LNA as the main one. Compile tested only.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru> Acked-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com> Link: https://msgid.link/20231211172502.25202-1-dmantipov@yandex.ru Signed-off-by: Sasha Levin <sashal@kernel.org>
To be used for adding asm functions to the ignore list. The "aw" is
needed to help the ELF section metadata match GCC-created sections.
Otherwise the linker creates duplicate sections instead of combining
them.
Add CPUID_LNX_5 to track cpufeatures' word 21, and add the appropriate
compile-time assert in KVM to prevent direct lookups on the features in
CPUID_LNX_5. KVM uses X86_FEATURE_* flags to manage guest CPUID, and so
must translate features that are scattered by Linux from the Linux-defined
bit to the hardware-defined bit, i.e. should never try to directly access
scattered features in guest CPUID.
Opportunistically add NR_CPUID_WORDS to enum cpuid_leafs, along with a
compile-time assert in KVM's CPUID infrastructure to ensure that future
additions update cpuid_leafs along with NCAPINTS.
No functional change intended.
Fixes: 7f274e609f3d ("x86/cpufeatures: Add new word for scattered features") Cc: Sandipan Das <sandipan.das@amd.com> Signed-off-by: Sean Christopherson <seanjc@google.com> Acked-by: Dave Hansen <dave.hansen@linux.intel.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Daniel Sneddon <daniel.sneddon@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Current MPTCP servers increment MPTcpExtMPCapableFallbackACK when they
accept non-MPC connections. As reported by Christoph, this is "surprising"
because the counter might become greater than MPTcpExtMPCapableSYNRX.
MPTcpExtMPCapableFallbackACK counter's name suggests it should only be
incremented when a connection was seen using MPTCP options, then a
fallback to TCP has been done. Let's do that by incrementing it when
the subflow context of an inbound MPC connection attempt is dropped.
Also, update mptcp_connect.sh kselftest, to ensure that the
above MIB does not increment in case a pure TCP client connects to a
MPTCP server.
Fixes: fc518953bc9c ("mptcp: add and use MIB counter infrastructure") Cc: stable@vger.kernel.org Reported-by: Christoph Paasch <cpaasch@apple.com> Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/449 Signed-off-by: Davide Caratti <dcaratti@redhat.com> Reviewed-by: Mat Martineau <martineau@kernel.org> Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Link: https://lore.kernel.org/r/20240329-upstream-net-20240329-fallback-mib-v1-1-324a8981da48@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The srso_alias_untrain_ret() dummy thunk in the !CONFIG_MITIGATION_SRSO
case is there only for the altenative in CALL_UNTRAIN_RET to have
a symbol to resolve.
However, testing with kernels which don't have CONFIG_MITIGATION_SRSO
enabled, leads to the warning in patch_return() to fire:
missing return thunk: srso_alias_untrain_ret+0x0/0x10-0x0: eb 0e 66 66 2e
WARNING: CPU: 0 PID: 0 at arch/x86/kernel/alternative.c:826 apply_returns (arch/x86/kernel/alternative.c:826
Put in a plain "ret" there so that gcc doesn't put a return thunk in
in its place which special and gets checked.
The original version of the mitigation would patch in the calls to the
untraining routines directly. That is, the alternative() in UNTRAIN_RET
will patch in the CALL to srso_alias_untrain_ret() directly.
However, even if commit e7c25c441e9e ("x86/cpu: Cleanup the untrain
mess") meant well in trying to clean up the situation, due to micro-
architectural reasons, the untraining routine srso_alias_untrain_ret()
must be the target of a CALL instruction and not of a JMP instruction as
it is done now.
Reshuffle the alternative macros to accomplish that.
Align system call table on 8 bytes. With sys_call_table entry size
of 8 bytes that eliminates the possibility of a system call pointer
crossing cache line boundary.
Modifying a MCA bank's MCA_CTL bits which control which error types to
be reported is done over
/sys/devices/system/machinecheck/
├── machinecheck0
│  ├── bank0
│  ├── bank1
│  ├── bank10
│  ├── bank11
...
sysfs nodes by writing the new bit mask of events to enable.
When the write is accepted, the kernel deletes all current timers and
reinits all banks.
Doing that in parallel can lead to initializing a timer which is already
armed and in the timer wheel, i.e., in use already:
ODEBUG: init active (active state 0) object: ffff888063a28000 object
type: timer_list hint: mce_timer_fn+0x0/0x240 arch/x86/kernel/cpu/mce/core.c:2642
WARNING: CPU: 0 PID: 8120 at lib/debugobjects.c:514
debug_print_object+0x1a0/0x2a0 lib/debugobjects.c:514
Fix that by grabbing the sysfs mutex as the rest of the MCA sysfs code
does.
In the following sequence:
1) of_platform_depopulate()
2) of_overlay_remove()
During the step 1, devices are destroyed and devlinks are removed.
During the step 2, OF nodes are destroyed but
__of_changeset_entry_destroy() can raise warnings related to missing
of_node_put():
ERROR: memory leak, expected refcount 1 instead of 2 ...
Indeed, during the devlink removals performed at step 1, the removal
itself releasing the device (and the attached of_node) is done by a job
queued in a workqueue and so, it is done asynchronously with respect to
function calls.
When the warning is present, of_node_put() will be called but wrongly
too late from the workqueue job.
In order to be sure that any ongoing devlink removals are done before
the of_node destruction, synchronize the of_changeset_destroy() with the
devlink removals.
The commit 80dd33cf72d1 ("drivers: base: Fix device link removal")
introduces a workqueue to release the consumer and supplier devices used
in the devlink.
In the job queued, devices are release and in turn, when all the
references to these devices are dropped, the release function of the
device itself is called.
Nothing is present to provide some synchronisation with this workqueue
in order to ensure that all ongoing releasing operations are done and
so, some other operations can be started safely.
For instance, in the following sequence:
1) of_platform_depopulate()
2) of_overlay_remove()
During the step 1, devices are released and related devlinks are removed
(jobs pushed in the workqueue).
During the step 2, OF nodes are destroyed but, without any
synchronisation with devlink removal jobs, of_overlay_remove() can raise
warnings related to missing of_node_put():
ERROR: memory leak, expected refcount 1 instead of 2
Indeed, the missing of_node_put() call is going to be done, too late,
from the workqueue job execution.
Introduce device_link_wait_removal() to offer a way to synchronize
operations waiting for the end of devlink removals (i.e. end of
workqueue jobs).
Also, as a flushing operation is done on the workqueue, the workqueue
used is moved from a system-wide workqueue to a local one.
This patch addresses an issue with the Panasonic CF-SZ6's existing quirk,
specifically its headset microphone functionality. Previously, the quirk
used ALC269_FIXUP_HEADSET_MODE, which does not support the CF-SZ6's design
of a single 3.5mm jack for both mic and audio output effectively. The
device uses pin 0x19 for the headset mic without jack detection.
Following verification on the CF-SZ6 and discussions with the original
patch author, i determined that the update to
ALC269_FIXUP_ASPIRE_HEADSET_MIC is the appropriate solution. This change
is custom-designed for the CF-SZ6's unique hardware setup, which includes
a single 3.5mm jack for both mic and audio output, connecting the headset
microphone to pin 0x19 without the use of jack detection.
Fixes: 0fca97a29b83 ("ALSA: hda/realtek - Add Panasonic CF-SZ6 headset jack quirk") Signed-off-by: I Gede Agastya Darma Laksana <gedeagas22@gmail.com> Cc: <stable@vger.kernel.org>
Message-ID: <20240401174602.14133-1-gedeagas22@gmail.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The myrb and myrs drivers use an odd way of implementing their sysfs files,
calling snprintf() with a fixed length of 32 bytes to print into a page
sized buffer. One of the strings is actually longer than 32 bytes, which
clang can warn about:
drivers/scsi/myrb.c:1906:10: error: 'snprintf' will always be truncated; specified size is 32, but format string expands to at least 34 [-Werror,-Wformat-truncation]
drivers/scsi/myrs.c:1089:10: error: 'snprintf' will always be truncated; specified size is 32, but format string expands to at least 34 [-Werror,-Wformat-truncation]
These could all be plain sprintf() without a length as the buffer is always
long enough. On the other hand, sysfs files should not be overly long
either, so just double the length to make sure the longest strings don't
get truncated here.
gcc warns about a memcpy() with overlapping pointers because of an
incorrect size calculation:
In file included from include/linux/string.h:369,
from drivers/ata/sata_sx4.c:66:
In function 'memcpy_fromio',
inlined from 'pdc20621_get_from_dimm.constprop' at drivers/ata/sata_sx4.c:962:2:
include/linux/fortify-string.h:97:33: error: '__builtin_memcpy' accessing 4294934464 bytes at offsets 0 and [16, 16400] overlaps 6442385281 bytes at offset -2147450817 [-Werror=restrict]
97 | #define __underlying_memcpy __builtin_memcpy
| ^
include/linux/fortify-string.h:620:9: note: in expansion of macro '__underlying_memcpy'
620 | __underlying_##op(p, q, __fortify_size); \
| ^~~~~~~~~~~~~
include/linux/fortify-string.h:665:26: note: in expansion of macro '__fortify_memcpy_chk'
665 | #define memcpy(p, q, s) __fortify_memcpy_chk(p, q, s, \
| ^~~~~~~~~~~~~~~~~~~~
include/asm-generic/io.h:1184:9: note: in expansion of macro 'memcpy'
1184 | memcpy(buffer, __io_virt(addr), size);
| ^~~~~~
The problem here is the overflow of an unsigned 32-bit number to a
negative that gets converted into a signed 'long', keeping a large
positive number.
Replace the complex calculation with a more readable min() variant
that avoids the warning.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Damien Le Moal <dlemoal@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
In snd_soc_info_volsw(), mask is generated by figuring out the index of
the most significant bit set in max and converting the index to a
bitmask through bit shift 1. Unintended wraparound occurs when max is an
integer value with msb bit set. Since the bit shift value 1 is treated
as an integer type, the left shift operation will wraparound and set
mask to 0 instead of all 1's. In order to fix this, we type cast 1 as
`1ULL` to prevent the wraparound.
Fixes: 7077148fb50a ("ASoC: core: Split ops out of soc-core.c") Signed-off-by: Stephen Lee <slee08177@gmail.com> Link: https://msgid.link/r/20240326010131.6211-1-slee08177@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Several Qualcomm Bluetooth controllers lack persistent storage for the
device address and instead one can be provided by the boot firmware
using the 'local-bd-address' devicetree property.
The Bluetooth bindings clearly states that the address should be
specified in little-endian order, but due to a long-standing bug in the
Qualcomm driver which reversed the address some boot firmware has been
providing the address in big-endian order instead.
The boot firmware in SC7180 Trogdor Chromebooks is known to be affected
so mark the 'local-bd-address' property as broken to maintain backwards
compatibility with older firmware when fixing the underlying driver bug.
Note that ChromeOS always updates the kernel and devicetree in lockstep
so that there is no need to handle backwards compatibility with older
devicetrees.
Fixes: 7ec3e67307f8 ("arm64: dts: qcom: sc7180-trogdor: add initial trogdor and lazor dt") Cc: stable@vger.kernel.org # 5.10 Cc: Rob Clark <robdclark@chromium.org> Reviewed-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Johan Hovold <johan+linaro@kernel.org> Acked-by: Bjorn Andersson <andersson@kernel.org> Reviewed-by: Bjorn Andersson <andersson@kernel.org> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The TX queue should be serviced each time the poll function is called,
even if the full RX work budget has been consumed. This prevents
starvation of the TX queue when RX bandwidth usage is high.
When rx-udp-gro-forwarding is enabled UDP packets might be GROed when
being forwarded. If such packets might land in a tunnel this can cause
various issues and udp_gro_receive makes sure this isn't the case by
looking for a matching socket. This is performed in
udp4/6_gro_lookup_skb but only in the current netns. This is an issue
with tunneled packets when the endpoint is in another netns. In such
cases the packets will be GROed at the UDP level, which leads to various
issues later on. The same thing can happen with rx-gro-list.
We saw this with geneve packets being GROed at the UDP level. In such
case gso_size is set; later the packet goes through the geneve rx path,
the geneve header is pulled, the offset are adjusted and frag_list skbs
are not adjusted with regard to geneve. When those skbs hit
skb_fragment, it will misbehave. Different outcomes are possible
depending on what the GROed skbs look like; from corrupted packets to
kernel crashes.
One example is a BUG_ON[1] triggered in skb_segment while processing the
frag_list. Because gso_size is wrong (geneve header was pulled)
skb_segment thinks there is "geneve header size" of data in frag_list,
although it's in fact the next packet. The BUG_ON itself has nothing to
do with the issue. This is only one of the potential issues.
Looking up for a matching socket in udp_gro_receive is fragile: the
lookup could be extended to all netns (not speaking about performances)
but nothing prevents those packets from being modified in between and we
could still not find a matching socket. It's OK to keep the current
logic there as it should cover most cases but we also need to make sure
we handle tunnel packets being GROed too early.
This is done by extending the checks in udp_unexpected_gso: GSO packets
lacking the SKB_GSO_UDP_TUNNEL/_CSUM bits and landing in a tunnel must
be segmented.
[1] kernel BUG at net/core/skbuff.c:4408!
RIP: 0010:skb_segment+0xd2a/0xf70
__udp_gso_segment+0xaa/0x560
Fixes: 9fd1ff5d2ac7 ("udp: Support UDP fraglist GRO/GSO.") Fixes: 36707061d6ba ("udp: allow forwarding of plain (non-fraglisted) UDP GRO packets") Signed-off-by: Antoine Tenart <atenart@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
To fix the regression introduced by commit 52424f974bc5, which causes
servers hang in very hard to reproduce conditions with resets races.
Using two sources for the information is the root cause.
In this function before the fix bumping v didn't mean bumping vf
pointer. But the code used this variables interchangeably, so stale vf
could point to different/not intended vf.
Remove redundant "v" variable and iterate via single VF pointer across
whole function instead to guarantee VF pointer validity.
Fixes: 52424f974bc5 ("i40e: Fix VF hang when reset is triggered on another VF") Signed-off-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com> Reviewed-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com> Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com> Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de> Tested-by: Rafal Romanowski <rafal.romanowski@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The bug usually affects untrusted VFs, because they are limited to 18 MACs,
it affects them badly, not letting to create MAC all filters.
Not stable to reproduce, it happens when VF user creates MAC filters
when other MACVLAN operations are happened in parallel.
But consequence is that VF can't receive desired traffic.
Fix counter to be bumped only for new or active filters.
Fixes: 621650cabee5 ("i40e: Refactoring VF MAC filters counting to make more reliable") Signed-off-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com> Reviewed-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com> Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de> Tested-by: Rafal Romanowski <rafal.romanowski@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
UDP GRO validates checksums and in udp4/6_gro_complete fraglist packets
are converted to CHECKSUM_UNNECESSARY to avoid later checks. However
this is an issue for CHECKSUM_PARTIAL packets as they can be looped in
an egress path and then their partial checksums are not fixed.
Different issues can be observed, from invalid checksum on packets to
traces like:
Fix this by only converting CHECKSUM_NONE packets to
CHECKSUM_UNNECESSARY by reusing __skb_incr_checksum_unnecessary. All
other checksum types are kept as-is, including CHECKSUM_COMPLETE as
fraglist packets being segmented back would have their skb->csum valid.
Fixes: 9fd1ff5d2ac7 ("udp: Support UDP fraglist GRO/GSO.") Signed-off-by: Antoine Tenart <atenart@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
syzkaller reported infinite recursive calls of fib6_dump_done() during
netlink socket destruction. [1]
From the log, syzkaller sent an AF_UNSPEC RTM_GETROUTE message, and then
the response was generated. The following recvmmsg() resumed the dump
for IPv6, but the first call of inet6_dump_fib() failed at kzalloc() due
to the fault injection. [0]
Here, fib6_dump_done() was set to nlk_sk(sk)->cb.done, and the next call
of inet6_dump_fib() set it to nlk_sk(sk)->cb.args[3]. syzkaller stopped
receiving the response halfway through, and finally netlink_sock_destruct()
called nlk_sk(sk)->cb.done().
fib6_dump_done() calls fib6_dump_end() and nlk_sk(sk)->cb.done() if it
is still not NULL. fib6_dump_end() rewrites nlk_sk(sk)->cb.done() by
nlk_sk(sk)->cb.args[3], but it has the same function, not NULL, calling
itself recursively and hitting the stack guard page.
To avoid the issue, let's set the destructor after kzalloc().
[0]:
FAULT_INJECTION: forcing a failure.
name failslab, interval 1, probability 0, space 0, times 0
CPU: 1 PID: 432110 Comm: syz-executor.3 Not tainted 6.8.0-12821-g537c2e91d354-dirty #11
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
Call Trace:
<TASK>
dump_stack_lvl (lib/dump_stack.c:117)
should_fail_ex (lib/fault-inject.c:52 lib/fault-inject.c:153)
should_failslab (mm/slub.c:3733)
kmalloc_trace (mm/slub.c:3748 mm/slub.c:3827 mm/slub.c:3992)
inet6_dump_fib (./include/linux/slab.h:628 ./include/linux/slab.h:749 net/ipv6/ip6_fib.c:662)
rtnl_dump_all (net/core/rtnetlink.c:4029)
netlink_dump (net/netlink/af_netlink.c:2269)
netlink_recvmsg (net/netlink/af_netlink.c:1988)
____sys_recvmsg (net/socket.c:1046 net/socket.c:2801)
___sys_recvmsg (net/socket.c:2846)
do_recvmmsg (net/socket.c:2943)
__x64_sys_recvmmsg (net/socket.c:3041 net/socket.c:3034 net/socket.c:3034)
The netdev CI runs in a VM and captures serial, so stdout and
stderr get combined. Because there's a missing new line in
stderr the test ends up corrupting KTAP:
# Successok 1 selftests: net: reuseaddr_conflict
which should have been:
# Success
ok 1 selftests: net: reuseaddr_conflict
Fixes: 422d8dc6fd3a ("selftest: add a reuseaddr test") Reviewed-by: Muhammad Usama Anjum <usama.anjum@collabora.com> Link: https://lore.kernel.org/r/20240329160559.249476-1-kuba@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Issue is that ip6erspan_rcv() (and erspan_rcv()) no longer make
sure erspan_base_hdr is present in skb linear part (skb->head)
before getting @ver field from it.
Add the missing pskb_may_pull() calls.
v2: Reload iph pointer in erspan_rcv() after pskb_may_pull()
because skb->head might have changed.
The driver should ensure that same priority is not mapped to multiple
rx queues. From DesignWare Cores Ethernet Quality-of-Service
Databook, section 17.1.29 MAC_RxQ_Ctrl2:
"[...]The software must ensure that the content of this field is
mutually exclusive to the PSRQ fields for other queues, that is,
the same priority is not mapped to multiple Rx queues[...]"
Previously rx_queue_priority() function was:
- clearing all priorities from a queue
- adding new priorities to that queue
After this patch it will:
- first assign new priorities to a queue
- then remove those priorities from all other queues
- keep other priorities previously assigned to that queue
Fixes: a8f5102af2a7 ("net: stmmac: TX and RX queue priority configuration") Fixes: 2142754f8b9c ("net: stmmac: Add MAC related callbacks for XGMAC2") Signed-off-by: Piotr Wejman <piotrwejman90@gmail.com> Link: https://lore.kernel.org/r/20240401192239.33942-1-piotrwejman90@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
syzkaller started using corpuses where a BPF tracing program deletes
elements from a sockmap/sockhash map. Because BPF tracing programs can be
invoked from any interrupt context, locks taken during a map_delete_elem
operation must be hardirq-safe. Otherwise a deadlock due to lock inversion
is possible, as reported by lockdep:
Locks in sockmap are hardirq-unsafe by design. We expects elements to be
deleted from sockmap/sockhash only in task (normal) context with interrupts
enabled, or in softirq context.
Detect when map_delete_elem operation is invoked from a context which is
_not_ hardirq-unsafe, that is interrupts are disabled, and bail out with an
error.
Note that map updates are not affected by this issue. BPF verifier does not
allow updating sockmap/sockhash from a BPF tracing program today.
Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface") Reported-by: xingwei lee <xrivendell7@gmail.com> Reported-by: yue sun <samsun1006219@gmail.com> Reported-by: syzbot+bc922f476bd65abbd466@syzkaller.appspotmail.com Reported-by: syzbot+d4066896495db380182e@syzkaller.appspotmail.com Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Tested-by: syzbot+d4066896495db380182e@syzkaller.appspotmail.com Acked-by: John Fastabend <john.fastabend@gmail.com> Closes: https://syzkaller.appspot.com/bug?extid=d4066896495db380182e Closes: https://syzkaller.appspot.com/bug?extid=bc922f476bd65abbd466 Link: https://lore.kernel.org/bpf/20240402104621.1050319-1-jakub@cloudflare.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If an load_nls_xxx() function fails a few lines above, the 'sbi->bdi_id' is
still 0.
So, in the error handling path, we will call ida_simple_remove(..., 0)
which is not allocated yet.
In order to prevent a spurious "ida_free called for id=0 which is not
allocated." message, tweak the error handling path and add a new label.
The buggy address belongs to the object at ffff88802cd73da0
which belongs to the cache kmalloc-8 of size 8
The buggy address is located 0 bytes inside of
allocated 1-byte region [ffff88802cd73da0, ffff88802cd73da1)
Memory state around the buggy address: ffff88802cd73c80: 07 fc fc fc 05 fc fc fc 05 fc fc fc fa fc fc fc ffff88802cd73d00: fa fc fc fc fa fc fc fc fa fc fc fc fa fc fc fc
>ffff88802cd73d80: fa fc fc fc 01 fc fc fc fa fc fc fc fa fc fc fc
^ ffff88802cd73e00: fa fc fc fc fa fc fc fc 05 fc fc fc 07 fc fc fc ffff88802cd73e80: 07 fc fc fc 07 fc fc fc 07 fc fc fc 07 fc fc fc
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Pablo Neira Ayuso <pablo@netfilter.org> Link: https://lore.kernel.org/r/20240404122051.2303764-1-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
nft_unregister_flowtable_type() within nf_flow_inet_module_exit() can
concurrent with __nft_flowtable_type_get() within nf_tables_newflowtable().
And thhere is not any protection when iterate over nf_tables_flowtables
list in __nft_flowtable_type_get(). Therefore, there is pertential
data-race of nf_tables_flowtables list entry.
Use list_for_each_entry_rcu() to iterate over nf_tables_flowtables list
in __nft_flowtable_type_get(), and use rcu_read_lock() in the caller
nft_flowtable_type_get() to protect the entire type query process.
Similar to 2c9f0293280e ("netfilter: nf_tables: flush pending destroy
work before netlink notifier") to address a race between exit_net and
the destroy workqueue.
The trace below shows an element to be released via destroy workqueue
while exit_net path (triggered via module removal) has already released
the set that is used in such transaction.
Before calling add partition or resize partition, there is no check
on whether the length is aligned with the logical block size.
If the logical block size of the disk is larger than 512 bytes,
then the partition size maybe not the multiple of the logical block size,
and when the last sector is read, bio_truncate() will adjust the bio size,
resulting in an IO error if the size of the read command is smaller than
the logical block size.If integrity data is supported, this will also
result in a null pointer dereference when calling bio_integrity_free.
Cc: <stable@vger.kernel.org> Signed-off-by: Min Li <min15.li@samsung.com> Reviewed-by: Damien Le Moal <dlemoal@kernel.org> Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Link: https://lore.kernel.org/r/20230629142517.121241-1-min15.li@samsung.com Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Ashwin Dayanand Kamat <ashwin.kamat@broadcom.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Sven reports an infinite loop in __alloc_pages_slowpath() for costly order
__GFP_RETRY_MAYFAIL allocations that are also GFP_NOIO. Such combination
can happen in a suspend/resume context where a GFP_KERNEL allocation can
have __GFP_IO masked out via gfp_allowed_mask.
Quoting Sven:
1. try to do a "costly" allocation (order > PAGE_ALLOC_COSTLY_ORDER)
with __GFP_RETRY_MAYFAIL set.
2. page alloc's __alloc_pages_slowpath tries to get a page from the
freelist. This fails because there is nothing free of that costly
order.
3. page alloc tries to reclaim by calling __alloc_pages_direct_reclaim,
which bails out because a zone is ready to be compacted; it pretends
to have made a single page of progress.
4. page alloc tries to compact, but this always bails out early because
__GFP_IO is not set (it's not passed by the snd allocator, and even
if it were, we are suspending so the __GFP_IO flag would be cleared
anyway).
5. page alloc believes reclaim progress was made (because of the
pretense in item 3) and so it checks whether it should retry
compaction. The compaction retry logic thinks it should try again,
because:
a) reclaim is needed because of the early bail-out in item 4
b) a zonelist is suitable for compaction
6. goto 2. indefinite stall.
(end quote)
The immediate root cause is confusing the COMPACT_SKIPPED returned from
__alloc_pages_direct_compact() (step 4) due to lack of __GFP_IO to be
indicating a lack of order-0 pages, and in step 5 evaluating that in
should_compact_retry() as a reason to retry, before incrementing and
limiting the number of retries. There are however other places that
wrongly assume that compaction can happen while we lack __GFP_IO.
To fix this, introduce gfp_compaction_allowed() to abstract the __GFP_IO
evaluation and switch the open-coded test in try_to_compact_pages() to use
it.
Also use the new helper in:
- compaction_ready(), which will make reclaim not bail out in step 3, so
there's at least one attempt to actually reclaim, even if chances are
small for a costly order
- in_reclaim_compaction() which will make should_continue_reclaim()
return false and we don't over-reclaim unnecessarily
- in __alloc_pages_slowpath() to set a local variable can_compact,
which is then used to avoid retrying reclaim/compaction for costly
allocations (step 5) if we can't compact and also to skip the early
compaction attempt that we do in some cases
Link: https://lkml.kernel.org/r/20240221114357.13655-2-vbabka@suse.cz Fixes: 3250845d0526 ("Revert "mm, oom: prevent premature OOM killer invocation for high order request"") Signed-off-by: Vlastimil Babka <vbabka@suse.cz> Reported-by: Sven van Ashbrook <svenva@chromium.org> Closes: https://lore.kernel.org/all/CAG-rBihs_xMKb3wrMO1%2B-%2Bp4fowP9oy1pa_OTkfxBzPUVOZF%2Bg@mail.gmail.com/ Tested-by: Karthikeyan Ramasubramanian <kramasub@chromium.org> Cc: Brian Geffon <bgeffon@google.com> Cc: Curtis Malainey <cujomalainey@chromium.org> Cc: Jaroslav Kysela <perex@perex.cz> Cc: Mel Gorman <mgorman@techsingularity.net> Cc: Michal Hocko <mhocko@kernel.org> Cc: Takashi Iwai <tiwai@suse.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since d794734c9bbf was also marked for -stable, let's back it out before
causing more damage.
Note that due to another upstream change the revert was not 100% automatic:
0a845e0f6348 mm/treewide: replace pud_large() with pud_leaf()
Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: <stable@vger.kernel.org> Cc: Russ Anderson <rja@hpe.com> Cc: Steve Wahl <steve.wahl@hpe.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Link: https://lore.kernel.org/all/3a1b9909-45ac-4f97-ad68-d16ef1ce99db@pavinjoseph.com/ Fixes: d794734c9bbf ("x86/mm/ident_map: Use gbpages only where full GB page should be mapped.") Signed-off-by: Steve Wahl <steve.wahl@hpe.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
io_uring: ensure '0' is returned on file registration success
A previous backport mistakenly removed code that cleared 'ret' to zero,
as the SCM logging was performed. Fix up the return value so we don't
return an errant error on fixed file registration.
Fixes: a6771f343af9 ("io_uring: drop any code related to SCM_RIGHTS") Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The eventfd_ctx trigger pointer of the vfio_fsl_mc_irq object is
initially NULL and may become NULL if the user sets the trigger
eventfd to -1. The interrupt handler itself is guaranteed that
trigger is always valid between request_irq() and free_irq(), but
the loopback testing mechanisms to invoke the handler function
need to test the trigger. The triggering and setting ioctl paths
both make use of igate and are therefore mutually exclusive.
The vfio-fsl-mc driver does not make use of irqfds, nor does it
support any sort of masking operations, therefore unlike vfio-pci
and vfio-platform, the flow can remain essentially unchanged.
Cc: Diana Craciun <diana.craciun@oss.nxp.com> Cc: <stable@vger.kernel.org> Fixes: cc0ee20bd969 ("vfio/fsl-mc: trigger an interrupt via eventfd") Reviewed-by: Kevin Tian <kevin.tian@intel.com> Reviewed-by: Eric Auger <eric.auger@redhat.com> Link: https://lore.kernel.org/r/20240308230557.805580-8-alex.williamson@redhat.com Signed-off-by: Alex Williamson <alex.williamson@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The vfio-platform SET_IRQS ioctl currently allows loopback triggering of
an interrupt before a signaling eventfd has been configured by the user,
which thereby allows a NULL pointer dereference.
Rather than register the IRQ relative to a valid trigger, register all
IRQs in a disabled state in the device open path. This allows mask
operations on the IRQ to nest within the overall enable state governed
by a valid eventfd signal. This decouples @masked, protected by the
@locked spinlock from @trigger, protected via the @igate mutex.
In doing so, it's guaranteed that changes to @trigger cannot race the
IRQ handlers because the IRQ handler is synchronously disabled before
modifying the trigger, and loopback triggering of the IRQ via ioctl is
safe due to serialization with trigger changes via igate.
For compatibility, request_irq() failures are maintained to be local to
the SET_IRQS ioctl rather than a fatal error in the open device path.
This allows, for example, a userspace driver with polling mode support
to continue to work regardless of moving the request_irq() call site.
This necessarily blocks all SET_IRQS access to the failed index.
Cc: Eric Auger <eric.auger@redhat.com> Cc: <stable@vger.kernel.org> Fixes: 57f972e2b341 ("vfio/platform: trigger an interrupt via eventfd") Reviewed-by: Kevin Tian <kevin.tian@intel.com> Reviewed-by: Eric Auger <eric.auger@redhat.com> Link: https://lore.kernel.org/r/20240308230557.805580-7-alex.williamson@redhat.com Signed-off-by: Alex Williamson <alex.williamson@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A vulnerability exists where the eventfd for INTx signaling can be
deconfigured, which unregisters the IRQ handler but still allows
eventfds to be signaled with a NULL context through the SET_IRQS ioctl
or through unmask irqfd if the device interrupt is pending.
Ideally this could be solved with some additional locking; the igate
mutex serializes the ioctl and config space accesses, and the interrupt
handler is unregistered relative to the trigger, but the irqfd path
runs asynchronous to those. The igate mutex cannot be acquired from the
atomic context of the eventfd wake function. Disabling the irqfd
relative to the eventfd registration is potentially incompatible with
existing userspace.
As a result, the solution implemented here moves configuration of the
INTx interrupt handler to track the lifetime of the INTx context object
and irq_type configuration, rather than registration of a particular
trigger eventfd. Synchronization is added between the ioctl path and
eventfd_signal() wrapper such that the eventfd trigger can be
dynamically updated relative to in-flight interrupts or irqfd callbacks.
In order to synchronize changes that can affect the thread callback,
introduce an interface to force a flush of the inject workqueue. The
irqfd pointer is only valid under spinlock, but the workqueue cannot
be flushed under spinlock. Therefore the flush work for the irqfd is
queued under spinlock. The vfio_irqfd_cleanup_wq workqueue is re-used
for queuing this work such that flushing the workqueue is also ordered
relative to shutdown.
Reviewed-by: Kevin Tian <kevin.tian@intel.com> Reviewed-by: Reinette Chatre <reinette.chatre@intel.com> Reviewed-by: Eric Auger <eric.auger@redhat.com> Link: https://lore.kernel.org/r/20240308230557.805580-4-alex.williamson@redhat.com Signed-off-by: Alex Williamson <alex.williamson@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Mask operations through config space changes to DisINTx may race INTx
configuration changes via ioctl. Create wrappers that add locking for
paths outside of the core interrupt code.
In particular, irq_type is updated holding igate, therefore testing
is_intx() requires holding igate. For example clearing DisINTx from
config space can otherwise race changes of the interrupt configuration.
This aligns interfaces which may trigger the INTx eventfd into two
camps, one side serialized by igate and the other only enabled while
INTx is configured. A subsequent patch introduces synchronization for
the latter flows.
Currently for devices requiring masking at the irqchip for INTx, ie.
devices without DisINTx support, the IRQ is enabled in request_irq()
and subsequently disabled as necessary to align with the masked status
flag. This presents a window where the interrupt could fire between
these events, resulting in the IRQ incrementing the disable depth twice.
This would be unrecoverable for a user since the masked flag prevents
nested enables through vfio.
Instead, invert the logic using IRQF_NO_AUTOEN such that exclusive INTx
is never auto-enabled, then unmask as required.
Cc: <stable@vger.kernel.org> Fixes: 89e1f7d4c66d ("vfio: Add PCI device driver") Reviewed-by: Kevin Tian <kevin.tian@intel.com> Reviewed-by: Eric Auger <eric.auger@redhat.com> Link: https://lore.kernel.org/r/20240308230557.805580-2-alex.williamson@redhat.com Signed-off-by: Alex Williamson <alex.williamson@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
* The code modified by this patch is guarded by IS_ERR(trans_private),
where trans_private is assigned as per the previous point in this analysis.
The only implementation of get_mr that I could locate is rds_ib_get_mr()
which can return an ERR_PTR if the conn (4th) argument is NULL.
* ret is set to PTR_ERR(trans_private).
rds_ib_get_mr can return ERR_PTR(-ENODEV) if the conn (4th) argument is NULL.
Thus ret may be -ENODEV in which case the code in question will execute.
Conclusion:
* cp may be NULL at the point where this patch adds a check;
this patch does seem to address a possible bug
Fixes: c055fc00c07b ("net/rds: fix WARNING in rds_conn_connect_if_down") Cc: stable@vger.kernel.org # v4.19+ Signed-off-by: Mahmoud Adam <mngyadam@amazon.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://lore.kernel.org/r/20240326153132.55580-1-mngyadam@amazon.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>