RFDS is a CPU vulnerability that may allow userspace to infer kernel
stale data previously used in floating point registers, vector registers
and integer registers. RFDS only affects certain Intel Atom processors.
Intel released a microcode update that uses VERW instruction to clear
the affected CPU buffers. Unlike MDS, none of the affected cores support
SMT.
Add RFDS bug infrastructure and enable the VERW based mitigation by
default, that clears the affected buffers just before exiting to
userspace. Also add sysfs reporting and cmdline parameter
"reg_file_data_sampling" to control the mitigation.
For details see:
Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst
[ pawan: - Resolved conflicts in sysfs reporting.
- s/ATOM_GRACEMONT/ALDERLAKE_N/ATOM_GRACEMONT is called
ALDERLAKE_N in 6.6. ]
Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently MMIO Stale Data mitigation for CPUs not affected by MDS/TAA is
to only deploy VERW at VMentry by enabling mmio_stale_data_clear static
branch. No mitigation is needed for kernel->user transitions. If such
CPUs are also affected by RFDS, its mitigation may set
X86_FEATURE_CLEAR_CPU_BUF to deploy VERW at kernel->user and VMentry.
This could result in duplicate VERW at VMentry.
Fix this by disabling mmio_stale_data_clear static branch when
X86_FEATURE_CLEAR_CPU_BUF is enabled.
Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
During VMentry VERW is executed to mitigate MDS. After VERW, any memory
access like register push onto stack may put host data in MDS affected
CPU buffers. A guest can then use MDS to sample host data.
Although likelihood of secrets surviving in registers at current VERW
callsite is less, but it can't be ruled out. Harden the MDS mitigation
by moving the VERW mitigation late in VMentry path.
Note that VERW for MMIO Stale Data mitigation is unchanged because of
the complexity of per-guest conditional VERW which is not easy to handle
that late in asm with no GPRs available. If the CPU is also affected by
MDS, VERW is unconditionally executed late in asm regardless of guest
having MMIO access.
Use EFLAGS.CF instead of EFLAGS.ZF to track whether to use VMRESUME versus
VMLAUNCH. Freeing up EFLAGS.ZF will allow doing VERW, which clobbers ZF,
for MDS mitigations as late as possible without needing to duplicate VERW
for both paths.
[ pawan: resolved merge conflict in __vmx_vcpu_run in backport. ]
Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Reviewed-by: Nikolay Borisov <nik.borisov@suse.com> Link: https://lore.kernel.org/all/20240213-delay-verw-v8-5-a6216d83edb7%40linux.intel.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The VERW mitigation at exit-to-user is enabled via a static branch
mds_user_clear. This static branch is never toggled after boot, and can
be safely replaced with an ALTERNATIVE() which is convenient to use in
asm.
Switch to ALTERNATIVE() to use the VERW mitigation late in exit-to-user
path. Also remove the now redundant VERW in exc_nmi() and
arch_exit_to_user_mode().
Mitigation for MDS is to use VERW instruction to clear any secrets in
CPU Buffers. Any memory accesses after VERW execution can still remain
in CPU buffers. It is safer to execute VERW late in return to user path
to minimize the window in which kernel data can end up in CPU buffers.
There are not many kernel secrets to be had after SWITCH_TO_USER_CR3.
Add support for deploying VERW mitigation after user register state is
restored. This helps minimize the chances of kernel data ending up into
CPU buffers after executing VERW.
Note that the mitigation at the new location is not yet enabled.
Corner case not handled
=======================
Interrupts returning to kernel don't clear CPUs buffers since the
exit-to-user path is expected to do that anyways. But, there could be
a case when an NMI is generated in kernel after the exit-to-user path
has cleared the buffers. This case is not handled and NMI returning to
kernel don't clear CPU buffers because:
1. It is rare to get an NMI after VERW, but before returning to user.
2. For an unprivileged user, there is no known way to make that NMI
less rare or target it.
3. It would take a large number of these precisely-timed NMIs to mount
an actual attack. There's presumably not enough bandwidth.
4. The NMI in question occurs after a VERW, i.e. when user state is
restored and most interesting data is already scrubbed. Whats left
is only the data that NMI touches, and that may or may not be of
any interest.
[ pawan: resolved conflict for hunk swapgs_restore_regs_and_return_to_usermode ]
MDS mitigation requires clearing the CPU buffers before returning to
user. This needs to be done late in the exit-to-user path. Current
location of VERW leaves a possibility of kernel data ending up in CPU
buffers for memory accesses done after VERW such as:
1. Kernel data accessed by an NMI between VERW and return-to-user can
remain in CPU buffers since NMI returning to kernel does not
execute VERW to clear CPU buffers.
2. Alyssa reported that after VERW is executed,
CONFIG_GCC_PLUGIN_STACKLEAK=y scrubs the stack used by a system
call. Memory accesses during stack scrubbing can move kernel stack
contents into CPU buffers.
3. When caller saved registers are restored after a return from
function executing VERW, the kernel stack accesses can remain in
CPU buffers(since they occur after VERW).
To fix this VERW needs to be moved very late in exit-to-user path.
In preparation for moving VERW to entry/exit asm code, create macros
that can be used in asm. Also make VERW patching depend on a new feature
flag X86_FEATURE_CLEAR_CPU_BUF.
[pawan: - Runtime patch jmp instead of verw in macro CLEAR_CPU_BUFFERS
due to lack of relative addressing support for relocations
in kernels < v6.5.
- Add UNWIND_HINT_EMPTY to avoid warning:
arch/x86/entry/entry.o: warning: objtool: mds_verw_sel+0x0: unreachable instruction]
Reported-by: Alyssa Milburn <alyssa.milburn@intel.com> Suggested-by: Andrew Cooper <andrew.cooper3@citrix.com> Suggested-by: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Link: https://lore.kernel.org/all/20240213-delay-verw-v8-1-a6216d83edb7%40linux.intel.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add a macro _ASM_RIP() to add a (%rip) suffix on 64 bits only. This is
useful for immediate memory references where one doesn't want gcc
to possibly use a register indirection as it may in the case of an "m"
constraint.
[ pawan: resolved merged conflict for __ASM_REGPFX ]
Signed-off-by: H. Peter Anvin (Intel) <hpa@zytor.com> Signed-off-by: Borislav Petkov <bp@suse.de> Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com> Link: https://lkml.kernel.org/r/20210910195910.2542662-3-hpa@zytor.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Presently stage2_apply_range() works on a batch of memory addressed by a
stage 2 root table entry for the VM. Depending on the IPA limit of the
VM and PAGE_SIZE of the host, this could address a massive range of
memory. Some examples:
4 level, 4K paging -> 512 GB batch size
3 level, 64K paging -> 4TB batch size
Unsurprisingly, working on such a large range of memory can lead to soft
lockups. When running dirty_log_perf_test:
Use the largest supported block mapping for the configured page size as
the batch granularity. In so doing the walker is guaranteed to visit a
leaf only once.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20221007234151.461779-3-oliver.upton@linux.dev Signed-off-by: Krister Johansen <kjlx@templeofstupid.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Work out the minimum page table level where KVM supports block mappings
at compile time. While at it, rewrite the comment around supported block
mappings to directly describe what KVM supports instead of phrasing in
terms of what it does not.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20221007234151.461779-2-oliver.upton@linux.dev Signed-off-by: Krister Johansen <kjlx@templeofstupid.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When about to transmit the function imx_uart_start_tx is called and in
some RS485 configurations this function will call imx_uart_stop_rx. The
problem is that imx_uart_stop_rx will enable loopback in order to
release the RS485 bus, but when loopback is enabled transmitted data
will just be looped to RX.
This patch fixes the above problem by not enabling loopback when about
to transmit.
This driver now works well when used for RS485 half duplex master
configurations.
Fixes: 79d0224f6bf2 ("tty: serial: imx: Handle RS485 DE signal active high") Cc: stable <stable@kernel.org> Signed-off-by: Rickard x Andersson <rickaran@axis.com> Tested-by: Christoph Niedermaier <cniedermaier@dh-electronics.com> Link: https://lore.kernel.org/r/20240221115304.509811-1-rickaran@axis.com Signed-off-by: Christoph Niedermaier <cniedermaier@dh-electronics.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
console_trylock_spinning() may takeover the console lock from a
schedulable context. Update @console_may_schedule to make sure it
reflects a trylock acquire.
Reported-by: Mukesh Ojha <quic_mojha@quicinc.com> Closes: https://lore.kernel.org/lkml/20240222090538.23017-1-quic_mojha@quicinc.com Fixes: dbdda842fe96 ("printk: Add console owner and waiter logic to load balance console writes") Signed-off-by: John Ogness <john.ogness@linutronix.de> Reviewed-by: Mukesh Ojha <quic_mojha@quicinc.com> Reviewed-by: Petr Mladek <pmladek@suse.com> Link: https://lore.kernel.org/r/875xybmo2z.fsf@jogness.linutronix.de Signed-off-by: Petr Mladek <pmladek@suse.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The swiotlb does not support a mapping size > swiotlb_max_mapping_size().
On the other hand, with a 64KB PAGE_SIZE configuration, it's observed that
an NVME device can map a size between 300KB~512KB, which certainly failed
the swiotlb mappings, though the default pool of swiotlb has many slots:
systemd[1]: Started Journal Service.
=> nvme 0000:00:01.0: swiotlb buffer is full (sz: 327680 bytes), total 32768 (slots), used 32 (slots)
note: journal-offline[392] exited with irqs disabled
note: journal-offline[392] exited with preempt_count 1
Since untrusted devices might go down the swiotlb pathway with dma-iommu,
these devices should not map a size larger than swiotlb_max_mapping_size.
To fix this bug, add iommu_dma_max_mapping_size() for untrusted devices to
take into account swiotlb_max_mapping_size() v.s. iova_rcache_range() from
the iommu_dma_opt_mapping_size().
Fixes: 82612d66d51d ("iommu: Allow the dma-iommu api to use bounce buffers") Link: https://lore.kernel.org/r/ee51a3a5c32cf885b18f6416171802669f4a718a.1707851466.git.nicolinc@nvidia.com Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
[will: Drop redundant is_swiotlb_active(dev) check] Signed-off-by: Will Deacon <will@kernel.org> Reviewed-by: Michael Kelley <mhklinux@outlook.com> Acked-by: Robin Murphy <robin.murphy@arm.com> Tested-by: Nicolin Chen <nicolinc@nvidia.com> Tested-by: Michael Kelley <mhklinux@outlook.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Add the IOMMU callback for DMA mapping API dma_opt_mapping_size(), which
allows the drivers to know the optimal mapping limit and thus limit the
requested IOVA lengths.
This value is based on the IOVA rcache range limit, as IOVAs allocated
above this limit must always be newly allocated, which may be quite slow.
Signed-off-by: John Garry <john.garry@huawei.com> Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Acked-by: Robin Murphy <robin.murphy@arm.com> Acked-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Christoph Hellwig <hch@lst.de>
Stable-dep-of: afc5aa46ed56 ("iommu/dma: Force swiotlb_max_mapping_size on an untrusted device") Signed-off-by: Sasha Levin <sashal@kernel.org>
Streaming DMA mapping involving an IOMMU may be much slower for larger
total mapping size. This is because every IOMMU DMA mapping requires an
IOVA to be allocated and freed. IOVA sizes above a certain limit are not
cached, which can have a big impact on DMA mapping performance.
Provide an API for device drivers to know this "optimal" limit, such that
they may try to produce mapping which don't exceed it.
Signed-off-by: John Garry <john.garry@huawei.com> Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Acked-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Christoph Hellwig <hch@lst.de>
Stable-dep-of: afc5aa46ed56 ("iommu/dma: Force swiotlb_max_mapping_size on an untrusted device") Signed-off-by: Sasha Levin <sashal@kernel.org>
Nicolin reports that swiotlb buffer allocations fail for an NVME device
behind an IOMMU using 64KiB pages. This is because we end up with a
minimum allocation alignment of 64KiB (for the IOMMU to map the buffer
safely) but a minimum DMA alignment mask corresponding to a 4KiB NVME
page (i.e. preserving the 4KiB page offset from the original allocation).
If the original address is not 4KiB-aligned, the allocation will fail
because swiotlb_search_pool_area() erroneously compares these unmasked
bits with the 64KiB-aligned candidate allocation.
Tweak swiotlb_search_pool_area() so that the DMA alignment mask is
reduced based on the required alignment of the allocation.
Fixes: 82612d66d51d ("iommu: Allow the dma-iommu api to use bounce buffers") Link: https://lore.kernel.org/r/cover.1707851466.git.nicolinc@nvidia.com Reported-by: Nicolin Chen <nicolinc@nvidia.com> Signed-off-by: Will Deacon <will@kernel.org> Reviewed-by: Michael Kelley <mhklinux@outlook.com> Tested-by: Nicolin Chen <nicolinc@nvidia.com> Tested-by: Michael Kelley <mhklinux@outlook.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Patch series "minmax: Relax type checks in min() and max()", v4.
The min() (etc) functions in minmax.h require that the arguments have
exactly the same types.
However when the type check fails, rather than look at the types and fix
the type of a variable/constant, everyone seems to jump on min_t(). In
reality min_t() ought to be rare - when something unusual is being done,
not normality.
The orginal min() (added in 2.4.9) replaced several inline functions and
included the type - so matched the implicit casting of the function call.
This was renamed min_t() in 2.4.10 and the current min() added. There is
no actual indication that the conversion of negatve values to large
unsigned values has ever been an actual problem.
A quick grep shows 5734 min() and 4597 min_t(). Having the casts on
almost half of the calls shows that something is clearly wrong.
If the wrong type is picked (and it is far too easy to pick the type of
the result instead of the larger input) then significant bits can get
discarded.
Pretty much the worst example is in the derived clamp_val(), consider:
unsigned char x = 200u;
y = clamp_val(x, 10u, 300u);
I also suspect that many of the min_t(u16, ...) are actually wrong. For
example copy_data() in printk_ringbuffer.c contains:
data_size = min_t(u16, buf_size, len);
Here buf_size is 'unsigned int' and len 'u16', pass a 64k buffer (can you
prove that doesn't happen?) and no data is returned. Apparantly it did -
and has since been fixed.
The only reason that most of the min_t() are 'fine' is that pretty much
all the values in the kernel are between 0 and INT_MAX.
Patch 1 adds umin(), this uses integer promotions to convert both
arguments to 'unsigned long long'. It can be used to compare a signed
type that is known to contain a non-negative value with an unsigned type.
The compiler typically optimises it all away. Added first so that it can
be referred to in patch 2.
Patch 2 replaces the 'same type' check with a 'same signedness' one. This
makes min(unsigned_int_var, sizeof()) be ok. The error message is also
improved and will contain the expanded form of both arguments (useful for
seeing how constants are defined).
Patch 3 just fixes some whitespace.
Patch 4 allows comparisons of 'unsigned char' and 'unsigned short' to
signed types. The integer promotion rules convert them both to 'signed
int' prior to the comparison so they can never cause a negative value be
converted to a large positive one.
Patch 5 (rewritted for v4) allows comparisons of unsigned values against
non-negative constant integer expressions. This makes
min(unsigned_int_var, 4) be ok.
The only common case that is still errored is the comparison of signed
values against unsigned constant integer expressions below __INT_MAX__.
Typcally min(int_val, sizeof (foo)), the real fix for this is casting the
constant: min(int_var, (int)sizeof (foo)).
With all the patches applied pretty much all the min_t() could be replaced
by min(), and most of the rest by umin(). However they all need careful
inspection due to code like:
sz = min_t(unsigned char, sz - 1, LIM - 1) + 1;
which converts 0 to LIM.
This patch (of 6):
umin() and umax() can be used when min()/max() errors a signed v unsigned
compare when the signed value is known to be non-negative.
Unlike min_t(some_unsigned_type, a, b) umin() will never mask off high
bits if an inappropriate type is selected.
The '+ 0u + 0ul + 0ull' may look strange.
The '+ 0u' is needed for 'signed int' on 64bit systems.
The '+ 0ul' is needed for 'signed long' on 32bit systems.
The '+ 0ull' is needed for 'signed long long'.
Link: https://lkml.kernel.org/r/b97faef60ad24922b530241c5d7c933c@AcuMS.aculab.com Link: https://lkml.kernel.org/r/41d93ca827a248698ec64bf57e0c05a5@AcuMS.aculab.com Signed-off-by: David Laight <david.laight@aculab.com> Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Jason A. Donenfeld <Jason@zx2c4.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthew Wilcox (Oracle) <willy@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Stable-dep-of: 51b30ecb73b4 ("swiotlb: Fix alignment checks when both allocation and DMA masks are present") Signed-off-by: Sasha Levin <sashal@kernel.org>
When a probe is registered at the trace_sys_enter() tracepoint, and that
probe changes the system call number, the old system call still gets
executed. This worked correctly until commit b6ec41346103 ("core/entry:
Report syscall correctly for trace and audit"), which removed the
re-evaluation of the syscall number after the trace point.
Restore the original semantics by re-evaluating the system call number
after trace_sys_enter().
The performance impact of this re-evaluation is minimal because it only
takes place when a trace point is active, and compared to the actual trace
point overhead the read from a cache hot variable is negligible.
Fixes: b6ec41346103 ("core/entry: Report syscall correctly for trace and audit") Signed-off-by: André Rösti <an.roesti@gmail.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lore.kernel.org/r/20240311211704.7262-1-an.roesti@gmail.com Signed-off-by: Sasha Levin <sashal@kernel.org>
The prescaler in the "Global Timer Control Register bit assignments" is
documented to use bits [15:8], which means that the maximum prescaler
register value is 0xff.
To align with ACPI 6.3+, since bit_width can be any 8-bit value, it
cannot be depended on to be always on a clean 8b boundary. This was
uncovered on the Cobalt 100 platform.
Instead, use access_width to determine the size and use the offset and
width to shift and mask the bits to read/write out. Make sure to add a
check for system memory since pcc redefines the access_width to
subspace id.
If access_width is not set, then fall back to using bit_width.
Signed-off-by: Jarred White <jarredwhite@linux.microsoft.com> Reviewed-by: Easwar Hariharan <eahariha@linux.microsoft.com> Cc: 5.15+ <stable@vger.kernel.org> # 5.15+
[ rjw: Subject and changelog edits, comment adjustments ] Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
[ eahariha: Backport to v5.15 by dropping SystemIO bits as
commit a2c8f92bea5f is not present ] Signed-off-by: Easwar Hariharan <eahariha@linux.microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
shutdown_pirq and startup_pirq are not taking the
irq_mapping_update_lock because they can't due to lock inversion. Both
are called with the irq_desc->lock being taking. The lock order,
however, is first irq_mapping_update_lock and then irq_desc->lock.
This opens multiple races:
- shutdown_pirq can be interrupted by a function that allocates an event
channel:
Assume here event channel e refers here to the same event channel
number.
After this race the evtchn_to_irq mapping for e is invalid (-1).
- __startup_pirq races with __unbind_from_irq in a similar way. Because
__startup_pirq doesn't take irq_mapping_update_lock it can grab the
evtchn that __unbind_from_irq is currently freeing and cleaning up. In
this case even though the event channel is allocated, its mapping can
be unset in evtchn_to_irq.
The fix is to first cleanup the mappings and then close the event
channel. In this way, when an event channel gets allocated it's
potential previous evtchn_to_irq mappings are guaranteed to be unset already.
This is also the reverse order of the allocation where first the event
channel is allocated and then the mappings are setup.
On a 5.10 kernel prior to commit 3fcdaf3d7634 ("xen/events: modify internal
[un]bind interfaces"), we hit a BUG like the following during probing of NVMe
devices. The issue is that during nvme_setup_io_queues, pci_free_irq
is called for every device which results in a call to shutdown_pirq.
With many nvme devices it's therefore likely to hit this race during
boot because there will be multiple calls to shutdown_pirq and
startup_pirq are running potentially in parallel.
If registering the platform device fails, the lookup table is
removed in the error path. On module removal we would try to
remove the lookup table again. Fix this by setting priv->lookup
only if registering the platform device was successful.
In addition free the memory allocated for the lookup table in
the error path.
The first kiocb_set_cancel_fn() argument may point at a struct kiocb
that is not embedded inside struct aio_kiocb. With the current code,
depending on the compiler, the req->ki_ctx read happens either before
the IOCB_AIO_RW test or after that test. Move the req->ki_ctx read such
that it is guaranteed that the IOCB_AIO_RW test happens first.
Reported-by: Eric Biggers <ebiggers@kernel.org> Cc: Benjamin LaHaise <ben@communityfibre.ca> Cc: Eric Biggers <ebiggers@google.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Avi Kivity <avi@scylladb.com> Cc: Sandeep Dhavale <dhavale@google.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Kent Overstreet <kent.overstreet@linux.dev> Cc: stable@vger.kernel.org Fixes: b820de741ae4 ("fs/aio: Restrict kiocb_set_cancel_fn() to I/O submitted via libaio") Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240304235715.3790858-1-bvanassche@acm.org Reviewed-by: Jens Axboe <axboe@kernel.dk> Reviewed-by: Eric Biggers <ebiggers@google.com> Signed-off-by: Christian Brauner <brauner@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is the same issue that was fixed for the VGA text buffer in commit 39cdb68c64d8 ("vt: fix memory overlapping when deleting chars in the
buffer"). The cure is also the same i.e. replace memcpy() with memmove()
due to the overlaping buffers.
Signed-off-by: Nicolas Pitre <nico@fluxnic.net> Fixes: 81732c3b2fed ("tty vt: Fix line garbage in virtual console on command line edition") Cc: stable <stable@kernel.org> Link: https://lore.kernel.org/r/sn184on2-3p0q-0qrq-0218-895349s4753o@syhkavp.arg Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If the remote uart device is not connected or not enabled after booting
up, the CTS line is high by default. At this time, if we enable the flow
control when opening the device(for example, using “stty -F /dev/ttyLP4
crtscts” command), there will be a pending idle preamble(first writing 0
and then writing 1 to UARTCTRL_TE will queue an idle preamble) that
cannot be sent out, resulting in the uart port fail to close(waiting for
TX empty), so the user space stty will have to wait for a long time or
forever.
This is an LPUART IP bug(idle preamble has higher priority than CTS),
here add a workaround patch to enable TX CTS after enabling UARTCTRL_TE,
so that the idle preamble does not get stuck due to CTS is deasserted.
Unused USB ports may have bogus location data in ACPI PLD tables.
This causes port peering failures as these unused USB2 and USB3 ports
location may match.
Due to these failures the driver prints a
"usb: port power management may be unreliable" warning, and
unnecessarily blocks port power off during runtime suspend.
This was debugged on a couple DELL systems where the unused ports
all returned zeroes in their location data.
Similar bugreports exist for other systems.
Don't try to peer or match ports that have connect type set to
USB_PORT_NOT_USED.
While connecting to a Linux host with CDC_NCM_NTB_DEF_SIZE_TX
set to 65536, it has been observed that we receive short packets,
which come at interval of 5-10 seconds sometimes and have block
length zero but still contain 1-2 valid datagrams present.
According to the NCM spec:
"If wBlockLength = 0x0000, the block is terminated by a
short packet. In this case, the USB transfer must still
be shorter than dwNtbInMaxSize or dwNtbOutMaxSize. If
exactly dwNtbInMaxSize or dwNtbOutMaxSize bytes are sent,
and the size is a multiple of wMaxPacketSize for the
given pipe, then no ZLP shall be sent.
wBlockLength= 0x0000 must be used with extreme care, because
of the possibility that the host and device may get out of
sync, and because of test issues.
wBlockLength = 0x0000 allows the sender to reduce latency by
starting to send a very large NTB, and then shortening it when
the sender discovers that there’s not sufficient data to justify
sending a large NTB"
However, there is a potential issue with the current implementation,
as it checks for the occurrence of multiple NTBs in a single
giveback by verifying if the leftover bytes to be processed is zero
or not. If the block length reads zero, we would process the same
NTB infintely because the leftover bytes is never zero and it leads
to a crash. Fix this by bailing out if block length reads zero.
Cc: stable@vger.kernel.org Fixes: 427694cfaafa ("usb: gadget: ncm: Handle decoding of multiple NTB's in unwrap call") Signed-off-by: Krishna Kurapati <quic_kriskura@quicinc.com> Reviewed-by: Maciej Żenczykowski <maze@google.com> Link: https://lore.kernel.org/r/20240228115441.2105585-1-quic_kriskura@quicinc.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The isd200 sub-driver in usb-storage uses the HEADS and SECTORS values
in the ATA ID information to calculate cylinder and head values when
creating a CDB for READ or WRITE commands. The calculation involves
division and modulus operations, which will cause a crash if either of
these values is 0. While this never happens with a genuine device, it
could happen with a flawed or subversive emulation, as reported by the
syzbot fuzzer.
Protect against this possibility by refusing to bind to the device if
either the ATA_ID_HEADS or ATA_ID_SECTORS value in the device's ID
information is 0. This requires isd200_Initialization() to return a
negative error code when initialization fails; currently it always
returns 0 (even when there is an error).
Error in mmu_interval_notifier_insert() can leave a NULL
notifier.mm pointer. Catch that and return early.
Fixes: ed29c2691188 ("drm/i915: Fix userptr so we do not have to worry about obj->mm.lock, v7.") Cc: <stable@vger.kernel.org> # v5.13+
[tursulin: Added Fixes and cc stable.] Cc: Andi Shyti <andi.shyti@linux.intel.com> Cc: Shawn Lee <shawn.c.lee@intel.com> Signed-off-by: Nirmoy Das <nirmoy.das@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240219125047.28906-1-nirmoy.das@intel.com Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
(cherry picked from commit db7bbd13f08774cde0332c705f042e327fe21e73) Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The .release() function does not get called until all readers of a file
descriptor are finished.
If a thread is blocked on reading a file descriptor in ring_buffer_wait(),
and another thread closes the file descriptor, it will not wake up the
other thread as ring_buffer_wake_waiters() is called by .release(), and
that will not get called until the .read() is finished.
The issue originally showed up in trace-cmd, but the readers are actually
other processes with their own file descriptors. So calling close() would wake
up the other tasks because they are blocked on another descriptor then the
one that was closed(). But there's other wake ups that solve that issue.
When a thread is blocked on a read, it can still hang even when another
thread closed its descriptor.
This is what the .flush() callback is for. Have the .flush() wake up the
readers.
Link: https://lore.kernel.org/linux-trace-kernel/20240308202432.107909457@goodmis.org Cc: stable@vger.kernel.org Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: linke li <lilinke99@qq.com> Cc: Rabin Vincent <rabin@rab.in> Fixes: f3ddb74ad0790 ("tracing: Wake up ring buffer waiters on closing of the file") Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Do the cache flush of converted pages in svm_register_enc_region() before
dropping kvm->lock to fix use-after-free issues where region and/or its
array of pages could be freed by a different task, e.g. if userspace has
__unregister_enc_region_locked() already queued up for the region.
Note, the "obvious" alternative of using local variables doesn't fully
resolve the bug, as region->pages is also dynamically allocated. I.e. the
region structure itself would be fine, but region->pages could be freed.
Flushing multiple pages under kvm->lock is unfortunate, but the entire
flow is a rare slow path, and the manual flush is only needed on CPUs that
lack coherency for encrypted memory.
Fixes: 19a23da53932 ("Fix unsynchronized access to sev members through svm_register_enc_region") Reported-by: Gabe Kirkpatrick <gkirkpatrick@google.com> Cc: Josh Eads <josheads@google.com> Cc: Peter Gonda <pgonda@google.com> Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20240217013430.2079561-1-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After a couple recent changes in LLVM, there is a warning (or error with
CONFIG_WERROR=y or W=e) from the compile time fortify source routines,
specifically the memset() in copy_to_user_tmpl().
In file included from net/xfrm/xfrm_user.c:14:
...
include/linux/fortify-string.h:438:4: error: call to '__write_overflow_field' declared with 'warning' attribute: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Werror,-Wattribute-warning]
438 | __write_overflow_field(p_size_field, size);
| ^
1 error generated.
While ->xfrm_nr has been validated against XFRM_MAX_DEPTH when its value
is first assigned in copy_templates() by calling validate_tmpl() first
(so there should not be any issue in practice), LLVM/clang cannot really
deduce that across the boundaries of these functions. Without that
knowledge, it cannot assume that the loop stops before i is greater than
XFRM_MAX_DEPTH, which would indeed result a stack buffer overflow in the
memset().
To make the bounds of ->xfrm_nr clear to the compiler and add additional
defense in case copy_to_user_tmpl() is ever used in a path where
->xfrm_nr has not been properly validated against XFRM_MAX_DEPTH first,
add an explicit bound check and early return, which clears up the
warning.
The VMBUS_RING_SIZE macro adds space for a ring buffer header to the
requested ring buffer size. The header size is always 1 page, and so
its size varies based on the PAGE_SIZE for which the kernel is built.
If the requested ring buffer size is a large power-of-2 size and the header
size is small, the resulting size is inefficient in its use of memory.
For example, a 512 Kbyte ring buffer with a 4 Kbyte page size results in
a 516 Kbyte allocation, which is rounded to up 1 Mbyte by the memory
allocator, and wastes 508 Kbytes of memory.
In such situations, the exact size of the ring buffer isn't that important,
and it's OK to allocate the 4 Kbyte header at the beginning of the 512
Kbytes, leaving the ring buffer itself with just 508 Kbytes. The memory
allocation can be 512 Kbytes instead of 1 Mbyte and nothing is wasted.
Update VMBUS_RING_SIZE to implement this approach for "large" ring buffer
sizes. "Large" is somewhat arbitrarily defined as 8 times the size of
the ring buffer header (which is of size PAGE_SIZE). For example, for
4 Kbyte PAGE_SIZE, ring buffers of 32 Kbytes and larger use the first
4 Kbytes as the ring buffer header. For 64 Kbyte PAGE_SIZE, ring buffers
of 512 Kbytes and larger use the first 64 Kbytes as the ring buffer
header. In both cases, smaller sizes add space for the header so
the ring size isn't reduced too much by using part of the space for
the header. For example, with a 64 Kbyte page size, we don't want
a 128 Kbyte ring buffer to be reduced to 64 Kbytes by allocating half
of the space for the header. In such a case, the memory allocation
is less efficient, but it's the best that can be done.
While the new algorithm slightly changes the amount of space allocated
for ring buffers by drivers that use VMBUS_RING_SIZE, the devices aren't
known to be sensitive to small changes in ring buffer size, so there
shouldn't be any effect.
Fixes: c1135c7fd0e9 ("Drivers: hv: vmbus: Introduce types of GPADL") Fixes: 6941f67ad37d ("hv_netvsc: Calculate correct ring size when PAGE_SIZE is not 4 Kbytes") Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218502 Cc: stable@vger.kernel.org Signed-off-by: Michael Kelley <mhklinux@outlook.com> Reviewed-by: Saurabh Sengar <ssengar@linux.microsoft.com> Reviewed-by: Dexuan Cui <decui@microsoft.com> Tested-by: Souradeep Chakrabarti <schakrabarti@linux.microsoft.com> Link: https://lore.kernel.org/r/20240229004533.313662-1-mhklinux@outlook.com Signed-off-by: Wei Liu <wei.liu@kernel.org>
Message-ID: <20240229004533.313662-1-mhklinux@outlook.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This set combination is weird: it allows for elements to be
added/deleted, but once bound to the rule it cannot be updated anymore.
Eventually, all elements expire, leading to an empty set which cannot
be updated anymore. Reject this flags combination.
Cc: stable@vger.kernel.org Fixes: 761da2935d6e ("netfilter: nf_tables: add set timeout API support") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
While the rhashtable set gc runs asynchronously, a race allows it to
collect elements from anonymous sets with timeouts while it is being
released from the commit path.
Mingi Cho originally reported this issue in a different path in 6.1.x
with a pipapo set with low timeouts which is not possible upstream since 7395dfacfff6 ("netfilter: nf_tables: use timestamp to check for set
element timeout").
Fix this by setting on the dead flag for anonymous sets to skip async gc
in this case.
According to 08e4c8c5919f ("netfilter: nf_tables: mark newset as dead on
transaction abort"), Florian plans to accelerate abort path by releasing
objects via workqueue, therefore, this sets on the dead flag for abort
path too.
Cc: stable@vger.kernel.org Fixes: 5f68718b34a5 ("netfilter: nf_tables: GC transaction API to avoid race with control plane") Reported-by: Mingi Cho <mgcho.minic@gmail.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
cpufreq: brcmstb-avs-cpufreq: fix up "add check for cpufreq_cpu_get's return value"
In commit d951cf510fb0 ("cpufreq: brcmstb-avs-cpufreq: add check for
cpufreq_cpu_get's return value"), build warnings occur because a
variable is created after some logic, resulting in:
drivers/cpufreq/brcmstb-avs-cpufreq.c: In function 'brcm_avs_cpufreq_get':
drivers/cpufreq/brcmstb-avs-cpufreq.c:486:9: error: ISO C90 forbids mixed
declarations and code [-Werror=declaration-after-statement]
486 | struct private_data *priv = policy->driver_data;
| ^~~~~~
cc1: all warnings being treated as errors
make[2]: *** [scripts/Makefile.build:289:
drivers/cpufreq/brcmstb-avs-cpufreq.o] Error 1
make[1]: *** [scripts/Makefile.build:552: drivers/cpufreq] Error 2
make[1]: *** Waiting for unfinished jobs....
make: *** [Makefile:1907: drivers] Error 2
7ee18d677989 ("x86/power: Make restore_processor_context() sane")
kmemleak reports this issue:
unreferenced object 0xf68241e0 (size 32):
comm "swapper/0", pid 1, jiffies 4294668610 (age 68.432s)
hex dump (first 32 bytes):
00 cc cc cc 29 10 01 c0 00 00 00 00 00 00 00 00 ....)...........
00 42 82 f6 cc cc cc cc cc cc cc cc cc cc cc cc .B..............
backtrace:
[<461c1d50>] __kmem_cache_alloc_node+0x106/0x260
[<ea65e13b>] __kmalloc+0x54/0x160
[<c3858cd2>] msr_build_context.constprop.0+0x35/0x100
[<46635aff>] pm_check_save_msr+0x63/0x80
[<6b6bb938>] do_one_initcall+0x41/0x1f0
[<3f3add60>] kernel_init_freeable+0x199/0x1e8
[<3b538fde>] kernel_init+0x1a/0x110
[<938ae2b2>] ret_from_fork+0x1c/0x28
Which is a false positive.
Reproducer:
- Run rsync of whole kernel tree (multiple times if needed).
- start a kmemleak scan
- Note this is just an example: a lot of our internal tests hit these.
The root cause is similar to the fix in:
b0b592cf0836 x86/pm: Fix false positive kmemleak report in msr_build_context()
ie. the alignment within the packed struct saved_context
which has everything unaligned as there is only "u16 gs;" at start of
struct where in the past there were four u16 there thus aligning
everything afterwards. The issue is with the fact that Kmemleak only
searches for pointers that are aligned (see how pointers are scanned in
kmemleak.c) so when the struct members are not aligned it doesn't see
them.
Testing:
We run a lot of tests with our CI, and after applying this fix we do not
see any kmemleak issues any more whilst without it we see hundreds of
the above report. From a single, simple test run consisting of 416 individual test
cases on kernel 5.10 x86 with kmemleak enabled we got 20 failures due to this,
which is quite a lot. With this fix applied we get zero kmemleak related failures.
[Why]
When mode switching is triggered there is momentary noise visible on
some HDMI TV or displays.
[How]
Wait for 2 frames to make sure we have enough time to send out AV mute
and sink receives a full frame.
Cc: Mario Limonciello <mario.limonciello@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Cc: stable@vger.kernel.org Reviewed-by: Wenjing Liu <wenjing.liu@amd.com> Acked-by: Wayne Lin <wayne.lin@amd.com> Signed-off-by: Leo Ma <hanghong.ma@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Otherwise after the GTT bo is released, the GTT and gart space is freed
but amdgpu_ttm_backend_unbind will not clear the gart page table entry
and leave valid mapping entry pointing to the stale system page. Then
if GPU access the gart address mistakely, it will read undefined value
instead page fault, harder to debug and reproduce the real issue.
Cc: stable@vger.kernel.org Signed-off-by: Philip Yang <Philip.Yang@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Previously, patches have been added to limit the reported count of SATA
ports for asm1064 and asm1166 SATA controllers, as those controllers do
report more ports than physically having.
While it is allowed to report more ports than physically having in CAP.NP,
it is not allowed to report more ports than physically having in the PI
(Ports Implemented) register, which is what these HBAs do.
(This is a AHCI spec violation.)
Unfortunately, it seems that the PMP implementation in these ASMedia HBAs
is also violating the AHCI and SATA-IO PMP specification.
What these HBAs do is that they do not report that they support PMP
(CAP.SPM (Supports Port Multiplier) is not set).
Instead, they have decided to add extra "virtual" ports in the PI register
that is used if a port multiplier is connected to any of the physical
ports of the HBA.
Enumerating the devices behind the PMP as specified in the AHCI and
SATA-IO specifications, by using PMP READ and PMP WRITE commands to the
physical ports of the HBA is not possible, you have to use the "virtual"
ports.
This is of course bad, because this gives us no way to detect the device
and vendor ID of the PMP actually connected to the HBA, which means that
we can not apply the proper PMP quirks for the PMP that is connected to
the HBA.
Limiting the port map will thus stop these controllers from working with
SATA Port Multipliers.
This patch reverts both patches for asm1064 and asm1166, so old behavior
is restored and SATA PMP will work again, but it will also reintroduce the
(minutes long) extra boot time for the ASMedia controllers that do not
have a PMP connected (either on the PCIe card itself, or an external PMP).
However, a longer boot time for some, is the lesser evil compared to some
other users not being able to detect their drives at all.
Fixes: 0077a504e1a4 ("ahci: asm1166: correct count of reported ports") Fixes: 9815e3961754 ("ahci: asm1064: correct count of reported ports") Cc: stable@vger.kernel.org Reported-by: Matt <cryptearth@googlemail.com> Signed-off-by: Conrad Kostecki <conikost@gentoo.org> Reviewed-by: Hans de Goede <hdegoede@redhat.com>
[cassel: rewrote commit message] Signed-off-by: Niklas Cassel <cassel@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The previous commit fixed a bug that led to a NULL peer->device being
dereferenced. It's actually easier and faster performance-wise to
instead get the device from ctx->wg. This semantically makes more sense
too, since ctx->wg->peer_allowedips.seq is compared with
ctx->allowedips_seq, basing them both in ctx. This also acts as a
defence in depth provision against freed peers.
Cc: stable@vger.kernel.org Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Reviewed-by: Jiri Pirko <jiri@nvidia.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
If all peers are removed via wg_peer_remove_all(), rather than setting
peer_list to empty, the peer is added to a temporary list with a head on
the stack of wg_peer_remove_all(). If a netlink dump is resumed and the
cursored peer is one that has been removed via wg_peer_remove_all(), it
will iterate from that peer and then attempt to dump freed peers.
Fix this by instead checking peer->is_dead, which was explictly created
for this purpose. Also move up the device_update_lock lockdep assertion,
since reading is_dead relies on that.
It can be reproduced by a small script like:
echo "Setting config..."
ip link add dev wg0 type wireguard
wg setconf wg0 /big-config
(
while true; do
echo "Showing config..."
wg showconf wg0 > /dev/null
done
) &
sleep 4
wg setconf wg0 <(printf "[Peer]\nPublicKey=$(wg genkey)\n")
Resulting in:
BUG: KASAN: slab-use-after-free in __lock_acquire+0x182a/0x1b20
Read of size 8 at addr ffff88811956ec70 by task wg/59
CPU: 2 PID: 59 Comm: wg Not tainted 6.8.0-rc2-debug+ #5
Call Trace:
<TASK>
dump_stack_lvl+0x47/0x70
print_address_description.constprop.0+0x2c/0x380
print_report+0xab/0x250
kasan_report+0xba/0xf0
__lock_acquire+0x182a/0x1b20
lock_acquire+0x191/0x4b0
down_read+0x80/0x440
get_peer+0x140/0xcb0
wg_get_device_dump+0x471/0x1130
Cc: stable@vger.kernel.org Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") Reported-by: Lillian Berry <lillian@star-ark.net> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Reviewed-by: Jiri Pirko <jiri@nvidia.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The __string() and __assign_str() helper macros of the TRACE_EVENT() macro
are going through some optimizations where only the source string of
__string() will be used and the __assign_str() source will be ignored and
later removed.
To make sure that there's no issues, a new check is added between the
__string() src argument and the __assign_str() src argument that does a
strcmp() to make sure they are the same string.
I'm working on restructuring the __string* macros so that it doesn't need
to recalculate the string twice. That is, it will save it off when
processing __string() and the __assign_str() will not need to do the work
again as it currently does.
Currently __string_len(item, src, len) doesn't actually use "src", but my
changes will require src to be correct as that is where the __assign_str()
will get its value from.
The event class nfsd_clid_class has:
__string_len(name, name, clp->cl_name.len)
But the second "name" does not exist and causes my changes to fail to
build. That second parameter should be: clp->cl_name.data.
Link: https://lore.kernel.org/linux-trace-kernel/20240222122828.3d8d213c@gandalf.local.home Cc: Neil Brown <neilb@suse.de> Cc: Olga Kornievskaia <kolga@netapp.com> Cc: Dai Ngo <Dai.Ngo@oracle.com> Cc: Tom Talpey <tom@talpey.com> Cc: stable@vger.kernel.org Fixes: d27b74a8675ca ("NFSD: Use new __string_len C macros for nfsd_clid_class") Acked-by: Chuck Lever <chuck.lever@oracle.com> Acked-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit 0499a78369ad ("ARM64: Dynamically allocate cpumasks and increase
supported CPUs to 512") changed the handling of cpumasks on ARM 64bit,
what resulted in the strange issues and warnings during cpufreq-dt
initialization on some big.LITTLE platforms.
This was caused by mixing OPPs between big and LITTLE cores, because
OPP-sharing information between big and LITTLE cores is computed on
cpumask, which in turn was not zeroed on allocation. Fix this by
switching to zalloc_cpumask_var() call.
Fix a bug where nilfs_get_block() returns a successful status when
searching and inserting the specified block both fail inconsistently. If
this inconsistent behavior is not due to a previously fixed bug, then an
unexpected race is occurring, so return a temporary error -EAGAIN instead.
This prevents callers such as __block_write_begin_int() from requesting a
read into a buffer that is not mapped, which would cause the BUG_ON check
for the BH_Mapped flag in submit_bh_wbc() to fail.
Link: https://lkml.kernel.org/r/20240313105827.5296-3-konishi.ryusuke@gmail.com Fixes: 1f5abe7e7dbc ("nilfs2: replace BUG_ON and BUG calls triggerable from ioctl") Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Patch series "nilfs2: fix kernel bug at submit_bh_wbc()".
This resolves a kernel BUG reported by syzbot. Since there are two
flaws involved, I've made each one a separate patch.
The first patch alone resolves the syzbot-reported bug, but I think
both fixes should be sent to stable, so I've tagged them as such.
This patch (of 2):
Syzbot has reported a kernel bug in submit_bh_wbc() when writing file data
to a nilfs2 file system whose metadata is corrupted.
There are two flaws involved in this issue.
The first flaw is that when nilfs_get_block() locates a data block using
btree or direct mapping, if the disk address translation routine
nilfs_dat_translate() fails with internal code -ENOENT due to DAT metadata
corruption, it can be passed back to nilfs_get_block(). This causes
nilfs_get_block() to misidentify an existing block as non-existent,
causing both data block lookup and insertion to fail inconsistently.
The second flaw is that nilfs_get_block() returns a successful status in
this inconsistent state. This causes the caller __block_write_begin_int()
or others to request a read even though the buffer is not mapped,
resulting in a BUG_ON check for the BH_Mapped flag in submit_bh_wbc()
failing.
This fixes the first issue by changing the return value to code -EINVAL
when a conversion using DAT fails with code -ENOENT, avoiding the
conflicting condition that leads to the kernel bug described above. Here,
code -EINVAL indicates that metadata corruption was detected during the
block lookup, which will be properly handled as a file system error and
converted to -EIO when passing through the nilfs2 bmap layer.
None of the callers of drm_panel_get_modes() expect it to return
negative error codes. Either they propagate the return value in their
struct drm_connector_helper_funcs .get_modes() hook (which is also not
supposed to return negative codes), or add it to other counts leading to
bogus values.
On the other hand, many of the struct drm_panel_funcs .get_modes() hooks
do return negative error codes, so handle them gracefully instead of
propagating further.
Return 0 for no modes, whatever the reason.
Cc: Neil Armstrong <neil.armstrong@linaro.org> Cc: Jessica Zhang <quic_jesszhan@quicinc.com> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: stable@vger.kernel.org Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org> Reviewed-by: Jessica Zhang <quic_jesszhan@quicinc.com> Acked-by: Thomas Zimmermann <tzimmermann@suse.de> Link: https://patchwork.freedesktop.org/patch/msgid/79f559b72d8c493940417304e222a4b04dfa19c4.1709913674.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Tests with hot-plugging crytpo cards on KVM guests with debug
kernel build revealed an use after free for the load field of
the struct zcrypt_card. The reason was an incorrect reference
handling of the zcrypt card object which could lead to a free
of the zcrypt card object while it was still in use.
The fix is simple: Before use of the queue not only the queue object
but also the card object needs to increase it's reference count
with a call to zcrypt_card_get(). Similar after use of the queue
not only the queue but also the card object's reference count is
decreased with zcrypt_card_put().
smp_call_function always runs its callback in hard IRQ context, even on
PREEMPT_RT, where spinlocks can sleep. So we need to use a raw spinlock
for cgr_lock to ensure we aren't waiting on a sleeping task.
Although this bug has existed for a while, it was not apparent until
commit ef2a8d5478b9 ("net: dpaa: Adjust queue depth on rate change")
which invokes smp_call_function_single via qman_update_cgr_safe every
time a link goes up or down.
Fixes: 96f413f47677 ("soc/fsl/qbman: fix issue in qman_delete_cgr_safe()") CC: stable@vger.kernel.org Reported-by: Vladimir Oltean <vladimir.oltean@nxp.com> Closes: https://lore.kernel.org/all/20230323153935.nofnjucqjqnz34ej@skbuf/ Reported-by: Steffen Trumtrar <s.trumtrar@pengutronix.de> Closes: https://lore.kernel.org/linux-arm-kernel/87wmsyvclu.fsf@pengutronix.de/ Signed-off-by: Sean Anderson <sean.anderson@linux.dev> Reviewed-by: Camelia Groza <camelia.groza@nxp.com> Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
This adds a function to update a CGR with new parameters. qman_create_cgr
can almost be used for this (with flags=0), but it's not suitable because
it also registers the callback function. The _safe variant was modeled off
of qman_cgr_delete_safe. However, we handle multiple arguments and a return
value.
Signed-off-by: Sean Anderson <sean.anderson@seco.com> Acked-by: Camelia Groza <camelia.groza@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Stable-dep-of: fbec4e7fed89 ("soc: fsl: qbman: Use raw spinlock for cgr_lock") Signed-off-by: Sasha Levin <sashal@kernel.org>
This breaks out/combines get_affine_portal and the cgr sanity check in
preparation for the next commit. No functional change intended.
Signed-off-by: Sean Anderson <sean.anderson@seco.com> Acked-by: Camelia Groza <camelia.groza@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Stable-dep-of: fbec4e7fed89 ("soc: fsl: qbman: Use raw spinlock for cgr_lock") Signed-off-by: Sasha Levin <sashal@kernel.org>
smp_call_function_single disables IRQs when executing the callback. To
prevent deadlocks, we must disable IRQs when taking cgr_lock elsewhere.
This is already done by qman_update_cgr and qman_delete_cgr; fix the
other lockers.
Fixes: 96f413f47677 ("soc/fsl/qbman: fix issue in qman_delete_cgr_safe()") CC: stable@vger.kernel.org Signed-off-by: Sean Anderson <sean.anderson@linux.dev> Reviewed-by: Camelia Groza <camelia.groza@nxp.com> Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Convert ring_buffer_wait() over to wait_event_interruptible(). The default
condition is to execute the wait loop inside __wait_event() just once.
This does not change the ring_buffer_wait() prototype yet, but
restructures the code so that it can take a "cond" and "data" parameter
and will call wait_event_interruptible() with a helper function as the
condition.
The helper function (rb_wait_cond) takes the cond function and data
parameters. It will first check if the buffer hit the watermark defined by
the "full" parameter and then call the passed in condition parameter. If
either are true, it returns true.
If rb_wait_cond() does not return true, it will set the appropriate
"waiters_pending" flag and returns false.
If a reader of the ring buffer is doing a poll, and waiting for the ring
buffer to hit a specific watermark, there could be a case where it gets
into an infinite ping-pong loop.
The writer will see full_waiters_pending and check if the ring buffer is
filled over the percentage of the shortest_full value. If it is, it calls
an irq_work to wake up all the waiters.
[ IRQ work ]
if (rbwork->wakeup_full) {
cpu_buffer->shortest_full = 0;
wakeup poll waiters;
[woken]
[ Wash, rinse, repeat! ]
In the poll, the shortest_full needs to be set before the
full_pending_waiters, as once that is set, the writer will compare the
current shortest_full (which is incorrect) to decide to call the irq_work,
which will reset the shortest_full (expecting the readers to update it).
Also move the setting of full_waiters_pending after the check if the ring
buffer has the required percentage filled. There's no reason to tell the
writer to wake up waiters if there are no waiters.
Link: https://lore.kernel.org/linux-trace-kernel/20240312131952.630922155@goodmis.org Cc: stable@vger.kernel.org Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Cc: Andrew Morton <akpm@linux-foundation.org> Fixes: 42fb0a1e84ff5 ("tracing/ring-buffer: Have polling block on watermark") Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The "shortest_full" variable is used to keep track of the waiter that is
waiting for the smallest amount on the ring buffer before being woken up.
When a tasks waits on the ring buffer, it passes in a "full" value that is
a percentage. 0 means wake up on any data. 1-100 means wake up from 1% to
100% full buffer.
As all waiters are on the same wait queue, the wake up happens for the
waiter with the smallest percentage.
The problem is that the smallest_full on the cpu_buffer that stores the
smallest amount doesn't get reset when all the waiters are woken up. It
does get reset when the ring buffer is reset (echo > /sys/kernel/tracing/trace).
This means that tasks may be woken up more often then when they want to
be. Instead, have the shortest_full field get reset just before waking up
all the tasks. If the tasks wait again, they will update the shortest_full
before sleeping.
Also add locking around setting of shortest_full in the poll logic, and
change "work" to "rbwork" to match the variable name for rb_irq_work
structures that are used in other places.
Link: https://lore.kernel.org/linux-trace-kernel/20240308202431.948914369@goodmis.org Cc: stable@vger.kernel.org Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: linke li <lilinke99@qq.com> Cc: Rabin Vincent <rabin@rab.in> Fixes: 2c2b0a78b3739 ("ring-buffer: Add percentage of ring buffer full to wake up reader") Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Stable-dep-of: 8145f1c35fa6 ("ring-buffer: Fix full_waiters_pending in poll") Signed-off-by: Sasha Levin <sashal@kernel.org>
The rb_watermark_hit() checks if the amount of data in the ring buffer is
above the percentage level passed in by the "full" variable. If it is, it
returns true.
But it also sets the "shortest_full" field of the cpu_buffer that informs
writers that it needs to call the irq_work if the amount of data on the
ring buffer is above the requested amount.
The rb_watermark_hit() always sets the shortest_full even if the amount in
the ring buffer is what it wants. As it is not going to wait, because it
has what it wants, there's no reason to set shortest_full.
Link: https://lore.kernel.org/linux-trace-kernel/20240312115641.6aa8ba08@gandalf.local.home Cc: stable@vger.kernel.org Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Fixes: 42fb0a1e84ff5 ("tracing/ring-buffer: Have polling block on watermark") Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
A task can wait on a ring buffer for when it fills up to a specific
watermark. The writer will check the minimum watermark that waiters are
waiting for and if the ring buffer is past that, it will wake up all the
waiters.
The waiters are in a wait loop, and will first check if a signal is
pending and then check if the ring buffer is at the desired level where it
should break out of the loop.
If a file that uses a ring buffer closes, and there's threads waiting on
the ring buffer, it needs to wake up those threads. To do this, a
"wait_index" was used.
Before entering the wait loop, the waiter will read the wait_index. On
wakeup, it will check if the wait_index is different than when it entered
the loop, and will exit the loop if it is. The waker will only need to
update the wait_index before waking up the waiters.
This had a couple of bugs. One trivial one and one broken by design.
The trivial bug was that the waiter checked the wait_index after the
schedule() call. It had to be checked between the prepare_to_wait() and
the schedule() which it was not.
The main bug is that the first check to set the default wait_index will
always be outside the prepare_to_wait() and the schedule(). That's because
the ring_buffer_wait() doesn't have enough context to know if it should
break out of the loop.
The loop itself is not needed, because all the callers to the
ring_buffer_wait() also has their own loop, as the callers have a better
sense of what the context is to decide whether to break out of the loop
or not.
Just have the ring_buffer_wait() block once, and if it gets woken up, exit
the function and let the callers decide what to do next.
Link: https://lore.kernel.org/all/CAHk-=whs5MdtNjzFkTyaUy=vHi=qwWgPi0JgTe6OYUYMNSRZfg@mail.gmail.com/ Link: https://lore.kernel.org/linux-trace-kernel/20240308202431.792933613@goodmis.org Cc: stable@vger.kernel.org Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: linke li <lilinke99@qq.com> Cc: Rabin Vincent <rabin@rab.in> Fixes: e30f53aad2202 ("tracing: Do not busy wait in buffer splice") Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Stable-dep-of: 761d9473e27f ("ring-buffer: Do not set shortest_full when full target is hit") Signed-off-by: Sasha Levin <sashal@kernel.org>
The commit message in commit fc9a77040b04 ("PCI: designware-ep: Configure
Resizable BAR cap to advertise the smallest size") claims that it modifies
the Resizable BAR capability to only advertise support for 1 MB size BARs.
However, the commit writes all zeroes to PCI_REBAR_CAP (the register which
contains the possible BAR sizes that a BAR be resized to).
According to the spec, it is illegal to not have a bit set in
PCI_REBAR_CAP, and 1 MB is the smallest size allowed.
Set bit 4 in PCI_REBAR_CAP, so that we actually advertise support for a
1 MB BAR size.
Before:
Capabilities: [2e8 v1] Physical Resizable BAR
BAR 0: current size: 1MB
BAR 1: current size: 1MB
BAR 2: current size: 1MB
BAR 3: current size: 1MB
BAR 4: current size: 1MB
BAR 5: current size: 1MB
After:
Capabilities: [2e8 v1] Physical Resizable BAR
BAR 0: current size: 1MB, supported: 1MB
BAR 1: current size: 1MB, supported: 1MB
BAR 2: current size: 1MB, supported: 1MB
BAR 3: current size: 1MB, supported: 1MB
BAR 4: current size: 1MB, supported: 1MB
BAR 5: current size: 1MB, supported: 1MB
Fixes: fc9a77040b04 ("PCI: designware-ep: Configure Resizable BAR cap to advertise the smallest size") Link: https://lore.kernel.org/linux-pci/20240307111520.3303774-1-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel@kernel.org> Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org> Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Cc: <stable@vger.kernel.org> # 5.2 Signed-off-by: Sasha Levin <sashal@kernel.org>
Clang enables -Wenum-enum-conversion and -Wenum-compare-conditional
under -Wenum-conversion. A recent change in Clang strengthened these
warnings and they appear frequently in common builds, primarily due to
several instances in common headers but there are quite a few drivers
that have individual instances as well.
include/linux/vmstat.h:508:43: warning: arithmetic between different enumeration types ('enum zone_stat_item' and 'enum numa_stat_item') [-Wenum-enum-conversion]
508 | return vmstat_text[NR_VM_ZONE_STAT_ITEMS +
| ~~~~~~~~~~~~~~~~~~~~~ ^
509 | item];
| ~~~~
Doing arithmetic between or returning two different types of enums could
be a bug, so each of the instance of the warning needs to be evaluated.
Unfortunately, as mentioned above, there are many instances of this
warning in many different configurations, which can break the build when
CONFIG_WERROR is enabled.
To avoid introducing new instances of the warnings while cleaning up the
disruption for the majority of users, disable these warnings for the
default build while leaving them on for W=1 builds.
This is because we're completing the nfs_direct_request twice in a row.
The source of this is when we have our commit requests to submit, we
process them and send them off, and then in the completion path for the
commit requests we have
if (nfs_commit_end(cinfo.mds))
nfs_direct_write_complete(dreq);
However since we're submitting asynchronous requests we sometimes have
one that completes before we submit the next one, so we end up calling
complete on the nfs_direct_request twice.
The only other place we use nfs_generic_commit_list() is in
__nfs_commit_inode, which wraps this call in a
nfs_commit_begin();
nfs_commit_end();
Which is a common pattern for this style of completion handling, one
that is also repeated in the direct code with get_dreq()/put_dreq()
calls around where we process events as well as in the completion paths.
Fix this by using the same pattern for the commit requests.
Before with my 200 node rocksdb stress running this warning would pop
every 10ish minutes. With my patch the stress test has been running for
several hours without popping.
PM runtime can be done simultaneously with AER error handling. Avoid that
by using pm_runtime_get_sync() before and pm_runtime_put() after reset in
pcie_do_recovery() for all recovering devices.
pm_runtime_get_sync() will increase dev->power.usage_count counter to
prevent any possible future request to runtime suspend a device. It will
also resume a device, if it was previously in D3hot state.
I tested with igc device by doing simultaneous aer_inject and rpm
suspend/resume via /sys/bus/pci/devices/PCI_ID/power/control and can
reproduce:
igc 0000:02:00.0: not ready 65535ms after bus reset; giving up
pcieport 0000:00:1c.2: AER: Root Port link has been reset (-25)
pcieport 0000:00:1c.2: AER: subordinate device reset failed
pcieport 0000:00:1c.2: AER: device recovery failed
igc 0000:02:00.0: Unable to change power state from D3hot to D0, device inaccessible
The problem disappears when this patch is applied.
When userland echoes 8bit characters to /dev/synth with e.g.
echo -e '\xe9' > /dev/synth
synth_write would get characters beyond 0x7f, and thus negative when
char is signed. When given to synth_buffer_add which takes a u16, this
would sign-extend and produce a U+ffxy character rather than U+xy.
Users thus get garbled text instead of accents in their output.
Let's fix this by making sure that we read unsigned characters.
This commit resolves an issue in the tegra-xudc USB gadget driver that
incorrectly fetched USB3 PHY instances. The problem stemmed from the
assumption of a one-to-one correspondence between USB2 and USB3 PHY
names and their association with physical USB ports in the device tree.
Previously, the driver associated USB3 PHY names directly with the USB3
instance number, leading to mismatches when mapping the physical USB
ports. For instance, if using USB3-1 PHY, the driver expect the
corresponding PHY name as 'usb3-1'. However, the physical USB ports in
the device tree were designated as USB2-0 and USB3-0 as we only have
one device controller, causing a misalignment.
This commit rectifies the issue by adjusting the PHY naming logic.
Now, the driver correctly correlates the USB2 and USB3 PHY instances,
allowing the USB2-0 and USB3-1 PHYs to form a physical USB port pair
while accurately reflecting their configuration in the device tree by
naming them USB2-0 and USB3-0, respectively.
The change ensures that the PHY and PHY names align appropriately,
resolving the mismatch between physical USB ports and their associated
names in the device tree.
Fixes: b4e19931c98a ("usb: gadget: tegra-xudc: Support multiple device modes") Cc: stable@vger.kernel.org Signed-off-by: Wayne Chang <waynec@nvidia.com> Reviewed-by: Jon Hunter <jonathanh@nvidia.com> Tested-by: Jon Hunter <jonathanh@nvidia.com> Link: https://lore.kernel.org/r/20240307030328.1487748-3-waynec@nvidia.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
This patch introduces a new API, tegra_xusb_padctl_get_port_number,
to the Tegra XUSB Pad Controller driver. This API is used to identify
the USB port that is associated with a given PHY.
The function takes a PHY pointer for either a USB2 PHY or USB3 PHY as input
and returns the corresponding port number. If the PHY pointer is invalid,
it returns -ENODEV.
Cc: stable@vger.kernel.org Signed-off-by: Wayne Chang <waynec@nvidia.com> Reviewed-by: Jon Hunter <jonathanh@nvidia.com> Tested-by: Jon Hunter <jonathanh@nvidia.com> Link: https://lore.kernel.org/r/20240307030328.1487748-2-waynec@nvidia.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
ida_alloc() and ida_free() should be preferred to the deprecated
ida_simple_get() and ida_simple_remove().
Note that the upper limit of ida_simple_get() is exclusive, but the one of
ida_alloc_range() is inclusive. So change this change allows one more
device. Previously address 0xFE was never used.
clang-16 warns about casting functions to incompatible types, as is done
here to call clk_disable_unprepare:
drivers/nvmem/meson-efuse.c:78:12: error: cast from 'void (*)(struct clk *)' to 'void (*)(void *)' converts to incompatible function type [-Werror,-Wcast-function-type-strict]
78 | (void(*)(void *))clk_disable_unprepare,
The pattern of getting, enabling and setting a disable callback for a
clock can be replaced with devm_clk_get_enabled(), which also fixes
this warning.
We observed a corruption during on-line resize of a file system that is
larger than 16 TiB with 4k block size. With having more then 2^32 blocks
resize_inode is turned off by default by mke2fs. The issue can be
reproduced on a smaller file system for convenience by explicitly
turning off resize_inode. An on-line resize across an 8 GiB boundary (the
size of a meta block group in this setup) then leads to a corruption:
dev=/dev/<some_dev> # should be >= 16 GiB
mkdir -p /corruption
/sbin/mke2fs -t ext4 -b 4096 -O ^resize_inode $dev $((2 * 2**21 - 2**15))
mount -t ext4 $dev /corruption
2^21 = 2^15*2^6 equals 8 GiB whereof 2^15 is the number of blocks per
block group and 2^6 are the number of block groups that make a meta
block group.
The last checksum might be different depending on how the file is laid
out across the physical blocks. The actual corruption occurs at physical
block 63*2^15 = 2064384 which would be the location of the backup of the
meta block group's block descriptor. During the on-line resize the file
system will be converted to meta_bg starting at s_first_meta_bg which is
2 in the example - meaning all block groups after 16 GiB. However, in
ext4_flex_group_add we might add block groups that are not part of the
first meta block group yet. In the reproducer we achieved this by
substracting the size of a whole block group from the point where the
meta block group would start. This must be considered when updating the
backup block group descriptors to follow the non-meta_bg layout. The fix
is to add a test whether the group to add is already part of the meta
block group or not.
Fixes: 01f795f9e0d67 ("ext4: add online resizing support for meta_bg and 64-bit file systems") Cc: <stable@vger.kernel.org> Signed-off-by: Maximilian Heyne <mheyne@amazon.de> Tested-by: Srivathsa Dara <srivathsa.d.dara@oracle.com> Reviewed-by: Srivathsa Dara <srivathsa.d.dara@oracle.com> Link: https://lore.kernel.org/r/20240215155009.94493-1-mheyne@amazon.de Signed-off-by: Theodore Ts'o <tytso@mit.edu> Signed-off-by: Sasha Levin <sashal@kernel.org>
Add of_match table for "ti,amc6821" compatible string.
This fixes automatic driver loading by userspace when using device-tree,
and if built as a module like major linux distributions do.
While devices probe just fine with i2c_device_id table, userspace can't
match the "ti,amc6821" compatible string from dt with the plain
"amc6821" device id. As a result, the kernel module can not be loaded.
Because sandboxing can be used as an opportunistic security measure,
user space may not log unsupported features. Let the system
administrator know if an application tries to use Landlock but failed
because it isn't enabled at boot time. This may be caused by boot
loader configurations with outdated "lsm" kernel's command-line
parameter.
The hwdb selection logic as a feature that allows it to mark some fields
as 'don't care'. If we match with such a field we memcpy(..)
the current etnaviv_chip_identity into ident.
This step can overwrite some id values read from the GPU with the
'don't care' value.
Fix this issue by restoring the affected values after the memcpy(..).
As this is crucial for user space to know when this feature works as
expected increment the minor version too.
Commit e7794c14fd73 ("mmc: rpmb: fixes pause retune on all RPMB
partitions.") added a mask check for 'part_type', but the mask used was
wrong leading to the code intended for rpmb also being executed for GP3.
On some MMCs (but not all) this would make gp3 partition inaccessible:
armadillo:~# head -c 1 < /dev/mmcblk2gp3
head: standard input: I/O error
armadillo:~# dmesg -c
[ 422.976583] mmc2: running CQE recovery
[ 423.058182] mmc2: running CQE recovery
[ 423.137607] mmc2: running CQE recovery
[ 423.137802] blk_update_request: I/O error, dev mmcblk2gp3, sector 0 op 0x0:(READ) flags 0x80700 phys_seg 4 prio class 0
[ 423.237125] mmc2: running CQE recovery
[ 423.318206] mmc2: running CQE recovery
[ 423.397680] mmc2: running CQE recovery
[ 423.397837] blk_update_request: I/O error, dev mmcblk2gp3, sector 0 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0
[ 423.408287] Buffer I/O error on dev mmcblk2gp3, logical block 0, async page read
the part_type values of interest here are defined as follow:
main 0
boot0 1
boot1 2
rpmb 3
gp0 4
gp1 5
gp2 6
gp3 7
so mask with EXT_CSD_PART_CONFIG_ACC_MASK (7) to correctly identify rpmb
There was previously a theoretical window where swapoff() could run and
teardown a swap_info_struct while a call to free_swap_and_cache() was
running in another thread. This could cause, amongst other bad
possibilities, swap_page_trans_huge_swapped() (called by
free_swap_and_cache()) to access the freed memory for swap_map.
This is a theoretical problem and I haven't been able to provoke it from a
test case. But there has been agreement based on code review that this is
possible (see link below).
Fix it by using get_swap_device()/put_swap_device(), which will stall
swapoff(). There was an extra check in _swap_info_get() to confirm that
the swap entry was not free. This isn't present in get_swap_device()
because it doesn't make sense in general due to the race between getting
the reference and swapoff. So I've added an equivalent check directly in
free_swap_and_cache().
Details of how to provoke one possible issue (thanks to David Hildenbrand
for deriving this):
--8<-----
__swap_entry_free() might be the last user and result in
"count == SWAP_HAS_CACHE".
swapoff->try_to_unuse() will stop as soon as soon as si->inuse_pages==0.
So the question is: could someone reclaim the folio and turn
si->inuse_pages==0, before we completed swap_page_trans_huge_swapped().
Imagine the following: 2 MiB folio in the swapcache. Only 2 subpages are
still references by swap entries.
Process 1 still references subpage 0 via swap entry.
Process 2 still references subpage 1 via swap entry.
Process 1 quits. Calls free_swap_and_cache().
-> count == SWAP_HAS_CACHE
[then, preempted in the hypervisor etc.]
Process 2 quits. Calls free_swap_and_cache().
-> count == SWAP_HAS_CACHE
Process 2 goes ahead, passes swap_page_trans_huge_swapped(), and calls
__try_to_reclaim_swap().
The general rule to use a swap entry is as follows.
When we get a swap entry, if there aren't some other ways to prevent
swapoff, such as the folio in swap cache is locked, page table lock is
held, etc., the swap entry may become invalid because of swapoff.
Then, we need to enclose all swap related functions with
get_swap_device() and put_swap_device(), unless the swap functions
call get/put_swap_device() by themselves.
Add the rule as comments of get_swap_device().
Link: https://lkml.kernel.org/r/20230529061355.125791-6-ying.huang@intel.com Signed-off-by: "Huang, Ying" <ying.huang@intel.com> Reviewed-by: David Hildenbrand <david@redhat.com> Reviewed-by: Yosry Ahmed <yosryahmed@google.com> Reviewed-by: Chris Li (Google) <chrisl@kernel.org> Cc: Hugh Dickins <hughd@google.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Matthew Wilcox <willy@infradead.org> Cc: Michal Hocko <mhocko@suse.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Tim Chen <tim.c.chen@linux.intel.com> Cc: Yang Shi <shy828301@gmail.com> Cc: Yu Zhao <yuzhao@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Stable-dep-of: 82b1c07a0af6 ("mm: swap: fix race between free_swap_and_cache() and swapoff()") Signed-off-by: Sasha Levin <sashal@kernel.org>
mac802154_llsec_key_del() can free resources of a key directly without
following the RCU rules for waiting before the end of a grace period. This
may lead to use-after-free in case llsec_lookup_key() is traversing the
list of keys in parallel with a key deletion:
Handle the proper resource release in the RCU callback function
mac802154_llsec_key_del_rcu().
Note that if llsec_lookup_key() finds a key, it gets a refcount via
llsec_key_get() and locally copies key id from key_entry (which is a
list element). So it's safe to call llsec_key_put() and free the list
entry after the RCU grace period elapses.
Found by Linux Verification Center (linuxtesting.org).
Fixes: 5d637d5aabd8 ("mac802154: add llsec structures and mutators") Cc: stable@vger.kernel.org Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru> Acked-by: Alexander Aring <aahringo@redhat.com>
Message-ID: <20240228163840.6667-1-pchelkin@ispras.ru> Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The lockdep assert is added by commit a448af25becf ("md/raid10: remove
rcu protection to access rdev from conf") in print_conf(). And I didn't
notice that dm-raid is calling "pers->hot_add_disk" without holding
'reconfig_mutex'.
"pers->hot_add_disk" read and write many fields that is protected by
'reconfig_mutex', and raid_resume() already grab the lock in other
contex. Hence fix this problem by protecting "pers->host_add_disk"
with the lock.
Fixes: 9092c02d9435 ("DM RAID: Add ability to restore transiently failed devices on resume") Fixes: a448af25becf ("md/raid10: remove rcu protection to access rdev from conf") Cc: stable@vger.kernel.org # v6.7+ Signed-off-by: Yu Kuai <yukuai3@huawei.com> Signed-off-by: Xiao Ni <xni@redhat.com> Acked-by: Mike Snitzer <snitzer@kernel.org> Signed-off-by: Song Liu <song@kernel.org> Link: https://lore.kernel.org/r/20240305072306.2562024-10-yukuai1@huaweicloud.com Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit 5459c0b70467 ("PCI/DPC: Quirk PIO log size for certain Intel Root
Ports") and commit 3b8803494a06 ("PCI/DPC: Quirk PIO log size for Intel Ice
Lake Root Ports") add quirks for Ice, Tiger and Alder Lake Root Ports.
System firmware for Raptor Lake still has the bug, so Linux logs the
warning below on several Raptor Lake systems like Dell Precision 3581 with
Intel Raptor Lake processor (0W18NX) system firmware/BIOS version 1.10.1.
pci 0000:00:07.0: [8086:a76e] type 01 class 0x060400
pci 0000:00:07.0: DPC: RP PIO log size 0 is invalid
pci 0000:00:07.1: [8086:a73f] type 01 class 0x060400
pci 0000:00:07.1: DPC: RP PIO log size 0 is invalid
Apply the quirk for Raptor Lake Root Ports as well.
This also enables the DPC driver to dump the RP PIO Log registers when DPC
is triggered.
Link: https://lore.kernel.org/r/20240305113057.56468-1-pmenzel@molgen.mpg.de Reported-by: Niels van Aert <nvaert1986@hotmail.com> Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218560 Signed-off-by: Paul Menzel <pmenzel@molgen.mpg.de> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com> Cc: <stable@vger.kernel.org> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Niels van Aert <nvaert1986@hotmail.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Some Root Ports on Intel Tiger Lake and Alder Lake systems support the RP
Extensions for DPC and the RP PIO Log registers but incorrectly advertise
an RP PIO Log Size of zero. This means the kernel complains that:
DPC: RP PIO log size 0 is invalid
and if DPC is triggered, the DPC driver will not dump the RP PIO Log
registers when it should.
This is caused by a BIOS bug and should be fixed the BIOS for future CPUs.
Add a quirk to set the correct RP PIO Log size for the affected Root Ports.
Intel DG2 discrete graphics PCIe endpoints advertise L1 acceptable exit
latency to be < 1us even though they can actually tolerate unlimited exit
latencies just fine. Quirk the L1 acceptable exit latency for these
endpoints to be unlimited so ASPM L1 can be enabled.
[bhelgaas: use FIELD_GET/FIELD_PREP, wordsmith comment & commit log] Link: https://lore.kernel.org/r/20220405093810.76613-1-mika.westerberg@linux.intel.com Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Stable-dep-of: 627c6db20703 ("PCI/DPC: Quirk PIO log size for Intel Raptor Lake Root Ports") Signed-off-by: Sasha Levin <sashal@kernel.org>
Per PCIe r5, sec 7.5.1.2.4, a device must not claim accesses to its
Expansion ROM unless both the Memory Space Enable and the Expansion ROM
Enable bit are set. But apparently some Intel I210 NICs don't work
correctly if the ROM BAR overlaps another BAR, even if the Expansion ROM is
disabled.
Michael reported that on a Kontron SMARC-sAL28 ARM64 system with U-Boot
v2021.01-rc3, the ROM BAR overlaps BAR 3, and networking doesn't work at
all:
BAR 0: 0x40000000 (32-bit, non-prefetchable) [size=1M]
BAR 3: 0x40200000 (32-bit, non-prefetchable) [size=16K]
ROM: 0x40200000 (disabled) [size=1M]
Previously, pci_std_update_resource() wrote the assigned ROM address to the
BAR only when the ROM was enabled. This meant that the I210 ROM BAR could
be left with an address assigned by firmware, which might overlap with
other BARs.
Quirk these I210 devices so pci_std_update_resource() always writes the
assigned address to the ROM BAR, whether or not the ROM is enabled.
A race condition between the .runtime_idle() callback and the .remove()
callback in the rtsx_pcr PCI driver leads to a kernel crash due to an
unhandled page fault [1].
The problem is that rtsx_pci_runtime_idle() is not expected to be running
after pm_runtime_get_sync() has been called, but the latter doesn't really
guarantee that. It only guarantees that the suspend and resume callbacks
will not be running when it returns.
However, if a .runtime_idle() callback is already running when
pm_runtime_get_sync() is called, the latter will notice that the runtime PM
status of the device is RPM_ACTIVE and it will return right away without
waiting for the former to complete. In fact, it cannot wait for
.runtime_idle() to complete because it may be called from that callback (it
arguably does not make much sense to do that, but it is not strictly
prohibited).
Thus in general, whoever is providing a .runtime_idle() callback needs
to protect it from running in parallel with whatever code runs after
pm_runtime_get_sync(). [Note that .runtime_idle() will not start after
pm_runtime_get_sync() has returned, but it may continue running then if it
has started earlier.]
One way to address that race condition is to call pm_runtime_barrier()
after pm_runtime_get_sync() (not before it, because a nonzero value of the
runtime PM usage counter is necessary to prevent runtime PM callbacks from
being invoked) to wait for the .runtime_idle() callback to complete should
it be running at that point. A suitable place for doing that is in
pci_device_remove() which calls pm_runtime_get_sync() before removing the
driver, so it may as well call pm_runtime_barrier() subsequently, which
will prevent the race in question from occurring, not just in the rtsx_pcr
driver, but in any PCI drivers providing .runtime_idle() callbacks.
projects / thirdparty / kernel / stable.git / log
