The result of that function is in essence boolean, so simplify to return the
result of the relevant expression. It also makes it follow the convention used
by __verify_patch_section().
Due to specific requirements while applying microcode patches on Zen1
and 2, the patch buffer mapping needs to be flushed from the TLB after
application. Do so.
If not, unnecessary and unnatural delays happen in the boot process.
Reported-by: Thomas De Schampheleire <thomas.de_schampheleire@nokia.com> Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de> Tested-by: Thomas De Schampheleire <thomas.de_schampheleire@nokia.com> Cc: <stable@kernel.org> # f1d84b59cbb9 ("x86/mm: Carve out INVLPG inline asm for use by others") Link: https://lore.kernel.org/r/ZyulbYuvrkshfsd2@antipodes Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, an unconditional cache flush is performed during every
microcode update. Although the original changelog did not mention
a specific erratum, this measure was primarily intended to address
a specific microcode bug, the load of which has already been blocked by
is_blacklisted(). Therefore, this cache flush is no longer necessary.
Additionally, the side effects of doing this have been overlooked. It
increases CPU rendezvous time during late loading, where the cache flush
takes between 1x to 3.5x longer than the actual microcode update.
Remove native_wbinvd() and update the erratum name to align with the
latest errata documentation, document ID 334163 Version 022US.
[ bp: Zap the flaky documentation URL. ]
Fixes: 91df9fdf5149 ("x86/microcode/intel: Writeback and invalidate caches before updating microcode") Reported-by: Yan Hua Wu <yanhua1.wu@intel.com> Reported-by: William Xie <william.xie@intel.com> Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com> Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de> Acked-by: Ashok Raj <ashok.raj@intel.com> Tested-by: Yan Hua Wu <yanhua1.wu@intel.com> Link: https://lore.kernel.org/r/20241001161042.465584-2-chang.seok.bae@intel.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This function should've been split a long time ago because it is used in
two paths:
1) On the late loading path, when the microcode is loaded through the
request_firmware interface
2) In the save_microcode_in_initrd() path which collects all the
microcode patches which are relevant for the current system before
the initrd with the microcode container has been jettisoned.
In that path, it is not really necessary to iterate over the nodes on
a system and match a patch however it didn't cause any trouble so it
was left for a later cleanup
However, that later cleanup was expedited by the fact that Jens was
enabling "Use L3 as a NUMA node" in the BIOS setting in his machine and
so this causes the NUMA CPU masks used in cpumask_of_node() to be
generated *after* 2) above happened on the first node. Which means, all
those masks were funky, wrong, uninitialized and whatnot, leading to
explosions when dereffing c->microcode in load_microcode_amd().
So split that function and do only the necessary work needed at each
stage.
Fixes: 94838d230a6c ("x86/microcode/AMD: Use the family,model,stepping encoded in the patch ID") Reported-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de> Tested-by: Jens Axboe <axboe@kernel.dk> Link: https://lore.kernel.org/r/91194406-3fdf-4e38-9838-d334af538f74@kernel.dk Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit in Fixes changed how a microcode patch is loaded on Zen and newer but
the patch matching needs to happen with different rigidity, depending on what
is being done:
1) When the patch is added to the patches cache, the stepping must be ignored
because the driver still supports different steppings per system
2) When the patch is matched for loading, then the stepping must be taken into
account because each CPU needs the patch matching its exact stepping
Take care of that by making the matching smarter.
Fixes: 94838d230a6c ("x86/microcode/AMD: Use the family,model,stepping encoded in the patch ID") Reported-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de> Tested-by: Jens Axboe <axboe@kernel.dk> Link: https://lore.kernel.org/r/91194406-3fdf-4e38-9838-d334af538f74@kernel.dk Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On Zen and newer, the family, model and stepping is part of the
microcode patch ID so that the equivalence table the driver has been
using, is not needed anymore.
So switch the driver to use that from now on.
The equivalence table in the microcode blob should still remain in case
there's need to pass some additional information to the kernel loader.
The AMD side of the loader issues the microcode revision for each
logical thread on the system, which can become really noisy on huge
machines. And doing that doesn't make a whole lot of sense - the
microcode revision is already in /proc/cpuinfo.
So in case one is interested in the theoretical support of mixed silicon
steppings on AMD, one can check there.
What is also missing on the AMD side - something which people have
requested before - is showing the microcode revision the CPU had
*before* the early update.
So abstract that up in the main code and have the BSP on each vendor
provide those revision numbers.
Then, dump them only once on driver init.
On Intel, do not dump the patch date - it is not needed.
Applying microcode late can be fatal for the running kernel when the
update changes functionality which is in use already in a non-compatible
way, e.g. by removing a CPUID bit.
There is no way for admins which do not have access to the vendors deep
technical support to decide whether late loading of such a microcode is
safe or not.
Intel has added a new field to the microcode header which tells the
minimal microcode revision which is required to be active in the CPU in
order to be safe.
Provide infrastructure for handling this in the core code and a command
line switch which allows to enforce it.
If the update is considered safe the kernel is not tainted and the annoying
warning message not emitted. If it's enforced and the currently loaded
microcode revision is not safe for late loading then the load is aborted.
Offline CPUs need to be parked in a safe loop when microcode update is
in progress on the primary CPU. Currently, offline CPUs are parked in
mwait_play_dead(), and for Intel CPUs, its not a safe instruction,
because the MWAIT instruction can be patched in the new microcode update
that can cause instability.
- Add a new microcode state 'UCODE_OFFLINE' to report status on per-CPU
basis.
- Force NMI on the offline CPUs.
Wake up offline CPUs while the update is in progress and then return
them back to mwait_play_dead() after microcode update is complete.
When SMT siblings are soft-offlined and parked in one of the play_dead()
variants they still react on NMI, which is problematic on affected Intel
CPUs. The default play_dead() variant uses MWAIT on modern CPUs, which is
not guaranteed to be safe when updated concurrently.
Right now late loading is prevented when not all SMT siblings are online,
but as they still react on NMI, it is possible to bring them out of their
park position into a trivial rendezvous handler.
Provide a function which allows to do that. I does sanity checks whether
the target is in the cpus_booted_once_mask and whether the APIC driver
supports it.
Mark X2APIC and XAPIC as capable, but exclude 32bit and the UV and NUMACHIP
variants as that needs feedback from the relevant experts.
The wait for control loop in which the siblings are waiting for the
microcode update on the primary thread must be protected against
instrumentation as instrumentation can end up in #INT3, #DB or #PF,
which then returns with IRET. That IRET reenables NMI which is the
opposite of what the NMI rendezvous is trying to achieve.
stop_machine() does not prevent the spin-waiting sibling from handling
an NMI, which is obviously violating the whole concept of rendezvous.
Implement a static branch right in the beginning of the NMI handler
which is nopped out except when enabled by the late loading mechanism.
The late loader enables the static branch before stop_machine() is
invoked. Each CPU has an nmi_enable in its control structure which
indicates whether the CPU should go into the update routine.
This is required to bridge the gap between enabling the branch and
actually being at the point where it is required to enter the loader
wait loop.
Each CPU which arrives in the stopper thread function sets that flag and
issues a self NMI right after that. If the NMI function sees the flag
clear, it returns. If it's set it clears the flag and enters the
rendezvous.
This is safe against a real NMI which hits in between setting the flag
and sending the NMI to itself. The real NMI will be swallowed by the
microcode update and the self NMI will then let stuff continue.
Otherwise this would end up with a spurious NMI.
The current all in one code is unreadable and really not suited for
adding future features like uniform loading with package or system
scope.
Provide a set of new control functions which split the handling of the
primary and secondary CPUs. These will replace the current rendezvous
all in one function in the next step. This is intentionally a separate
change because diff makes an complete unreadable mess otherwise.
So the flow separates the primary and the secondary CPUs into their own
functions which use the control field in the per CPU ucode_ctrl struct.
Add a per CPU control field to ucode_ctrl and define constants for it
which are going to be used to control the loading state machine.
In theory this could be a global control field, but a global control does
not cover the following case:
15 primary CPUs load microcode successfully
1 primary CPU fails and returns with an error code
With global control the sibling of the failed CPU would either try again or
the whole operation would be aborted with the consequence that the 15
siblings do not invoke the apply path and end up with inconsistent software
state. The result in dmesg would be inconsistent too.
There are two additional fields added and initialized:
ctrl_cpu and secondaries. ctrl_cpu is the CPU number of the primary thread
for now, but with the upcoming uniform loading at package or system scope
this will be one CPU per package or just one CPU. Secondaries hands the
control CPU a CPU mask which will be required to release the secondary CPUs
out of the wait loop.
Preparatory change for implementing a properly split control flow for
primary and secondary CPUs.
- The return value is pointless as this is a strict boolean.
- It's way simpler to count down from num_online_cpus() and check for
zero.
- The timeout argument is pointless as this is always one second.
- Touching the NMI watchdog every 100ns does not make any sense, neither
does checking every 100ns. This is really not a hotpath operation.
Preload the atomic counter with the number of online CPUs and simplify the
whole timeout logic. Delay for one microsecond and touch the NMI watchdog
once per millisecond.
reload_store() is way too complicated. Split the inner workings out and
make the following enhancements:
- Taint the kernel only when the microcode was actually updated. If. e.g.
the rendezvous fails, then nothing happened and there is no reason for
tainting.
- Return useful error codes
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de> Reviewed-by: Nikolay Borisov <nik.borisov@suse.com> Link: https://lore.kernel.org/r/20231002115903.145048840@linutronix.de Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On CPUs where microcode loading is not NMI-safe the SMT siblings which
are parked in one of the play_dead() variants still react to NMIs.
So if an NMI hits while the primary thread updates the microcode the
resulting behaviour is undefined. The default play_dead() implementation on
modern CPUs is using MWAIT which is not guaranteed to be safe against
a microcode update which affects MWAIT.
Take the cpus_booted_once_mask into account to detect this case and
refuse to load late if the vendor specific driver does not advertise
that late loading is NMI safe.
AMD stated that this is safe, so mark the AMD driver accordingly.
This requirement will be partially lifted in later changes.
Scheduling work on all CPUs to collect the microcode information is just
another extra step for no value. Let the CPU hotplug callback registration
do it.
Now that the microcode cache is initialized before the APs are brought
up, there is no point in scanning builtin/initrd microcode during AP
loading.
Convert the AP loader to utilize the cache, which in turn makes the CPU
hotplug callback which applies the microcode after initrd/builtin is
gone, obsolete as the early loading during late hotplug operations
including the resume path depends now only on the cache.
There is no reason to scan builtin/initrd microcode on each AP.
Cache the builtin/initrd microcode in an early initcall so that the
early AP loader can utilize the cache.
The existing fs initcall which invoked save_microcode_in_initrd_amd() is
still required to maintain the initrd_gone flag. Rename it accordingly.
This will be removed once the AP loader code is converted to use the
cache.
Microcode is applied on the APs during early bringup. There is no point
in trying to apply the microcode again during the hotplug operations and
neither at the point where the microcode device is initialized.
Collect CPU info and microcode revision in setup_online_cpu() for now.
This will move to the CPU hotplug callback later.
[ bp: Leave the starting notifier for the following scenario:
- boot, late load, suspend to disk, resume
without the starting notifier, only the last core manages to update the
microcode upon resume:
For the future, one should check whether potential unification of
the CPU init path could cover the resume path too so that this can
be simplified even more.
tglx: This is caused by the odd handling of APs which try to find the
microcode blob in builtin or initrd instead of caching the microcode
blob during early init before the APs are brought up. Will be cleaned
up in a later step. ]
Nothing needs struct ucode_cpu_info. Make it take struct cpu_signature,
let it return a boolean and simplify the implementation. Rename it now
that the silly name clash with collect_cpu_info() is gone.
There are situations where the late microcode is loaded into memory but
is not applied:
1) The rendezvous fails
2) The microcode is rejected by the CPUs
If any of this happens then the pointer which was updated at firmware
load time is stale and subsequent CPU hotplug operations either fail to
update or create inconsistent microcode state.
Save the loaded microcode in a separate pointer before the late load is
attempted and when successful, update the hotplug pointer accordingly
via a new microcode_ops callback.
Remove the pointless fallback in the loader to a microcode pointer which
is never populated.
- It scans the builtin/initrd for microcode not only on the BSP, but also
on all APs during early boot and then later in the boot process it
scans again to duplicate and save the microcode before initrd goes
away.
That's a pointless exercise because this can be simply done before
bringing up the APs when the memory allocator is up and running.
- Saving the microcode from within the scan loop is completely
non-obvious and a left over of the microcode cache.
This can be done at the call site now which makes it obvious.
Rework the code so that only the BSP scans the builtin/initrd microcode
once during early boot and save it away in an early initcall for later
use.
[ bp: Test and fold in a fix from tglx ontop which handles the need to
distinguish what save_microcode() does depending on when it is
called:
- when on the BSP during early load, it needs to find a newer
revision than the one currently loaded on the BSP
- later, before SMP init, it still runs on the BSP and gets the BSP
revision just loaded and uses that revision to know which patch
to save for the APs. For that it needs to find the exact one as
on the BSP.
]
Sanitize the microcode scan loop, fixup printks and move the loading
function for builtin microcode next to the place where it is used and mark
it __init.
Mixed steppings aren't supported on Intel CPUs. Only one microcode patch
is required for the entire system. The caching of microcode blobs which
match the family and model is therefore pointless and in fact is
dysfunctional as CPU hotplug updates use only a single microcode blob,
i.e. the one where *intel_ucode_patch points to.
Remove the microcode cache and make it an AMD local feature.
[ tglx:
- save only at the end. Otherwise random microcode ends up in the
pointer for early loading
- free the ucode patch pointer in save_microcode_patch() only
after kmemdup() has succeeded, as reported by Andrew Cooper ]
32-bit loads microcode before paging is enabled. The commit which
introduced that has zero justification in the changelog. The cover
letter has slightly more content, but it does not give any technical
justification either:
"The problem in current microcode loading method is that we load a
microcode way, way too late; ideally we should load it before turning
paging on. This may only be practical on 32 bits since we can't get
to 64-bit mode without paging on, but we should still do it as early
as at all possible."
Handwaving word salad with zero technical content.
Someone claimed in an offlist conversation that this is required for
curing the ATOM erratum AAE44/AAF40/AAG38/AAH41. That erratum requires
an microcode update in order to make the usage of PSE safe. But during
early boot, PSE is completely irrelevant and it is evaluated way later.
Neither is it relevant for the AP on single core HT enabled CPUs as the
microcode loading on the AP is not doing anything.
On dual core CPUs there is a theoretical problem if a split of an
executable large page between enabling paging including PSE and loading
the microcode happens. But that's only theoretical, it's practically
irrelevant because the affected dual core CPUs are 64bit enabled and
therefore have paging and PSE enabled before loading the microcode on
the second core. So why would it work on 64-bit but not on 32-bit?
The erratum:
"AAG38 Code Fetch May Occur to Incorrect Address After a Large Page is
Split Into 4-Kbyte Pages
Problem: If software clears the PS (page size) bit in a present PDE
(page directory entry), that will cause linear addresses mapped through
this PDE to use 4-KByte pages instead of using a large page after old
TLB entries are invalidated. Due to this erratum, if a code fetch uses
this PDE before the TLB entry for the large page is invalidated then it
may fetch from a different physical address than specified by either the
old large page translation or the new 4-KByte page translation. This
erratum may also cause speculative code fetches from incorrect addresses."
The practical relevance for this is exactly zero because there is no
splitting of large text pages during early boot-time, i.e. between paging
enable and microcode loading, and neither during CPU hotplug.
IOW, this load microcode before paging enable is yet another voodoo
programming solution in search of a problem. What's worse is that it causes
at least two serious problems:
1) When stackprotector is enabled, the microcode loader code has the
stackprotector mechanics enabled. The read from the per CPU variable
__stack_chk_guard is always accessing the virtual address either
directly on UP or via %fs on SMP. In physical address mode this
results in an access to memory above 3GB. So this works by chance as
the hardware returns the same value when there is no RAM at this
physical address. When there is RAM populated above 3G then the read
is by chance the same as nothing changes that memory during the very
early boot stage. That's not necessarily true during runtime CPU
hotplug.
2) When function tracing is enabled, the relevant microcode loader
functions and the functions invoked from there will call into the
tracing code and evaluate global and per CPU variables in physical
address mode. What could potentially go wrong?
Cure this and move the microcode loading after the early paging enable, use
the new temporary initrd mapping and remove the gunk in the microcode
loader which is required to handle physical address mode.
UART controllers without flow control seem to behave unstable
in case DMA is enabled. The issues were indicated in the message:
https://lore.kernel.org/linux-arm-kernel/CAMdYzYpXtMocCtCpZLU_xuWmOp2Ja_v0Aj0e6YFNRA-yV7u14g@mail.gmail.com/
In case of PX30-uQ7 Ringneck SoM, it was noticed that after couple
of hours of UART communication, the CPU stall was occurring,
leading to the system becoming unresponsive.
After disabling the DMA, extensive UART communication tests for
up to two weeks were performed, and no issues were further
observed.
The flow control pins for uart5 are not available on PX30-uQ7
Ringneck, as configured by pinctrl-0, so the DMA nodes were
removed on SoM dtsi.
Cc: stable@vger.kernel.org Fixes: c484cf93f61b ("arm64: dts: rockchip: add PX30-µQ7 (Ringneck) SoM with Haikou baseboard") Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de> Signed-off-by: Lukasz Czechowski <lukasz.czechowski@thaumatec.com> Link: https://lore.kernel.org/r/20250121125604.3115235-3-lukasz.czechowski@thaumatec.com Signed-off-by: Heiko Stuebner <heiko@sntech.de>
[ conflict resolution due to missing (cosmetic) backport of 4eee627ea59304cdd66c5d4194ef13486a6c44fc] Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Intel idle driver is preferred over the ACPI processor idle driver,
but fails to implement the work around for Core2 generation CPUs, where
the TSC stops in C2 and deeper C-states. This causes stalls and boot
delays, when the clocksource watchdog does not catch the unstable TSC
before the CPU goes deep idle for the first time.
The ACPI driver marks the TSC unstable when it detects that the CPU
supports C2 or deeper and the CPU does not have a non-stop TSC.
Add the equivivalent work around to the Intel idle driver to cure that.
Fixes: 18734958e9bf ("intel_idle: Use ACPI _CST for processor models without C-state tables") Reported-by: Fab Stz <fabstz-it@yahoo.fr> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Fab Stz <fabstz-it@yahoo.fr> Cc: All applicable <stable@vger.kernel.org> Closes: https://lore.kernel.org/all/10cf96aa-1276-4bd4-8966-c890377030c3@yahoo.fr Link: https://patch.msgid.link/87bjupfy7f.ffs@tglx Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Before this patch the NETDEV_XDP_ACT_NDO_XMIT XDP feature flag is set by
default as part of driver initialization, and is never cleared. However,
this flag differs from others in that it is used as an indicator for
whether the driver is ready to perform the ndo_xdp_xmit operation as
part of an XDP_REDIRECT. Kernel helpers
xdp_features_(set|clear)_redirect_target exist to convey this meaning.
This patch ensures that the netdev is only reported as a redirect target
when XDP queues exist to forward traffic.
Fixes: 39a7f4aa3e4a ("gve: Add XDP REDIRECT support for GQI-QPL format") Cc: stable@vger.kernel.org Reviewed-by: Praveen Kaligineedi <pkaligineedi@google.com> Reviewed-by: Jeroen de Borst <jeroendb@google.com> Signed-off-by: Joshua Washington <joshwash@google.com> Link: https://patch.msgid.link/20250214224417.1237818-1-joshwash@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Joshua Washington <joshwash@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When using rtla timerlat with userspace threads (-u or -U), rtla
disables the OSNOISE_WORKLOAD option in
/sys/kernel/tracing/osnoise/options. This option is not re-enabled in a
subsequent run with kernel-space threads, leading to rtla collecting no
results if the previous run exited abnormally:
$ rtla timerlat top -u
^\Quit (core dumped)
$ rtla timerlat top -k -d 1s
Timer Latency
0 00:00:01 | IRQ Timer Latency (us) | Thread Timer Latency (us)
CPU COUNT | cur min avg max | cur min avg max
The issue persists until OSNOISE_WORKLOAD is set manually by running:
$ echo OSNOISE_WORKLOAD > /sys/kernel/tracing/osnoise/options
Set OSNOISE_WORKLOAD when running rtla with kernel-space threads if
available to fix the issue.
Cc: stable@vger.kernel.org Cc: John Kacur <jkacur@redhat.com> Cc: Luis Goncalves <lgoncalv@redhat.com> Link: https://lore.kernel.org/20250107144823.239782-4-tglozar@redhat.com Fixes: cdca4f4e5e8e ("rtla/timerlat_top: Add timerlat user-space support") Signed-off-by: Tomas Glozar <tglozar@redhat.com> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
[ params->kernel_workload does not exist in 6.6, use
!params->user_top ] Signed-off-by: Tomas Glozar <tglozar@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When using rtla timerlat with userspace threads (-u or -U), rtla
disables the OSNOISE_WORKLOAD option in
/sys/kernel/tracing/osnoise/options. This option is not re-enabled in a
subsequent run with kernel-space threads, leading to rtla collecting no
results if the previous run exited abnormally:
The signal context of certain RISC-V extensions will be appended after
struct __riscv_extra_ext_header, which already includes an empty context
header. Therefore, there is no need to preserve a separate hdr for the
END of signal context.
Fixes: 8ee0b41898fa ("riscv: signal: Add sigcontext save/restore for vector") Signed-off-by: Yong-Xuan Wang <yongxuan.wang@sifive.com> Reviewed-by: Zong Li <zong.li@sifive.com> Reviewed-by: Andy Chiu <AndybnAC@gmail.com> Reviewed-by: Alexandre Ghiti <alexghiti@rivosinc.com> Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20241220083926.19453-2-yongxuan.wang@sifive.com Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Make sure the compare value in the lr/sc loop is sign extended to match
what lr.w does. Fortunately, due to the compiler keeping the register
contents sign extended anyway the lack of the explicit extension didn't
result in wrong code so far, but this cannot be relied upon.
When working on OpenRISC support for restartable sequences I noticed
and fixed these two issues with the riscv support bits.
1 The 'inc' argument to RSEQ_ASM_OP_R_DEREF_ADDV was being implicitly
passed to the macro. Fix this by adding 'inc' to the list of macro
arguments.
2 The inline asm input constraints for 'inc' and 'off' use "er", The
riscv gcc port does not have an "e" constraint, this looks to be
copied from the x86 port. Fix this by just using an "r" constraint.
I have compile tested this only for riscv. However, the same fixes I
use in the OpenRISC rseq selftests and everything passes with no issues.
David reported a warning observed while loop testing kexec jump:
Interrupts enabled after irqrouter_resume+0x0/0x50
WARNING: CPU: 0 PID: 560 at drivers/base/syscore.c:103 syscore_resume+0x18a/0x220
kernel_kexec+0xf6/0x180
__do_sys_reboot+0x206/0x250
do_syscall_64+0x95/0x180
The corresponding interrupt flag trace:
hardirqs last enabled at (15573): [<ffffffffa8281b8e>] __up_console_sem+0x7e/0x90
hardirqs last disabled at (15580): [<ffffffffa8281b73>] __up_console_sem+0x63/0x90
That means __up_console_sem() was invoked with interrupts enabled. Further
instrumentation revealed that in the interrupt disabled section of kexec
jump one of the syscore_suspend() callbacks woke up a task, which set the
NEED_RESCHED flag. A later callback in the resume path invoked
cond_resched() which in turn led to the invocation of the scheduler:
This is a long standing problem, which probably got more visible with
the recent printk changes. Something does a task wakeup and the
scheduler sets the NEED_RESCHED flag. cond_resched() sees it set and
invokes schedule() from a completely bogus context. The scheduler
enables interrupts after context switching, which causes the above
warning at the end.
Quite some of the code paths in syscore_suspend()/resume() can result in
triggering a wakeup with the exactly same consequences. They might not
have done so yet, but as they share a lot of code with normal operations
it's just a question of time.
The problem only affects the PREEMPT_NONE and PREEMPT_VOLUNTARY scheduling
models. Full preemption is not affected as cond_resched() is disabled and
the preemption check preemptible() takes the interrupt disabled flag into
account.
Cure the problem by adding a corresponding check into cond_resched().
Reported-by: David Woodhouse <dwmw@amazon.co.uk> Suggested-by: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@kernel.org> Tested-by: David Woodhouse <dwmw@amazon.co.uk> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: stable@vger.kernel.org Closes: https://lore.kernel.org/all/7717fe2ac0ce5f0a2c43fdab8b11f4483d54a2a4.camel@infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The problem is the additional read in the slow path (after it
decremented to RCUREF_NOREF) which can happen after the counter has been
marked RCUREF_DEAD.
Prevent this by reusing the return value of the decrement. Now every "final"
put uses RCUREF_NOREF in the slow path and attempts the final cmpxchg() to
RCUREF_DEAD.
[ bigeasy: Add changelog ]
Fixes: ee1ee6db07795 ("atomics: Provide rcuref - scalable reference counting") Reported-by: kernel test robot <oliver.sang@intel.com> Debugged-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: stable@vger.kernel.org Closes: https://lore.kernel.org/oe-lkp/202412311453.9d7636a2-lkp@intel.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the kernel, there are architectures (x86, arm64) that perform
boot-time relocation (for KASLR) without relying on PIE codegen. In this
case, all const global objects are emitted into .rodata, including const
objects with fields that will be fixed up by the boot-time relocation
code. This implies that .rodata (and .text in some cases) need to be
writable at boot, but they will usually be mapped read-only as soon as
the boot completes.
When using PIE codegen, the compiler will emit const global objects into
.data.rel.ro rather than .rodata if the object contains fields that need
such fixups at boot-time. This permits the linker to annotate such
regions as requiring read-write access only at load time, but not at
execution time (in user space), while keeping .rodata truly const (in
user space, this is important for reducing the CoW footprint of dynamic
executables).
This distinction does not matter for the kernel, but it does imply that
const data will end up in writable memory if the .data.rel.ro sections
are not treated in a special way, as they will end up in the writable
.data segment by default.
Before this patch, if the checksum was not used, the subflow was only
reset if map_data_len was != 0. If there were no MPTCP options or an
invalid mapping, map_data_len was not set to the data len, and then the
subflow was not reset as it should have been, leaving the MPTCP
connection in a wrong fallback mode.
This map_data_len condition has been introduced to handle the reception
of the infinite mapping. Instead, a new dedicated mapping error could
have been returned and treated as a special case. However, the commit 31bf11de146c ("mptcp: introduce MAPPING_BAD_CSUM") has been introduced
by Paolo Abeni soon after, and backported later on to stable. It better
handle the csum case, and it means the exception for valid_csum_seen in
subflow_can_fallback(), plus this one for the infinite mapping in
subflow_check_data_avail(), are no longer needed.
In other words, the code can be simplified there: a fallback should only
be done if msk->allow_infinite_fallback is set. This boolean is set to
false once MPTCP-specific operations acting on the whole MPTCP
connection vs the initial path have been done, e.g. a second path has
been created, or an MPTCP re-injection -- yes, possible even with a
single subflow. The subflow_can_fallback() helper can then be dropped,
and replaced by this single condition.
This also makes the code clearer: a fallback should only be done if it
is possible to do so.
While at it, no need to set map_data_len to 0 in get_mapping_status()
for the infinite mapping case: it will be set to skb->len just after, at
the end of subflow_check_data_avail(), and not read in between.
Fixes: f8d4bcacff3b ("mptcp: infinite mapping receiving") Cc: stable@vger.kernel.org Reported-by: Chester A. Unal <chester.a.unal@xpedite-tech.com> Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/544 Acked-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Tested-by: Chester A. Unal <chester.a.unal@xpedite-tech.com> Link: https://patch.msgid.link/20250224-net-mptcp-misc-fixes-v1-2-f550f636b435@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Indeed the PM can try to send a RM_ADDR over a msk without acquiring
first the msk socket lock.
The bugged code-path comes from an early optimization: when there
are no subflows, the PM should (usually) not send RM_ADDR
notifications.
The above statement is incorrect, as without locks another process
could concurrent create a new subflow and cause the RM_ADDR generation.
Additionally the supposed optimization is not very effective even
performance-wise, as most mptcp sockets should have at least one
subflow: the MPC one.
Address the issue removing the buggy code path, the existing "slow-path"
will handle correctly even the edge case.
Fixes: b6c08380860b ("mptcp: remove addr and subflow in PM netlink") Cc: stable@vger.kernel.org Reported-by: syzbot+cd3ce3d03a3393ae9700@syzkaller.appspotmail.com Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/546 Signed-off-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Link: https://patch.msgid.link/20250224-net-mptcp-misc-fixes-v1-1-f550f636b435@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In exynos5_usbdrd_{pipe3,utmi}_set_refclk(), the masks
PHYCLKRST_MPLL_MULTIPLIER_MASK and PHYCLKRST_SSC_REFCLKSEL_MASK are not
inverted when applied to the register values. Fix it.
Observed VBUS_OVERRIDE & ID_OVERRIDE might be programmed
with unexpected value prior to XUSB PADCTL driver, this
could also occur in virtualization scenario.
For example, UEFI firmware programs ID_OVERRIDE=GROUNDED to set
a type-c port to host mode and keeps the value to kernel.
If the type-c port is connected a usb host, below errors can be
observed right after usb host mode driver gets probed. The errors
would keep until usb role class driver detects the type-c port
as device mode and notifies usb device mode driver to set both
ID_OVERRIDE and VBUS_OVERRIDE to correct value by XUSB PADCTL
driver.
[ 173.765814] usb usb3-port2: Cannot enable. Maybe the USB cable is bad?
[ 173.765837] usb usb3-port2: config error
Taking virtualization into account, asserting XUSB PADCTL
reset would break XUSB functions used by other guest OS,
hence only reset VBUS & ID OVERRIDE of the port in
utmi_phy_init.
There is an off-by-one issue for the err_chained_bd path, it will free
one more tx_swbd than expected. But there is no such issue for the
err_map_data path. To fix this off-by-one issue and make the two error
handling consistent, the increment of 'i' and 'count' remain in sync
and enetc_unwind_tx_frame() is called for error handling.
Fixes: fb8629e2cbfc ("net: enetc: add support for software TSO") Cc: stable@vger.kernel.org Suggested-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Wei Fang <wei.fang@nxp.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com> Link: https://patch.msgid.link/20250224111251.1061098-9-wei.fang@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is an issue with one-step timestamp based on UDP/IP. The peer will
discard the sync packet because of the wrong UDP checksum. For ENETC v1,
the software needs to update the UDP checksum when updating the
originTimestamp field, so that the hardware can correctly update the UDP
checksum when updating the correction field. Otherwise, the UDP checksum
in the sync packet will be wrong.
Fixes: 7294380c5211 ("enetc: support PTP Sync packet one-step timestamping") Cc: stable@vger.kernel.org Signed-off-by: Wei Fang <wei.fang@nxp.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://patch.msgid.link/20250224111251.1061098-6-wei.fang@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When creating a TSO header, if the skb is VLAN tagged, the extended BD
will be used and the 'count' should be increased by 2 instead of 1.
Otherwise, when an error occurs, less tx_swbd will be freed than the
actual number.
Fixes: fb8629e2cbfc ("net: enetc: add support for software TSO") Cc: stable@vger.kernel.org Suggested-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Wei Fang <wei.fang@nxp.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com> Link: https://patch.msgid.link/20250224111251.1061098-3-wei.fang@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Syzbot reports [1] a warning in usb_submit_urb() triggered by
inconsistencies between expected and actually present endpoints
in gl620a driver. Since genelink_bind() does not properly
verify whether specified eps are in fact provided by the device,
in this case, an artificially manufactured one, one may get a
mismatch.
Fix the issue by resorting to a usbnet utility function
usbnet_get_endpoints(), usually reserved for this very problem.
Check for endpoints and return early before proceeding further if
any are missing.
According to the chip manual, the I2C register access type of
Loongson-2K2000/LS7A is "B", so we can only access registers in byte
form (readb()/writeb()).
Although Loongson-2K0500/Loongson-2K1000 do not have similar
constraints, register accesses in byte form also behave correctly.
Also, in hardware, the frequency division registers are defined as two
separate registers (high 8-bit and low 8-bit), so we just access them
directly as bytes.
Fixes: 015e61f0bffd ("i2c: ls2x: Add driver for Loongson-2K/LS7A I2C controller") Co-developed-by: Hongliang Wang <wanghongliang@loongson.cn> Signed-off-by: Hongliang Wang <wanghongliang@loongson.cn> Signed-off-by: Binbin Zhou <zhoubinbin@loongson.cn> Cc: stable@vger.kernel.org # v6.3+ Reviewed-by: Andy Shevchenko <andy@kernel.org> Signed-off-by: Andi Shyti <andi.shyti@kernel.org> Link: https://lore.kernel.org/r/20250220125612.1910990-1-zhoubinbin@loongson.cn Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The customer reports that there is a soft lockup issue related to
the i2c driver. After checking, the i2c module was doing a tx transfer
and the bmc machine reboots in the middle of the i2c transaction, the i2c
module keeps the status without being reset.
Due to such an i2c module status, the i2c irq handler keeps getting
triggered since the i2c irq handler is registered in the kernel booting
process after the bmc machine is doing a warm rebooting.
The continuous triggering is stopped by the soft lockup watchdog timer.
Disable the interrupt enable bit in the i2c module before calling
devm_request_irq to fix this issue since the i2c relative status bit
is read-only.
[Why]
DC is not using amdgpu_irq_get/put to manage the HPD interrupt refcounts.
So when amdgpu_irq_gpu_reset_resume_helper() reprograms all of the IRQs,
HPD gets disabled.
[How]
Use amdgpu_irq_get/put() for HPD init/fini in DM in order to sync refcounts
Cc: Mario Limonciello <mario.limonciello@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Mario Limonciello <mario.limonciello@amd.com> Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com> Signed-off-by: Roman Li <Roman.Li@amd.com> Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit f3dde2ff7fcaacd77884502e8f572f2328e9c745) Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[Why]
PSR-SU may cause some glitching randomly on several panels.
[How]
Temporarily disable the PSR-SU and fallback to PSR1 for
all eDP panels.
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/3388 Cc: Mario Limonciello <mario.limonciello@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Sun peng Li <sunpeng.li@amd.com> Signed-off-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Roman Li <roman.li@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 6deeefb820d0efb0b36753622fb982d03b37b3ad) Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
$ perf record -e cpu_core/instructions/ppp -F 120
Error:
The sys_perf_event_open() syscall returned with 22 (Invalid argument)
for event (cpu_core/instructions/ppp).
"dmesg | grep -i perf" may provide additional information.
The limit_period() check avoids a low sampling period on a counter. It
doesn't intend to limit the frequency.
The check in the x86_pmu_hw_config() should be limited to non-freq mode.
The attr.sample_period and attr.sample_freq are union. The
attr.sample_period should not be used to indicate the frequency mode.
The perf_iterate_ctx() function performs RCU list traversal but
currently lacks RCU read lock protection. This causes lockdep warnings
when running perf probe with unshare(1) under CONFIG_PROVE_RCU_LIST=y:
WARNING: suspicious RCU usage
kernel/events/core.c:8168 RCU-list traversed in non-reader section!!
This protection was previously present but was removed in commit bd2756811766 ("perf: Rewrite core context handling"). Add back the
necessary rcu_read_lock()/rcu_read_unlock() pair around
perf_iterate_ctx() call in perf_event_exec().
[ mingo: Use scoped_guard() as suggested by Peter ]
This fixes a regression introduced a few weeks ago in stable kernels
6.12.14 and 6.13.3. The internal microphone on ASUS Vivobook N705UD /
X705UD laptops is broken: the microphone appears in userspace (e.g.
Gnome settings) but no sound is detected.
I bisected it to commit 3b4309546b48 ("ALSA: hda: Fix headset detection
failure due to unstable sort").
I figured out the cause:
1. The initial pins enabled for the ALC256 driver are:
cfg->inputs == {
{ pin=0x19, type=AUTO_PIN_MIC,
is_headset_mic=1, is_headphone_mic=0, has_boost_on_pin=1 },
{ pin=0x1a, type=AUTO_PIN_MIC,
is_headset_mic=0, is_headphone_mic=0, has_boost_on_pin=1 } }
2. Since 2017 and commits c1732ede5e8 ("ALSA: hda/realtek - Fix headset
and mic on several ASUS laptops with ALC256") and 28e8af8a163 ("ALSA:
hda/realtek: Fix mic and headset jack sense on ASUS X705UD"), the
quirk ALC256_FIXUP_ASUS_MIC is also applied to ASUS X705UD / N705UD
laptops.
This added another internal microphone on pin 0x13:
cfg->inputs == {
{ pin=0x13, type=AUTO_PIN_MIC,
is_headset_mic=0, is_headphone_mic=0, has_boost_on_pin=1 },
{ pin=0x19, type=AUTO_PIN_MIC,
is_headset_mic=1, is_headphone_mic=0, has_boost_on_pin=1 },
{ pin=0x1a, type=AUTO_PIN_MIC,
is_headset_mic=0, is_headphone_mic=0, has_boost_on_pin=1 } }
I don't know what this pin 0x13 corresponds to. To the best of my
knowledge, these laptops have only one internal microphone.
3. Before 2025 and commit 3b4309546b48 ("ALSA: hda: Fix headset
detection failure due to unstable sort"), the sort function would let
the microphone of pin 0x1a (the working one) *before* the microphone
of pin 0x13 (the phantom one).
4. After this commit 3b4309546b48, the fixed sort function puts the
working microphone (pin 0x1a) *after* the phantom one (pin 0x13). As
a result, no sound is detected anymore.
It looks like the quirk ALC256_FIXUP_ASUS_MIC is not needed anymore for
ASUS Vivobook X705UD / N705UD laptops. Without it, everything works
fine:
- the internal microphone is detected and records actual sound,
- plugging in a jack headset is detected and can record actual sound
with it,
- unplugging the jack headset makes the system go back to internal
microphone and can record actual sound.
Check whether denominator expression x * (x - 1) * 1000 mod {2^32, 2^64}
produce zero and skip stddev computation in that case.
For now don't care about rec->counter * rec->counter overflow because
rec->time * rec->time overflow will likely happen earlier.
Cc: stable@vger.kernel.org Cc: Wen Yang <wenyang@linux.alibaba.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Link: https://lore.kernel.org/20250206090156.1561783-1-kniv@yandex-team.ru Fixes: e31f7939c1c27 ("ftrace: Avoid potential division by zero in function profiler") Signed-off-by: Nikolay Kuratov <kniv@yandex-team.ru> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ret = create_actions(); (return -EINVAL)
if (ret)
goto out_unreg;
[..]
ret = hist_trigger_enable(data, ...) {
list_add_tail_rcu(&data->list, &file->triggers); <<<---- SKIPPED!!! (this is important!)
[..]
out_unreg:
event_hist_unregister(.., data) {
cmd_ops->unreg(.., data, ..) [hist_unregister_trigger()] {
list_for_each_entry(iter, &file->triggers, list) {
if (!hist_trigger_match(data, iter, named_data, false)) <- never matches
continue;
[..]
test = iter;
}
if (test && test->ops->free) <<<-- test is NULL
test->ops->free(test) [event_hist_trigger_free()] {
[..]
if (data->name)
del_named_trigger(data) {
list_del(&data->named_list); <<<<-- NEVER gets removed!
}
}
}
}
[..]
kfree(data); <<<-- frees item but it is still on list
The next time a hist with name is registered, it causes an u-a-f bug and
the kernel can crash.
Move the code around such that if event_trigger_register() succeeds, the
next thing called is hist_trigger_enable() which adds it to the list.
A bunch of actions is called if get_named_trigger_data() returns false.
But that doesn't need to be called after event_trigger_register(), so it
can be moved up, allowing event_trigger_register() to be called just
before hist_trigger_enable() keeping them together and allowing the
file->triggers to be properly populated.
Cc: stable@vger.kernel.org Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Link: https://lore.kernel.org/20250227163944.1c37f85f@gandalf.local.home Fixes: 067fe038e70f6 ("tracing: Add variable reference handling to hist triggers") Reported-by: Tomas Glozar <tglozar@redhat.com> Tested-by: Tomas Glozar <tglozar@redhat.com> Reviewed-by: Tom Zanussi <zanussi@kernel.org> Closes: https://lore.kernel.org/all/CAP4=nvTsxjckSBTz=Oe_UYh8keD9_sZC4i++4h72mJLic4_W4A@mail.gmail.com/ Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When an invalid function ID of an SBI extension is used we should
return not-supported, not invalid-param. Also, when we see that at
least one hartid constructed from the base and mask parameters is
invalid, then we should return invalid-param. Finally, rather than
relying on overflowing a left shift to result in zero and then using
that zero in a condition which [correctly] skips sending an IPI (but
loops unnecessarily), explicitly check for overflow and exit the loop
immediately.
Fixes: 5f862df5585c ("RISC-V: KVM: Add v0.1 replacement SBI extensions defined in v0.2") Signed-off-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Anup Patel <anup@brainfault.org> Link: https://lore.kernel.org/r/20250217084506.18763-10-ajones@ventanamicro.com Signed-off-by: Anup Patel <anup@brainfault.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
"Not stopped" means started or suspended so we need to check for
a single state in order to have a chance to check for each state.
Also, we need to use target_vcpu when checking for the suspend
state.
Documentation/virt/kvm/locking.rst advises that kvm->lock should be
acquired outside vcpu->mutex and kvm->srcu. However, when KVM/RISC-V
handling SBI_EXT_HSM_HART_START, the lock ordering is vcpu->mutex,
kvm->srcu then kvm->lock.
Although the lockdep checking no longer complains about this after commit f0f44752f5f6 ("rcu: Annotate SRCU's update-side lockdep dependencies"),
it's necessary to replace kvm->lock with a new dedicated lock to ensure
only one hart can execute the SBI_EXT_HSM_HART_START call for the target
hart simultaneously.
Additionally, this patch also rename "power_off" to "mp_state" with two
possible values. The vcpu->mp_state_lock also protects the access of
vcpu->mp_state.
I still have some Soekris net4826 in a Community Wireless Network I
volunteer with. These devices use an AMD SC1100 SoC. I am running
OpenWrt on them, which uses a patched kernel, that naturally has
evolved over time. I haven't updated the ones in the field in a
number of years (circa 2017), but have one in a test bed, where I have
intermittently tried out test builds.
A few years ago, I noticed some trouble, particularly when "warm
booting", that is, doing a reboot without removing power, and noticed
the device was hanging after the kernel message:
[ 0.081615] Working around Cyrix MediaGX virtual DMA bugs.
If I removed power and then restarted, it would boot fine, continuing
through the message above, thusly:
[ 0.081615] Working around Cyrix MediaGX virtual DMA bugs.
[ 0.090076] Enable Memory-Write-back mode on Cyrix/NSC processor.
[ 0.100000] Enable Memory access reorder on Cyrix/NSC processor.
[ 0.100070] Last level iTLB entries: 4KB 0, 2MB 0, 4MB 0
[ 0.110058] Last level dTLB entries: 4KB 0, 2MB 0, 4MB 0, 1GB 0
[ 0.120037] CPU: NSC Geode(TM) Integrated Processor by National Semi (family: 0x5, model: 0x9, stepping: 0x1)
[...]
In order to continue using modern tools, like ssh, to interact with
the software on these old devices, I need modern builds of the OpenWrt
firmware on the devices. I confirmed that the warm boot hang was still
an issue in modern OpenWrt builds (currently using a patched linux
v6.6.65).
Last night, I decided it was time to get to the bottom of the warm
boot hang, and began bisecting. From preserved builds, I narrowed down
the bisection window from late February to late May 2019. During this
period, the OpenWrt builds were using 4.14.x. I was able to build
using period-correct Ubuntu 18.04.6. After a number of bisection
iterations, I identified a kernel bump from 4.14.112 to 4.14.113 as
the commit that introduced the warm boot hang.
So, I tried reverting just that kernel change on top of the breaking
OpenWrt commit, and my warm boot hang went away.
Presumably, the warm boot hang is due to some register not getting
cleared in the same way that a loss of power does. That is
approximately as much as I understand about the problem.
More poking/prodding and coaching from Jonas Gorski, it looks
like this test patch fixes the problem on my board: Tested against
v6.6.67 and v4.14.113.
Fixes: 18fb053f9b82 ("x86/cpu/cyrix: Use correct macros for Cyrix calls on Geode processors") Debugged-by: Jonas Gorski <jonas.gorski@gmail.com> Signed-off-by: Russell Senior <russell@personaltelco.net> Signed-off-by: Ingo Molnar <mingo@kernel.org> Link: https://lore.kernel.org/r/CAHP3WfOgs3Ms4Z+L9i0-iBOE21sdMk5erAiJurPjnrL9LSsgRA@mail.gmail.com Cc: Matthew Whitehead <tedheadster@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
The cause is that zero pfn is set to the PTE without increasing the RSS
count in mfill_atomic_pte_zeropage() and the refcount of zero folio does
not increase accordingly. Then, the operation on the same pfn is performed
in uprobe_write_opcode()->__replace_page() to unconditional decrease the
RSS count and old_folio's refcount.
Therefore, two bugs are introduced:
1. The RSS count is incorrect, when process exit, the check_mm() report
error "Bad rss-count".
2. The reserved folio (zero folio) is freed when folio->refcount is zero,
then free_pages_prepare->free_page_is_bad() report error
"Bad page state".
There is more, the following warning could also theoretically be triggered:
Syskaller triggers a warning due to prev_epc->pmu != next_epc->pmu in
perf_event_swap_task_ctx_data(). vmcore shows that two lists have the same
perf_event_pmu_context, but not in the same order.
The problem is that the order of pmu_ctx_list for the parent is impacted by
the time when an event/PMU is added. While the order for a child is
impacted by the event order in the pinned_groups and flexible_groups. So
the order of pmu_ctx_list in the parent and child may be different.
To fix this problem, insert the perf_event_pmu_context to its proper place
after iteration of the pmu_ctx_list.
The follow testcase can trigger above warning:
# perf record -e cycles --call-graph lbr -- taskset -c 3 ./a.out &
# perf stat -e cpu-clock,cs -p xxx // xxx is the pid of a.out
pid = fork();
if (pid == -1) {
printf("fork error\n");
return;
}
if (pid == 0) {
while (1) {
count++;
}
} else {
while (1) {
count++;
}
}
}
The testcase first opens an LBR event, so it will allocate task_ctx_data,
and then open tracepoint and software events, so the parent context will
have 3 different perf_event_pmu_contexts. On inheritance, child ctx will
insert the perf_event_pmu_context in another order and the warning will
trigger.
IEP driver supports both perout and pps signal generation
but perout feature is faulty with half-cooked support
due to some missing configuration. Remove perout
support from the driver and reject perout requests with
"not supported" error code.
As all sources of concurrency in hardware register access occur in
non-interrupt context eliminate spinlock-based synchronization and
rely on the mutex-based synchronization that is already present.
Signed-off-by: Diogo Ivo <diogo.ivo@siemens.com> Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Stable-dep-of: 54e1b4becf5e ("net: ti: icss-iep: Reject perout generation request") Signed-off-by: Sasha Levin <sashal@kernel.org>
This patch mitigates the two-reallocations issue with rpl_iptunnel by
providing the dst_entry (in the cache) to the first call to
skb_cow_head(). As a result, the very first iteration would still
trigger two reallocations (i.e., empty cache), while next iterations
would only trigger a single reallocation.
Performance tests before/after applying this patch, which clearly shows
there is no impact (it even shows improvement):
- before: https://ibb.co/nQJhqwc
- after: https://ibb.co/4ZvW6wV
Signed-off-by: Justin Iurman <justin.iurman@uliege.be> Cc: Alexander Aring <aahringo@redhat.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: 13e55fbaec17 ("net: ipv6: fix dst ref loop on input in rpl lwt") Signed-off-by: Sasha Levin <sashal@kernel.org>
This patch mitigates the two-reallocations issue with seg6_iptunnel by
providing the dst_entry (in the cache) to the first call to
skb_cow_head(). As a result, the very first iteration would still
trigger two reallocations (i.e., empty cache), while next iterations
would only trigger a single reallocation.
Performance tests before/after applying this patch, which clearly shows
the improvement:
- before: https://ibb.co/3Cg4sNH
- after: https://ibb.co/8rQ350r
Signed-off-by: Justin Iurman <justin.iurman@uliege.be> Cc: David Lebrun <dlebrun@google.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: c64a0727f9b1 ("net: ipv6: fix dst ref loop on input in seg6 lwt") Signed-off-by: Sasha Levin <sashal@kernel.org>
Add static inline dst_dev_overhead() function to include/net/dst.h. This
helper function is used by ioam6_iptunnel, rpl_iptunnel and
seg6_iptunnel to get the dev's overhead based on a cache entry
(dst_entry). If the cache is empty, the default and generic value
skb->mac_len is returned. Otherwise, LL_RESERVED_SPACE() over dst's dev
is returned.
Signed-off-by: Justin Iurman <justin.iurman@uliege.be> Cc: Alexander Lobakin <aleksander.lobakin@intel.com> Cc: Vadim Fedorenko <vadim.fedorenko@linux.dev> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: c64a0727f9b1 ("net: ipv6: fix dst ref loop on input in seg6 lwt") Signed-off-by: Sasha Levin <sashal@kernel.org>
In certain cases, napi_get_frags() returns an skb that points to an old
received fragment, This skb may have its skb->ip_summed, csum, and other
fields set from previous fragment handling.
Some network drivers set skb->ip_summed to either CHECKSUM_COMPLETE or
CHECKSUM_UNNECESSARY when getting skb from napi_get_frags(), while
others only set skb->ip_summed when RX checksum offload is enabled on
the device, and do not set any value for skb->ip_summed when hardware
checksum offload is disabled, assuming that the skb->ip_summed
initiated to zero by napi_reuse_skb, ionic driver for example will
ignore/unset any value for the ip_summed filed if HW checksum offload is
disabled, and if we have a situation where the user disables the
checksum offload during a traffic that could lead to the following
errors shown in the kernel logs:
<IRQ>
dump_stack_lvl+0x34/0x48
__skb_gro_checksum_complete+0x7e/0x90
tcp6_gro_receive+0xc6/0x190
ipv6_gro_receive+0x1ec/0x430
dev_gro_receive+0x188/0x360
? ionic_rx_clean+0x25a/0x460 [ionic]
napi_gro_frags+0x13c/0x300
? __pfx_ionic_rx_service+0x10/0x10 [ionic]
ionic_rx_service+0x67/0x80 [ionic]
ionic_cq_service+0x58/0x90 [ionic]
ionic_txrx_napi+0x64/0x1b0 [ionic]
__napi_poll+0x27/0x170
net_rx_action+0x29c/0x370
handle_softirqs+0xce/0x270
__irq_exit_rcu+0xa3/0xc0
common_interrupt+0x80/0xa0
</IRQ>
This inconsistency sometimes leads to checksum validation issues in the
upper layers of the network stack.
To resolve this, this patch clears the skb->ip_summed value for each
reused skb in by napi_reuse_skb(), ensuring that the caller is responsible
for setting the correct checksum status. This eliminates potential
checksum validation issues caused by improper handling of
skb->ip_summed.
Fixes: 76620aafd66f ("gro: New frags interface to avoid copying shinfo") Signed-off-by: Mohammad Heib <mheib@redhat.com> Reviewed-by: Shannon Nelson <shannon.nelson@amd.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://patch.msgid.link/20250225112852.2507709-1-mheib@redhat.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Recently a bug was discovered where the server had entered TCP_ESTABLISHED
state, but the upper layers were not notified.
The same 5-tuple packet may be processed by different CPUSs, so two
CPUs may receive different ack packets at the same time when the
state is TCP_NEW_SYN_RECV.
In that case, req->ts_recent in tcp_check_req may be changed concurrently,
which will probably cause the newsk's ts_recent to be incorrectly large.
So that tcp_validate_incoming will fail. At this point, newsk will not be
able to enter the TCP_ESTABLISHED.
The cpu2's skb or a newly received skb will call tcp_v4_do_rcv to get
the newsk into the TCP_ESTABLISHED state, but at this point it is no
longer possible to notify the upper layer application. A notification
mechanism could be added here, but the fix is more complex, so the
current fix is used.
In tcp_check_req, req->ts_recent is used to assign a value to
tcp_sk(child)->rx_opt.ts_recent, so removing the change in req->ts_recent
and changing tcp_sk(child)->rx_opt.ts_recent directly after owning the
req fixes this bug.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Wang Hai <wanghai38@huawei.com> Reviewed-by: Jason Xing <kerneljasonxing@gmail.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
projects / thirdparty / kernel / stable.git / log
