在 2024/10/17 17:39, Jonathan Cameron 写道:
On Mon, 14 Oct 2024 16:42:38 +0800
Shuai Xue <xueshuai@xxxxxxxxxxxxxxxxx> wrote:
Synchronous error was detected as a result of user-space process accessing
a 2-bit uncorrected error. The CPU will take a synchronous error exception
such as Synchronous External Abort (SEA) on Arm64. The kernel will queue a
memory_failure() work which poisons the related page, unmaps the page, and
then sends a SIGBUS to the process, so that a system wide panic can be
avoided.
However, no memory_failure() work will be queued when abnormal synchronous
errors occur. These errors can include situations such as invalid PA,
unexpected severity, no memory failure config support, invalid GUID
section, etc. In such case, the user-space process will trigger SEA again.
This loop can potentially exceed the platform firmware threshold or even
trigger a kernel hard lockup, leading to a system reboot.
Fix it by performing a force kill if no memory_failure() work is queued
for synchronous errors.
Signed-off-by: Shuai Xue <xueshuai@xxxxxxxxxxxxxxxxx>
Reviewed-by: Jarkko Sakkinen <jarkko@xxxxxxxxxx>
The subtle cases in here are the various other forms of delayed handling
buried in some of the record handling that don't set queued.
I've been through them all and have convinced myself that either
hey should never be synchronous or that there is no attempt to
recover in kernel today (non memory things such as CXL protocol
collapse, which might I guess be detected synchronously on a read
- though I'd expect poison and a memory error first) so the correct
thing to do is what you have here.
Fiddly code though with a lot of paths, so more eyes welcome!
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@xxxxxxxxxx>
+CC linux-cxl for info.
Thanks :)
Best Regards,
Shuai