Andrew Donnellan <a...@linux.ibm.com> writes:
> On 29/5/20 4:14 pm, Daniel Axtens wrote:
>> syzkaller is picking up a bunch of crashes that look like this:
>>
>> Unrecoverable exception 380 at c00000000037ed60 (msr=8000000000001031)
>> Oops: Unrecoverable exception, sig: 6 [#1]
>> LE PAGE_SIZE=64K MMU=Hash SMP NR_CPUS=2048 NUMA pSeries
>> Modules linked in:
>> CPU: 0 PID: 874 Comm: syz-executor.0 Not tainted
>> 5.7.0-rc7-syzkaller-00016-gb0c3ba31be3e #0
>> NIP: c00000000037ed60 LR: c00000000004bac8 CTR: c000000000030990
>> REGS: c0000000555a7230 TRAP: 0380 Not tainted
>> (5.7.0-rc7-syzkaller-00016-gb0c3ba31be3e)
>> MSR: 8000000000001031 <SF,ME,IR,DR,LE> CR: 48222882 XER: 20000000
>> CFAR: c00000000004bac4 IRQMASK: 0
>> GPR00: c00000000004bb68 c0000000555a74c0 c0000000024b3500 0000000000000005
>> GPR04: 0000000000000000 0000000000000000 c00000000004bb88 c008000000910000
>> GPR08: 00000000000b0000 c00000000004bac8 0000000000016000 c000000002503500
>> GPR12: c000000000030990 c000000003190000 00000000106a5898 00000000106a0000
>> GPR16: 00000000106a5890 c000000007a92000 c000000008180e00 c000000007a8f700
>> GPR20: c000000007a904b0 0000000010110000 c00000000259d318 5deadbeef0000100
>> GPR24: 5deadbeef0000122 c000000078422700 c000000009ee88b8 c000000078422778
>> GPR28: 0000000000000001 800000000280b033 0000000000000000 c0000000555a75a0
>> NIP [c00000000037ed60] __sanitizer_cov_trace_pc+0x40/0x50
>> LR [c00000000004bac8] interrupt_exit_kernel_prepare+0x118/0x310
>> Call Trace:
>> [c0000000555a74c0] [c00000000004bb68]
>> interrupt_exit_kernel_prepare+0x1b8/0x310 (unreliable)
>> [c0000000555a7530] [c00000000000f9a8] interrupt_return+0x118/0x1c0
>> --- interrupt: 900 at __sanitizer_cov_trace_pc+0x0/0x50
>> ...<random previous call chain>...
>>
>> That looks like the KCOV helper accessing memory that's not safe to
>> access in the interrupt handling context.
>>
>> Do not instrument the new syscall entry/exit code with KCOV, GCOV or
>> UBSAN.
>>
>> Cc: Nicholas Piggin <npig...@gmail.com>
>> Fixes: 68b34588e202 ("powerpc/64/sycall: Implement syscall entry/exit logic
>> in C")
>> Signed-off-by: Daniel Axtens <d...@axtens.net>
>
> This seems reasonable - I've verified that this does indeed suppress the
> kcov trace calls.
Thanks.
> Acked-by: Andrew Donnellan <a...@linux.ibm.com>
>
> (does this need to be tagged for stable? the Fixes: commit is in 5.6 but
> we're at 5.7-rc7 at this point...)
No. The Fixed commit is based on v5.6-rc2, but it didn't go in until v5.7-rc1:
$ git describe --contains --match 'v*' 68b34588e202
v5.7-rc1~35^2~46
I plan to send it to Linus before the v5.7 release.
cheers
