* Re: [PATCH v2 0/3][request for stable 3.10 inclusion] vm: make stack guard page errors return VM_FAULT_SIGSEGV rather than SIGBUS
[not found] <1425035663-9349-1-git-send-email-shengyong1@huawei.com>
@ 2015-04-26 10:49 ` Greg KH
0 siblings, 0 replies; only message in thread
From: Greg KH @ 2015-04-26 10:49 UTC (permalink / raw)
To: Sheng Yong; +Cc: stable
On Fri, Feb 27, 2015 at 11:14:20AM +0000, Sheng Yong wrote:
> Hi, Greg,
>
> V2:
> Backport commit 7fb08eca ("x86: mm: move mmap_sem unlock from
> mm_fault_error() to caller)" to solve lock problem on x86.
>
> V1:
> The commit 88b5d12 ("mm: propagate error from stack expansion even for
> guard page") for 3.10-stable makes the testcase get "bus error", while
> it had "segmentation fault" in older kernels.
>
> ===== testcase =====
> int main(void)
> {
> main();
> return 0;
> }
> ====================
>
> These 2 patches fix it by adding a VM_FAULT_SIGSEGV in all architecture
> fault handlers.
>
> [PATCH 1/3] is asjusted to ignore un-supported arch and driver in 3.10.
> And some files are separated or renamed in newer kernel, so [PATCH 1/3]
> is also adjusted to add the VM_FAULT_SIGSEGV handling in the original
> files in 3.10.
>
> thanks,
> Sheng
I think I've now queued these up properly, but can you test to verify I
got it correct?
thanks,
greg k-h
^ permalink raw reply [flat|nested] only message in thread