Re: [RFC PATCH] x86/mm/fault: Allow stack access below %rsp

[Waiman Long]

On 11/02/2018 04:11 PM, Andy Lutomirski wrote:
> On Fri, Nov 2, 2018 at 12:50 PM Waiman Long <longman@xxxxxxxxxx> wrote:
>> On 11/02/2018 03:44 PM, Dave Hansen wrote:
>>> On 11/2/18 12:40 PM, Waiman Long wrote:
>>>> The 64k+ limit check is kind of arbitrary. So the check is now removed
>>>> to just let expand_stack() decide if a segmentation fault should happen.
>>> With the 64k check removed, what's the next limit that we bump into?  Is
>>> it just the stack_guard_gap space above the next-lowest VMA?
>> I think it is both the stack_guard_gap space above the next lowest VMA
>> and the rlimit(RLIMIT_STACK).
>>
> Did the non-working programs ever work?  Because, if not, I say let them fail.

The program was working before on older kernel. After the backport of
the commit 1be7107fbe18 ("mm: larger stack guard gap, between vmas"),
the program stopped working. I believe it is the removal of the
check_stack_guard_page() function which did expand the stack. So the 64k
check was actually not functional because of the stack expansion before
hand, but now it does.

Anyway, I see the current 64k check a very crude check on an userspace
error. I think there are quite a number of more powerful userspace tools
out there that can do this kind of check more effectively. Kernel isn't
the right place to do that.

Cheers,
Longman