Daniel Axtens writes:
> Hi Michael,
>
> I tested v1 of this. I ran the test from the bug with a range of stack
> sizes, in a loop, for several hours and didn't see any crashes/signal
> delivery failures.
>
> I retested v2 for a few minutes just to be sure, and I ran stress-ng's
> stack, stackmmap
Hi Michael,
I tested v1 of this. I ran the test from the bug with a range of stack
sizes, in a loop, for several hours and didn't see any crashes/signal
delivery failures.
I retested v2 for a few minutes just to be sure, and I ran stress-ng's
stack, stackmmap and bad-altstack stressors to make su
We have powerpc specific logic in our page fault handling to decide if
an access to an unmapped address below the stack pointer should expand
the stack VMA.
The logic aims to prevent userspace from doing bad accesses below the
stack pointer. However as long as the stack is < 1MB in size, we allow
