[PATCH] Tests for minimal signal handler functionality in MINSIGSTKSZ space.

[Florian Weimer]

* Zack Weinberg:

> It also occurs to me that on some architectures MINSIGSTKSZ is less
> than a page; on those architectures, the rounding done in xsigstack.c
> means we aren't _really_ testing this stuff in MINSIGSTKSZ space.
> Since overflow is much more of a concern than underflow, what do
> people think of adjusting the code in xsigstack.c so that the area
> actually passed to sigaltstack will not be rounded and will be right
> up against the guard in the direction of overflow?  This would mean
> xsigstack.c has to know which direction the stack grows, but I think
> we already have internal macros for that,

_STACK_GROWS_DOWN, _STACK_GROWS_UP, and NEED_SEPARATE_REGISTER_STACK on
ia64.  But ia64 uses stack sizes which are a multiple of the page size,
so you are lucky. 8-)

Thanks,
Florian