Apply Cysignals fix for Cygwin exception handler on sigaltstack

[embray]

This adds a version of the patch from sagemath/cysignals#108 modified to apply to cysignals 1.8.1.

Since the issue only affects Cygwin I thought for now it would be easier (as in #27267) to apply the patch locally, rather than wait on Jeroen to approve the upstream PR, make a new release of cysignals, and make Sage depend on that release.

This fix certainly resolves the blocker issue well-enough for me for the time being. Fixing this is a prerequisite for fixing #27214.

CC: @jdemeyer

Component: porting: Cygwin

Author: Erik Bray

Branch/Commit: u/embray/cygwin/patch-cysignals-altstack-exception-handler @ f95dcff

Issue created by migration from https://trac.sagemath.org/ticket/27384

[embray]

New commits:

f95dcff

Trac #27384: Patch cysignals with fix for Cygwin's sigaltstack exception

[embray]

Branch: u/embray/cygwin/patch-cysignals-altstack-exception-handler

[embray]

Commit: f95dcff

[embray]

Author: Erik Bray

[jdemeyer]

comment:2

Don't you need sagemath/cysignals@8913e0b also?

It's not a big deal to make a new cysignals release, just not right now.

[embray]

comment:3

Replying to @jdemeyer:

Don't you need sagemath/cysignals@8913e0b also?

I thought so at first too, but the change that that was fixing was after 1.8.1 so it's not relevant right now (I will need it for any future cysignals update in Sage)

It's not a big deal to make a new cysignals release, just not right now.

I know it's not a big deal. But as I explained (and I think you're agreeing) it's easier to just include the patch in the Sage SPKG for now.

[jdemeyer]

comment:4

Can this be closed in favour of #27070?

[embray]

comment:5

Sure, as long as Cysignals 1.10 includes the fix (and doesn't break anything else). I will review the changelog to make sure there isn't anything else suspect.