This is the mail archive of the
mailing list for the Cygwin project.
Re: TEST RELEASE: Cygwin 2.1.0-0.4
- From: Ken Brown <kbrown at cornell dot edu>
- To: cygwin at cygwin dot com
- Date: Wed, 15 Jul 2015 10:24:42 -0400
- Subject: Re: TEST RELEASE: Cygwin 2.1.0-0.4
- Authentication-results: sourceware.org; auth=none
- References: <20150705213417 dot GH2918 at calimero dot vinschen dot de> <70AD8D85-7D30-4894-857B-4FD12E55BBBE at etr-usa dot com> <55A58727 dot 6060807 at redhat dot com> <55A5C073 dot 1020401 at cornell dot edu> <20150715075117 dot GA19844 at calimero dot vinschen dot de> <20150715124404 dot GP16720 at calimero dot vinschen dot de>
On 7/15/2015 8:44 AM, Corinna Vinschen wrote:
On Jul 15 09:51, Corinna Vinschen wrote:
On Jul 14 22:07, Ken Brown wrote:
Doing SIGSEGV pass 1.
Stack overflow 1 caught.
Doing SIGSEGV pass 2.
Stack overflow 2 caught.
Any idea why these two tests have been skipped? That means the
HAVE_STACK_OVERFLOW_RECOVERY autoconf test failed. You removed cygwin
from the explicit
mingw* | cygwin*) sv_cv_have_stack_overflow_recovery=yes ;;
which is the right thing to do, but that means CFG_LEAVE has been
set to leave-none.c, apparently.
I haven't much time to look into that right now, but will later today if
you don't beat me to it.
Got it. What's needed is a Cygwin-specific fault-*.h file which exposes
how to fetch the stack pointer register from mcontext_t. As you can see
from the plethora of fault-*.h files in the src subdir, this is highly
Here's the set of files you need to rebuild libsigsegv for Cygwin 2.1.0,
with all tests running and passing on i686 and x86_64. No other patch
So, who of you is going to maintain it?
I'll defer to Eric, but if he doesn't want it then I'll take it.
Problem reports: http://cygwin.com/problems.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple