vos: Avoid LWP stack overflow error on SIGINT

vsproc's method of longjmp'ing out of a signal handler can thoroughly
confuse LWP. At the very least, the information on our stack becomes
incorrect, since we're no longer using the stack supplied by LWP. At the
very least, we seem to (almost) always fail the sanity check checking
for stack overflow. To prevent the test from happening, NULL the stack
pointer before the longjmp so we can attempt recovery.

FIXES 33360

Change-Id: Ie910cb1cc8564ac119c3b751fd8295bf312a3ce7
Reviewed-on: http://gerrit.openafs.org/1559
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
This commit is contained in:
Andrew Deason 2010-03-10 14:15:58 -06:00 committed by Derrick Brashear
parent bf2d0db680
commit 9af8d46e77

View File

@ -1146,8 +1146,16 @@ int interrupt = 0;
static void *
do_interrupt(void * unused)
{
if (interrupt)
if (interrupt) {
#ifndef AFS_PTHREAD_ENV
/* Avoid LWP from getting confused that our stack has suddenly
* changed. This will avoid some sanity checks, but until a better way
* is found, the only alternative is always crashing and burning on at
* least the stack-overflow check. */
lwp_cpptr->stack = NULL;
#endif
longjmp(env, 0);
}
fprintf(STDOUT, "\nSIGINT handler: vos move operation in progress\n");
fprintf(STDOUT,