d711a485a0
Valgrind did not handle the following sysarch calls, and thus crashed
immediately
o get_fsbase
o set_fsbase
o get_gsbase
o set_fsbase
- Valgrind no longer dumps core when its client application does. [2]
Valgrind-snapshot does contain for safing the client's core dump,
but it does not work (it resulted in a crash) and has therefore been
disabled.
- Fix crash with valgrind-snapshot when the application to debug
does not exists or cannot be accessed.
- Fix symlink farms for threading libraries. FreeBSD 4-7 are
handled. [2]
- Bump PORTREVISION
Reported by: Ulrich Spoerlein <q@galgenberg.net> [1],
Chris Gilbert <Chris@lainos.org> [2]
PR: ports/86007 [1]
26 lines
923 B
C
26 lines
923 B
C
--- coregrind/vg_signals.c.orig Sun May 9 13:00:57 2004
|
|
+++ coregrind/vg_signals.c Mon Sep 19 18:18:07 2005
|
|
@@ -1374,6 +1374,12 @@
|
|
#define NT_PRXFPREG 0x46e62b7f /* copied from gdb5.1/include/elf/common.h */
|
|
#endif /* NT_PRXFPREG */
|
|
|
|
+// FIXME: Needed for FreeBSD 4, same ugly hack as in coregrind/ume.c XXX
|
|
+#ifndef ELFMAG
|
|
+#define ELFMAG "\177ELF"
|
|
+#define SELFMAG 4
|
|
+#endif
|
|
+
|
|
/* If true, then this Segment may be mentioned in the core */
|
|
static Bool may_dump(const Segment *seg)
|
|
{
|
|
@@ -1900,8 +1906,8 @@
|
|
if (core) {
|
|
static struct vki_rlimit zero = { 0, 0 };
|
|
|
|
- make_coredump(tid, info, corelim.rlim_cur);
|
|
-
|
|
+ // make_coredump(tid, info, corelim.rlim_cur);
|
|
+ VG_(message) (Vg_UserMsg, "Core dumping is broken on FreeBSD. Please re-run valgrind after fixing the crash.");
|
|
/* make sure we don't get a confusing kernel-generated coredump */
|
|
VG_(setrlimit)(VKI_RLIMIT_CORE, &zero);
|
|
}
|