https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69278
Bug ID: 69278
Summary: Confusion option handling for -sanitize-recovery=alll
Product: gcc
Version: 6.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: sanitizer
Assignee: unassigned at gcc dot gnu.org
Reporter: marxin at gcc dot gnu.org
CC: dodji at gcc dot gnu.org, dvyukov at gcc dot gnu.org,
jakub at gcc dot gnu.org, kcc at gcc dot gnu.org
Target Milestone: ---
As I've just read opts.c, passing -fsanitize-recovery=all enables recovery mode
for all options except SANITIZE_USER_ADDRESS, SANITIZE_THREAD, SANITIZE_LEAK.
/* Handle both -fsanitize and -fno-sanitize cases. */
if (value && spec[i].flag == ~0U)
{
if (code == OPT_fsanitize_)
error_at (loc, "-fsanitize=all option is not valid");
else
*flag |= ~(SANITIZE_USER_ADDRESS | SANITIZE_THREAD | SANITIZE_LEAK);
AFAIK that's because of one cannot enable -fsanitize=leak with one of
-fsanitize=address or -fsanitize=thread.
However, following command line contains just one and I would expect to enable
recovery mode for -fsanitize=address:
$ g++ -fsanitize=address missing-store.C -g -fsanitize-recover=all
$ ASAN_OPTIONS="halt_on_error=0" ./a.out
...
==20905==ABORTING
OTOH:
$ g++ -fsanitize=address missing-store.C -g -fsanitize-recover=address
$ ASAN_OPTIONS="halt_on_error=0" ./a.out && echo $?
0
Maybe we should enable recovery for all and mask it at the end of opts parsing
with flag_sanitize?
If it's a good approach, I can come up with a patch.
Thanks,
Martin
