This is the mail archive of the
glibc-bugs@sourceware.org
mailing list for the glibc project.
[Bug ports/14238] New: tilegx: robust mutexes hang
- From: "cmetcalf at tilera dot com" <sourceware-bugzilla at sourceware dot org>
- To: glibc-bugs at sources dot redhat dot com
- Date: Thu, 14 Jun 2012 17:35:56 +0000
- Subject: [Bug ports/14238] New: tilegx: robust mutexes hang
- Auto-submitted: auto-generated
http://sourceware.org/bugzilla/show_bug.cgi?id=14238
Bug #: 14238
Summary: tilegx: robust mutexes hang
Product: glibc
Version: 2.15
Status: NEW
Severity: normal
Priority: P2
Component: ports
AssignedTo: unassigned@sourceware.org
ReportedBy: cmetcalf@tilera.com
CC: carlos@systemhalted.org, roland@gnu.org
Classification: Unclassified
The glibc testsuite reports five unit tests failing:
nptl/tst-robust1.out
nptl/tst-robust3.out
nptl/tst-robust4.out
nptl/tst-robust6.out
nptl/tst-robust9.out
I looked at the first one, and it seems like it is hanging in the LOCK(&m1)
call in tst-robust1.c, where we are checking that the lock will exit and return
with errno = EOWNERDEAD if the owner dies with the lock held. In fact it stays
stuck in the lock code, here:
#0 __lll_robust_lock_wait (futex=0x23a08, private=128)
at ../nptl/sysdeps/unix/sysv/linux/lowlevelrobustlock.c:52
#1 0x000000aaaaabb580 in __lll_robust_lock (mutex=0x23a08)
at ../sysdeps/unix/sysv/linux/tile/nptl/lowlevellock.h:180
#2 __pthread_mutex_lock_full (mutex=0x23a08) at pthread_mutex_lock.c:220
[...]
Note that these tests all pass on tilepro (and x86_64, etc.), and note that the
"robustpi" tests all pass on both tile platforms.
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.