summaryrefslogtreecommitdiffstats
path: root/share/man/man3
diff options
context:
space:
mode:
authorpkelsey <pkelsey@FreeBSD.org>2015-05-28 16:18:41 +0000
committerpkelsey <pkelsey@FreeBSD.org>2015-05-28 16:18:41 +0000
commit8a477638bd5fc6f57f630e19dd2a8f1458f45526 (patch)
treeacd43d69dbcc6b4fd0bb189c97ea071df36a8615 /share/man/man3
parenta82f33e60c7dfa71ecd32bfbaf1a5f930fcecb96 (diff)
downloadFreeBSD-src-8a477638bd5fc6f57f630e19dd2a8f1458f45526.zip
FreeBSD-src-8a477638bd5fc6f57f630e19dd2a8f1458f45526.tar.gz
Provide an unambiguous description of the potential hazard in calling
pthread_setspecific(3) from a key destructor. Reviewed by: jhb Approved by: jmallett (mentor) MFC after: 3 days
Diffstat (limited to 'share/man/man3')
-rw-r--r--share/man/man3/pthread_setspecific.35
1 files changed, 4 insertions, 1 deletions
diff --git a/share/man/man3/pthread_setspecific.3 b/share/man/man3/pthread_setspecific.3
index 1ea3002..3153c99 100644
--- a/share/man/man3/pthread_setspecific.3
+++ b/share/man/man3/pthread_setspecific.3
@@ -67,7 +67,10 @@ is undefined.
The
.Fn pthread_setspecific
function may be called from a thread-specific data destructor function,
-however this may result in lost storage or infinite loops.
+however this may result in lost storage or infinite loops if doing so
+causes non-NULL key values to remain after
+.Bq PTHREAD_DESTRUCTOR_ITERATIONS
+iterations of destructor calls have been made.
.Sh RETURN VALUES
If successful, the
.Fn pthread_setspecific
OpenPOWER on IntegriCloud