jwoolley 01/10/01 11:46:40
Modified: . CHANGES
file_io/unix open.c
Log:
Fix a potential segfault on Unix when apr_file_close() is called and the
file_cleanup fails. Previously in that situation the cleanup would not
be deregistered and therefore it would get run again when the pool went
away. Similar behavior in the directory cleanup logic caused segfaults
at some point in the past.
Reviewed by: Ryan Bloom, Bill Rowe
Revision Changes Path
1.164 +5 -0 apr/CHANGES
Index: CHANGES
===================================================================
RCS file: /home/cvs/apr/CHANGES,v
retrieving revision 1.163
retrieving revision 1.164
diff -u -d -u -r1.163 -r1.164
--- CHANGES 2001/09/28 15:22:34 1.163
+++ CHANGES 2001/10/01 18:46:40 1.164
@@ -1,5 +1,10 @@
Changes with APR b1
+ *) Fix potential segfault when closing a file on Unix. If
+ apr_file_close() was called and it failed, it would not
+ deregister the file cleanup. Therefore the cleanup would
+ be run again later on a now-invalid descriptor. [Cliff Woolley]
+
*) Introduce apr_pool_lock for debugging, in combination with
ALLOC_USE_MALLOC + DEBUG_WITH_MPROTECT. Only implemented
on Win32 today, very effective for debugging pool constness.
1.85 +2 -7 apr/file_io/unix/open.c
Index: open.c
===================================================================
RCS file: /home/cvs/apr/file_io/unix/open.c,v
retrieving revision 1.84
retrieving revision 1.85
diff -u -d -u -r1.84 -r1.85
--- open.c 2001/08/10 21:04:47 1.84
+++ open.c 2001/10/01 18:46:40 1.85
@@ -174,13 +174,8 @@
APR_DECLARE(apr_status_t) apr_file_close(apr_file_t *file)
{
- apr_status_t rv;
-
- if ((rv = apr_unix_file_cleanup(file)) == APR_SUCCESS) {
- apr_pool_cleanup_kill(file->cntxt, file, apr_unix_file_cleanup);
- return APR_SUCCESS;
- }
- return rv;
+ apr_pool_cleanup_kill(file->cntxt, file, apr_unix_file_cleanup);
+ return apr_unix_file_cleanup(file);
}
APR_DECLARE(apr_status_t) apr_file_remove(const char *path, apr_pool_t *cont)
