Bug 29073 - mod_ssl segfault in bio_flush while in ssl_io_filter_cleanup
Summary: mod_ssl segfault in bio_flush while in ssl_io_filter_cleanup
Status: CLOSED DUPLICATE of bug 27945
Alias: None
Product: Apache httpd-2
Classification: Unclassified
Component: mod_ssl (show other bugs)
Version: 2.0.49
Hardware: Sun Solaris
: P3 critical (vote)
Target Milestone: ---
Assignee: Apache HTTPD Bugs Mailing List
Depends on:
Reported: 2004-05-18 21:22 UTC by mark
Modified: 2004-11-16 19:05 UTC (History)
0 users


Note You need to log in before you can comment on or make changes to this bug.
Description mark 2004-05-18 21:22:56 UTC
openssl version 0.9.7d
solaris 2.7 
worker mpm
MAY be related to PR #28577

httpd segfaults after random number of accesses.
here is the gdb stack trace capture while running:
gdb bin/httpd
gdb> run -X -f conf/httpd.conf -DSSL

. (launches LWPs)

Program received signal SIGSEGV, Segmentation fault.
allocator_free (allocator=0x2db658, node=0x0) at apr_pools.c:324
324     apr_pools.c: No such file or directory.
(gdb) bt
#0  allocator_free (allocator=0x2db658, node=0x0) at apr_pools.c:324
#1  0xff167ca0 in apr_pool_clear (pool=0x32f868) at apr_pools.c:713
#2  0x80440 in core_output_filter (f=0x31a270, b=0x32f8a0) at core.c:4163
#3  0x784a8 in ap_pass_brigade (next=0x31a270, bb=0x32f8a0) at util_filter.c:511
#4  0x41954 in bio_filter_out_flush (bio=0x1) at ssl_engine_io.c:154
#5  0x41a88 in bio_filter_out_write (bio=0xffffffff, in=0x79e438 "\025\003", inl=23)
    at ssl_engine_io.c:220
#6  0xff29447c in BIO_write () from /usr/local/ssl/lib/libcrypto.so.0.9.7
#7  0xff357634 in ssl3_write_pending () from /usr/local/ssl/lib/libssl.so.0.9.7
#8  0xff358264 in ssl3_dispatch_alert () from /usr/local/ssl/lib/libssl.so.0.9.7
#9  0xff355b3c in ssl3_shutdown () from /usr/local/ssl/lib/libssl.so.0.9.7
#10 0xff35e1f4 in SSL_shutdown () from /usr/local/ssl/lib/libssl.so.0.9.7
#11 0x4ebb0 in SSL_smart_shutdown (ssl=0x74eb08) at ssl_util_ssl.c:188
#12 0x42848 in ssl_filter_io_shutdown (filter_ctx=0x2f8b88, c=0x2f87f8, abortive=0)
        at ssl_engine_io.c:955
#13 0x4293c in ssl_io_filter_cleanup (data=0x2f8b88) at ssl_engine_io.c:996
#14 0xff168538 in run_cleanups (cref=0x2f8700) at apr_pools.c:1951
#15 0xff167c5c in apr_pool_clear (pool=0x2f86f0) at apr_pools.c:693
#16 0x67e5c in worker_thread (thd=0x125000, dummy=0x2f86f0) at worker.c:847
#17 0xff1635bc in dummy_worker (opaque=0x125000) at thread.c:88
Comment 1 mark 2004-05-25 17:16:48 UTC
FYI: this does not happen with prefork mpm, only with worker.
Comment 2 Joe Orton 2004-05-25 19:15:18 UTC
Try this patch:


*** This bug has been marked as a duplicate of 27945 ***