Stefan Sperling wrote on Fri, Jun 24, 2011 at 19:27:44 +0200: > I don't know if this is already known, so I am reporting it > here to get confirmation. > > (From #svn-dev): > <stsp> run svn log --diff -r1139384 for fun > <stsp> with serf > <stsp> Ctrl-C out in the middle when it prints diff text > <stsp> -> core dump > <stsp> #0 0x0000000204cef469 in destroy_request (request=0x2083f47b8) > <stsp> at ./outgoing.c:309 > <stsp> 309 serf_debug__closed_conn(request->req_bkt->allocator); > <stsp> (gdb) p *request->req_bkt > <stsp> Cannot access memory at address 0x20d41f1b8 > <stsp> is this known?
On Debian, this raises a valgrind error virtually instantaneously (without having to wait for any output): [[[ % $prefix/apache2/build/libtool --mode=execute \ valgrind $svn log --diff -r1139384 ^/ ... ==15085== Conditional jump or move depends on uninitialised value(s) ==15085== at 0x4629A2E: inflateReset2 (in /usr/lib/libz.so.1.2.3.4) ==15085== by 0x4629B0C: inflateInit2_ (in /usr/lib/libz.so.1.2.3.4) ==15085== by 0x428E453: serf_deflate_read (deflate_buckets.c:201) ==15085== by 0x428D177: serf_response_read (response_buckets.c:387) ==15085== by 0x427B39A: svn_ra_serf__handle_xml_parser (util.c:1536) ==15085== by 0x427BC1E: handle_response (util.c:1804) ==15085== by 0x427BD0F: handle_response_cb (util.c:1835) ==15085== by 0x428A387: serf__process_connection (outgoing.c:829) ==15085== by 0x42889B5: serf_event_trigger (context.c:206) ==15085== by 0x4288B68: serf_context_run (context.c:269) ==15085== by 0x42796C5: svn_ra_serf__context_run_wait (util.c:668) ==15085== by 0x426C280: svn_ra_serf__wait_for_props (property.c:622) ]]]