[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: [PATCH] Fix segfault when gnome keyring lookup fails

From: Philip Martin <philip_at_codematters.co.uk>
Date: Thu, 12 Jul 2018 19:52:48 +0100

Jan Palus <jpalus_at_fastmail.com> writes:

> On 12.07.2018 15:08, Daniel Shahaf wrote:
>> Thanks for the patch; however, I don't think that's the correct fix.
> To be honest I wasn't sure of that either but on the other hand I thought that
> having initialized state won't hurt.
>> Does the following patch fix the problem? Speaking of which, could
>> you explain
>> how to reproduce the problem?
> As expected the patch does fix the issue, however the question about reproducer
> is somewhat problematic.

A memory tool is one way to detect such problems. With valgrind I get:

==14195== Conditional jump or move depends on uninitialised value(s)
==14195== at 0x5A1C043: svn_auth__simple_creds_cache_get (simple_providers.c:246)
==14195== by 0xB938E79: simple_gnome_keyring_first_creds (gnome_keyring.c:437)
==14195== by 0x5A05BA6: svn_auth_first_credentials (auth.c:290)
==14195== by 0x6F40D56: svn_ra_serf__credentials_callback (util.c:1202)
==14195== by 0x9285093: serf__handle_basic_auth (auth_basic.c:97)
==14195== by 0x9284F34: handle_auth_headers (auth.c:200)
==14195== by 0x9284F34: dispatch_auth (auth.c:312)
==14195== by 0x9284F34: serf__handle_auth_response (auth.c:369)
==14195== by 0x927BE3F: handle_response (outgoing.c:929)
==14195== by 0x927BE3F: read_from_connection (outgoing.c:1136)
==14195== by 0x927BE3F: serf__process_connection (outgoing.c:1257)
==14195== by 0x927A73D: serf_event_trigger (context.c:231)
==14195== by 0x927A878: serf_context_run (context.c:305)
==14195== by 0x6F44680: svn_ra_serf__context_run (util.c:910)
==14195== by 0x6F419B2: svn_ra_serf__context_run_wait (util.c:981)
==14195== by 0x6F41A53: svn_ra_serf__context_run_one (util.c:1021)

Received on 2018-07-12 20:52:59 CEST

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.