From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Wed, 30 May 2018 21:53:36 +0300 From: Konstantin Osipov Subject: Re: [PATCH] vinyl: fix false-positive assertion at exit Message-ID: <20180530185336.GE18850@atlas> References: <6237e2c7bd915c08b4ca396c9a0a14e1700e2fea.1527600966.git.vdavydov.dev@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <6237e2c7bd915c08b4ca396c9a0a14e1700e2fea.1527600966.git.vdavydov.dev@gmail.com> To: Vladimir Davydov Cc: tarantool-patches@freelists.org List-ID: * Vladimir Davydov [18/05/29 18:20]: > latch_destroy() and fiber_cond_destroy() are basically no-op. All they > do is check that latch/cond is not used. When a global latch or cond > object is destroyed at exit, it may still have users and this is OK as > we don't stop fibers at exit. In vinyl this results in the following > false-positive assertion failures: > > src/latch.h:81: latch_destroy: Assertion `l->owner == NULL' failed. > > src/fiber_cond.c:49: fiber_cond_destroy: Assertion `rlist_empty(&c->waiters)' failed. > I'm ok with the patch, but could you, please, once again, try to reduce the test case execution time? You could easily do that by increasing padding and reducing the number of rows in the loop - each written row is a WAL write, and WAL is the biggest bottleneck in a single-fiber write scenario. Thanks! -- Konstantin Osipov, Moscow, Russia, +7 903 626 22 32 http://tarantool.io - www.twitter.com/kostja_osipov