This is the mail archive of the
mailing list for the glibc project.
Re: [PATCH] Speedup nptl/tst-mutex9 and nptl/tst-mutexpi9.
- From: Carlos O'Donell <carlos at redhat dot com>
- To: Stefan Liebler <stli at linux dot ibm dot com>, GNU C Library <libc-alpha at sourceware dot org>
- Date: Thu, 12 Sep 2019 10:08:39 -0400
- Subject: Re: [PATCH] Speedup nptl/tst-mutex9 and nptl/tst-mutexpi9.
- References: <email@example.com>
On 9/12/19 8:15 AM, Stefan Liebler wrote:
> each test runs for 6s. do_test_clock is called three times where it
> sleeps for two seconds and is then checking if the child process
> was killed by SIGALRM. The child process tries to lock a mutex
> which times out after 0.5s and is setting up alarm(1) before it blocks
> due to pthread_mutex_lock.
> This patch removes the sleep(2) and reduces the timeout for the first
> pthread_mutex_[timed|clock]lock call.
> As the nptl tests run in sequence, this patch saves roughly 6s of
> runtime for "make check".
> * nptl/tst-mutex9.c (do_test_clock): Remove sleep and
> reduce timeout.
Same comment as I made with "[PATCH] Speedup various nptl/tst-cancel20 and nptl/tst-cancel21 tests."