This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: Lock elision test results
- From: Dominik Vogt <vogt at linux dot vnet dot ibm dot com>
- To: libc-alpha at sourceware dot org
- Cc: Andi Kleen <andi at firstfloor dot org>
- Date: Thu, 4 Jul 2013 10:10:31 +0200
- Subject: Re: Lock elision test results
- References: <20130614102653 dot GA21917 at linux dot vnet dot ibm dot com> <1372767484 dot 22198 dot 4505 dot camel at triegel dot csb> <20130703065355 dot GA4522 at linux dot vnet dot ibm dot com> <1372849432 dot 14172 dot 20 dot camel at triegel dot csb> <20130703141503 dot GK6123 at two dot firstfloor dot org>
- Reply-to: libc-alpha at sourceware dot org
On Wed, Jul 03, 2013 at 04:15:03PM +0200, Andi Kleen wrote:
> First I must say I'm pretty sceptic on the idea of assembling a
> micro benchmark suite for lock elision. It is unlikely to give
> you a true picture of the performance, as small programs tend to not
> model real applications well for something as complex as transactions.
It's not a benchmark suite but meant to shed light on the
performance of certain corner cases and identify areas where
performance needs or can be improved. The current tests are
designed to find cases with poor performance on z. If I knew more
about the tsx implementation I could try to identify cases that
show bad performance there too.
I think identifying nad understanding problematic situations is
important for a good overall performance.
Ciao
Dominik ^_^ ^_^
--
Dominik Vogt
IBM Germany