This is the mail archive of the
mailing list for the glibc project.
Re: Tunables for 2.23?
- From: Siddhesh Poyarekar <sid at reserved-bit dot com>
- To: "Paul E. Murphy" <murphyp at linux dot vnet dot ibm dot com>
- Cc: "libc-alpha at sourceware dot org" <libc-alpha at sourceware dot org>, Carlos O'Donell <carlos at redhat dot com>, "munroesj at linux dot vnet dot ibm dot com" <munroesj at linux dot vnet dot ibm dot com>, Tulio Magno Quites Machado Filho <tuliom at linux dot vnet dot ibm dot com>, Andi Kleen <andi at firstfloor dot org>
- Date: Mon, 11 Jan 2016 03:48:34 +0530
- Subject: Re: Tunables for 2.23?
- Authentication-results: sourceware.org; auth=none
- References: <568C2AE6 dot 90002 at linux dot vnet dot ibm dot com>
On Tue, Jan 05, 2016 at 02:43:18PM -0600, Paul E. Murphy wrote:
> Siddhesh has been working on Tunables on his siddhesh/tunables branch. I'd done
> some extra work to extend it to support dynamic TLE . This is needed to get
> good TLE performance from power8 hardware.
> I'd like to see this cleaned up and included in the next release. Is this still
I have pushed a reworked version of the two patches to the branch,
which may break the patches y'all have on top of the branch. As
Roland suggested, I have split the feature into two parts, one that
implements the tunables framework (this time with all the works:
configure option, README, etc.) and another that initializes the
I have implemented initialization of the tunable list from a single
environment variable GLIBC_TUNABLES because that is the method I like
best right now. I believe this second patch is where the differences
in opinion are (some prefer one envvar per tunable, Roland seemed to
want another solution that doesn't involve envvars at all, etc.) so we
might want to focus our efforts on getting the first patch into 2.23
while we converge on the approach we prefer for the second patch.
It's pretty late here (or early, depending on how you look at it), so
I'll work on the patch submission tomorrow.