This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [RFC] Automatic nightly builds?
- From: "Joseph S. Myers" <joseph at codesourcery dot com>
- To: "Frank Ch. Eigler" <fche at redhat dot com>
- Cc: =?utf-8?B?T25kxZllaiBCw61sa2E=?= <neleai at seznam dot cz>, Carlos O'Donell <carlos at redhat dot com>, GNU C Library <libc-alpha at sourceware dot org>, Marcus Shawcroft <marcus dot shawcroft at linaro dot org>, "H.J. Lu" <hjl dot tools at gmail dot com>, Thomas Schwinge <thomas at codesourcery dot com>, Roland McGrath <roland at hack dot frob dot com>, Andreas Krebbel <krebbel at linux dot vnet dot ibm dot com>, David Miller <davem at davemloft dot net>, Mike Frysinger <vapier at gentoo dot org>, Chris Metcalf <cmetcalf at tilera dot com>, Andreas Schwab <schwab at linux-m68k dot org>
- Date: Sun, 21 Jul 2013 21:03:08 +0000
- Subject: Re: [RFC] Automatic nightly builds?
- References: <51E5C427 dot 60304 at redhat dot com> <20130720095751 dot GA7273 at domone> <Pine dot LNX dot 4 dot 64 dot 1307201155550 dot 3231 at digraph dot polyomino dot org dot uk> <y0msiz77rcx dot fsf at fche dot csb>
On Sun, 21 Jul 2013, Frank Ch. Eigler wrote:
> "Joseph S. Myers" <joseph@codesourcery.com> writes:
>
> > [...]
> > Diffs in .out files are not reliable indications of changes to testsuite
> > failures. Instead, we should get in some version of Tomas Dohnalek's
> > patch to log PASS / FAIL results for individual tests. At that point, you
> > have something for which comparisons are reasonably reliable and useful,
> > even if not perfect.
>
> As an experiment, some of y'all might try emailing a copy of your
> dejagnu .sum files to dejazilla@elastic.org, to get them stored in a
Without Tomas Dohnalek's changes, there isn't anything looking passably
like a DejaGnu .sum file at all. (With those changes, generating a
DejaGnu-like header and footer would be easy enough. There are plenty of
other glibc testsuite issues as described at length in
<http://sourceware.org/ml/libc-alpha/2013-07/msg00256.html>, including
wanting PASS/FAIL structured information for individual assertions within
tests, but just PASS/FAIL information for tests run from makefiles - or
even just for the subset of tests run with generic makefile rules rather
than with their own special rules, which I think is all Tomas's patches
actually handle - would be a very good start.)
--
Joseph S. Myers
joseph@codesourcery.com