This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug target/20360] 20021014-1.c fails on account of unsupported build option
- From: "tprince at computer dot org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 13 Mar 2005 04:07:04 -0000
- Subject: [Bug target/20360] 20021014-1.c fails on account of unsupported build option
- References: <20050307143259.20360.tprince@computer.org>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Additional Comments From tprince at computer dot org 2005-03-13 04:07 -------
Subject: Re: 20021014-1.c fails on account of
unsupported build option
At 09:28 AM 3/12/2005, pinskia at gcc dot gnu dot org wrote:
>------- Additional Comments From pinskia at gcc dot gnu dot
>org 2005-03-12 17:28 -------
>(In reply to comment #2)
> > If I understand the purpose of the test, it was to check whether profiling
> > is working correctly, as far as being capable of building and running
> > without a failure report. I've not found an explanation why certain
> > targets, like cygwin, support only -pg, others only -p, and others
> > both. The main point is to stop reporting a spurious error when running
> > the testsuite.
>
>No it is just testing -p and not -pg, there is a difference but I don't
>remember what.
IIRC, the original code generation problem for which the test was
instituted was the same for -p and -pg, for those targets which support
both. Certain targets support only one or the other at link time, and
choosing the wrong one produces a link time error. I don't care whether
the policy is to disable the test for targets which don't support -p, or to
switch over to -pg. I think general policy is to not declare a failure for
a test which can't be supported for the target, and that policy should
extend to cygwin.
Tim Prince
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20360