This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH,testsuite]Make test pr36227 reflect target pointer size
- From: "Joseph S. Myers" <joseph at codesourcery dot com>
- To: NightStrike <nightstrike at gmail dot com>
- Cc: Andy H <hutchinsonandy at aim dot com>, gcc-patches at gcc dot gnu dot org, janis187 at us dot ibm dot com, Kai Tietz <ktietz70 at googlemail dot com>, Kai Tietz <Kai dot Tietz at onevision dot com>
- Date: Sun, 22 Jun 2008 16:33:44 +0000 (UTC)
- Subject: Re: [PATCH,testsuite]Make test pr36227 reflect target pointer size
- References: <485E49A5.5000903@aim.com> <b609cb3b0806220923p5256530fn48093d72f0cfde1e@mail.gmail.com>
On Sun, 22 Jun 2008, NightStrike wrote:
> This is related to the size_t / intptr_t discussions that keep coming
> up. Is there a gcc coding convention that says somewhere -- "Use
> intptr_t for pointer sizes"? There's still a lot of assumptions that
> long = pointer size, and how those are fixed should be consistent.
> This patch proposes one way, Kai's various patches in other areas have
> proposed another way, and discussions on those things I think have
> pointed to using autoconf to figure out of intptr_t is available and
> use that before doing anything else.
Use of autoconf is more or less necessary for the host. This is a
testcase patch, and requirements for testcases are completely different
from those for host code or target libraries, and testcases do not need to
follow coding conventions (in fact, I think GDB recommends that GDB
testcases do not follow a uniform style). For existing testcases, what's
important is to change as little as possible what the test tests on
targets where it's already a correct test.
--
Joseph S. Myers
joseph@codesourcery.com