This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: C ASSEMBLER_NAME patch
- From: Bob Wilson <bwilson at tensilica dot com>
- To: Geoffrey Keating <geoffk at apple dot com>
- Cc: gkeating at apple dot com, gcc-patches at gcc dot gnu dot org
- Date: Thu, 24 Jul 2003 14:50:05 -0700
- Subject: Re: C ASSEMBLER_NAME patch
- Organization: Tensilica, Inc.
- References: <05C9B266-BE1D-11D7-9D66-0030657EA24A@apple.com>
On Thursday 24 July 2003 14:23, Geoffrey Keating wrote:
> Could you please double-check that you're using a version of the
> compiler with this actual patch in it?
>
> Unless you're trying to build glibc with intermodule analysis, you
> should not be seeing the behaviour you're seeing.
I'm using the gcc mainline sources as of yesterday (July 23). I configured
with --target=xtensa-linux on a RedHat 7.3 machine and did nothing special to
enable intermodule analysis. I have not made any changes to the glibc
makefiles, either.
I agree that I should not be seeing this behavior :-)
What exactly doesn't make sense to you? An earlier version of your patch
caused static functions to be renamed even when intermodule analysis is not
being used, right? As I understood it, your patch on July 17 was supposed to
fix that so that the static functions wouldn't be renamed when compiling only
one translation unit. It's not working in this case. The symptoms I am
seeing appear to match exactly the case you described in the patch as an
"ongoing problem".
Whether you fix that or not, the second and more immediate issue is that the
modified function names are not being used consistently (e.g., in the aliased
function declaration).
--Bob