|
Message
From: Matjaz Breskvar <phoenix@f...>
Date: Wed, 12 Nov 2003 13:22:48 +0100
Subject: Re: [openrisc] Optimization Problems
* pablo.huerta@a... (pablo.huerta@a...) wrote:
> Hi all!
>
> I have read some messages about optimization problems when using GCC
> 3.2.3. In
> http://www.teisa.unican.es/~csanchez/guides/double_modes_bug_explanation.shtml
> there is an explanation on how we solved a bug with optimization
> problems in GCC 2.9.5 that is not solved yet in GCC 3.2.3, reading
> this may help OpenCores GCC maintainers to solve this problems.
i'll look into this and fix it.
> If any GCC maintainer needs help to solve that problem, I can help
> him. I have also found some problems in file or32.S, in the
> __udivsi3 function. I think that functions in or32.S file should be
> revised. I don't have write acces to CVS so if any maintainer could
> help me to solve that bugs please contact me.
i'm defenetly interested in fixing any/all bugs in OpenRISC toolchain.
please do post any bugs you've found with proposed bugfixes.
> Also we have already developed the port of GCC 3.4 and will be ready
> for downloading in that web page soon:
> http://www.teisa.unican.es/~csanchez/download1.shtml A port of eCos 2.0
> is also ready to be published. I will post a message in the forum when
> everythingh is ready to download.
please make sure that your gcc-3.4 port uses the same ABI
than ours. one quick test would be to compile uclinux from opencores
cvs. i think it's in all ours best interest to have one unified toolchain
to avoid duplicating efforts.
regards,
p.
|
 |
Browse
