<div class="gmail_extra"><div class="gmail_quote">On Mon, Apr 23, 2012 at 12:14 AM, Evgeniy Stepanov <span dir="ltr"><<a href="mailto:eugeni.stepanov@gmail.com" target="_blank">eugeni.stepanov@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">On Sun, Apr 22, 2012 at 5:41 AM, Hal Finkel <<a href="mailto:hfinkel@anl.gov">hfinkel@anl.gov</a>> wrote:<br>
> Evgeniy,<br>
><br>
> Unfortunately, we need to be much more careful here. While gas for ARM<br>
> does support -mcpu, -march and -mfpu, this is not true for other ISAs.<br>
> Looking at the man page, it seems that the supported flags are:<br>
><br>
> ARM: -mcpu=, -march=, -mfpu=<br>
> i386: -march= (-mtune=: should we map -mcpu= to -mtune=?)<br>
> MIPS: (same as i386)<br>
> PowerPC: supports cpu, but as -m<cpu> (meaning -m403|-m405|-mppc64|...)<br>
> SPARC: supports arch, but as -xarch=<br>
><br>
> How do you think we should fix this?<br>
<br>
</div>Well, we can make the set of forwarded options dependent on the target.<br></blockquote><div><br></div><div>Yea, we need each target to manage the translation of these options.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
But what's the failing use case? I mean, why do you pass these options<br>
to the compiler in the first place?<br></blockquote><div><br></div><div>-mcpu is a supported driver option? It should DTRT for every platform, and that platform's assembler.</div></div></div>