Re: [PATCH] MIPS32 DSP intrinsics

[Mark Mitchell <mark at codesourcery dot com>]

Eric Christopher wrote:
> > FWIW, I was kind-of sympathetic to your position ;)
>
>
> thanks. :)
>
>
> > So I'm reluctant to allow the code in at this stage as well.  I'm still
> > happy to review the patch though.  Hopefully then the final version can
> > be applied soon after the secrecy is lifted.

I don't think it's reasonable to require people to contribute simulator 
patches in order to check in architecture support.  There are lots of 
GCC ports without (free) simulators.

Similarly, having publicly available ISA manuals should not be a 
requirement.  Sometimes, for example, manuals are written as single 
documents with ISA documentation mixed in with very proprietary stuff, 
and it's not reasonable to ask people to split them up.  Or, sometimes, 
people people are reverse-engineering ISAs.

In either case, it's great to encourage these things, but we know that 
users will want the compiler support, even if these things aren't 
necessarily available.  If there's no simulator or manual, I think it's 
reasonable to say that it's not your fault if you change something that 
breaks such a platform; you couldn't reasonably test it.  But, simply 
refusing to accept the code seems like a case of cutting off our noses 
to spite our faces.

--
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com
(916) 791-8304