[Gc]: libatomic-ops patch for Renesas SH
hans.boehm at hp.com
Fri Oct 2 14:25:43 PDT 2009
> From: Ivan Maidanski
> Nobuhiro Iwamatsu <iwamatsu at nigauri.org> wrote:
> > Hi, all.
> > I am now trying to run linux on Renesas SH CPU(sh4).
> > Because I am porting new architecture(sh4) to Debian GNU/Linux.
> > Current libatomic-ops does not support this CPU.
> > I port to this, and do BTS to Debian BTS of libatomic-ops.
> > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=547136
> > Can you have comment?
> Two things to refine:
> - copyright (Hans, what's your opinion? is it ok?);
For substantially new files, I think this is OK. There is a down side to having multiple copyright holders, in that it makes it even harder to adjust the copyright if we needed to do that at some point. Thus there's some small chance that such files might get dropped in that case, if we couldn't track down the author. But I think the situation is essentially identical to that for mips.h in the same directory.
> - AO_test_and_set_full should return values within
> AO_TS_VAL_t type (i.e., AO_TS_CLEAR(0) and AO_TS_SET(0xff)
> instead of 0 and 0x80).
Unless I'm missing something, this looks to me like it's not broken. The AO_TS_SET and AO_TS_CLEAR values are allowed to be hardware dependent. But I admit, it's not clear to me why the nonstandard value helps here. Does anything break if the redeinition of AO_TS_SET is just removed?
> > If without a problem, would you apply this?
> Yes, if the above things would be solved somehow.
> > I got comment from Hans about this patch.
> > 2009/10/2 Boehm, Hans <hans.boehm at hp.com>:
> > > I don't see a problem with the patch. Could you post it
> to gc at linux.hpl.hp.com so others also get a chance to look at
> it? You can add my comment if you like.
> > >
> > > Thanks.
> > >
> > > Hans
> > Best regards,
> > Nobuhiro
> Gc mailing list
> Gc at linux.hpl.hp.com
More information about the Gc