Re: [PATCH 2/3 v2] sh: fix DMA driver's descriptor chaining and cookie assignment

[Guennadi Liakhovetski <g.liakhovetski@gmx.de>]

On Thu, 17 Dec 2009, Nobuhiro Iwamatsu wrote:

> HI,
> 
> 2009/12/17 Guennadi Liakhovetski <g.liakhovetski@gmx.de>:
> > Hello Iwamatsu-san
> >
> > On Thu, 17 Dec 2009, Nobuhiro Iwamatsu wrote:
> >
> >> Hi,
> >>
> >> 2009/12/12 Nobuhiro Iwamatsu <iwamatsu@nigauri.org>:
> >> > Hi,
> >> >
> >> > Oh, I was going to check an old patch.
> >> >
> >> > 2009/12/11 Paul Mundt <lethal@linux-sh.org>:
> >> >> On Thu, Dec 10, 2009 at 06:35:11PM +0100, Guennadi Liakhovetski wrote:
> >> >>> The SH DMA driver wrongly assigns negative cookies to transfer descriptors,
> >> >>> also, its chaining of partial descriptors is broken. The latter problem is
> >> >>> usually invisible, because maximum transfer size per chunk is 16M, but if you
> >> >>> artificially set this limit lower, the driver fails. Since cookies are also
> >> >>> used in chunk management, both these problems are fixed in one patch. As side
> >> >>> effects a possible memory leak, when descriptors are prepared, but not
> >> >>> submitted, and multiple races have also been fixed.
> >> >>>
> >> >>> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> >> >>
> >> >> Looks good to me.
> >> > Me too.
> >> >
> >> > But I want to check other CPU of sh. I will do next week.
> >> > Please wait "Acked-by".
> >> >
> >> >
> >>
> >> I tested this patch. But I can not apply to Paul's git/HEAD.
> >> I re-create patch and I attached.
> >> This version of patch work fine.
> >> Guennadi, could you test this patch on sh7724?
> >
> > Thanks for testing. I looked at your version of my patch, and I don't
> > agree with it. You needlessly changed indentation in it and also forgot to
> > include a "From: " line in it to indicate its original authorship.
> > Besides, I think, Dan will be applying this patch on top of current next
> > or Linus' tree. Of course, it can be, that my version of the patch does
> > not apply to those trees either now. So, Dan, if you cannot merge my
> > version of this patch yourself, please, let me know, I'll rebase it, but
> > please do not use this version.
> 
> Yes, you are right.
> And sorry, I am not going to seize your patch by force.

Sure, I never assumed the opposite, that's why I said "you forgot." Same 
thing happened to me too recently, fortunately, I noticed it early enough 
to push a correct version. The actual problem with your version was 
white-space changes.

> I sent the patch which I revised at hand.

Thanks
Guennadi
---
Guennadi Liakhovetski