patch reverts on linux-next - ioremap_uc() for atyfb

patch reverts on linux-next - ioremap_uc() for atyfb

[Luis R. Rodriguez]

Andrew,

I got a notice from Ingo on July 21 that one of my patches, "x86/mm,
asm-generic: Add IOMMU ioremap_uc() variant default" was merged into tip. It
was merged a long with other patches, for example:

http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=3cc2dac5be3f23414a4efdee0b26d79bed297cac

I wrote this patch after Boris had my atyfb series bake on his tree
as his tree receives 0-day tests. Then this patch for example makes use of
ioremap_uc():

"drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC"

I noticed though that on top there's a revert of that same patch:

http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=4c090fb7209d523ef4cedb354192a190edd0d166

Revert "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC" akpm-base
This reverts commit 3cc2dac5be3f23414a4efdee0b26d79bed297cac.

It doesn't explain why this was reverted though. Is it OK for things be
reverted like this ? Is it understood by others ? It was a bit of a surprise to
me though as I was not able to verify things were going through to linux-next.
Since Boris was also on vacation and since my trees do not get 0-day-tests
it also meant I realied on the chain for issues to be found. I'll fix the
fact that my trees do not get 0-day tests but it seems we should probably
only put so many dev trees on 0-day test, I'll check with Fengguang Wu if
he has bandwidth to put some of my trees.

I was just not sure what was going on since although I got a notice from Ingo
the patch was merged into tip I did not happen to see it until *today* on
linux-next.

I guess because it also trickled through other test machines today Stephen
Rothwell found one issue with this patch for missing iorenmap_uc() calls for
architectures that do not include asm-generic/io.h. Since I got this report
today I sent a prompt follow up fix up as soon as I saw and understood the
report but the report did come with a delay as the patch was being reverted and
I hadn't gotten any notice of it being excluded from linux-next or why. It got
through today, but not sure why, and I think my patch fixes the issue.

What criteria is being used for patches to be reverted to your tree? Since
your tree gets merged to linux-next it means outbound users cannot test
the 'real' linux-next, but since we get no notification it unfortunatley
also means we can't know about the issues unless all dev trees are on
0-day testing which of course seems a bit overkill right now.

Sorry if I screwed up, just want to know what is proper here and if I need
to get my trees tested with 0-day bot I'll try to make arrangements with
Fengguang Wu if his machines have coverage.

  Luis

Re: patch reverts on linux-next - ioremap_uc() for atyfb

[Andrew Morton]

On Tue, 28 Jul 2015 22:05:49 +0200 "Luis R. Rodriguez" <mcgrof@suse.com> wrote:

> I got a notice from Ingo on July 21 that one of my patches, "x86/mm,
> asm-generic: Add IOMMU ioremap_uc() variant default" was merged into tip. It
> was merged a long with other patches, for example:
> 
> http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=3cc2dac5be3f23414a4efdee0b26d79bed297cac
> 
> I wrote this patch after Boris had my atyfb series bake on his tree
> as his tree receives 0-day tests. Then this patch for example makes use of
> ioremap_uc():
> 
> "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC"
> 
> I noticed though that on top there's a revert of that same patch:
> 
> http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=4c090fb7209d523ef4cedb354192a190edd0d166
> 
> Revert "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC" akpm-base
> This reverts commit 3cc2dac5be3f23414a4efdee0b26d79bed297cac.

Stephen did this while preparing the linux-next release.  Note that he
reverted it "for today".




Begin forwarded message:

Date: Tue, 28 Jul 2015 15:33:03 +1000
From: Stephen Rothwell <sfr@canb.auug.org.au>
To: Thomas Gleixner <tglx@linutronix.de>, Ingo Molnar <mingo@elte.hu>, "H. Peter Anvin" <hpa@zytor.com>, Peter Zijlstra <peterz@infradead.org>
Cc: linux-next@vger.kernel.org, linux-kernel@vger.kernel.org, "Luis R. Rodriguez" <mcgrof@suse.com>
Subject: linux-next: build failure after merge of the tip tree


Hi all,

After merging the tip tree, today's linux-next build (powerpc
allyesconfig) failed like this:

drivers/video/fbdev/aty/atyfb_base.c: In function 'atyfb_setup_generic':
drivers/video/fbdev/aty/atyfb_base.c:3447:2: error: implicit declaration of function 'ioremap_uc' [-Werror=implicit-function-declaration]
  par->ati_regbase = ioremap_uc(info->fix.mmio_start, 0x1000);
  ^
drivers/video/fbdev/aty/atyfb_base.c:3447:19: warning: assignment makes pointer from integer without a cast
  par->ati_regbase = ioremap_uc(info->fix.mmio_start, 0x1000);
                   ^

Caused by commits

  3cc2dac5be3f ("drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC")
  8c7ea50c010b ("x86/mm, asm-generic: Add IOMMU ioremap_uc() variant default")

The latter defines ioremap_uc() for x86 and those architectures that
use asm-generic/io.h - which is not all of them :-( .  The former commit
then uses ioremap_uc().

I have reverted commit 3cc2dac5be3f (and 7d89a3cb159a that follows it)
for today.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

Re: patch reverts on linux-next - ioremap_uc() for atyfb

[Fengguang Wu]

HI Luis,

On Tue, Jul 28, 2015 at 10:05:49PM +0200, Luis R. Rodriguez wrote:
> Andrew,
> 
> I got a notice from Ingo on July 21 that one of my patches, "x86/mm,
> asm-generic: Add IOMMU ioremap_uc() variant default" was merged into tip. It
> was merged a long with other patches, for example:
> 
> http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=3cc2dac5be3f23414a4efdee0b26d79bed297cac
> 
> I wrote this patch after Boris had my atyfb series bake on his tree
> as his tree receives 0-day tests. Then this patch for example makes use of
> ioremap_uc():
> 
> "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC"
> 
> I noticed though that on top there's a revert of that same patch:
> 
> http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=4c090fb7209d523ef4cedb354192a190edd0d166
> 
> Revert "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC" akpm-base
> This reverts commit 3cc2dac5be3f23414a4efdee0b26d79bed297cac.
> 
> It doesn't explain why this was reverted though. Is it OK for things be
> reverted like this ? Is it understood by others ? It was a bit of a surprise to
> me though as I was not able to verify things were going through to linux-next.
> Since Boris was also on vacation and since my trees do not get 0-day-tests
> it also meant I realied on the chain for issues to be found. I'll fix the
> fact that my trees do not get 0-day tests but it seems we should probably
> only put so many dev trees on 0-day test, I'll check with Fengguang Wu if
> he has bandwidth to put some of my trees.

Would you tell me your git URL? I'm always seeking to add more dev
trees to the 0-day test pool. Bandwidth is not a problem you need to
worry.

Thanks,
Fengguang

Re: patch reverts on linux-next - ioremap_uc() for atyfb

[Ingo Molnar]

* Luis R. Rodriguez <mcgrof@suse.com> wrote:

> Andrew,
> 
> I got a notice from Ingo on July 21 that one of my patches, "x86/mm,
> asm-generic: Add IOMMU ioremap_uc() variant default" was merged into tip. It
> was merged a long with other patches, for example:
> 
> http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=3cc2dac5be3f23414a4efdee0b26d79bed297cac
> 
> I wrote this patch after Boris had my atyfb series bake on his tree
> as his tree receives 0-day tests. Then this patch for example makes use of
> ioremap_uc():
> 
> "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC"
> 
> I noticed though that on top there's a revert of that same patch:
> 
> http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=4c090fb7209d523ef4cedb354192a190edd0d166
> 
> Revert "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC" akpm-base
> This reverts commit 3cc2dac5be3f23414a4efdee0b26d79bed297cac.
> 
> It doesn't explain why this was reverted though. Is it OK for things be
> reverted like this ?

The revert Stephen has done were only temporary, done for the throw-away 
linux-next integration branch, to make sure linux-next builds on the architectures 
your patch broke.

I've applied your fix to -tip, and after some testing will push it out to 
linux-next - at that point the temporary reverts will not be needed in linux-next 
anymore and will be dropped. They won't ever go upstream.

Thanks,

	Ingo

Re: patch reverts on linux-next - ioremap_uc() for atyfb

[Luis R. Rodriguez]

On Wed, Jul 29, 2015 at 08:59:18AM +0800, Fengguang Wu wrote:
> HI Luis,
> 
> On Tue, Jul 28, 2015 at 10:05:49PM +0200, Luis R. Rodriguez wrote:
> > Andrew,
> > 
> > I got a notice from Ingo on July 21 that one of my patches, "x86/mm,
> > asm-generic: Add IOMMU ioremap_uc() variant default" was merged into tip. It
> > was merged a long with other patches, for example:
> > 
> > http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=3cc2dac5be3f23414a4efdee0b26d79bed297cac
> > 
> > I wrote this patch after Boris had my atyfb series bake on his tree
> > as his tree receives 0-day tests. Then this patch for example makes use of
> > ioremap_uc():
> > 
> > "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC"
> > 
> > I noticed though that on top there's a revert of that same patch:
> > 
> > http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=4c090fb7209d523ef4cedb354192a190edd0d166
> > 
> > Revert "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC" akpm-base
> > This reverts commit 3cc2dac5be3f23414a4efdee0b26d79bed297cac.
> > 
> > It doesn't explain why this was reverted though. Is it OK for things be
> > reverted like this ? Is it understood by others ? It was a bit of a surprise to
> > me though as I was not able to verify things were going through to linux-next.
> > Since Boris was also on vacation and since my trees do not get 0-day-tests
> > it also meant I realied on the chain for issues to be found. I'll fix the
> > fact that my trees do not get 0-day tests but it seems we should probably
> > only put so many dev trees on 0-day test, I'll check with Fengguang Wu if
> > he has bandwidth to put some of my trees.
> 
> Would you tell me your git URL? I'm always seeking to add more dev
> trees to the 0-day test pool. Bandwidth is not a problem you need to
> worry.

Will do! Thanks. I'll send you the URL next.

  Luis

Re: patch reverts on linux-next - ioremap_uc() for atyfb

[Luis R. Rodriguez]

On Tue, Jul 28, 2015 at 01:44:28PM -0700, Andrew Morton wrote:
> On Tue, 28 Jul 2015 22:05:49 +0200 "Luis R. Rodriguez" <mcgrof@suse.com> wrote:
> 
> > I got a notice from Ingo on July 21 that one of my patches, "x86/mm,
> > asm-generic: Add IOMMU ioremap_uc() variant default" was merged into tip. It
> > was merged a long with other patches, for example:
> > 
> > http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=3cc2dac5be3f23414a4efdee0b26d79bed297cac
> > 
> > I wrote this patch after Boris had my atyfb series bake on his tree
> > as his tree receives 0-day tests. Then this patch for example makes use of
> > ioremap_uc():
> > 
> > "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC"
> > 
> > I noticed though that on top there's a revert of that same patch:
> > 
> > http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/video/fbdev/aty/atyfb_base.c?id=4c090fb7209d523ef4cedb354192a190edd0d166
> > 
> > Revert "drivers/video/fbdev/atyfb: Replace MTRR UC hole with strong UC" akpm-base
> > This reverts commit 3cc2dac5be3f23414a4efdee0b26d79bed297cac.
> 
> Stephen did this while preparing the linux-next release.  Note that he
> reverted it "for today".

Sorry for the noise folks... I'll be testing all my patches moving forward
behidn the 0-day bot prior to posting.

 Luis