Patchwork [GIT,PULL] ARM: OMAP: DSS hwmod/reset fixes for 3.2

login
register
mail settings
Submitter Paul Walmsley
Date Oct. 7, 2011, 12:07 a.m.
Message ID <alpine.DEB.2.00.1110061806150.4611@utopia.booyaka.com>
Download mbox
Permalink /patch/118187/
State New
Headers show

Pull-request

git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2

Comments

Paul Walmsley - Oct. 7, 2011, 12:07 a.m.
Hi Tony,

The following changes since commit be73246058737beec52ae232bcab7776332a9e06:

  ARM: OMAP2+: Remove custom init_irq for remaining boards (2011-09-26 17:50:37 -0700)

are available in the git repository at:
  git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2

Archit Taneja (1):
      ARM: OMAP2PLUS: DSS: Ensure DSS works correctly if display is enabled in bootloader

Tomi Valkeinen (9):
      ARM: OMAP2xxx: HWMOD: Fix DSS reset
      ARM: OMAP2xxx: HWMOD: fix DSS clock data
      ARM: OMAP3: HWMOD: Fix DSS reset
      ARM: OMAP3: HWMOD: fix DSS clock data
      ARM: OMAP4: HWMOD: remove extra clocks
      ARM: OMAP4: HWMOD: Add HWMOD_CONTROL_OPT_CLKS_IN_RESET for dss_core
      ARM: OMAP4: HWMOD: fix DSS clock data
      ARM: OMAP2/3: HWMOD: Add SYSS_HAS_RESET_STATUS for dss
      ARM: OMAP: HWMOD: Unify DSS resets for OMAPs

 arch/arm/mach-omap2/Makefile                       |    5 +-
 arch/arm/mach-omap2/display.c                      |  160 ++++++++++++++++++++
 arch/arm/mach-omap2/display.h                      |   29 ++++
 arch/arm/mach-omap2/omap_hwmod_2420_data.c         |   17 ++-
 arch/arm/mach-omap2/omap_hwmod_2430_data.c         |   17 ++-
 .../mach-omap2/omap_hwmod_2xxx_3xxx_ipblock_data.c |    5 +-
 arch/arm/mach-omap2/omap_hwmod_3xxx_data.c         |   37 ++++-
 arch/arm/mach-omap2/omap_hwmod_44xx_data.c         |   24 ++--
 arch/arm/mach-omap2/omap_hwmod_common_data.c       |    4 +
 arch/arm/mach-omap2/omap_hwmod_common_data.h       |    4 +
 arch/arm/plat-omap/include/plat/common.h           |    3 +
 include/video/omapdss.h                            |    7 -
 12 files changed, 277 insertions(+), 35 deletions(-)
 create mode 100644 arch/arm/mach-omap2/display.h
Tony Lindgren - Nov. 5, 2011, 12:35 a.m.
* Paul Walmsley <paul@pwsan.com> [111006 16:33]:
> Hi Tony,
> 
> The following changes since commit be73246058737beec52ae232bcab7776332a9e06:
> 
>   ARM: OMAP2+: Remove custom init_irq for remaining boards (2011-09-26 17:50:37 -0700)
> 
> are available in the git repository at:
>   git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2

Looks like you should take a look at the merge conflicts with
this one, maybe rebase it on 994c0e992522c123298b4a91b72f5e67ba2d1123.

Regards,

Tony
 
> Archit Taneja (1):
>       ARM: OMAP2PLUS: DSS: Ensure DSS works correctly if display is enabled in bootloader
> 
> Tomi Valkeinen (9):
>       ARM: OMAP2xxx: HWMOD: Fix DSS reset
>       ARM: OMAP2xxx: HWMOD: fix DSS clock data
>       ARM: OMAP3: HWMOD: Fix DSS reset
>       ARM: OMAP3: HWMOD: fix DSS clock data
>       ARM: OMAP4: HWMOD: remove extra clocks
>       ARM: OMAP4: HWMOD: Add HWMOD_CONTROL_OPT_CLKS_IN_RESET for dss_core
>       ARM: OMAP4: HWMOD: fix DSS clock data
>       ARM: OMAP2/3: HWMOD: Add SYSS_HAS_RESET_STATUS for dss
>       ARM: OMAP: HWMOD: Unify DSS resets for OMAPs
> 
>  arch/arm/mach-omap2/Makefile                       |    5 +-
>  arch/arm/mach-omap2/display.c                      |  160 ++++++++++++++++++++
>  arch/arm/mach-omap2/display.h                      |   29 ++++
>  arch/arm/mach-omap2/omap_hwmod_2420_data.c         |   17 ++-
>  arch/arm/mach-omap2/omap_hwmod_2430_data.c         |   17 ++-
>  .../mach-omap2/omap_hwmod_2xxx_3xxx_ipblock_data.c |    5 +-
>  arch/arm/mach-omap2/omap_hwmod_3xxx_data.c         |   37 ++++-
>  arch/arm/mach-omap2/omap_hwmod_44xx_data.c         |   24 ++--
>  arch/arm/mach-omap2/omap_hwmod_common_data.c       |    4 +
>  arch/arm/mach-omap2/omap_hwmod_common_data.h       |    4 +
>  arch/arm/plat-omap/include/plat/common.h           |    3 +
>  include/video/omapdss.h                            |    7 -
>  12 files changed, 277 insertions(+), 35 deletions(-)
>  create mode 100644 arch/arm/mach-omap2/display.h
Paul Walmsley - Nov. 8, 2011, 7:50 a.m.
On Fri, 4 Nov 2011, Tony Lindgren wrote:

> * Paul Walmsley <paul@pwsan.com> [111006 16:33]:
> > Hi Tony,
> > 
> > The following changes since commit be73246058737beec52ae232bcab7776332a9e06:
> > 
> >   ARM: OMAP2+: Remove custom init_irq for remaining boards (2011-09-26 17:50:37 -0700)
> > 
> > are available in the git repository at:
> >   git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2
> 
> Looks like you should take a look at the merge conflicts with
> this one, maybe rebase it on 994c0e992522c123298b4a91b72f5e67ba2d1123.

Do you want this series for 3.2-rc or 3.3 at this point?


- Paul
Tomi Valkeinen - Nov. 8, 2011, 10:08 a.m.
On Tue, 2011-11-08 at 00:50 -0700, Paul Walmsley wrote:
> On Fri, 4 Nov 2011, Tony Lindgren wrote:
> 
> > * Paul Walmsley <paul@pwsan.com> [111006 16:33]:
> > > Hi Tony,
> > > 
> > > The following changes since commit be73246058737beec52ae232bcab7776332a9e06:
> > > 
> > >   ARM: OMAP2+: Remove custom init_irq for remaining boards (2011-09-26 17:50:37 -0700)
> > > 
> > > are available in the git repository at:
> > >   git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2
> > 
> > Looks like you should take a look at the merge conflicts with
> > this one, maybe rebase it on 994c0e992522c123298b4a91b72f5e67ba2d1123.
> 
> Do you want this series for 3.2-rc or 3.3 at this point?

Seems that at least HDMI output is broken due to this series missing.

 Tomi
Paul Walmsley - Nov. 8, 2011, 10:33 a.m.
On Tue, 8 Nov 2011, Tomi Valkeinen wrote:

> On Tue, 2011-11-08 at 00:50 -0700, Paul Walmsley wrote:
> > On Fri, 4 Nov 2011, Tony Lindgren wrote:
> > 
> > > * Paul Walmsley <paul@pwsan.com> [111006 16:33]:
> > > > The following changes since commit be73246058737beec52ae232bcab7776332a9e06:
> > > > 
> > > >   ARM: OMAP2+: Remove custom init_irq for remaining boards (2011-09-26 17:50:37 -0700)
> > > > 
> > > > are available in the git repository at:
> > > >   git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2
> > > 
> > > Looks like you should take a look at the merge conflicts with
> > > this one, maybe rebase it on 994c0e992522c123298b4a91b72f5e67ba2d1123.
> > 
> > Do you want this series for 3.2-rc or 3.3 at this point?
> 
> Seems that at least HDMI output is broken due to this series missing.

OK thanks, 3.2-rc it is.

The merge conflicts have been fixed, they were trivial.  Tomi, care to 
do a quick test boot with the updated branch

git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2rc

?

My testbed is down at the moment and won't have the chance to deal with it 
for a few more days.


- Paul
Tomi Valkeinen - Nov. 8, 2011, 12:14 p.m.
On Tue, 2011-11-08 at 03:33 -0700, Paul Walmsley wrote:
> On Tue, 8 Nov 2011, Tomi Valkeinen wrote:
> 
> > On Tue, 2011-11-08 at 00:50 -0700, Paul Walmsley wrote:
> > > On Fri, 4 Nov 2011, Tony Lindgren wrote:
> > > 
> > > > * Paul Walmsley <paul@pwsan.com> [111006 16:33]:
> > > > > The following changes since commit be73246058737beec52ae232bcab7776332a9e06:
> > > > > 
> > > > >   ARM: OMAP2+: Remove custom init_irq for remaining boards (2011-09-26 17:50:37 -0700)
> > > > > 
> > > > > are available in the git repository at:
> > > > >   git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2
> > > > 
> > > > Looks like you should take a look at the merge conflicts with
> > > > this one, maybe rebase it on 994c0e992522c123298b4a91b72f5e67ba2d1123.
> > > 
> > > Do you want this series for 3.2-rc or 3.3 at this point?
> > 
> > Seems that at least HDMI output is broken due to this series missing.
> 
> OK thanks, 3.2-rc it is.
> 
> The merge conflicts have been fixed, they were trivial.  Tomi, care to 
> do a quick test boot with the updated branch
> 
> git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2rc
> 

Thanks, the patches in your branch look ok and work ok for me.

 Tomi
Tony Lindgren - Nov. 14, 2011, 8:46 p.m.
* Tomi Valkeinen <tomi.valkeinen@ti.com> [111108 03:40]:
> On Tue, 2011-11-08 at 03:33 -0700, Paul Walmsley wrote:
> > On Tue, 8 Nov 2011, Tomi Valkeinen wrote:
> > 
> > > On Tue, 2011-11-08 at 00:50 -0700, Paul Walmsley wrote:
> > > > On Fri, 4 Nov 2011, Tony Lindgren wrote:
> > > > 
> > > > > * Paul Walmsley <paul@pwsan.com> [111006 16:33]:
> > > > > > The following changes since commit be73246058737beec52ae232bcab7776332a9e06:
> > > > > > 
> > > > > >   ARM: OMAP2+: Remove custom init_irq for remaining boards (2011-09-26 17:50:37 -0700)
> > > > > > 
> > > > > > are available in the git repository at:
> > > > > >   git://git.pwsan.com/linux-2.6 hwmod_dss_fixes_3.2
> > > > > 
> > > > > Looks like you should take a look at the merge conflicts with
> > > > > this one, maybe rebase it on 994c0e992522c123298b4a91b72f5e67ba2d1123.
> > > > 
> > > > Do you want this series for 3.2-rc or 3.3 at this point?
> > > 
> > > Seems that at least HDMI output is broken due to this series missing.
> > 
> > OK thanks, 3.2-rc it is.

I suggest we only merge the ones with "fix" in the subject during the -rc
cycle. The others add new features for the reset, so they should wait until
v3.3 merge window.

Regards,

Tony
Paul Walmsley - Nov. 14, 2011, 9:40 p.m.
On Mon, 14 Nov 2011, Tony Lindgren wrote:

> I suggest we only merge the ones with "fix" in the subject during the -rc
> cycle. The others add new features for the reset, so they should wait until
> v3.3 merge window.

Most of the other patches in the branch discuss what they fix in the patch 
descriptions, rather than the subject lines.  Do you want the subject 
lines rewritten to highlight this?


- Paul
Tony Lindgren - Nov. 14, 2011, 10:04 p.m.
* Paul Walmsley <paul@pwsan.com> [111114 13:05]:
> On Mon, 14 Nov 2011, Tony Lindgren wrote:
> 
> > I suggest we only merge the ones with "fix" in the subject during the -rc
> > cycle. The others add new features for the reset, so they should wait until
> > v3.3 merge window.
> 
> Most of the other patches in the branch discuss what they fix in the patch 
> descriptions, rather than the subject lines.  Do you want the subject 
> lines rewritten to highlight this?

Well we should only merge major bugs and regressions during the -rc cycle.
It seems that at least "OMAP: HWMOD: Unify DSS resets for OMAPs" has high
flame potential based on that? Maybe we should wait until the next merge
window to start removing some of these workarounds?

Tony
Tomi Valkeinen - Nov. 15, 2011, 7:06 a.m.
On Mon, 2011-11-14 at 14:04 -0800, Tony Lindgren wrote:
> * Paul Walmsley <paul@pwsan.com> [111114 13:05]:
> > On Mon, 14 Nov 2011, Tony Lindgren wrote:
> > 
> > > I suggest we only merge the ones with "fix" in the subject during the -rc
> > > cycle. The others add new features for the reset, so they should wait until
> > > v3.3 merge window.
> > 
> > Most of the other patches in the branch discuss what they fix in the patch 
> > descriptions, rather than the subject lines.  Do you want the subject 
> > lines rewritten to highlight this?
> 
> Well we should only merge major bugs and regressions during the -rc cycle.
> It seems that at least "OMAP: HWMOD: Unify DSS resets for OMAPs" has high
> flame potential based on that? Maybe we should wait until the next merge
> window to start removing some of these workarounds?

I'm not sure what you mean...

The "Unify DSS resets..." patch itself fixes only a minor bug: "This
causes causes the HWMOD reset to fail for dss_dispc and dss_rfbi". But
it makes the next patch, "Ensure DSS works correctly...", possible. And
that patch fixes a bigger bug, DSS or the whole device hanging if the
bootloader had enabled the display (which I believe is the default on
the shipped beagles/pandas).

There's one patch in the series that is not a fix, though: "ARM: OMAP4:
HWMOD: remove extra clocks". I'm fine with queuing that for the next
merge window. But most of the patches have now somehow missed two merge
windows, and I'd really like to get them in.

 Tomi
Tony Lindgren - Nov. 19, 2011, 12:06 a.m.
* Tomi Valkeinen <tomi.valkeinen@ti.com> [111114 22:31]:
> On Mon, 2011-11-14 at 14:04 -0800, Tony Lindgren wrote:
> > * Paul Walmsley <paul@pwsan.com> [111114 13:05]:
> > > On Mon, 14 Nov 2011, Tony Lindgren wrote:
> > > 
> > > > I suggest we only merge the ones with "fix" in the subject during the -rc
> > > > cycle. The others add new features for the reset, so they should wait until
> > > > v3.3 merge window.
> > > 
> > > Most of the other patches in the branch discuss what they fix in the patch 
> > > descriptions, rather than the subject lines.  Do you want the subject 
> > > lines rewritten to highlight this?
> > 
> > Well we should only merge major bugs and regressions during the -rc cycle.
> > It seems that at least "OMAP: HWMOD: Unify DSS resets for OMAPs" has high
> > flame potential based on that? Maybe we should wait until the next merge
> > window to start removing some of these workarounds?
> 
> I'm not sure what you mean...
> 
> The "Unify DSS resets..." patch itself fixes only a minor bug: "This
> causes causes the HWMOD reset to fail for dss_dispc and dss_rfbi". But
> it makes the next patch, "Ensure DSS works correctly...", possible. And
> that patch fixes a bigger bug, DSS or the whole device hanging if the
> bootloader had enabled the display (which I believe is the default on
> the shipped beagles/pandas).
> 
> There's one patch in the series that is not a fix, though: "ARM: OMAP4:
> HWMOD: remove extra clocks". I'm fine with queuing that for the next
> merge window. But most of the patches have now somehow missed two merge
> windows, and I'd really like to get them in.

OK let's try to get them in then. Hopefully after these DSS will be
more independent from the core omap changes.

Tony