At Sun, 23 Aug 2009 11:16:57 +0100, Mark Brown wrote:

The following changes since commit a0a499c5792b8656cd51e11d5e0db9fb21640f58: Eduardo Valentin (1): ASoC: OMAP: Use DMA operating mode of McBSP

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound-2.6.git for-2.6.32

Hm, this causes merge errors regarding wm97xx battery driver. Also, when I merge with 2.6.31-rc7, it gives more merge errors in arch code. Do they depend on other tree changes like linux-arm? thanks, Takashi

Ben Dooks (4): ASoC: S3C24XX: Add audio core and tlv320aic23 for Simtec boards ASoC: tlv320aic3x: Change to use device model ASoC: tlv320aic3x: fixup board device changes ASoC: S3C24XX: Support for Simtec Hermes boards

Kuninori Morimoto (2): ASoC: Add ak4642/ak4643 codec support ASoC: Add FSI-AK4642 sound support for SuperH

Marek Vasut (5): power_supply: Make wm97xx-core pass platform_data to battery, ts power_supply: Convert WM97xx battery to use platform_data power_supply: Let wm97xx-battery use irq to detect charger state [ARM] pxa: Convert all platforms to pass platform_data to wm97xx battery ASoC: Pass correct platform data from pxa2xx-ac97

Mark Brown (1): ASoC: Add DAPM widget power decision debugfs files

Roel Kluin (1): ASoC: free socdev if init_card() fails in wm9705_soc_probe()

arch/arm/mach-pxa/mioa701.c | 10 +- arch/arm/mach-pxa/palmld.c | 19 +- arch/arm/mach-pxa/palmt5.c | 19 +- arch/arm/mach-pxa/palmte2.c | 20 +- arch/arm/mach-pxa/palmtx.c | 19 +- arch/arm/mach-pxa/palmz72.c | 20 +- arch/arm/mach-pxa/treo680.c | 1 - arch/arm/plat-s3c/include/plat/audio-simtec.h | 37 ++ drivers/input/touchscreen/wm97xx-core.c | 3 + drivers/power/wm97xx_battery.c | 56 ++- include/linux/wm97xx.h | 18 + include/linux/wm97xx_batt.h | 26 -- include/sound/soc-dapm.h | 1 + include/sound/soc.h | 1 + sound/arm/pxa2xx-ac97.c | 4 +- sound/soc/codecs/Kconfig | 4 + sound/soc/codecs/Makefile | 2 + sound/soc/codecs/ak4642.c | 502 ++++++++++++++++++++++++ sound/soc/codecs/ak4642.h | 20 + sound/soc/codecs/tlv320aic3x.c | 221 ++++++----- sound/soc/codecs/tlv320aic3x.h | 2 - sound/soc/codecs/wm9705.c | 2 +- sound/soc/davinci/davinci-evm.c | 12 +- sound/soc/omap/n810.c | 12 +- sound/soc/pxa/pxa2xx-ac97.c | 4 +- sound/soc/s3c24xx/Kconfig | 19 + sound/soc/s3c24xx/Makefile | 7 + sound/soc/s3c24xx/s3c24xx_simtec.c | 394 +++++++++++++++++++ sound/soc/s3c24xx/s3c24xx_simtec.h | 22 + sound/soc/s3c24xx/s3c24xx_simtec_hermes.c | 153 +++++++ sound/soc/s3c24xx/s3c24xx_simtec_tlv320aic23.c | 137 +++++++ sound/soc/s6000/s6105-ipcam.c | 12 +- sound/soc/sh/Kconfig | 8 + sound/soc/sh/Makefile | 2 + sound/soc/sh/fsi-ak4642.c | 107 +++++ sound/soc/soc-core.c | 8 + sound/soc/soc-dapm.c | 86 ++++ 37 files changed, 1784 insertions(+), 206 deletions(-) create mode 100644 arch/arm/plat-s3c/include/plat/audio-simtec.h delete mode 100644 include/linux/wm97xx_batt.h create mode 100644 sound/soc/codecs/ak4642.c create mode 100644 sound/soc/codecs/ak4642.h create mode 100644 sound/soc/s3c24xx/s3c24xx_simtec.c create mode 100644 sound/soc/s3c24xx/s3c24xx_simtec.h create mode 100644 sound/soc/s3c24xx/s3c24xx_simtec_hermes.c create mode 100644 sound/soc/s3c24xx/s3c24xx_simtec_tlv320aic23.c create mode 100644 sound/soc/sh/fsi-ak4642.c

At Sun, 23 Aug 2009 18:38:28 +0100, Mark Brown wrote:

On Sun, Aug 23, 2009 at 07:06:17PM +0200, Takashi Iwai wrote:

...

Hm, this causes merge errors regarding wm97xx battery driver.

Causes merge errors with which tree? It's a straight fast forward aginst topic/asoc.

Merging your changes into topic/asoc, then merging into master or for-next. So, it seems conflicting with the latest upstream.

...

Also, when I merge with 2.6.31-rc7, it gives more merge errors in arch code.

...

Do they depend on other tree changes like linux-arm?

I suspect they're at most only textual. In any case, this is all causing far too much trouble so I've just dropped all of the patches that are causing issues.

Marek, I suggest that you submit the patch to use IRQs against the power tree. I also suggest that you submit a version of the battery platform data changes against the power tree which avoids making the platforms instabuggy then when that's applied submit changes to Eric to use the new platform data, making it clear that the machine changes will need to wait for both power and ALSA trees to be merged during the merge window. This is the approach that's been suggested previously - it makes life much easier if changes go through the relevant trees.

The following changes since commit a0a499c5792b8656cd51e11d5e0db9fb21640f58: Eduardo Valentin (1): ASoC: OMAP: Use DMA operating mode of McBSP

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound-2.6.git for-2.6.32

Thanks, pulled now. Takashi

Ben Dooks (4): ASoC: S3C24XX: Add audio core and tlv320aic23 for Simtec boards ASoC: tlv320aic3x: Change to use device model ASoC: tlv320aic3x: fixup board device changes ASoC: S3C24XX: Support for Simtec Hermes boards

Kuninori Morimoto (2): ASoC: Add ak4642/ak4643 codec support ASoC: Add FSI-AK4642 sound support for SuperH

Marek Vasut (1): ASoC: Pass correct platform data from pxa2xx-ac97

Mark Brown (1): ASoC: Add DAPM widget power decision debugfs files

Roel Kluin (1): ASoC: free socdev if init_card() fails in wm9705_soc_probe()

arch/arm/plat-s3c/include/plat/audio-simtec.h | 37 ++ include/sound/soc-dapm.h | 1 + include/sound/soc.h | 1 + sound/arm/pxa2xx-ac97.c | 4 +- sound/soc/codecs/Kconfig | 4 + sound/soc/codecs/Makefile | 2 + sound/soc/codecs/ak4642.c | 502 ++++++++++++++++++++++++ sound/soc/codecs/ak4642.h | 20 + sound/soc/codecs/tlv320aic3x.c | 221 ++++++----- sound/soc/codecs/tlv320aic3x.h | 2 - sound/soc/codecs/wm9705.c | 2 +- sound/soc/davinci/davinci-evm.c | 12 +- sound/soc/omap/n810.c | 12 +- sound/soc/pxa/pxa2xx-ac97.c | 4 +- sound/soc/s3c24xx/Kconfig | 19 + sound/soc/s3c24xx/Makefile | 7 + sound/soc/s3c24xx/s3c24xx_simtec.c | 394 +++++++++++++++++++ sound/soc/s3c24xx/s3c24xx_simtec.h | 22 + sound/soc/s3c24xx/s3c24xx_simtec_hermes.c | 153 +++++++ sound/soc/s3c24xx/s3c24xx_simtec_tlv320aic23.c | 137 +++++++ sound/soc/s6000/s6105-ipcam.c | 12 +- sound/soc/sh/Kconfig | 8 + sound/soc/sh/Makefile | 2 + sound/soc/sh/fsi-ak4642.c | 107 +++++ sound/soc/soc-core.c | 8 + sound/soc/soc-dapm.c | 86 ++++ 26 files changed, 1661 insertions(+), 118 deletions(-) create mode 100644 arch/arm/plat-s3c/include/plat/audio-simtec.h create mode 100644 sound/soc/codecs/ak4642.c create mode 100644 sound/soc/codecs/ak4642.h create mode 100644 sound/soc/s3c24xx/s3c24xx_simtec.c create mode 100644 sound/soc/s3c24xx/s3c24xx_simtec.h create mode 100644 sound/soc/s3c24xx/s3c24xx_simtec_hermes.c create mode 100644 sound/soc/s3c24xx/s3c24xx_simtec_tlv320aic23.c create mode 100644 sound/soc/sh/fsi-ak4642.c

Dne Po 24. srpna 2009 09:37:16 Takashi Iwai napsal(a):

At Sun, 23 Aug 2009 18:38:28 +0100,

Mark Brown wrote:

...

On Sun, Aug 23, 2009 at 07:06:17PM +0200, Takashi Iwai wrote:

...

Hm, this causes merge errors regarding wm97xx battery driver.

Causes merge errors with which tree? It's a straight fast forward aginst topic/asoc.

Merging your changes into topic/asoc, then merging into master or for-next. So, it seems conflicting with the latest upstream.

Those patches were made against sound-2.6 from Mark ... I assume some additional stuff was pushed into the kernel after Mark updated his tree.

...

...

Also, when I merge with 2.6.31-rc7, it gives more merge errors in arch code.

Do they depend on other tree changes like linux-arm?

I suspect they're at most only textual. In any case, this is all causing far too much trouble so I've just dropped all of the patches that are causing issues.

Marek, I suggest that you submit the patch to use IRQs against the power tree.

Where would this one be ?

...

I also suggest that you submit a version of the battery platform data changes against the power tree which avoids making the platforms instabuggy then when that's applied submit changes to Eric to use the new platform data, making it clear that the machine changes will need to wait for both power and ALSA trees to be merged during the merge window. This is the approach that's been suggested previously - it makes life much easier if changes go through the relevant trees.

If those changes go through one tree, it'll make them consistent and avoid breakage (because if some of those changes go through the power tree, some through Erics tree ... all of those trees would be broken for some time).

...

The following changes since commit a0a499c5792b8656cd51e11d5e0db9fb21640f58: Eduardo Valentin (1): ASoC: OMAP: Use DMA operating mode of McBSP

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound-2.6.git for-2.6.32

Thanks, pulled now.

Takashi

I rebased those patches against your tree -- sound-2.6/master -- and attached is the patch that had problems with merging (fixed version). The other patches apply fine.

Dne Pá 28. srpna 2009 15:49:49 Mark Brown napsal(a):

On Fri, 2009-08-28 at 02:29 +0200, Marek Vasut wrote:

...

Dne Po 24. srpna 2009 09:37:16 Takashi Iwai napsal(a):

...

Merging your changes into topic/asoc, then merging into master or for-next. So, it seems conflicting with the latest upstream.

Those patches were made against sound-2.6 from Mark ... I assume some additional stuff was pushed into the kernel after Mark updated his tree.

Yes, there have been changes in mainline since the last time it was merged up onto the ASoC branch.

Yes, that part of patch that conflicted in Takashi's tree was the same one I had to fix for your tree iirc.

...

...

...

Marek, I suggest that you submit the patch to use IRQs against the power tree.

Where would this one be ?

MAINTAINERS says:

POWER SUPPLY CLASS/SUBSYSTEM and DRIVERS P: Anton Vorontsov M: cbou@mail.ru P: David Woodhouse M: dwmw2@infradead.org T: git git://git.infradead.org/battery-2.6.git S: Maintained F: include/linux/power_supply.h F: drivers/power/power_supply*

I think you noticed it's not very up-to-date (.31-rc1, last update from 30/07/2009).

...

...

...

I also suggest that you submit a version of the battery platform data changes against the power tree which avoids making the platforms instabuggy then when that's applied submit changes to Eric to use the new platform data, making it clear that the machine changes will need to wait for both power and ALSA trees to be merged during the merge window. This is the approach that's been suggested previously - it makes life much easier if changes go through the relevant trees.

If those changes go through one tree, it'll make them consistent and avoid breakage (because if some of those changes go through the power tree, some through Erics tree ... all of those trees would be broken for some time).

As previously explained these issues would be pretty much dealt with if you were to ensure that changing the battery driver did not require immediate updates to the boards. This is the standard technique used to do this sort of change which affects many trees and generally works pretty well. As things stand the arch/arm bit shouldn't be split from the drivers/power change to the platform data.

...

I rebased those patches against your tree -- sound-2.6/master -- and attached is the patch that had problems with merging (fixed version). The other patches apply fine.

Please also check that the resulting branch will merge down cleanly into -next - that's what they'll end up getting merged into.

Also, I've asked you several times to stop sending multiple patches in one e-mail.

I mailed only one.

As I've said before the fact that you are submitting patches like this means that they require a special manual workflow to apply. Have you tried using git format-patch and git send-email to send patches? I have suggested this to you several times without any response. These programs will take care of the overwhelming majority of the technical issues with your submissions - there are example configurations available for use with gmail. Yeah, I need to look into that ... it's not like Im ignoring it.