| Message ID | 20170222022929.10540-5-afaerber@suse.de (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Feb 22, 2017 at 03:29:24AM +0100, Andreas Färber wrote: This is patch 4 of 8, what's going on with dependencies here? Please use subject lines matching the style for the subsystem. This makes it easier for people to identify relevant patches. > config SND_MMP_SOC > - bool "Soc Audio for Marvell MMP chips" > - depends on ARCH_MMP > + bool "SoC Audio for Marvell MMP chips" > + depends on ARCH_MMP && ARM There are no obvious build time architecture dependencies for this driver, shouldn't we have an || COMPILE_TEST in here?
Am 22.02.2017 um 19:47 schrieb Mark Brown: > On Wed, Feb 22, 2017 at 03:29:24AM +0100, Andreas Färber wrote: > > This is patch 4 of 8, what's going on with dependencies here? No dependencies on previous patches. 5/8 adds ARCH_MMP to arm64. It compiled without this (just Kconfig warnings), so this could go through the appropriate sound tree as a cleanup. > Please use subject lines matching the style for the subsystem. This > makes it easier for people to identify relevant patches. Should I make it "ASoC: pxa: ..." then? SND_MMP_SOC selects SND_ARM, which in sound/arm/Kconfig says 'Drivers that are implemented on ASoC can be found in "ALSA for SoC audio support" section.', which confused me about whether or not this is an ASoC compliant driver if it selects a non-ASoC symbol. >> config SND_MMP_SOC >> - bool "Soc Audio for Marvell MMP chips" >> - depends on ARCH_MMP >> + bool "SoC Audio for Marvell MMP chips" >> + depends on ARCH_MMP && ARM > > There are no obvious build time architecture dependencies for this > driver, shouldn't we have an || COMPILE_TEST in here? I'm the wrong person to answer that - the symbol relies on architecture-specific SND_ARM, and my understanding was that the code still relies on platform data, therefore it did not seem worthwhile to just disable inapplicable dependencies such as MMP_SRAM (arch/arm/mach-mmp/; superseded by DT mmio-sram) or SND_ARM. Regards, Andreas
On Thu, 23 Feb 2017 15:11:46 +0100, Andreas Färber wrote: > > Am 22.02.2017 um 19:47 schrieb Mark Brown: > > On Wed, Feb 22, 2017 at 03:29:24AM +0100, Andreas Färber wrote: > > > > This is patch 4 of 8, what's going on with dependencies here? > > No dependencies on previous patches. 5/8 adds ARCH_MMP to arm64. > > It compiled without this (just Kconfig warnings), so this could go > through the appropriate sound tree as a cleanup. > > > Please use subject lines matching the style for the subsystem. This > > makes it easier for people to identify relevant patches. > > Should I make it "ASoC: pxa: ..." then? SND_MMP_SOC selects SND_ARM, > which in sound/arm/Kconfig says 'Drivers that are implemented on ASoC > can be found in "ALSA for SoC audio support" section.', which confused > me about whether or not this is an ASoC compliant driver if it selects a > non-ASoC symbol. It's a bug that selects SND_ARM there. I guess this was wrongly taken from the old example of AC97 ARM SoC code, which was fixed later. ASoC stuff is basically irrelevant from the legacy sound/arm/*. thanks, Takashi
diff --git a/sound/soc/pxa/Kconfig b/sound/soc/pxa/Kconfig index 823b5a2..f8e1c06 100644 --- a/sound/soc/pxa/Kconfig +++ b/sound/soc/pxa/Kconfig @@ -8,8 +8,8 @@ config SND_PXA2XX_SOC to select the audio interfaces to support below. config SND_MMP_SOC - bool "Soc Audio for Marvell MMP chips" - depends on ARCH_MMP + bool "SoC Audio for Marvell MMP chips" + depends on ARCH_MMP && ARM select MMP_SRAM select SND_SOC_GENERIC_DMAENGINE_PCM select SND_ARM
SND_MMP_SOC selects SND_ARM, which is limited to ARM, as well as MMP_SRAM, which is a non-DT driver in arch/arm/mach-arm/. Avoid Kconfig warnings by not making this driver available on ARM64. While at it, tidy the title (Soc -> SoC). Signed-off-by: Andreas Färber <afaerber@suse.de> --- v2: new sound/soc/pxa/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)