| Message ID | 1517318506-14664-1-git-send-email-geert@linux-m68k.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Tue, Jan 30, 2018 at 02:21:46PM +0100, Geert Uytterhoeven wrote: > If NO_DMA=y: > > ERROR: "bad_dma_ops" [drivers/iio/adc/stm32-dfsdm-adc.ko] undefined! > > Add a dependency on HAS_DMA to fix this. Once again I have to ask why we're doing this per driver and not having the few architectures that have this issue set something up.
Hi Mark, On Tue, Jan 30, 2018 at 4:03 PM, Mark Brown <broonie@kernel.org> wrote: > On Tue, Jan 30, 2018 at 02:21:46PM +0100, Geert Uytterhoeven wrote: >> If NO_DMA=y: >> >> ERROR: "bad_dma_ops" [drivers/iio/adc/stm32-dfsdm-adc.ko] undefined! >> >> Add a dependency on HAS_DMA to fix this. > > Once again I have to ask why we're doing this per driver and not having > the few architectures that have this issue set something up. Because doing so would lead us to depends on COMPILE_TEST || HAS_DMA in many places. Is that any better? I agree we could indeed use DMA dummies in cases where a driver uses: depends on ARCH_FOO || ARCH_BAR || COMPILE_TEST depends on HAS_DMA But we also have generic options like: config ATA_BMDMA bool "ATA BMDMA support" depends on HAS_DMA And even architecture-specific drivers for SoCs that may or not have DMA: config HASH_DEV_STM32 tristate "Support for STM32 hash accelerators" depends on ARCH_STM32 depends on HAS_DMA Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
On Tue, Jan 30, 2018 at 04:31:24PM +0100, Geert Uytterhoeven wrote: > On Tue, Jan 30, 2018 at 4:03 PM, Mark Brown <broonie@kernel.org> wrote: > > Once again I have to ask why we're doing this per driver and not having > > the few architectures that have this issue set something up. > Because doing so would lead us to > depends on COMPILE_TEST || HAS_DMA > in many places. Is that any better? Honestly I'd just not have any dependency here, I'm not sure the DMA dependencies are really covering a good percentage of what's really likely to be found on these platforms.
diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig index 39e3b345a6c84033..d976c64107c8e9fd 100644 --- a/drivers/iio/adc/Kconfig +++ b/drivers/iio/adc/Kconfig @@ -683,6 +683,7 @@ config STM32_DFSDM_CORE config STM32_DFSDM_ADC tristate "STMicroelectronics STM32 dfsdm adc" depends on (ARCH_STM32 && OF) || COMPILE_TEST + depends on HAS_DMA select STM32_DFSDM_CORE select REGMAP_MMIO select IIO_BUFFER_HW_CONSUMER
If NO_DMA=y: ERROR: "bad_dma_ops" [drivers/iio/adc/stm32-dfsdm-adc.ko] undefined! Add a dependency on HAS_DMA to fix this. Fixes: e2e6771c64625165 ("IIO: ADC: add STM32 DFSDM sigma delta ADC support") Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> --- drivers/iio/adc/Kconfig | 1 + 1 file changed, 1 insertion(+)