On Fri, Sep 23, 2016 at 10:02:23AM +0100, Jon Hunter wrote: > > On 23/09/16 09:14, Arnd Bergmann wrote: > > On Friday, September 23, 2016 9:14:42 AM CEST Vinod Koul wrote: > >> On Wed, Sep 21, 2016 at 11:57:54PM +0200, Arnd Bergmann wrote: > >>> On Wednesday, September 21, 2016 11:46:28 PM CEST Arnd Bergmann wrote: > >>>>> @@ -482,7 +482,7 @@ config TEGRA20_APB_DMA > >>>>> > >>>>> config TEGRA210_ADMA > >>>>> bool "NVIDIA Tegra210 ADMA support" > >>>>> - depends on ARCH_TEGRA_210_SOC > >>>>> + depends on ARCH_TEGRA_210_SOC || (ARM && COMPILE_TEST) > >>>>> select DMA_ENGINE > >>>>> select DMA_VIRTUAL_CHANNELS > >>>>> select PM_CLK > >>>> > >>>> I think we have to drop the 'select PM_CLK' here: > >>>> > >>>> warning: (ARM_GIC_PM && TEGRA_ACONNECT && TEGRA210_ADMA) selects PM_CLK which has unmet direct dependencies (PM && HAVE_CLK) > >>>> 0xA55EBB4C Wed Sep 21 18:31:56 CEST 2016 failed > >>>> drivers/base/power/clock_ops.o: In function `pm_clk_create': > >>>> clock_ops.c:(.text.pm_clk_create+0x0): undefined reference to `dev_pm_get_subsys_data' > >>>> drivers/base/power/clock_ops.o: In function `pm_clk_destroy': > >>>> clock_ops.c:(.text.pm_clk_destroy+0xb8): undefined reference to `dev_pm_put_subsys_data' > >>>> drivers/base/power/clock_ops.o: In function `pm_clk_notify': > >>>> clock_ops.c:(.text.pm_clk_notify+0x44): undefined reference to `dev_pm_get_subsys_data' > >>>> /git/arm-soc/Makefile:972: recipe for target 'vmlinux' failed > >>>> > >>>> Basically you can never 'select' that symbol unless it's > >>>> already enabled, so there is no harm in removing that line. > >>>> > >>> > >>> No, it actually needs to be 'depends on PM && HAVE_CLK'. Still testing > >>> if that is sufficient. > >> > >> Can you confirm if that's sufficient. I would like to merge the series.. > >> > > > > > > Yes, I've built a few hundred randconfigs overnight, everything succeeds with > > that now. > > Please note that 'depends on PM && HAVE_CLK' is equivalent to just > 'depends on PM_CLK' and so why not use this? This is what I was > suggesting as an alternative to the fix for patch 26/30 [0]. That's true, I am comitting with 'depends on PM_CLK' now. Since the other patch doesn't seem to be required, am dropping that one. Thanks -- ~Vinod -- To unsubscribe from this list: send the line "unsubscribe dmaengine" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
