From mboxrd@z Thu Jan 1 00:00:00 1970 From: jic23@cam.ac.uk (Jonathan Cameron) Date: Tue, 09 Mar 2010 12:19:04 +0000 Subject: pxa: Couple of build issues (missing dependencies) In-Reply-To: References: <4B953A83.7020705@cam.ac.uk> Message-ID: <4B963CB8.1070206@cam.ac.uk> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 03/09/10 03:12, Eric Miao wrote: > Correct Ian Molton's email address. > > On Tue, Mar 9, 2010 at 11:05 AM, Eric Miao wrote: >> On Tue, Mar 9, 2010 at 1:57 AM, Jonathan Cameron wrote: >>> Whilst testing the uncompress patch just sent, I tried a full build >>> of pxa platforms by just enabling them along side my normal couple. >>> This is more or less a vanilla tree as of earlier today (+ a couple >>> of uncompress related patches from Eric's tree). I haven't chased >>> these down terribly carefully but thought I might as well bring them >>> to the lists attention. >>> >>> Sharpsl_pm calls max1111_read_channel if CONFIG_CORGI_SSP_DEPRECATED >>> is not set. There is no dependancy to enforce this being available. >>> (none obvious in the pxa devel tree either). >>> >> >> diff --git a/arch/arm/mach-pxa/Kconfig b/arch/arm/mach-pxa/Kconfig >> index 38fbd0a..912bdfd 100644 >> --- a/arch/arm/mach-pxa/Kconfig >> +++ b/arch/arm/mach-pxa/Kconfig >> @@ -454,6 +454,13 @@ config PXA_SHARPSL >> config SHARPSL_PM >> bool >> select APM_EMULATION >> + select SHARPSL_PM_MAX1111 >> + >> +config SHARPSL_PM_MAX1111 >> + bool >> + depends on !CORGI_SSP_DEPRECATED >> + select HWMON >> + select SENSORS_MAX1111 >> >> config CORGI_SSP_DEPRECATED >> bool >> >> >> An ugly patch above should address this problem, but a real fix would >> be to have a sharpsl-battery.c driver. >> >>> >>> FB_W100 is selected by MACH_E740 leading to the following which >>> I would guess indicate some missing dependencies as well. >>> Looks like it doesn't actually drag in frame buffer support as I'd >>> imagine was intended: >> >> MACH_E740 should not have selected that option, as user has the >> flexibility to include support of FB_W100 or not. Apparently, the build >> error is caused by FB_W100 being selected, while FB is made to >> module. >> >> I'm yet not sure why MACH_E740 mandates FB_W100, and as Ian is >> no longer active on this series, I'm going to drop this 'select'. >> >> As well as for MACH_HIMALAYA. Zbynek CC'ed. >> >> diff --git a/arch/arm/mach-pxa/Kconfig b/arch/arm/mach-pxa/Kconfig >> index 912bdfd..5b6ee46 100644 >> --- a/arch/arm/mach-pxa/Kconfig >> +++ b/arch/arm/mach-pxa/Kconfig >> @@ -272,7 +272,6 @@ config MACH_H5000 >> config MACH_HIMALAYA >> bool "HTC Himalaya Support" >> select CPU_PXA26x >> - select FB_W100 >> >> config MACH_MAGICIAN >> bool "Enable HTC Magician Support" >> @@ -554,7 +553,6 @@ config MACH_E740 >> bool "Toshiba e740" >> default y >> depends on ARCH_PXA_ESERIES >> - select FB_W100 >> help >> Say Y here if you intend to run this kernel on a Toshiba >> e740 family PDA. >> @@ -563,7 +561,6 @@ config MACH_E750 >> bool "Toshiba e750" >> default y >> depends on ARCH_PXA_ESERIES >> - select FB_W100 >> help >> Say Y here if you intend to run this kernel on a Toshiba >> e750 family PDA. >> @@ -580,7 +577,6 @@ config MACH_E800 >> bool "Toshiba e800" >> default y >> depends on ARCH_PXA_ESERIES >> - select FB_W100 >> help >> Say Y here if you intend to run this kernel on a Toshiba >> e800 family PDA. >> Both look fine to me. Acked-by: Jonathan Cameron Thanks for sorting out the patches. Jonathan