From mboxrd@z Thu Jan 1 00:00:00 1970 From: Heiko =?ISO-8859-1?Q?St=FCbner?= Subject: Re: [PATCH v14 0/3] Add drm driver for Rockchip Socs Date: Thu, 27 Nov 2014 23:17:54 +0100 Message-ID: <3170288.yzWC52s6pC@diego> References: <1416447994-9921-1-git-send-email-mark.yao@rock-chips.com> <5476CDF7.60208@rock-chips.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: Sender: linux-doc-owner@vger.kernel.org To: Daniel Kurtz Cc: Mark yao , Arnd Bergmann , Dave Airlie , Joerg Roedel , Boris BREZILLON , Rob Clark , Daniel Vetter , Rob Herring , Pawel Moll , Mark Rutland , Ian Campbell , Kumar Gala , Randy Dunlap , Grant Likely , Greg Kroah-Hartman , John Stultz , Rom Lemarchand , "devicetree@vger.kernel.org" , linux-doc@vger.kernel.org, LKML , dri-devel , linux-api@vger.kernel.org, "open list:ARM/Rockchip SoC..." List-Id: linux-api@vger.kernel.org Am Donnerstag, 27. November 2014, 14:05:00 schrieb Daniel Kurtz: > On Thu, Nov 27, 2014 at 2:08 AM, Mark yao w= rote: > > On 2014=E5=B9=B411=E6=9C=8827=E6=97=A5 10:12, Dave Airlie wrote: > >>> Hi Dave > >>>=20 > >>> Do you mean that I need send you a branch, based on drm-next= , merge > >>> with > >>>=20 > >>> iommu tree and rockchip drm? > >>=20 > >> Yes, grab drm-next, git pull the arm/rockchip branch from Joerg's > >> tree, put rockchip drm > >> patches on top, send me pull request. > >>=20 > >> I'll validate it then. > >>=20 > >> Dave. > >=20 > > Hi Dave > >=20 > > I have send a pull request to you, with Joerg's iommu arm/rockc= hip > > branch. > > =20 > > I got a same problem when use "make multi_v7_defconfig" as Heik= o said: > > drivers/video/fbdev/Kconfig:5:error: recursive dependency > > detected! > > drivers/video/fbdev/Kconfig:5: symbol FB is selected by > > DRM_KMS_FB_HELPER drivers/gpu/drm/Kconfig:34: symbol > > DRM_KMS_FB_HELPER is selected by DRM_ROCKCHIP > > drivers/gpu/drm/rockchip/Kconfig:1: symbol DRM_ROCKCHIP > > depends on ARM_DMA_USE_IOMMU arch/arm/Kconfig:95: symbol > > ARM_DMA_USE_IOMMU is selected by VIDEO_OMAP3 > > drivers/media/platform/Kconfig:96: symbol VIDEO_OMAP3 de= pends > > on VIDEO_V4L2 drivers/media/v4l2-core/Kconfig:6: symbol > > VIDEO_V4L2 depends on I2C drivers/i2c/Kconfig:7: symbol = I2C is > > selected by FB_DDC > > drivers/video/fbdev/Kconfig:59: symbol FB_DDC is selecte= d by > > FB_CYBER2000_DDC drivers/video/fbdev/Kconfig:374: symbol > > FB_CYBER2000_DDC depends on FB_CYBER2000 > > drivers/video/fbdev/Kconfig:362: symbol FB_CYBER2000 dep= ends > > on FB> =20 > > I was confused how to solve the recursive dependency, remove de= pends > > on ARM_DMA_USE_IOMMU & IOMMU_API? > The "depends on ARM_DMA_USE_IOMMU & IOMMU_API" was suggested by Arnd > Bergmann during code review (originally they were selected). >=20 > Removing them definitely fixes the dependency recursion. > Also, since they are both already selected by ROCKCHIP_IOMMU, > everything will build correctly. > So, this sounds good to me, but I am no expert on Kconfig. personally I'd think the omap v4l stuff is at fault here. It is selecti= ng=20 ARM_DMA_USE_IOMMU and OMAP_IOMMU instead of depending on them. And OMAP_IOMMU is a regular driver option which can be selected through= =20 Kconfig, so I don't think it should get selected but enabled by choice. It would be interesting to know if the Exynos drm is also affected by t= his, as=20 it is using a similar depency-structure by depending on the iommu and=20 ARM_DMA_USE_IOMMU. Heiko