On Wed, Dec 30, 2020 at 9:35 AM Liu, Yongxin <yongxin....@windriver.com> wrote: > > > > > -----Original Message----- > > From: linux-yocto@lists.yoctoproject.org <linux- > > yo...@lists.yoctoproject.org> On Behalf Of Zhang, Qiang > > Sent: Wednesday, December 30, 2020 4:24 PM > > To: Bruce Ashfield <bruce.ashfi...@gmail.com> > > Cc: linux-yocto@lists.yoctoproject.org > > Subject: [linux-yocto] 回复: Question on i915: adust helper to 'm' > > > > > > > > ________________________________________ > > 发件人: Bruce Ashfield <bruce.ashfi...@gmail.com> > > 发送时间: 2020年12月23日 0:08 > > 收件人: Zhang, Qiang > > 抄送: linux-yocto@lists.yoctoproject.org > > 主题: Re: Question on i915: adust helper to 'm' > > > > On Tue, Dec 22, 2020 at 1:06 AM Zhang, Qiang <qiang.zh...@windriver.com> > > wrote: > > > > > > Hello Bruce > > > > > > about commit 27ad9c754f20bdc59d281e6bc7c9b2a96bbed0ee > > > There is warning here: > > > > > > > > > [NOTE]: 'CONFIG_DRM_KMS_HELPER' last val (m) and .config val (y) do not > > match > > > [INFO]: CONFIG_DRM_KMS_HELPER : y ## .config: > > 4070 :configs/standard/intel-x86/features/i915/i915.cfg (m) > > > [INFO]: raw config text: > > > > > > config DRM_KMS_HELPER > > > tristate > > > depends on DRM && HAS_IOMEM > > > help > > > CRTC helpers for KMS drivers. > > > > > > Config 'DRM_KMS_HELPER' has the following Direct dependencies > > (DRM_KMS_HELPER=y): > > > DRM(=y) && HAS_IOMEM(=y) > > > Parent dependencies are: > > > HAS_IOMEM [y] DRM [y] > > > [INFO]: config 'CONFIG_DRM_KMS_HELPER' was set, but it wasn't > > > assignable, check (parent) dependencies > > > > > > [INFO]: selection details for 'CONFIG_DRM_KMS_HELPER': > > > Symbols currently y-selecting this symbol: > > > - DRM_FBDEV_EMULATION > > > - DRM_VIRTIO_GPU > > > > > > Symbols currently m-selecting this symbol: > > > - DRM_RADEON > > > - DRM_NOUVEAU > > > - DRM_I915 > > > - DRM_MGAG200 > > > > > > > > > I find this will trigger Warning: > > > > > > Symbol: DRM_FBDEV_EMULATION [=y] > > > Type : bool > > > Selects: DRM_KMS_HELPER [=y] && DRM_KMS_FB_HELPER [=y] > > > > > > > >You need to override the settings for that in the relevant BSPs. > > > > > >I just sent another fix to the tools that won't always show the > > >warning, but the reality is that those settings are used by multiple > > >kernel types and we end up with different values based on BSPs >and the > > >kernel type. > > > > > >So to really silence the warning, we should set the value in the BSP, > > >since it knows the proper setting. That was always the design of the > > >base values -> BSP, we are only now finally getting the right >warnings > > >from the tools. > > > > > >Bruce > > > > Hello Bruce > > > > this warnning happend in intel-x86-64 bsp, DRM_FBDEV_EMULATION=y will > > select DRM_KMS_HELPER=y, and depend CONFIG_DRM=y and type is bool, > > > > I set CONFIG_DRM =m in x86-64 bsp, this DRM_KMS_HELPER warning, but trigger > > another warning about DRM_VIRTIO_GPU. > > I want to know what you suggest > > If I understand Bruce correctly, you can add intel-x86-drm-overrides.cfg in > intel-x86 > bsp to overwrite CONFIG_DRM_KMS_HELPER in features/i915/i915.cfg.
Correct. The issue with some of these values is that they are selected differently in the various kernel types (standard vs tiny vs rt), so the base configuration value cannot always match the final one in the .config in all configurations. The warning is a good thing, in that it alerts us to the fact that we aren't getting that base fragments value. If the warned value is correct (and it almost always it), you can just set that value in your BSPs fragments and the last value will now match the .config value and the warning goes away. Alternatively, you have to re-work the fragments and split them, adjusting all the dependency values to get the warned option to the value you want. As you saw, this can lead to other warnings which you have to correct. (I did an example of this for the yocto summit kernel lab, but we never did present it a few months ago). (There is also a 3rd option of simply declaring the option as "non-hw" in your BSP (or the base fragment), but you need the very latest tools for that to work .. I wouldn't recommend this option, but it is sometimes appropriate). If you are STILL seeing warning after those efforts, if you send me the exact configuration, I'll do some local runs and see if there's a bug that has slipped in. Bruce > > Thanks, > Yongxin > > > > > > > Qiang > > > > > Thanks > > > Qiang > > > > > > > > -- > > - Thou shalt not follow the NULL pointer, for chaos and madness await thee > > at its end > > - "Use the force Harry" - Gandalf, Star Trek II -- - Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end - "Use the force Harry" - Gandalf, Star Trek II
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#9325): https://lists.yoctoproject.org/g/linux-yocto/message/9325 Mute This Topic: https://lists.yoctoproject.org/mt/79312804/21656 Group Owner: linux-yocto+ow...@lists.yoctoproject.org Unsubscribe: https://lists.yoctoproject.org/g/linux-yocto/unsub [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
