mesa-pvr: correct INSANE_SKIP for mesa-megadriver

In the original file this was copied from, $PN == mesa. But in this
context, $PN == mesa-pvr, while the PACKAGES and FILES logic still
explicitly refers to the mesa-megadriver package. Hence the build
fails

ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: QA Issue: non -dev/-dbg/nativesdk- package mesa-megadriver contains symlink .so '/usr/lib/dri/swrast_dri.so' [dev-so]
ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: QA Issue: non -dev/-dbg/nativesdk- package mesa-megadriver contains symlink .so '/usr/lib/dri/virtio_gpu_dri.so' [dev-so]
ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: QA Issue: non -dev/-dbg/nativesdk- package mesa-megadriver contains symlink .so '/usr/lib/dri/kms_swrast_dri.so' [dev-so]
ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: QA Issue: non -dev/-dbg/nativesdk- package mesa-megadriver contains symlink .so '/usr/lib/dri/zink_dri.so' [dev-so]
ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: Fatal QA errors were found, failing task.
ERROR: Logfile of failure stored in: [...]/mesa-pvr/25.2.8/temp/log.do_package_qa.3911784
ERROR: Task ([...]/meta-ti/meta-ti-bsp/recipes-graphics/mesa/mesa-pvr_25.2.8.bb:do_package_qa) failed with exit code '1'

Spell the package override correctly for the INSANE_SKIP suppression
to have effect.

Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
---
 meta-ti-bsp/recipes-graphics/mesa/mesa-pvr-25.inc | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

meta-ti / na / 20260717124847.2123825-1-ravi

PRC Results: PASS

=========================================================
  check-yocto-patches: PASS
=========================================================
Patches
----------------------------------------
WARN - [PATCH] mesa-pvr: correct INSANE_SKIP for mesa-megadriver
    WARN: Missing branches specifier [master/scarthgap/XXXXX]. (META-2)
        patch:43
            Subject: [PATCH] mesa-pvr: correct INSANE_SKIP for mesa-megadriver
    
    WARN: Commit message does not include file/recipe name: mesa-pvr: correct INSANE_SKIP for mesa-megadriver. (COMMIT-MESSAGE-2)
        patch
    
    For details on the above errors/warnings visit: https://lists.yoctoproject.org/g/meta-ti/wiki/40887



=========================================================
  apply-yocto-patch: PASS
=========================================================
master
=====================
Summary:
- Patch Series: [PATCH] mesa-pvr: correct INSANE_SKIP for mesa-megadriver
- Submitter: From: Rasmus Villemoes <ravi@prevas.dk>
- Date: Date: Fri, 17 Jul 2026 14:48:47 +0200
- Num Patches: 1
- Mailing List (public inbox) Commit SHA: a291efa408b21134a86fa9d228db103b580b9db9

Applied to:
- Repository: lcpd-prc-meta-ti
- Base Branch: master-wip
- Commit Author: Ryan Eatmon <reatmon@ti.com>
- Commit Subject: XXX: Temp update to ti-linux-fw branch and SRCREV
- Commit SHA: ba3ac1c0a1753bfc6dcecc132bf7de38851673aa

Patches
----------------------------------------
All patches applied



=========================================================
  check-yocto-repo: PASS
=========================================================
master
=====================
PASS



=========================================================
  yocto-check-layers: PASS
=========================================================
master - PASS
=====================
All checks passed

On 7/17/2026 7:48 AM, Rasmus Villemoes via lists.yoctoproject.org wrote:
> In the original file this was copied from, $PN == mesa. But in this
> context, $PN == mesa-pvr, while the PACKAGES and FILES logic still
> explicitly refers to the mesa-megadriver package. Hence the build
> fails
> 

Good catch.  I wonder why we are not seeing these QA errors in our 
builds.  Are you turning something special on?


> ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: QA Issue: non -dev/-dbg/nativesdk- package mesa-megadriver contains symlink .so '/usr/lib/dri/swrast_dri.so' [dev-so]
> ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: QA Issue: non -dev/-dbg/nativesdk- package mesa-megadriver contains symlink .so '/usr/lib/dri/virtio_gpu_dri.so' [dev-so]
> ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: QA Issue: non -dev/-dbg/nativesdk- package mesa-megadriver contains symlink .so '/usr/lib/dri/kms_swrast_dri.so' [dev-so]
> ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: QA Issue: non -dev/-dbg/nativesdk- package mesa-megadriver contains symlink .so '/usr/lib/dri/zink_dri.so' [dev-so]
> ERROR: mesa-pvr-2_25.2.8-r1 do_package_qa: Fatal QA errors were found, failing task.
> ERROR: Logfile of failure stored in: [...]/mesa-pvr/25.2.8/temp/log.do_package_qa.3911784
> ERROR: Task ([...]/meta-ti/meta-ti-bsp/recipes-graphics/mesa/mesa-pvr_25.2.8.bb:do_package_qa) failed with exit code '1'
> 
> Spell the package override correctly for the INSANE_SKIP suppression
> to have effect.
> 
> Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
> ---
>   meta-ti-bsp/recipes-graphics/mesa/mesa-pvr-25.inc | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/meta-ti-bsp/recipes-graphics/mesa/mesa-pvr-25.inc b/meta-ti-bsp/recipes-graphics/mesa/mesa-pvr-25.inc
> index 71e9895b..2ec48c48 100644
> --- a/meta-ti-bsp/recipes-graphics/mesa/mesa-pvr-25.inc
> +++ b/meta-ti-bsp/recipes-graphics/mesa/mesa-pvr-25.inc
> @@ -331,7 +331,7 @@ FILES:${PN}-tools = "${bindir} ${datadir}"
>   ALLOW_EMPTY:${PN}-tools = "1"
>   
>   # All DRI drivers are symlinks to libdril_dri.so
> -INSANE_SKIP:${PN}-megadriver += "dev-so"
> +INSANE_SKIP:mesa-megadriver += "dev-so"
>   
>   # OpenCL ICDs package also ship correspondig .so files, there is no -dev package
>   INSANE_SKIP:libopencl-mesa += "dev-so"
> 
> 
> 
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#20158): https://lists.yoctoproject.org/g/meta-ti/message/20158
> Mute This Topic: https://lists.yoctoproject.org/mt/120313933/6551054
> Group Owner: meta-ti+owner@lists.yoctoproject.org
> Unsubscribe: https://lists.yoctoproject.org/g/meta-ti/unsub [reatmon@ti.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>

On Fri, Jul 17 2026, Ryan Eatmon <reatmon@ti.com> wrote:

> On 7/17/2026 7:48 AM, Rasmus Villemoes via lists.yoctoproject.org wrote:
>> In the original file this was copied from, $PN == mesa. But in this
>> context, $PN == mesa-pvr, while the PACKAGES and FILES logic still
>> explicitly refers to the mesa-megadriver package. Hence the build
>> fails
>> 
>
> Good catch.  I wonder why we are not seeing these QA errors in our
> builds.  Are you turning something special on?

No, not that I know of. I hit this immediately when trying to update
from whinlatter to wrynose, but I don't think we have any special flags
set anywhere that should increase bitbake warning/pedantic level or
something like that.

Rasmus

On Mon, Jul 20, 2026 at 09:14:42AM +0200, Rasmus Villemoes via lists.yoctoproject.org wrote:
> On Fri, Jul 17 2026, Ryan Eatmon <reatmon@ti.com> wrote:
> 
> > On 7/17/2026 7:48 AM, Rasmus Villemoes via lists.yoctoproject.org wrote:
> >> In the original file this was copied from, $PN == mesa. But in this
> >> context, $PN == mesa-pvr, while the PACKAGES and FILES logic still
> >> explicitly refers to the mesa-megadriver package. Hence the build
> >> fails
> >> 
> >
> > Good catch.  I wonder why we are not seeing these QA errors in our
> > builds.  Are you turning something special on?
> 
> No, not that I know of. I hit this immediately when trying to update
> from whinlatter to wrynose, but I don't think we have any special flags
> set anywhere that should increase bitbake warning/pedantic level or
> something like that.

It's not about warning or pedantic level. It's rather about how you have 
mesa and the entire distro configured. Because we don't have any of those 
4 libraries in /usr/lib/dri/ being built, as we only build for Wayland, but 
not X11.

PACKAGECONFIG="gallium video-codecs vulkan wayland opengl egl gles gbm virgl opencl libclc gallium-llvm zink xmlconfig gallium-llvm"
DISTRO_FEATURES="pam opencl systemd usrmerge 3g acl alsa bluetooth debuginfod ext2 gobject-introspection-data ipv4 ipv6 ldconfig multiarch nfc nfs opengl pci pulseaudio seccomp vfat vulkan wayland wifi xattr zeroconf"
DISTRO_FEATURES_OPTED_OUT="ptest x11"

Might want to compare these to your settings.

Anyway, your fix is correct - this is just an explanation for why we are not 
seeing the error in our builds.