ci/db410c: Fix networking so we get artifacts from our jobs.
authorEric Anholt <eric@anholt.net>
Sat, 7 Nov 2020 17:32:02 +0000 (09:32 -0800)
committerMarge Bot <eric+marge@anholt.net>
Wed, 11 Nov 2020 17:22:47 +0000 (17:22 +0000)
Seems some sort of linux change (bugfix?) resulted in the db410cs
selecting device mode for the db410cs due to the micro cable being
plugged in (fastboot runs them in device mode), so we weren't finding
the network and getting artifacts out.

Closes: #3728
Acked-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6971>

.gitlab-ci.yml
.gitlab-ci/container/lava_build.sh

index ff9d273..1081793 100644 (file)
@@ -414,7 +414,7 @@ arm64_test:
   extends:
     - .use-arm_test-base
   variables:
-    FDO_DISTRIBUTION_TAG: &arm64_test "2020-11-09-deqp-version"
+    FDO_DISTRIBUTION_TAG: &arm64_test "2020-11-09-db410c-net"
 
 .use-arm64_test:
   variables:
index ecb160d..e42363e 100755 (executable)
@@ -158,6 +158,10 @@ if [ -n "$INSTALL_KERNEL_MODULES" ]; then
     sed -i 's/=m/=n/g' ${DEFCONFIG}
 fi
 
+# Force db410c to host mode instead of OTG (which is otherwise selected by
+# default due to our micro cable for fastboot)
+sed -i 's/dr_mode = "otg"/dr_mode = "host"/' arch/arm64/boot/dts/qcom/apq8016-sbc.dtsi
+
 ./scripts/kconfig/merge_config.sh ${DEFCONFIG} ../.gitlab-ci/${KERNEL_ARCH}.config
 make ${KERNEL_IMAGE_NAME}
 for image in ${KERNEL_IMAGE_NAME}; do