On Fri, Nov 26, 2021 at 05:27:24PM +0100, Thomas Huth wrote:
The device-crash-test script has been quite neglected in the past,
so that it bit-rot quite often. Let's add CI jobs that run this
script for at least some targets, so that this script does not
regress that easily anymore.
Signed-off-by: Thomas Huth <thuth@redhat.com>
---
.gitlab-ci.d/buildtest.yml | 23 +++++++++++++++++++++++
1 file changed, 23 insertions(+)
diff --git a/.gitlab-ci.d/buildtest.yml b/.gitlab-ci.d/buildtest.yml
index 71d0f407ad..7e1cb0b3c2 100644
--- a/.gitlab-ci.d/buildtest.yml
+++ b/.gitlab-ci.d/buildtest.yml
@@ -100,6 +100,17 @@ avocado-system-debian:
IMAGE: debian-amd64
MAKE_CHECK_ARGS: check-avocado
+crash-test-debian:
+ extends: .native_test_job_template
+ needs:
+ - job: build-system-debian
+ artifacts: true
+ variables:
+ IMAGE: debian-amd64
+ script:
+ - cd build
+ - scripts/device-crash-test -q ./qemu-system-i386
+
build-system-fedora:
extends: .native_build_job_template
needs:
@@ -134,6 +145,18 @@ avocado-system-fedora:
IMAGE: fedora
MAKE_CHECK_ARGS: check-avocado
+crash-test-fedora:
+ extends: .native_test_job_template
+ needs:
+ - job: build-system-fedora
+ artifacts: true
+ variables:
+ IMAGE: fedora
+ script:
+ - cd build
+ - scripts/device-crash-test -q ./qemu-system-ppc
+ - scripts/device-crash-test -q ./qemu-system-riscv32
I'm curious why you picked all the 32-bit architecture targets to test
this for, since 95% of usage these days will be on the 64-bit targets ?
I guess both 32-bit & 64-bit targets probably have the same set of
devices built mostly, but still wondering if there was a reason for
your choice.