ci/traces: Make sure we have no pre-existing traces-db before starting.

bare-metal can reboot boards into an existing rootfs on intermittent
device failure, but traces-db doesn't do any sanity-checking of the local
downloads of traces and would proceed to just trying to replay them.

Nuke any existing trace db so that it re-downloads every time, same as
LAVA or docker container tests do.

Fixes: #5585
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15440>

.gitlab-ci/piglit/piglit-traces.sh

@@ -188,6 +188,13 @@ fi
 
 ci-fairy minio login $MINIO_ARGS --token-file "${CI_JOB_JWT_FILE}"
 
+# The replayer doesn't do any size or checksum verification for the traces in
+# the replayer db, so if we had to restart the system due to intermittent device
+# errors (or tried to cache replayer-db between runs, which would be nice to
+# have), you could get a corrupted local trace that would spuriously fail the
+# run.
+rm -rf replayer-db
+
 eval $RUN_CMD
 
 if [ $? -ne 0 ]; then