ci: Reinstate Panfrost CI

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3118>

.gitlab-ci/deqp-panfrost-t760-skips.txt

@@ -8,6 +8,3 @@ dEQP-GLES[0-9]*.stress
 
 # These are really slow on tiling architectures (including llvmpipe).
 dEQP-GLES[0-9]*.functional.flush_finish
-
-# See T860 skip file
-dEQP-GLES2.functional.texture.vertex.2d.filtering.nearest_mipmap_nearest_nearest_repeat

.gitlab-ci/deqp-panfrost-t860-skips.txt

@@ -11,10 +11,3 @@ dEQP-GLES[0-9]*.functional.flush_finish
 
 # XXX: Why does this flake?
 dEQP-GLES2.functional.clipping.triangle_vertex.clip_three.clip_neg_x_neg_z_and_pos_x_pos_z_and_neg_x_neg_y_pos_z
-
-# XXX: This fails on CI since adding a new field in
-# 20b2b70da884df33eb970c5fc7714362a829bb84 but passes locally; failure depends
-# where in the struct the stack_size field is added. It's not clear what kind
-# of undefined behaviour is at play, but I don't want that commit to be blocked
-# since it fixes actual bugs.
-dEQP-GLES2.functional.texture.vertex.cube.filtering.linear_nearest_clamp

.gitlab-ci/lava-gitlab-ci.yml

@@ -173,14 +173,14 @@ panfrost-t720-test:arm64:
   tags:
     - lava-sun50i-h6-pine-h64
 
-.panfrost-t760-test:armhf:
+panfrost-t760-test:armhf:
   extends: .lava-test:armhf
   variables:
     DEVICE_TYPE: rk3288-veyron-jaq
   tags:
     - lava-rk3288-veyron-jaq
 
-.panfrost-t860-test:arm64:
+panfrost-t860-test:arm64:
   extends: .lava-test:arm64
   variables:
     DEVICE_TYPE: rk3399-gru-kevin