Skip to contentSkip to conversation

test/gem_ctx_param - Use the last I915_CONTEXT_PARAM+1 for invalid tests

PATCH In review Marta Lofstedt 2 participants 3 messages Jan 11, 2018

lore
Share & tools
lore
#0 lore raw

PATCH 1/2 test/gem_ctx_param - Use the last I915_CONTEXT_PARAM+1 for invalid tests

The invalid-param-[get|set] exploits the last I915_CONTEXT_PARAM + 1, to check for ABI extentsions. However, the last param was set to I915_CONTEXT_PARAM_BANNABLE, so when I915_CONTEXT_PARAM_PRIORITY was added to the next enum, the test started failing.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103107

---
 tests/gem_ctx_param.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tests/gem_ctx_param.c +1 -1
index c20ae1ee..75d2a20f 100644
@@ -140,7 +140,7 @@ igt_main
 	 * to catch ABI extensions. Don't "fix" this testcase without adding all
 	 * the tests for the new param first.
 	 */
-	arg.param = I915_CONTEXT_PARAM_BANNABLE + 1;
+	arg.param = I915_CONTEXT_PARAM_PRIORITY + 1;
 
 	igt_subtest("invalid-param-get") {
 		arg.ctx_id = ctx;
Signature · 6 lines
2.11.0

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#1 lore raw
Quoted 17 lines · Marta Lofstedt

The invalid-param-[get|set] exploits the last I915_CONTEXT_PARAM + 1, to check for ABI extentsions. However, the last param was set to I915_CONTEXT_PARAM_BANNABLE, so when I915_CONTEXT_PARAM_PRIORITY was added to the next enum, the test started failing.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103107

---
  tests/gem_ctx_param.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
tests/gem_ctx_param.c +0 -0
index c20ae1ee..75d2a20f 100644

tests/gem_ctx_param.c igt_main · line 146

Quoted 8 lines
tests/gem_ctx_param.c +1 -1
@@ -140,7 +140,7 @@ igt_main
  	 * to catch ABI extensions. Don't "fix" this testcase without adding all
  	 * the tests for the new param first.
  	 */
-	arg.param = I915_CONTEXT_PARAM_BANNABLE + 1;
+	arg.param = I915_CONTEXT_PARAM_PRIORITY + 1;
 
  	igt_subtest("invalid-param-get") {
  		arg.ctx_id = ctx;

We tried before, last time AFAIR https://patchwork.freedesktop.org/patch/187073/, but again AFAIR I think Chris objected and would rather remove negative tests in general. Shrug, an impasse. :I

Regards,

Tvrtko
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#2 lore raw
Quoted 37 lines

-----Original Message-----
From: Tvrtko Ursulin [mailto:[email protected]]
Sent: Thursday, January 11, 2018 1:46 PM
To: Lofstedt, Marta <[email protected]>; intel- [email protected]
Cc: Chris Wilson <[email protected]>
Subject: Re: [Intel-gfx] [PATCH i-g-t 1/2] test/gem_ctx_param - Use the last I915_CONTEXT_PARAM+1 for invalid tests

Quoted 3 lines · Marta Lofstedt

The invalid-param-[get|set] exploits the last I915_CONTEXT_PARAM + 1, to check for ABI extentsions. However, the last param was set to I915_CONTEXT_PARAM_BANNABLE, so when

I915_CONTEXT_PARAM_PRIORITY was

Quoted 13 lines

added to the next enum, the test started failing.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103107

---
  tests/gem_ctx_param.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
tests/gem_ctx_param.c index renamed from tests/gem_ctx_param.c +0 -0

c20ae1ee..75d2a20f 100644

tests/gem_ctx_param.c igt_main · line 140

Quoted 1 line
tests/gem_ctx_param.c +0 -0
@@ -140,7 +140,7 @@ igt_main
  	 * to catch ABI extensions. Don't "fix" this testcase without

adding all

Quoted 5 lines
tests/gem_ctx_param.c +1 -1
  	 * the tests for the new param first.
  	 */
-	arg.param = I915_CONTEXT_PARAM_BANNABLE + 1;
+	arg.param = I915_CONTEXT_PARAM_PRIORITY + 1;
 
Quoted 9 lines

tests/gem_ctx_param.c igt_main · line 146

Quoted 3 lines
tests/gem_ctx_param.c +0 -0
  	igt_subtest("invalid-param-get") {
  		arg.ctx_id = ctx;

We tried before, last time AFAIR
https://patchwork.freedesktop.org/patch/187073/, but again AFAIR I think Chris objected and would rather remove negative tests in general. Shrug, an impasse. :I

Well, then maybe V2 will be to remove these test ;)

Quoted 3 lines

Regards,

Tvrtko

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx