[PATCH v2] tests/intel/xe_debugfs: Extend gt test to check few debugfs entries

Lucas De Marchi lucas.demarchi at intel.com
Thu Dec 19 14:46:05 UTC 2024


On Thu, Dec 19, 2024 at 09:26:09AM +0530, Pravalika Gurram wrote:
>Read and dump  below debugfs entries.
>ggtt
>register-save-restore
>workarounds
>default_lrc_rcs
>default_lrc_ccs
>default_lrc_bcs
>default_lrc_vcs
>default_lrc_vecs
>hwconfig
>
>Reviewed-by: Matthew Brost <matthew.brost at intel.com>
>Signed-off-by: Pravalika Gurram <pravalika.gurram at intel.com>
>---
> tests/intel/xe_debugfs.c | 50 ++++++++++++----------------------------
> 1 file changed, 15 insertions(+), 35 deletions(-)
>
>diff --git a/tests/intel/xe_debugfs.c b/tests/intel/xe_debugfs.c
>index 700575031..a31de41a4 100644
>--- a/tests/intel/xe_debugfs.c
>+++ b/tests/intel/xe_debugfs.c
>@@ -162,6 +162,16 @@ test_gt(int fd, int gt_id)
> 		"pat",
> 		"mocs",
> //		"force_reset"
>+		"ggtt",
>+		"register-save-restore",
>+		"workarounds",
>+		"default_lrc_rcs",
>+		"default_lrc_ccs",
>+		"default_lrc_bcs",
>+		"default_lrc_vcs",
>+		"default_lrc_vecs",
>+		"hwconfig"
>+
> 	};
> 	static const char * const expected_files_uc[] = {
> 		"huc_info",
>@@ -170,41 +180,11 @@ test_gt(int fd, int gt_id)
> //		"guc_ct_selftest"
> 	};
>
>-	sprintf(name, "gt%d/hw_engines", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);
>-
>-	sprintf(name, "gt%d/sa_info", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);
>-
>-	sprintf(name, "gt%d/steering", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);
>-
>-	sprintf(name, "gt%d/topology", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);
>-
>-	sprintf(name, "gt%d/pat", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);
>-
>-	sprintf(name, "gt%d/mocs", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);
>-
>-	sprintf(name, "gt%d/uc/guc_info", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);
>-
>-	sprintf(name, "gt%d/uc/huc_info", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);

these last 3 would be dropped from the dump. I don't think that's
intentional, is it?

Ideally this would be more than 1 patch:
1) replace the copy and paste with the loop
2) add more files to be dumped

Also, it seems we are not actually running this
"test" as part of BAT. I think that more than actually checking the
asserts this would be useful to check the state of the hardware
in the beginning of a BAT execution.

Lucas De Marchi

>-
>-	sprintf(name, "gt%d/uc/guc_log", gt_id);
>-	igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>-	igt_debugfs_dump(fd, name);
>+	for (int i = 0; i < ARRAY_SIZE(expected_files); i++) {
>+		sprintf(name, "gt%d/%s", gt_id, expected_files[i]);
>+		igt_assert(igt_debugfs_exists(fd, name, O_RDONLY));
>+		igt_debugfs_dump(fd, name);
>+	}
>
> 	sprintf(name, "/gt%d", gt_id);
> 	validate_entries(fd, name, expected_files, ARRAY_SIZE(expected_files));
>-- 
>2.34.1
>


More information about the igt-dev mailing list