Intel-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: ville.syrjala@linux.intel.com
To: intel-gfx@lists.freedesktop.org
Subject: [PATCH] tests/gem_fenced_exec_thrash: Test with > max fences
Date: Thu, 11 Apr 2013 20:43:40 +0300	[thread overview]
Message-ID: <1365702220-20155-1-git-send-email-ville.syrjala@linux.intel.com> (raw)
In-Reply-To: <20130409130500.GM3792@cantiga.alporthouse.com>

From: Ville Syrjälä <ville.syrjala@linux.intel.com>

Make sure the kernel returns EDEADLK when the number of fences is
exceeded for gen2-3. For gen4+ the test makes sure the kernel ignores
the EXEC_OBJECT_NEEDS_FENCE flag.

Note that I changed the code not to round the num_fences to an even
number. Not sure why that was there, and if there's a reason for it,
we need to add it back.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 tests/gem_fenced_exec_thrash.c | 38 +++++++++++++++++++++++++-------------
 1 file changed, 25 insertions(+), 13 deletions(-)

diff --git a/tests/gem_fenced_exec_thrash.c b/tests/gem_fenced_exec_thrash.c
index ca88c53..c8a2c58 100644
--- a/tests/gem_fenced_exec_thrash.c
+++ b/tests/gem_fenced_exec_thrash.c
@@ -38,6 +38,7 @@
 #include <i915_drm.h>
 
 #include "drmtest.h"
+#include "intel_gpu_tools.h"
 
 #define WIDTH 1024
 #define HEIGHT 1024
@@ -47,8 +48,6 @@
 
 #define MAX_FENCES 32
 
-#define MI_BATCH_BUFFER_END	(0xA<<23)
-
 /*
  * Testcase: execbuf fence accounting
  *
@@ -95,7 +94,7 @@ static int get_num_fences(int fd)
 	printf ("total %d fences\n", val);
 	assert(val > 4);
 
-	return val - 2;
+	return val;
 }
 
 static void fill_reloc(struct drm_i915_gem_relocation_entry *reloc, uint32_t handle)
@@ -106,23 +105,19 @@ static void fill_reloc(struct drm_i915_gem_relocation_entry *reloc, uint32_t han
 	reloc->write_domain = 0;
 }
 
-int
-main(int argc, char **argv)
+static void run_test(int fd, int num_fences, int expected_errno)
 {
 	struct drm_i915_gem_execbuffer2 execbuf[2];
-	struct drm_i915_gem_exec_object2 exec[2][2*MAX_FENCES+1];
-	struct drm_i915_gem_relocation_entry reloc[2*MAX_FENCES];
+	struct drm_i915_gem_exec_object2 exec[2][2*MAX_FENCES+3];
+	struct drm_i915_gem_relocation_entry reloc[2*MAX_FENCES+2];
 
-	int fd = drm_open_any();
-	int i, n, num_fences;
+	int i, n;
 	int loop = 1000;
 
 	memset(execbuf, 0, sizeof(execbuf));
 	memset(exec, 0, sizeof(exec));
 	memset(reloc, 0, sizeof(reloc));
 
-	num_fences = get_num_fences(fd) & ~1;
-	assert(num_fences <= MAX_FENCES);
 	for (n = 0; n < 2*num_fences; n++) {
 		uint32_t handle = tiled_bo_create(fd);
 		exec[1][2*num_fences - n-1].handle = exec[0][n].handle = handle;
@@ -148,13 +143,30 @@ main(int argc, char **argv)
 		ret = drmIoctl(fd,
 			       DRM_IOCTL_I915_GEM_EXECBUFFER2,
 			       &execbuf[0]);
-		assert(ret == 0);
+		assert(expected_errno ?
+		       ret < 0 && errno == expected_errno :
+		       ret == 0);
 
 		ret = drmIoctl(fd,
 			       DRM_IOCTL_I915_GEM_EXECBUFFER2,
 			       &execbuf[1]);
-		assert(ret == 0);
+		assert(expected_errno ?
+		       ret < 0 && errno == expected_errno :
+		       ret == 0);
 	} while (--loop);
+}
+
+int
+main(int argc, char **argv)
+{
+	int fd = drm_open_any();
+	int num_fences = get_num_fences(fd);
+	uint32_t devid = intel_get_drm_devid(fd);
+
+	assert(num_fences <= MAX_FENCES);
+
+	run_test(fd, num_fences - 2, 0);
+	run_test(fd, num_fences + 1, intel_gen(devid) >= 4 ? 0 : EDEADLK);
 
 	close(fd);
 
-- 
1.8.1.5

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

  reply	other threads:[~2013-04-11 17:43 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-04-09 12:25 [PATCH i-g-t 0/3] Fence register stuff ville.syrjala
2013-04-09 12:25 ` [PATCH i-g-t 1/3] tests/gem_fenced_exec_thrash: Increase MAX_FENCES to 32 ville.syrjala
2013-04-09 13:05   ` Chris Wilson
2013-04-11 17:43     ` ville.syrjala [this message]
2013-04-11 18:22       ` [PATCH] tests/gem_fenced_exec_thrash: Test with > max fences Daniel Vetter
2013-04-09 12:25 ` [PATCH i-g-t 2/3] tests: Use gem_available_fences() ville.syrjala
2013-04-09 12:25 ` [PATCH i-g-t 3/3] tests/gem_tiling_max_stride: Add a test for max fence stride ville.syrjala
2013-04-09 13:06   ` Chris Wilson
2013-04-09 13:41     ` [PATCH i-g-t v2] " ville.syrjala
2013-04-09 13:51       ` Chris Wilson
2013-04-09 14:45         ` [PATCH i-g-t v4] " ville.syrjala
2013-04-09 18:06           ` Chris Wilson
2013-04-09 18:21             ` Daniel Vetter

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1365702220-20155-1-git-send-email-ville.syrjala@linux.intel.com \
    --to=ville.syrjala@linux.intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox