Re: [Intel-gfx] [PATCH i-g-t 2/4] tests/gem_scheduler: Add gem_scheduler test

2016-03-01 Thread Morton, Derek J 
>
>
>-Original Message-
>From: Ceraolo Spurio, Daniele 
>Sent: Wednesday, February 17, 2016 12:37 PM
>To: Morton, Derek J <derek.j.mor...@intel.com>; intel-gfx@lists.freedesktop.org
>Subject: Re: [Intel-gfx] [PATCH i-g-t 2/4] tests/gem_scheduler: Add 
>gem_scheduler test
>
>Hi,
>
>first round of comments inline. A few things will probably have to change 
>based on the comments on the previous patch in the series so I'll have a 
>better look at the related logic once the new series is up.
>
>Regards,
>Daniele
>
>
>On 12/02/16 09:38, Derek Morton wrote:
>> This is intended to test the scheduler behaviour is correct.
>> The subtests are
>> -basic
>> Tests that batch buffers of the same priority submitted to a ring 
>> execute in the order they are submitted.
>> -read
>> Submits a batch buffer with a read dependency to a buffer object to a 
>> ring which is held in the scheduler queue by a long running batch 
>> buffer. Submit batch buffers to other rings that have a read 
>> dependency to the same buffer object. Ensure they execute before the 
>> batch buffer being held up behind the long running batch buffer.
>> -write
>> Submits a batch buffer with a write dependency to a buffer object to a 
>> ring which is held in the scheduler queue by a long running batch 
>> buffer. Submit batch buffers to other rings that have a write 
>> dependency to the same buffer object. Submit batch buffers with no 
>> interdependencies to all rings. Ensure the batch buffers that have 
>> write dependencies are executed in submission order but the batch 
>> buffers without interdependencies do not get held up.
>>
>> Signed-off-by: Derek Morton <derek.j.mor...@intel.com>
>> ---
>>   tests/Makefile.sources |   1 +
>>   tests/gem_scheduler.c  | 409 
>> +
>>   2 files changed, 410 insertions(+)
>>   create mode 100644 tests/gem_scheduler.c
>>
>> diff --git a/tests/Makefile.sources b/tests/Makefile.sources index 
>> df92586..439f62c 100644
>> --- a/tests/Makefile.sources
>> +++ b/tests/Makefile.sources
>> @@ -66,6 +66,7 @@ TESTS_progs_M = \
>>  gem_request_retire \
>>  gem_reset_stats \
>>  gem_ringfill \
>> +gem_scheduler \
>>  gem_set_tiling_vs_blt \
>>  gem_softpin \
>>  gem_stolen \
>> diff --git a/tests/gem_scheduler.c b/tests/gem_scheduler.c new file 
>> mode 100644 index 000..4824c13
>> --- /dev/null
>> +++ b/tests/gem_scheduler.c
>> @@ -0,0 +1,409 @@
>> +/*
>> + * Copyright © 2016 Intel Corporation
>> + *
>> + * Permission is hereby granted, free of charge, to any person 
>> +obtaining a
>> + * copy of this software and associated documentation files (the 
>> +"Software"),
>> + * to deal in the Software without restriction, including without 
>> +limitation
>> + * the rights to use, copy, modify, merge, publish, distribute, 
>> +sublicense,
>> + * and/or sell copies of the Software, and to permit persons to whom 
>> +the
>> + * Software is furnished to do so, subject to the following conditions:
>> + *
>> + * The above copyright notice and this permission notice (including 
>> +the next
>> + * paragraph) shall be included in all copies or substantial portions 
>> +of the
>> + * Software.
>> + *
>> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, 
>> +EXPRESS OR
>> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF 
>> +MERCHANTABILITY,
>> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT 
>> +SHALL
>> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES 
>> +OR OTHER
>> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, 
>> +ARISING
>> + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR 
>> +OTHER DEALINGS
>> + * IN THE SOFTWARE.
>> + *
>> + * Authors:
>> + *Derek Morton <derek.j.mor...@intel.com>
>> + *
>> + */
>> +
>> +#include "igt.h"
>> +#include 
>> +#include 
>> +#include 
>> +#include 
>> +#include 
>> +#include 
>> +#include 
>> +#include 
>> +#include 
>> +#include 
>> +
>> +IGT_TEST_DESCRIPTION("Check scheduler behaviour. Basic tests ensure 
>> independant "
>> + "batch buffers of the same priority are executed in "
>> + "submission order. Read-read tests ensure

Re: [Intel-gfx] [PATCH i-g-t 2/4] tests/gem_scheduler: Add gem_scheduler test

2016-02-17 Thread Daniele Ceraolo Spurio 

Hi,

first round of comments inline. A few things will probably have to 
change based on the comments on the previous patch in the series so I'll 
have a better look at the related logic once the new series is up.


Regards,
Daniele


On 12/02/16 09:38, Derek Morton wrote:

This is intended to test the scheduler behaviour is correct.
The subtests are
-basic
Tests that batch buffers of the same priority submitted to a ring
execute in the order they are submitted.
-read
Submits a batch buffer with a read dependency to a buffer object to
a ring which is held in the scheduler queue by a long running batch
buffer. Submit batch buffers to other rings that have a read dependency
to the same buffer object. Ensure they execute before the batch buffer
being held up behind the long running batch buffer.
-write
Submits a batch buffer with a write dependency to a buffer object to
a ring which is held in the scheduler queue by a long running batch
buffer. Submit batch buffers to other rings that have a write dependency
to the same buffer object. Submit batch buffers with no interdependencies
to all rings. Ensure the batch buffers that have write dependencies are
executed in submission order but the batch buffers without interdependencies
do not get held up.

Signed-off-by: Derek Morton 
---
  tests/Makefile.sources |   1 +
  tests/gem_scheduler.c  | 409 +
  2 files changed, 410 insertions(+)
  create mode 100644 tests/gem_scheduler.c

diff --git a/tests/Makefile.sources b/tests/Makefile.sources
index df92586..439f62c 100644
--- a/tests/Makefile.sources
+++ b/tests/Makefile.sources
@@ -66,6 +66,7 @@ TESTS_progs_M = \
gem_request_retire \
gem_reset_stats \
gem_ringfill \
+   gem_scheduler \
gem_set_tiling_vs_blt \
gem_softpin \
gem_stolen \
diff --git a/tests/gem_scheduler.c b/tests/gem_scheduler.c
new file mode 100644
index 000..4824c13
--- /dev/null
+++ b/tests/gem_scheduler.c
@@ -0,0 +1,409 @@
+/*
+ * Copyright © 2016 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Authors:
+ *Derek Morton 
+ *
+ */
+
+#include "igt.h"
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+IGT_TEST_DESCRIPTION("Check scheduler behaviour. Basic tests ensure independant 
"
+ "batch buffers of the same priority are executed in "
+ "submission order. Read-read tests ensure "
+ "batch buffers with a read dependency to the same buffer "
+ "object do not block each other. Write-write dependency "
+ "tests ensure batch buffers with a write dependency to a "
+ "buffer object will be executed in submission order but "
+ "will not block execution of other independant batch "
+ "buffers.");
+
+#define SEC_TO_NSEC (1000 * 1000 * 1000)
+
+struct ring {
+   const char *name;
+   int id;
+} rings[] = {
+   { "render", I915_EXEC_RENDER },
+   { "bsd",I915_EXEC_BSD },


Is BSD1/BSD2 difference of any importance for the tests?


+   { "blt",I915_EXEC_BLT },
+   { "vebox",  I915_EXEC_VEBOX },
+};


This is a slight duplication of intel_execution_engines. However, I'm 
not sure that intel_execution_engines is suitable here considering that 
it doesn't seem you want to check both the vanilla I915_EXEC_BSD case 
and the BSD1/BSD2 case



+
+#define NBR_RINGS (sizeof(rings)/sizeof(struct ring))
+
+/* Basic test. Check batch buffers of the same priority and with no 
dependencies
+ * are executed in the order they are submitted.
+ */
+#define NBR_BASIC_FDs (3)
+static void run_test_basic(int in_flight, int ringid)
+{
+   int fd[NBR_BASIC_FDs];
+   int loop;
+

[Intel-gfx] [PATCH i-g-t 2/4] tests/gem_scheduler: Add gem_scheduler test

2016-02-12 Thread Derek Morton 
This is intended to test the scheduler behaviour is correct.
The subtests are
-basic
Tests that batch buffers of the same priority submitted to a ring
execute in the order they are submitted.
-read
Submits a batch buffer with a read dependency to a buffer object to
a ring which is held in the scheduler queue by a long running batch
buffer. Submit batch buffers to other rings that have a read dependency
to the same buffer object. Ensure they execute before the batch buffer
being held up behind the long running batch buffer.
-write
Submits a batch buffer with a write dependency to a buffer object to
a ring which is held in the scheduler queue by a long running batch
buffer. Submit batch buffers to other rings that have a write dependency
to the same buffer object. Submit batch buffers with no interdependencies
to all rings. Ensure the batch buffers that have write dependencies are
executed in submission order but the batch buffers without interdependencies
do not get held up.

Signed-off-by: Derek Morton 
---
 tests/Makefile.sources |   1 +
 tests/gem_scheduler.c  | 409 +
 2 files changed, 410 insertions(+)
 create mode 100644 tests/gem_scheduler.c

diff --git a/tests/Makefile.sources b/tests/Makefile.sources
index df92586..439f62c 100644
--- a/tests/Makefile.sources
+++ b/tests/Makefile.sources
@@ -66,6 +66,7 @@ TESTS_progs_M = \
gem_request_retire \
gem_reset_stats \
gem_ringfill \
+   gem_scheduler \
gem_set_tiling_vs_blt \
gem_softpin \
gem_stolen \
diff --git a/tests/gem_scheduler.c b/tests/gem_scheduler.c
new file mode 100644
index 000..4824c13
--- /dev/null
+++ b/tests/gem_scheduler.c
@@ -0,0 +1,409 @@
+/*
+ * Copyright © 2016 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Authors:
+ *Derek Morton 
+ *
+ */
+
+#include "igt.h"
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+IGT_TEST_DESCRIPTION("Check scheduler behaviour. Basic tests ensure 
independant "
+ "batch buffers of the same priority are executed in "
+ "submission order. Read-read tests ensure "
+ "batch buffers with a read dependency to the same buffer "
+ "object do not block each other. Write-write dependency "
+ "tests ensure batch buffers with a write dependency to a "
+ "buffer object will be executed in submission order but "
+ "will not block execution of other independant batch "
+ "buffers.");
+
+#define SEC_TO_NSEC (1000 * 1000 * 1000)
+
+struct ring {
+   const char *name;
+   int id;
+} rings[] = {
+   { "render", I915_EXEC_RENDER },
+   { "bsd",I915_EXEC_BSD },
+   { "blt",I915_EXEC_BLT },
+   { "vebox",  I915_EXEC_VEBOX },
+};
+
+#define NBR_RINGS (sizeof(rings)/sizeof(struct ring))
+
+/* Basic test. Check batch buffers of the same priority and with no 
dependencies
+ * are executed in the order they are submitted.
+ */
+#define NBR_BASIC_FDs (3)
+static void run_test_basic(int in_flight, int ringid)
+{
+   int fd[NBR_BASIC_FDs];
+   int loop;
+   drm_intel_bufmgr *bufmgr[NBR_BASIC_FDs];
+   uint32_t *delay_buf, *ts1_buf, *ts2_buf;
+   struct intel_batchbuffer *ts1_bb, *ts2_bb;
+   struct intel_batchbuffer **in_flight_bbs;
+   uint32_t calibrated_1s;
+   drm_intel_bo *delay_bo, *ts1_bo, *ts2_bo;
+
+   in_flight_bbs = malloc(in_flight * sizeof(struct intel_batchbuffer *));
+   igt_assert(in_flight_bbs);
+
+   /* Need multiple i915 fd's. Scheduler will not change execution order of
+* batch buffers from the same context.
+*/
+   for(loop=0; loop < NBR_BASIC_FDs; loop++)