* [PATCH i-g-t] gem_ppgtt: Test VMA leak on context destruction
@ 2015-09-11 14:31 Tvrtko Ursulin
2015-09-18 11:17 ` Thomas Wood
0 siblings, 1 reply; 4+ messages in thread
From: Tvrtko Ursulin @ 2015-09-11 14:31 UTC (permalink / raw)
To: Intel-gfx
From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Test that VMAs associated with a context are cleaned up when
contexts are destroyed.
In practice this emulates the leak seen between fbcon and X server.
Every time the X server exits we gain one VMA on the fbcon frame
buffer object as externally visible via for example
/sys/kernel/debug/dri/0/i915_gem_gtt.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
tests/gem_ppgtt.c | 80 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 80 insertions(+)
diff --git a/tests/gem_ppgtt.c b/tests/gem_ppgtt.c
index 4f6df063214a..363f9d701585 100644
--- a/tests/gem_ppgtt.c
+++ b/tests/gem_ppgtt.c
@@ -265,6 +265,83 @@ static void flink_and_close(void)
close(fd2);
}
+static int grep_name(char *match, int to_match)
+{
+ int fdd, ret, matched;
+
+ fdd = open("/sys/kernel/debug/dri/0/i915_gem_gtt", O_RDONLY);
+ igt_assert(fdd >= 0);
+
+ matched = 0;
+ do {
+ char ch;
+
+ ret = read(fdd, &ch, 1);
+ if (ret == 0)
+ break;
+ igt_assert(ret == 1);
+
+ if (ch == match[matched])
+ matched++;
+ else
+ matched = 0;
+ } while (matched < to_match);
+
+ close(fdd);
+
+ return matched;
+}
+
+static void flink_and_exit(void)
+{
+ uint32_t fd, fd2;
+ uint32_t bo, flinked_bo, name;
+ char match[100];
+ int matched, to_match;
+ int retry = 0;
+ const int retries = 50;
+
+ fd = drm_open_any();
+ igt_require(uses_full_ppgtt(fd));
+
+ bo = gem_create(fd, 4096);
+ name = gem_flink(fd, bo);
+
+ to_match = snprintf(match, sizeof(match), "(name: %u)", name);
+ igt_assert(to_match < sizeof(match));
+
+ fd2 = drm_open_any();
+
+ flinked_bo = gem_open(fd2, name);
+ exec_and_get_offset(fd2, flinked_bo);
+ gem_sync(fd2, flinked_bo);
+
+ /* Verify looking for string works OK. */
+ matched = grep_name(match, to_match);
+ igt_assert_eq(matched, to_match);
+
+ gem_close(fd2, flinked_bo);
+
+ /* Close the context. */
+ close(fd2);
+
+retry:
+ /* Give cleanup some time to run. */
+ usleep(100000);
+
+ /* The flinked bo VMA should have been cleared now, so list of VMAs
+ * in debugfs should not contain the one for the imported object.
+ */
+ matched = grep_name(match, to_match);
+ if (matched >= to_match && retry++ < retries)
+ goto retry;
+
+ igt_assert_lt(matched, to_match);
+
+ gem_close(fd, bo);
+ close(fd);
+}
+
#define N_CHILD 8
int main(int argc, char **argv)
{
@@ -297,5 +374,8 @@ int main(int argc, char **argv)
igt_subtest("flink-and-close-vma-leak")
flink_and_close();
+ igt_subtest("flink-and-exit-vma-leak")
+ flink_and_exit();
+
igt_exit();
}
--
2.5.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH i-g-t] gem_ppgtt: Test VMA leak on context destruction
2015-09-11 14:31 [PATCH i-g-t] gem_ppgtt: Test VMA leak on context destruction Tvrtko Ursulin
@ 2015-09-18 11:17 ` Thomas Wood
2015-09-23 12:36 ` Tvrtko Ursulin
0 siblings, 1 reply; 4+ messages in thread
From: Thomas Wood @ 2015-09-18 11:17 UTC (permalink / raw)
To: Tvrtko Ursulin; +Cc: Intel Graphics Development
On 11 September 2015 at 15:31, Tvrtko Ursulin
<tvrtko.ursulin@linux.intel.com> wrote:
> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>
> Test that VMAs associated with a context are cleaned up when
> contexts are destroyed.
>
> In practice this emulates the leak seen between fbcon and X server.
> Every time the X server exits we gain one VMA on the fbcon frame
> buffer object as externally visible via for example
> /sys/kernel/debug/dri/0/i915_gem_gtt.
>
> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> ---
> tests/gem_ppgtt.c | 80 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 80 insertions(+)
>
> diff --git a/tests/gem_ppgtt.c b/tests/gem_ppgtt.c
> index 4f6df063214a..363f9d701585 100644
> --- a/tests/gem_ppgtt.c
> +++ b/tests/gem_ppgtt.c
> @@ -265,6 +265,83 @@ static void flink_and_close(void)
> close(fd2);
> }
>
> +static int grep_name(char *match, int to_match)
This could be added to igt_debugfs.c, as it might be useful for other tests.
> +{
> + int fdd, ret, matched;
> +
> + fdd = open("/sys/kernel/debug/dri/0/i915_gem_gtt", O_RDONLY);
igt_debugfs_open would be more robust here, as it checks various
locations for the debugfs and also attempts to mount it if it can't be
found.
> + igt_assert(fdd >= 0);
> +
> + matched = 0;
> + do {
> + char ch;
> +
> + ret = read(fdd, &ch, 1);
> + if (ret == 0)
> + break;
> + igt_assert(ret == 1);
> +
> + if (ch == match[matched])
> + matched++;
> + else
> + matched = 0;
> + } while (matched < to_match);
igt_debugfs_fopen is also available, which would allow the use of
getline and strstr here instead, as a slightly simpler implementation.
> +
> + close(fdd);
> +
> + return matched;
> +}
> +
> +static void flink_and_exit(void)
> +{
> + uint32_t fd, fd2;
> + uint32_t bo, flinked_bo, name;
> + char match[100];
> + int matched, to_match;
> + int retry = 0;
> + const int retries = 50;
> +
> + fd = drm_open_any();
> + igt_require(uses_full_ppgtt(fd));
> +
> + bo = gem_create(fd, 4096);
> + name = gem_flink(fd, bo);
> +
> + to_match = snprintf(match, sizeof(match), "(name: %u)", name);
> + igt_assert(to_match < sizeof(match));
> +
> + fd2 = drm_open_any();
> +
> + flinked_bo = gem_open(fd2, name);
> + exec_and_get_offset(fd2, flinked_bo);
> + gem_sync(fd2, flinked_bo);
> +
> + /* Verify looking for string works OK. */
> + matched = grep_name(match, to_match);
> + igt_assert_eq(matched, to_match);
> +
> + gem_close(fd2, flinked_bo);
> +
> + /* Close the context. */
> + close(fd2);
> +
> +retry:
> + /* Give cleanup some time to run. */
> + usleep(100000);
> +
> + /* The flinked bo VMA should have been cleared now, so list of VMAs
> + * in debugfs should not contain the one for the imported object.
> + */
> + matched = grep_name(match, to_match);
> + if (matched >= to_match && retry++ < retries)
> + goto retry;
> +
> + igt_assert_lt(matched, to_match);
> +
> + gem_close(fd, bo);
> + close(fd);
> +}
> +
> #define N_CHILD 8
> int main(int argc, char **argv)
> {
> @@ -297,5 +374,8 @@ int main(int argc, char **argv)
> igt_subtest("flink-and-close-vma-leak")
> flink_and_close();
>
> + igt_subtest("flink-and-exit-vma-leak")
> + flink_and_exit();
> +
> igt_exit();
> }
> --
> 2.5.1
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH i-g-t] gem_ppgtt: Test VMA leak on context destruction
2015-09-18 11:17 ` Thomas Wood
@ 2015-09-23 12:36 ` Tvrtko Ursulin
2015-09-23 16:16 ` Thomas Wood
0 siblings, 1 reply; 4+ messages in thread
From: Tvrtko Ursulin @ 2015-09-23 12:36 UTC (permalink / raw)
To: Thomas Wood; +Cc: Intel Graphics Development
Hi,
On 09/18/2015 12:17 PM, Thomas Wood wrote:
> On 11 September 2015 at 15:31, Tvrtko Ursulin
> <tvrtko.ursulin@linux.intel.com> wrote:
>> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>
>> Test that VMAs associated with a context are cleaned up when
>> contexts are destroyed.
>>
>> In practice this emulates the leak seen between fbcon and X server.
>> Every time the X server exits we gain one VMA on the fbcon frame
>> buffer object as externally visible via for example
>> /sys/kernel/debug/dri/0/i915_gem_gtt.
>>
>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>> ---
>> tests/gem_ppgtt.c | 80 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
>> 1 file changed, 80 insertions(+)
>>
>> diff --git a/tests/gem_ppgtt.c b/tests/gem_ppgtt.c
>> index 4f6df063214a..363f9d701585 100644
>> --- a/tests/gem_ppgtt.c
>> +++ b/tests/gem_ppgtt.c
>> @@ -265,6 +265,83 @@ static void flink_and_close(void)
>> close(fd2);
>> }
>>
>> +static int grep_name(char *match, int to_match)
>
> This could be added to igt_debugfs.c, as it might be useful for other tests.
I couldn't force myself to do it.
>
>> +{
>> + int fdd, ret, matched;
>> +
>> + fdd = open("/sys/kernel/debug/dri/0/i915_gem_gtt", O_RDONLY);
>
> igt_debugfs_open would be more robust here, as it checks various
> locations for the debugfs and also attempts to mount it if it can't be
> found.
Done.
>
>
>> + igt_assert(fdd >= 0);
>> +
>> + matched = 0;
>> + do {
>> + char ch;
>> +
>> + ret = read(fdd, &ch, 1);
>> + if (ret == 0)
>> + break;
>> + igt_assert(ret == 1);
>> +
>> + if (ch == match[matched])
>> + matched++;
>> + else
>> + matched = 0;
>> + } while (matched < to_match);
>
> igt_debugfs_fopen is also available, which would allow the use of
> getline and strstr here instead, as a slightly simpler implementation.
Also done. Ah not via fopen but open+fdopen... drat..
You want v3 or you can live with v2?
Regards,
Tvrtko
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH i-g-t] gem_ppgtt: Test VMA leak on context destruction
2015-09-23 12:36 ` Tvrtko Ursulin
@ 2015-09-23 16:16 ` Thomas Wood
0 siblings, 0 replies; 4+ messages in thread
From: Thomas Wood @ 2015-09-23 16:16 UTC (permalink / raw)
To: Tvrtko Ursulin; +Cc: Intel Graphics Development
On 23 September 2015 at 13:36, Tvrtko Ursulin
<tvrtko.ursulin@linux.intel.com> wrote:
>
> Hi,
>
> On 09/18/2015 12:17 PM, Thomas Wood wrote:
>>
>> On 11 September 2015 at 15:31, Tvrtko Ursulin
>> <tvrtko.ursulin@linux.intel.com> wrote:
>>>
>>> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>>
>>> Test that VMAs associated with a context are cleaned up when
>>> contexts are destroyed.
>>>
>>> In practice this emulates the leak seen between fbcon and X server.
>>> Every time the X server exits we gain one VMA on the fbcon frame
>>> buffer object as externally visible via for example
>>> /sys/kernel/debug/dri/0/i915_gem_gtt.
>>>
>>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>> ---
>>> tests/gem_ppgtt.c | 80
>>> +++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>> 1 file changed, 80 insertions(+)
>>>
>>> diff --git a/tests/gem_ppgtt.c b/tests/gem_ppgtt.c
>>> index 4f6df063214a..363f9d701585 100644
>>> --- a/tests/gem_ppgtt.c
>>> +++ b/tests/gem_ppgtt.c
>>> @@ -265,6 +265,83 @@ static void flink_and_close(void)
>>> close(fd2);
>>> }
>>>
>>> +static int grep_name(char *match, int to_match)
>>
>>
>> This could be added to igt_debugfs.c, as it might be useful for other
>> tests.
>
>
> I couldn't force myself to do it.
I've sent a follow up patch to move the function into the library and
use it in gem_ppgtt and gem_tiled_swapping.
>
>>
>>> +{
>>> + int fdd, ret, matched;
>>> +
>>> + fdd = open("/sys/kernel/debug/dri/0/i915_gem_gtt", O_RDONLY);
>>
>>
>> igt_debugfs_open would be more robust here, as it checks various
>> locations for the debugfs and also attempts to mount it if it can't be
>> found.
>
>
> Done.
>
>>
>>
>>> + igt_assert(fdd >= 0);
>>> +
>>> + matched = 0;
>>> + do {
>>> + char ch;
>>> +
>>> + ret = read(fdd, &ch, 1);
>>> + if (ret == 0)
>>> + break;
>>> + igt_assert(ret == 1);
>>> +
>>> + if (ch == match[matched])
>>> + matched++;
>>> + else
>>> + matched = 0;
>>> + } while (matched < to_match);
>>
>>
>> igt_debugfs_fopen is also available, which would allow the use of
>> getline and strstr here instead, as a slightly simpler implementation.
>
>
> Also done. Ah not via fopen but open+fdopen... drat..
The follow up patch I've sent addresses this and a few other minor issues.
>
> You want v3 or you can live with v2?
Unfortunately another version will be needed, as drm_open_any has been
replaced by drm_open_driver.
>
> Regards,
>
> Tvrtko
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2015-09-23 16:16 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-09-11 14:31 [PATCH i-g-t] gem_ppgtt: Test VMA leak on context destruction Tvrtko Ursulin
2015-09-18 11:17 ` Thomas Wood
2015-09-23 12:36 ` Tvrtko Ursulin
2015-09-23 16:16 ` Thomas Wood
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.