From: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>, intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH v4 12/16] drm/i915: Exercise request cancellation using a mock selftest
Date: Thu, 23 Feb 2017 12:21:02 +0000 [thread overview]
Message-ID: <e01cb421-1cfb-71d1-20b3-5e9fcba2b363@linux.intel.com> (raw)
In-Reply-To: <20170223074422.4125-13-chris@chris-wilson.co.uk>
On 23/02/2017 07:44, Chris Wilson wrote:
> Add a mock selftest to preempt a request and check that we cancel it,
> requeue the request and then complete its execution.
>
> v2: Error leaks no more.
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
> drivers/gpu/drm/i915/selftests/i915_gem_request.c | 64 +++++++++++++++++++++++
> drivers/gpu/drm/i915/selftests/mock_request.c | 19 +++++++
> drivers/gpu/drm/i915/selftests/mock_request.h | 2 +
> 3 files changed, 85 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_request.c b/drivers/gpu/drm/i915/selftests/i915_gem_request.c
> index 9d056a86723d..42bdeac93324 100644
> --- a/drivers/gpu/drm/i915/selftests/i915_gem_request.c
> +++ b/drivers/gpu/drm/i915/selftests/i915_gem_request.c
> @@ -26,6 +26,7 @@
>
> #include "../i915_selftest.h"
>
> +#include "mock_context.h"
> #include "mock_gem_device.h"
>
> static int igt_add_request(void *arg)
> @@ -181,12 +182,75 @@ static int igt_fence_wait(void *arg)
> return err;
> }
>
> +static int igt_request_rewind(void *arg)
> +{
> + struct drm_i915_private *i915 = arg;
> + struct drm_i915_gem_request *request, *vip;
> + struct i915_gem_context *ctx[2];
> + int err = -EINVAL;
> +
> + mutex_lock(&i915->drm.struct_mutex);
> + ctx[0] = mock_context(i915, "A");
> + request = mock_request(i915->engine[RCS], ctx[0], 2 * HZ);
> + if (!request) {
> + err = -ENOMEM;
> + goto err_context_0;
> + }
> +
> + i915_gem_request_get(request);
> + i915_add_request(request);
> +
> + ctx[1] = mock_context(i915, "B");
> + vip = mock_request(i915->engine[RCS], ctx[1], 0);
> + if (!vip) {
> + err = -ENOMEM;
> + goto err_context_1;
> + }
> +
> + /* Simulate preemption by manual reordering */
> + if (!mock_cancel_request(request)) {
> + pr_err("failed to cancel request (already executed)!\n");
> + i915_add_request(vip);
> + goto err_context_1;
> + }
> + i915_gem_request_get(vip);
> + i915_add_request(vip);
> + request->engine->submit_request(request);
> +
> + mutex_unlock(&i915->drm.struct_mutex);
> +
> + if (i915_wait_request(vip, 0, HZ) == -ETIME) {
> + pr_err("timed out waiting for high priority request, vip.seqno=%d, current seqno=%d\n",
> + vip->global_seqno, intel_engine_get_seqno(i915->engine[RCS]));
> + goto err;
> + }
> +
> + if (i915_gem_request_completed(request)) {
> + pr_err("low priority request already completed\n");
> + goto err;
> + }
> +
> + err = 0;
> +err:
> + i915_gem_request_put(vip);
> + mutex_lock(&i915->drm.struct_mutex);
> +err_context_1:
> + mock_context_close(ctx[1]);
> + i915_gem_request_put(request);
> +err_context_0:
> + mock_context_close(ctx[0]);
> + mock_device_flush(i915);
> + mutex_unlock(&i915->drm.struct_mutex);
> + return err;
> +}
> +
> int i915_gem_request_mock_selftests(void)
> {
> static const struct i915_subtest tests[] = {
> SUBTEST(igt_add_request),
> SUBTEST(igt_wait_request),
> SUBTEST(igt_fence_wait),
> + SUBTEST(igt_request_rewind),
> };
> struct drm_i915_private *i915;
> int err;
> diff --git a/drivers/gpu/drm/i915/selftests/mock_request.c b/drivers/gpu/drm/i915/selftests/mock_request.c
> index e23242d1b88a..0e8d2e7f8c70 100644
> --- a/drivers/gpu/drm/i915/selftests/mock_request.c
> +++ b/drivers/gpu/drm/i915/selftests/mock_request.c
> @@ -22,6 +22,7 @@
> *
> */
>
> +#include "mock_engine.h"
> #include "mock_request.h"
>
> struct drm_i915_gem_request *
> @@ -42,3 +43,21 @@ mock_request(struct intel_engine_cs *engine,
>
> return &mock->base;
> }
> +
> +bool mock_cancel_request(struct drm_i915_gem_request *request)
> +{
> + struct mock_request *mock = container_of(request, typeof(*mock), base);
> + struct mock_engine *engine =
> + container_of(request->engine, typeof(*engine), base);
> + bool was_queued;
> +
> + spin_lock_irq(&engine->hw_lock);
> + was_queued = !list_empty(&mock->link);
> + list_del_init(&mock->link);
> + spin_unlock_irq(&engine->hw_lock);
> +
> + if (was_queued)
> + i915_gem_request_unsubmit(request);
> +
> + return was_queued;
> +}
> diff --git a/drivers/gpu/drm/i915/selftests/mock_request.h b/drivers/gpu/drm/i915/selftests/mock_request.h
> index cc76d4f4eb4e..4dea74c8e96d 100644
> --- a/drivers/gpu/drm/i915/selftests/mock_request.h
> +++ b/drivers/gpu/drm/i915/selftests/mock_request.h
> @@ -41,4 +41,6 @@ mock_request(struct intel_engine_cs *engine,
> struct i915_gem_context *context,
> unsigned long delay);
>
> +bool mock_cancel_request(struct drm_i915_gem_request *request);
> +
> #endif /* !__MOCK_REQUEST__ */
>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Regards,
Tvrtko
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2017-02-23 12:21 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-02-23 7:44 Preemption preparation pass phaw Chris Wilson
2017-02-23 7:44 ` [PATCH v4 01/16] drm/i915: Check against the signaled bit for fences/requests Chris Wilson
2017-02-23 7:44 ` [PATCH v4 02/16] drm/i915: Keep a global seqno per-engine Chris Wilson
2017-02-23 7:44 ` [PATCH v4 03/16] drm/i915: Move reserve_seqno() next to unreserve_seqno() Chris Wilson
2017-02-23 7:44 ` [PATCH v4 04/16] drm/i915: Use a local to shorten req->i915->gpu_error.wait_queue Chris Wilson
2017-02-23 7:44 ` [PATCH v4 05/16] drm/i915: Add ourselves to the gpu error waitqueue for the entire wait Chris Wilson
2017-02-23 7:44 ` [PATCH v4 06/16] drm/i915: Inline __i915_gem_request_wait_for_execute() Chris Wilson
2017-02-23 7:44 ` [PATCH v4 07/16] drm/i915: Deconstruct execute fence Chris Wilson
2017-02-23 7:44 ` [PATCH v4 08/16] drm/i915: Protect the request->global_seqno with the engine->timeline lock Chris Wilson
2017-02-23 7:44 ` [PATCH v4 09/16] drm/i915: Take a reference whilst processing the signaler request Chris Wilson
2017-02-23 7:44 ` [PATCH v4 10/16] drm/i915: Allow a request to be cancelled Chris Wilson
2017-02-23 7:44 ` [PATCH v4 11/16] drm/i915: Remove the preempted request from the execution queue Chris Wilson
2017-02-23 12:11 ` Tvrtko Ursulin
2017-02-23 7:44 ` [PATCH v4 12/16] drm/i915: Exercise request cancellation using a mock selftest Chris Wilson
2017-02-23 12:21 ` Tvrtko Ursulin [this message]
2017-02-23 7:44 ` [PATCH v4 13/16] drm/i915: Replace reset_wait_queue with default_wake_function Chris Wilson
2017-02-23 7:44 ` [PATCH v4 14/16] drm/i915: Refactor direct GPU reset from request waiters Chris Wilson
2017-02-23 12:23 ` Tvrtko Ursulin
2017-02-23 7:44 ` [PATCH v4 15/16] drm/i915: Immediately process a reset before starting waiting Chris Wilson
2017-02-23 12:35 ` Tvrtko Ursulin
2017-02-23 12:41 ` Chris Wilson
2017-02-23 12:45 ` Tvrtko Ursulin
2017-02-23 7:44 ` [PATCH v4 16/16] drm/i915: Remove one level of indention from wait-for-execute Chris Wilson
2017-02-23 8:18 ` ✓ Fi.CI.BAT: success for series starting with [v4,01/16] drm/i915: Check against the signaled bit for fences/requests Patchwork
2017-02-23 15:21 ` Chris Wilson
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=e01cb421-1cfb-71d1-20b3-5e9fcba2b363@linux.intel.com \
--to=tvrtko.ursulin@linux.intel.com \
--cc=chris@chris-wilson.co.uk \
--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