intel-gfx.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: meghanelogal <megha.i.nelogal@intel.com>
To: Daniel Vetter <daniel@ffwll.ch>
Cc: paulo.r.zanoni@intel.com, daniel.vetter@ffwll.ch,
	intel-gfx@lists.freedesktop.org, rodrigo.vivi@intel.com
Subject: Re: [PATCH v2] kms_atomic : Added subtest for Single Pipe DBUF validation
Date: Wed, 26 Oct 2016 14:10:49 +0530	[thread overview]
Message-ID: <58106C11.2010905@intel.com> (raw)
In-Reply-To: <20161026063931.u5bqrexg5xd3zyfo@phenom.ffwll.local>

Daniel,

I have already added platform check and is valid for platforms>=gen 9
This subtest is for Watermark/DDB kernel patches floated by Mahesh Kumar ..
     reference -> https://patchwork.freedesktop.org/series/12222/

Regards,
Megha

On Wednesday 26 October 2016 12:09 PM, Daniel Vetter wrote:
> On Tue, Oct 25, 2016 at 02:40:00PM +0530, meghanelogal wrote:
>> Existing DDB algorithm divide the DDB wrt data rate,
>> hence the planes with the less height but same width
>> will be allocated less blocks and watermark are based
>> on width which requires more DDB. With this data the flip
>> may fail.
>>
>> In new DDB algorithm, the DDB is divided based on
>> watermark requirement.
>>
>> In this subtest, dividing the htotal/200 will allocate
>> ~2-4 blocks out of total(512/896), flip may fail with the
>> exisiting algorithm.But with the new algorithm it will pass.
>>
>> Signed-off-by: Megha Nelogal <megha.i.nelogal@intel.com>
> Pretty sure I mentioned this already, but how does this test work? It
> doesn't check for a specific platform/hw combo to make sure it's
> assumptions are valid, and you can't just assume that this will always
> work on all drivers/platforms.
>
> tbh not entirely sure of the value this provides here ...
> -Daniel
>
>> ---
>>   tests/kms_atomic.c | 82 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>   1 file changed, 82 insertions(+)
>>
>> diff --git a/tests/kms_atomic.c b/tests/kms_atomic.c
>> index f27ee46..bfa4b35 100644
>> --- a/tests/kms_atomic.c
>> +++ b/tests/kms_atomic.c
>> @@ -1280,6 +1280,60 @@ static void atomic_invalid_params(struct kms_atomic_crtc_state *crtc,
>>   	do_ioctl_err(desc->fd, DRM_IOCTL_MODE_ATOMIC, &ioc, EFAULT);
>>   }
>>   
>> +static void validate_dbuf(struct kms_atomic_crtc_state *crtc,
>> +				struct kms_atomic_plane_state **plane_array,
>> +				int plane_count)
>> +{
>> +	int i;
>> +	struct kms_atomic_desc *desc = crtc->state->desc;
>> +
>> +	/* for active crtc do modeset on the native resolution */
>> +	drmModeAtomicReq *req = drmModeAtomicAlloc();
>> +	struct drm_mode_modeinfo *mode = crtc->mode.data;
>> +	struct kms_atomic_plane_state plane;
>> +	struct igt_fb fb;
>> +
>> +	crtc_populate_req(crtc, req);
>> +
>> +	/* Add plane data to the structure...*/
>> +	uint32_t crtc_x = 0;
>> +	uint32_t crtc_y = 0;
>> +
>> +	for (i = 0; i < plane_count; i++) {
>> +		plane = *plane_array[i];
>> +		uint32_t format = plane_get_igt_format(&plane);
>> +
>> +		igt_require(format != 0);
>> +		plane.src_x = 0;
>> +		plane.src_y = 0;
>> +		plane.src_w = (mode->hdisplay / (1)) << 16;
>> +		plane.crtc_x = crtc_x;
>> +		plane.crtc_y = crtc_y;
>> +		plane.crtc_w = (mode->hdisplay) / (1);
>> +		plane.crtc_h = (mode->vdisplay) / (i + 1);
>> +
>> +		plane.crtc_id = crtc->obj;
>> +
>> +		if (i%2 == 0) {
>> +			plane.src_h = (mode->vdisplay / (1)) << 16;
>> +			plane.crtc_h = (mode->vdisplay) / (1);
>> +		}
>> +
>> +		if (i%2 == 1) {
>> +			plane.src_h = ceil((mode->vdisplay / 200) << 16);
>> +			plane.crtc_h = ceil((mode->vdisplay / 200));
>> +		}
>> +
>> +		plane.fb_id = igt_create_fb(plane.state->desc->fd,
>> +						plane.crtc_w, plane.crtc_h,
>> +						format, I915_TILING_NONE, &fb);
>> +		plane_populate_req(&plane, req);
>> +	}
>> +	do_atomic_commit(desc->fd, req, ATOMIC_RELAX_NONE);
>> +	drmModeAtomicFree(req);
>> +}
>> +
>> +
>>   igt_main
>>   {
>>   	struct kms_atomic_desc desc;
>> @@ -1373,6 +1427,34 @@ igt_main
>>   		atomic_state_free(scratch);
>>   	}
>>   
>> +	igt_subtest("validate_dbuf") {
>> +		int gen;
>> +
>> +		gen = intel_gen(intel_get_drm_devid(desc.fd));
>> +		igt_require(gen >= 9);
>> +
>> +		struct kms_atomic_state *scratch = atomic_state_dup(current);
>> +		struct kms_atomic_crtc_state *crtc = find_crtc(scratch, true);
>> +		struct kms_atomic_plane_state *plane;
>> +		struct kms_atomic_connector_state *connector =
>> +				find_connector(scratch, crtc);
>> +		struct kms_atomic_plane_state **plane_array = NULL;
>> +		/* Enabling Two planes */
>> +		plane_array = calloc(2, sizeof(struct kms_atomic_plane_state *));
>> +		int plane_count = 0;
>> +
>> +		igt_require(crtc);
>> +		plane = find_plane(scratch, PLANE_TYPE_PRIMARY, crtc);
>> +		igt_require(plane);
>> +		plane_array[plane_count] = plane; plane_count++;
>> +		plane = find_plane(scratch, PLANE_TYPE_OVERLAY, crtc);
>> +		igt_require(plane);
>> +		plane_array[plane_count] = plane; plane_count++;
>> +		igt_require(connector);
>> +		validate_dbuf(crtc, plane_array, plane_count);
>> +		atomic_state_free(scratch);
>> +	}
>> +
>>   	atomic_state_free(current);
>>   
>>   	igt_fixture
>> -- 
>> 1.9.1
>>

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

      reply	other threads:[~2016-10-26  8:53 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <mailto:20161013131629.GO20761@phenom.ffwll.local>
2016-10-25  9:10 ` [PATCH v2] kms_atomic : Added subtest for Single Pipe DBUF validation meghanelogal
2016-10-25 10:10   ` Chris Wilson
2016-10-26  6:39   ` Daniel Vetter
2016-10-26  8:40     ` meghanelogal [this message]

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=58106C11.2010905@intel.com \
    --to=megha.i.nelogal@intel.com \
    --cc=daniel.vetter@ffwll.ch \
    --cc=daniel@ffwll.ch \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=paulo.r.zanoni@intel.com \
    --cc=rodrigo.vivi@intel.com \
    /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;
as well as URLs for NNTP newsgroup(s).