[PATCH i-g-t 1/4] tests/kms_flip: Use human readable pipe and connector names

[PATCH i-g-t 1/4] tests/kms_flip: Use human readable pipe and connector names

[ville.syrjala]

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

Print the pipes and connectors in a human readable form instead of using
the integer IDs.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 tests/kms_flip.c | 28 +++++++++++++++++++++-------
 1 file changed, 21 insertions(+), 7 deletions(-)

diff --git a/tests/kms_flip.c b/tests/kms_flip.c
index e1b5856..649678c 100644
--- a/tests/kms_flip.c
+++ b/tests/kms_flip.c
@@ -158,6 +158,7 @@ struct test_output {
 	drmModeConnector *kconnector[4];
 	uint32_t _connector[4];
 	uint32_t _crtc[4];
+	int _pipe[4];
 	int count; /* 1:1 mapping between crtc:connector */
 	int flags;
 	int pipe; /* primary pipe for vblank */
@@ -1102,6 +1103,7 @@ static void connector_find_preferred_mode(uint32_t connector_id, int crtc_idx,
 	o->kconnector[0] = config.connector;
 	o->kencoder[0] = config.encoder;
 	o->_crtc[0] = config.crtc->crtc_id;
+	o->_pipe[0] = config.pipe;
 	o->kmode[0] = config.default_mode;
 	o->mode_valid = 1;
 
@@ -1170,11 +1172,13 @@ found:
 	o->kconnector[0] = config[0].connector;
 	o->kencoder[0] = config[0].encoder;
 	o->_crtc[0] = config[0].crtc->crtc_id;
+	o->_pipe[0] = config[0].pipe;
 	o->kmode[0] = *mode[0];
 
 	o->kconnector[1] = config[1].connector;
 	o->kencoder[1] = config[1].encoder;
 	o->_crtc[1] = config[1].crtc->crtc_id;
+	o->_pipe[1] = config[1].pipe;
 	o->kmode[1] = *mode[1];
 
 	drmModeFreeCrtc(config[0].crtc);
@@ -1350,22 +1354,32 @@ static void run_test_on_crtc_set(struct test_output *o, int *crtc_idxs,
 	switch (crtc_count) {
 	case 1:
 		connector_find_preferred_mode(o->_connector[0], crtc_idxs[0], o);
+		if (!o->mode_valid)
+			return;
 		snprintf(test_name, sizeof(test_name),
-			  "%s on crtc %d, connector %d",
-			  igt_subtest_name(), o->_crtc[0], o->_connector[0]);
+			 "%s on pipe %s, connector %s-%d",
+			 igt_subtest_name(),
+			 kmstest_pipe_name(o->_pipe[0]),
+			 kmstest_connector_type_str(o->kconnector[0]->connector_type),
+			 o->kconnector[0]->connector_type_id);
 		break;
 	case 2:
 		connector_find_compatible_mode(crtc_idxs[0], crtc_idxs[1], o);
+		if (!o->mode_valid)
+			return;
 		snprintf(test_name, sizeof(test_name),
-			 "%s on crtc %d:%d, connector %d:%d",
-			 igt_subtest_name(), o->_crtc[0], o->_crtc[1],
-			 o->_connector[0], o->_connector[1]);
+			 "%s on pipe %s:%s, connector %s-%d:%s-%d",
+			 igt_subtest_name(),
+			 kmstest_pipe_name(o->_pipe[0]),
+			 kmstest_pipe_name(o->_pipe[1]),
+			 kmstest_connector_type_str(o->kconnector[0]->connector_type),
+			 o->kconnector[0]->connector_type_id,
+			 kmstest_connector_type_str(o->kconnector[1]->connector_type),
+			 o->kconnector[1]->connector_type_id);
 		break;
 	default:
 		igt_assert(0);
 	}
-	if (!o->mode_valid)
-		return;
 
 	igt_assert_eq(o->count, crtc_count);
 
-- 
2.4.10

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

[PATCH i-g-t 2/4] tests/kms_flip: Disable all pipes before each test

[ville.syrjala]

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

Currently kms_flip leaks the state of the pipes from one subtest to the
next. Meaning a single pipe test can actually have two or more pipes
actually up and running, and similarly a two pipe test can have three
pipes running.

This is particularly nasty on IVB since one of the pipes still running
but not actually part of the test maybe have reserved the shared FDI
lanes, thus preventing one of the pipes taking part in the test from
being enabled.

To avoid such problems explicitly disable all pipes before each
subtests.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 tests/kms_flip.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/tests/kms_flip.c b/tests/kms_flip.c
index 649678c..632f264 100644
--- a/tests/kms_flip.c
+++ b/tests/kms_flip.c
@@ -849,6 +849,22 @@ static bool is_hung(int fd)
 	return errno == EIO;
 }
 
+static void disable_crtcs(void)
+{
+	int n;
+
+	for (n = 0; n < resources->count_crtcs; n++) {
+		drmModeCrtc *crtc = drmModeGetCrtc(drm_fd, resources->crtcs[n]);
+
+		int ret = drmModeSetCrtc(drm_fd, crtc->crtc_id,
+					 0, 0, 0,
+					 0, 0, 0);
+		igt_assert(ret == 0);
+
+		drmModeFreeCrtc(crtc);
+	}
+}
+
 static int set_mode(struct test_output *o, uint32_t fb, int x, int y)
 {
 	int n;
@@ -1425,6 +1441,8 @@ static void run_test_on_crtc_set(struct test_output *o, int *crtc_idxs,
 	for (i = 0; i < o->count; i++)
 		kmstest_dump_mode(&o->kmode[i]);
 
+	disable_crtcs();
+
 	if (set_mode(o, o->fb_ids[0], 0, 0)) {
 		/* We may fail to apply the mode if there are hidden
 		 * constraints, such as bandwidth on the third pipe.
-- 
2.4.10

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

[PATCH i-g-t 3/4] tests/kms_flip: Modeset pipes in reverse order

[ville.syrjala]

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

To make more multi-pipe tests run on IVB, do the modesets in the reverse
order (ie. pipe C first, pipe A last). This way pipe B can't reserve the
2 shared FDI lanes before pipe C is set up.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 tests/kms_flip.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/kms_flip.c b/tests/kms_flip.c
index 632f264..23dadad 100644
--- a/tests/kms_flip.c
+++ b/tests/kms_flip.c
@@ -869,7 +869,7 @@ static int set_mode(struct test_output *o, uint32_t fb, int x, int y)
 {
 	int n;
 
-	for (n = 0; n < o->count; n++) {
+	for (n = o->count - 1; n >= 0; n--) {
 		if (fb == 0) {
 			int ret = drmModeSetCrtc(drm_fd, o->_crtc[n],
 						 0, 0, 0,
-- 
2.4.10

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

[PATCH i-g-t 4/4] tests/kms_flip: Add a note that the test was skipped when modeset fails

[ville.syrjala]

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

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 tests/kms_flip.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tests/kms_flip.c b/tests/kms_flip.c
index 23dadad..af1ccfb 100644
--- a/tests/kms_flip.c
+++ b/tests/kms_flip.c
@@ -1449,6 +1449,7 @@ static void run_test_on_crtc_set(struct test_output *o, int *crtc_idxs,
 		 */
 		igt_assert_f(crtc_count > 1 || crtc_idxs[0] < 2,
 			     "set_mode may only fail on the 3rd pipe or in multiple crtc tests\n");
+		igt_info("\n%s: SKIPPED\n\n", test_name);
 		goto out;
 	}
 	igt_assert(fb_is_bound(o, o->fb_ids[0]));
-- 
2.4.10

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

Re: [PATCH i-g-t 2/4] tests/kms_flip: Disable all pipes before each test

[Paulo Zanoni]

2015-11-11 15:32 GMT-02:00  <ville.syrjala@linux.intel.com>:
> From: Ville Syrjälä <ville.syrjala@linux.intel.com>
>
> Currently kms_flip leaks the state of the pipes from one subtest to the
> next. Meaning a single pipe test can actually have two or more pipes
> actually up and running, and similarly a two pipe test can have three
> pipes running.
>
> This is particularly nasty on IVB since one of the pipes still running
> but not actually part of the test maybe have reserved the shared FDI
> lanes, thus preventing one of the pipes taking part in the test from
> being enabled.
>
> To avoid such problems explicitly disable all pipes before each
> subtests.
>
> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> ---
>  tests/kms_flip.c | 18 ++++++++++++++++++
>  1 file changed, 18 insertions(+)
>
> diff --git a/tests/kms_flip.c b/tests/kms_flip.c
> index 649678c..632f264 100644
> --- a/tests/kms_flip.c
> +++ b/tests/kms_flip.c
> @@ -849,6 +849,22 @@ static bool is_hung(int fd)
>         return errno == EIO;
>  }
>
> +static void disable_crtcs(void)
> +{
> +       int n;
> +
> +       for (n = 0; n < resources->count_crtcs; n++) {
> +               drmModeCrtc *crtc = drmModeGetCrtc(drm_fd, resources->crtcs[n]);
> +
> +               int ret = drmModeSetCrtc(drm_fd, crtc->crtc_id,

You could have used resources->crtcs[n] do avoid the GetCrtc() call.

Anyway, you can avoid this new function and just call
kmstest_unset_all_crtcs(drm_fd, resources) from igt_kms.h.

> +                                        0, 0, 0,
> +                                        0, 0, 0);
> +               igt_assert(ret == 0);
> +
> +               drmModeFreeCrtc(crtc);
> +       }
> +}
> +
>  static int set_mode(struct test_output *o, uint32_t fb, int x, int y)
>  {
>         int n;
> @@ -1425,6 +1441,8 @@ static void run_test_on_crtc_set(struct test_output *o, int *crtc_idxs,
>         for (i = 0; i < o->count; i++)
>                 kmstest_dump_mode(&o->kmode[i]);
>
> +       disable_crtcs();
> +
>         if (set_mode(o, o->fb_ids[0], 0, 0)) {
>                 /* We may fail to apply the mode if there are hidden
>                  * constraints, such as bandwidth on the third pipe.
> --
> 2.4.10
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx



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

Re: [PATCH i-g-t 2/4] tests/kms_flip: Disable all pipes before each test

[Ville Syrjälä]

On Wed, Nov 11, 2015 at 03:41:16PM -0200, Paulo Zanoni wrote:
> 2015-11-11 15:32 GMT-02:00  <ville.syrjala@linux.intel.com>:
> > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> >
> > Currently kms_flip leaks the state of the pipes from one subtest to the
> > next. Meaning a single pipe test can actually have two or more pipes
> > actually up and running, and similarly a two pipe test can have three
> > pipes running.
> >
> > This is particularly nasty on IVB since one of the pipes still running
> > but not actually part of the test maybe have reserved the shared FDI
> > lanes, thus preventing one of the pipes taking part in the test from
> > being enabled.
> >
> > To avoid such problems explicitly disable all pipes before each
> > subtests.
> >
> > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > ---
> >  tests/kms_flip.c | 18 ++++++++++++++++++
> >  1 file changed, 18 insertions(+)
> >
> > diff --git a/tests/kms_flip.c b/tests/kms_flip.c
> > index 649678c..632f264 100644
> > --- a/tests/kms_flip.c
> > +++ b/tests/kms_flip.c
> > @@ -849,6 +849,22 @@ static bool is_hung(int fd)
> >         return errno == EIO;
> >  }
> >
> > +static void disable_crtcs(void)
> > +{
> > +       int n;
> > +
> > +       for (n = 0; n < resources->count_crtcs; n++) {
> > +               drmModeCrtc *crtc = drmModeGetCrtc(drm_fd, resources->crtcs[n]);
> > +
> > +               int ret = drmModeSetCrtc(drm_fd, crtc->crtc_id,
> 
> You could have used resources->crtcs[n] do avoid the GetCrtc() call.

I thought we freed those. Maybe not.

> 
> Anyway, you can avoid this new function and just call
> kmstest_unset_all_crtcs(drm_fd, resources) from igt_kms.h.

OK. That's nice.

> 
> > +                                        0, 0, 0,
> > +                                        0, 0, 0);
> > +               igt_assert(ret == 0);
> > +
> > +               drmModeFreeCrtc(crtc);
> > +       }
> > +}
> > +
> >  static int set_mode(struct test_output *o, uint32_t fb, int x, int y)
> >  {
> >         int n;
> > @@ -1425,6 +1441,8 @@ static void run_test_on_crtc_set(struct test_output *o, int *crtc_idxs,
> >         for (i = 0; i < o->count; i++)
> >                 kmstest_dump_mode(&o->kmode[i]);
> >
> > +       disable_crtcs();
> > +
> >         if (set_mode(o, o->fb_ids[0], 0, 0)) {
> >                 /* We may fail to apply the mode if there are hidden
> >                  * constraints, such as bandwidth on the third pipe.
> > --
> > 2.4.10
> >
> > _______________________________________________
> > Intel-gfx mailing list
> > Intel-gfx@lists.freedesktop.org
> > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
> 
> 
> 
> -- 
> Paulo Zanoni

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

Re: [PATCH i-g-t 2/4] tests/kms_flip: Disable all pipes before each test

[Ville Syrjälä]

On Wed, Nov 11, 2015 at 07:49:40PM +0200, Ville Syrjälä wrote:
> On Wed, Nov 11, 2015 at 03:41:16PM -0200, Paulo Zanoni wrote:
> > 2015-11-11 15:32 GMT-02:00  <ville.syrjala@linux.intel.com>:
> > > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > >
> > > Currently kms_flip leaks the state of the pipes from one subtest to the
> > > next. Meaning a single pipe test can actually have two or more pipes
> > > actually up and running, and similarly a two pipe test can have three
> > > pipes running.
> > >
> > > This is particularly nasty on IVB since one of the pipes still running
> > > but not actually part of the test maybe have reserved the shared FDI
> > > lanes, thus preventing one of the pipes taking part in the test from
> > > being enabled.
> > >
> > > To avoid such problems explicitly disable all pipes before each
> > > subtests.
> > >
> > > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > ---
> > >  tests/kms_flip.c | 18 ++++++++++++++++++
> > >  1 file changed, 18 insertions(+)
> > >
> > > diff --git a/tests/kms_flip.c b/tests/kms_flip.c
> > > index 649678c..632f264 100644
> > > --- a/tests/kms_flip.c
> > > +++ b/tests/kms_flip.c
> > > @@ -849,6 +849,22 @@ static bool is_hung(int fd)
> > >         return errno == EIO;
> > >  }
> > >
> > > +static void disable_crtcs(void)
> > > +{
> > > +       int n;
> > > +
> > > +       for (n = 0; n < resources->count_crtcs; n++) {
> > > +               drmModeCrtc *crtc = drmModeGetCrtc(drm_fd, resources->crtcs[n]);
> > > +
> > > +               int ret = drmModeSetCrtc(drm_fd, crtc->crtc_id,
> > 
> > You could have used resources->crtcs[n] do avoid the GetCrtc() call.
> 
> I thought we freed those. Maybe not.

Ah misread the whole thing. Yes, there's no point in this id->struct->id
dance I'm doing.

> 
> > 
> > Anyway, you can avoid this new function and just call
> > kmstest_unset_all_crtcs(drm_fd, resources) from igt_kms.h.
> 
> OK. That's nice.
> 
> > 
> > > +                                        0, 0, 0,
> > > +                                        0, 0, 0);
> > > +               igt_assert(ret == 0);
> > > +
> > > +               drmModeFreeCrtc(crtc);
> > > +       }
> > > +}
> > > +
> > >  static int set_mode(struct test_output *o, uint32_t fb, int x, int y)
> > >  {
> > >         int n;
> > > @@ -1425,6 +1441,8 @@ static void run_test_on_crtc_set(struct test_output *o, int *crtc_idxs,
> > >         for (i = 0; i < o->count; i++)
> > >                 kmstest_dump_mode(&o->kmode[i]);
> > >
> > > +       disable_crtcs();
> > > +
> > >         if (set_mode(o, o->fb_ids[0], 0, 0)) {
> > >                 /* We may fail to apply the mode if there are hidden
> > >                  * constraints, such as bandwidth on the third pipe.
> > > --
> > > 2.4.10
> > >
> > > _______________________________________________
> > > Intel-gfx mailing list
> > > Intel-gfx@lists.freedesktop.org
> > > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
> > 
> > 
> > 
> > -- 
> > Paulo Zanoni
> 
> -- 
> Ville Syrjälä
> Intel OTC
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

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

[PATCH i-g-t v2 2/4] tests/kms_flip: Disable all pipes before each test

[ville.syrjala]

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

Currently kms_flip leaks the state of the pipes from one subtest to the
next. Meaning a single pipe test can actually have two or more pipes
actually up and running, and similarly a two pipe test can have three
pipes running.

This is particularly nasty on IVB since one of the pipes still running
but not actually part of the test maybe have reserved the shared FDI
lanes, thus preventing one of the pipes taking part in the test from
being enabled.

To avoid such problems explicitly disable all pipes before each
subtests.

v2: Use kmstest_unset_all_crtcs() (Paulo)

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 tests/kms_flip.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tests/kms_flip.c b/tests/kms_flip.c
index 649678c..8bba121 100644
--- a/tests/kms_flip.c
+++ b/tests/kms_flip.c
@@ -1425,6 +1425,8 @@ static void run_test_on_crtc_set(struct test_output *o, int *crtc_idxs,
 	for (i = 0; i < o->count; i++)
 		kmstest_dump_mode(&o->kmode[i]);
 
+	kmstest_unset_all_crtcs(drm_fd, resources);
+
 	if (set_mode(o, o->fb_ids[0], 0, 0)) {
 		/* We may fail to apply the mode if there are hidden
 		 * constraints, such as bandwidth on the third pipe.
-- 
2.4.10

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