[i-g-t,v3] tests/core_getparams: Create new test core_getparams

Commit Message

[e-mail] March 13, 2015, 12:26 a.m. UTC

From: Jeff McGee <[e-mail]>

New test core_getparams consists of 2 subtests, each one testing
the ability of userspace to query the correct value of a GT config
attribute: subslice total or EU total. drm/i915 implementation of
these queries is required for Cherryview and Gen9+ devices (non-
simulated).

v2: Duplicate small amount of new libdrm functionality to avoid
    bumping libdrm version requirement (Daniel). Convert some
    igt_asserts to the appropriate comparison variants. Add a
    test description.
v3: Actually use the LOCAL GETPARAM defines. Otherwise can't build
    against older libdrm as intended by v2.

For: VIZ-4636
Signed-off-by: Jeff McGee <[e-mail]>
---
 tests/.gitignore       |   1 +
 tests/Makefile.sources |   1 +
 tests/core_getparams.c | 167 +++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 169 insertions(+)
 create mode 100644 tests/core_getparams.c

Comments

Daniel Vetter March 13, 2015, 8:58 a.m. UTC | #1

On Thu, Mar 12, 2015 at 05:26:25PM -0700, [e-mail] wrote:
> From: Jeff McGee <[e-mail]>
> 
> New test core_getparams consists of 2 subtests, each one testing
> the ability of userspace to query the correct value of a GT config
> attribute: subslice total or EU total. drm/i915 implementation of
> these queries is required for Cherryview and Gen9+ devices (non-
> simulated).
> 
> v2: Duplicate small amount of new libdrm functionality to avoid
>     bumping libdrm version requirement (Daniel). Convert some
>     igt_asserts to the appropriate comparison variants. Add a
>     test description.
> v3: Actually use the LOCAL GETPARAM defines. Otherwise can't build
>     against older libdrm as intended by v2.
> 
> For: VIZ-4636
> Signed-off-by: Jeff McGee <[e-mail]>
> ---
>  tests/.gitignore       |   1 +
>  tests/Makefile.sources |   1 +
>  tests/core_getparams.c | 167 +++++++++++++++++++++++++++++++++++++++++++++++++
>  3 files changed, 169 insertions(+)
>  create mode 100644 tests/core_getparams.c
> 
> diff --git a/tests/.gitignore b/tests/.gitignore
> index 426cc67..c742308 100644
> --- a/tests/.gitignore
> +++ b/tests/.gitignore
> @@ -1,6 +1,7 @@
>  # Please keep sorted alphabetically
>  core_get_client_auth
>  core_getclient
> +core_getparams
>  core_getstats
>  core_getversion
>  drm_import_export
> diff --git a/tests/Makefile.sources b/tests/Makefile.sources
> index 51e8376..999c8f8 100644
> --- a/tests/Makefile.sources
> +++ b/tests/Makefile.sources
> @@ -15,6 +15,7 @@ NOUVEAU_TESTS_M = \
>  
>  TESTS_progs_M = \
>  	core_get_client_auth \
> +	core_getparams \

Sorry I missed this little one: core_ is for drm core stuff shared by all
drivers (i.e. even those not supporting gem or kms). getparam is an i915
specific thing, so imo better to have drv_ as a prefix.

For our naming conventions and igt test sections see the docs at

http://people.freedesktop.org/~danvet/igt/

For review I think signing up someone from the beignet team might be best.
Cheers, Daniel

>  	drv_suspend \
>  	drv_hangman \
>  	gem_bad_reloc \
> diff --git a/tests/core_getparams.c b/tests/core_getparams.c
> new file mode 100644
> index 0000000..2855d06
> --- /dev/null
> +++ b/tests/core_getparams.c
> @@ -0,0 +1,167 @@
> +/*
> + * Copyright © 2014 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:
> + *    Jeff McGee <[e-mail]>
> + *
> + */
> +
> +#include <unistd.h>
> +#include <errno.h>
> +#include <xf86drm.h>
> +#include <i915_drm.h>
> +#include "drmtest.h"
> +#include "intel_chipset.h"
> +#include "intel_bufmgr.h"
> +
> +IGT_TEST_DESCRIPTION("Tests the export of parameters via DRM_IOCTL_I915_GETPARAM\n");
> +
> +int drm_fd;
> +int devid;
> +
> +static void
> +init(void)
> +{
> +	drm_fd = drm_open_any();
> +	devid = intel_get_drm_devid(drm_fd);
> +}
> +
> +static void
> +deinit(void)
> +{
> +	close(drm_fd);
> +}
> +
> +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL	33
> +#define LOCAL_I915_PARAM_EU_TOTAL	34
> +
> +static int
> +getparam(int param, int *value)
> +{
> +	drm_i915_getparam_t gp;
> +	int ret;
> +
> +	memset(&gp, 0, sizeof(gp));
> +	gp.value = value;
> +	gp.param = param;
> +	ret = drmIoctl(drm_fd, DRM_IOCTL_I915_GETPARAM, &gp);
> +	if (ret)
> +		return -errno;
> +
> +	return 0;
> +}
> +
> +static void
> +subslice_total(void)
> +{
> +	unsigned int subslice_total = 0;
> +	int ret;
> +
> +	ret = getparam(LOCAL_I915_PARAM_SUBSLICE_TOTAL, (int*)&subslice_total);
> +
> +	if (ret) {
> +		/*
> +		 * These devices are not required to implement the
> +		 * interface. If they do not, -ENODEV must be returned.
> +		*/
> +		if ((intel_gen(devid) < 8) ||
> +		    IS_BROADWELL(devid) ||
> +		    igt_run_in_simulation()) {
> +			igt_assert_eq(ret, -ENODEV);
> +			igt_info("subslice total: unknown\n");
> +		/*
> +		 * All other devices must implement the interface, so
> +		 * fail them if we are here.
> +		*/
> +		} else {
> +			igt_assert_neq(ret, EINVAL); /* request not recognized? */
> +			igt_assert_neq(ret, ENODEV); /* device not supported? */
> +			igt_assert_eq(ret, 0); /* other error? */
> +		}
> +	} else {
> +		/*
> +		 * On success, just make sure the returned count value is
> +		 * non-zero. The validity of the count value for the given
> +		 * device is not checked.
> +		*/
> +		igt_assert_neq(subslice_total, 0);
> +		igt_info("subslice total: %u\n", subslice_total);
> +	}
> +}
> +
> +static void
> +eu_total(void)
> +{
> +	unsigned int eu_total = 0;
> +	int ret;
> +
> +	ret = getparam(LOCAL_I915_PARAM_EU_TOTAL, (int*)&eu_total);
> +
> +	if (ret) {
> +		/*
> +		 * These devices are not required to implement the
> +		 * interface. If they do not, -ENODEV must be returned.
> +		*/
> +		if ((intel_gen(devid) < 8) ||
> +		    IS_BROADWELL(devid) ||
> +		    igt_run_in_simulation()) {
> +			igt_assert_eq(ret, -ENODEV);
> +			igt_info("EU total: unknown\n");
> +		/*
> +		 * All other devices must implement the interface, so
> +		 * fail them if we are here.
> +		*/
> +		} else {
> +			igt_assert_neq(ret, EINVAL); /* request not recognized? */
> +			igt_assert_neq(ret, ENODEV); /* device not supported? */
> +			igt_assert_eq(ret, 0); /* other error? */
> +		}
> +	} else {
> +		/*
> +		 * On success, just make sure the returned count value is
> +		 * non-zero. The validity of the count value for the given
> +		 * device is not checked.
> +		*/
> +		igt_assert_neq(eu_total, 0);
> +		igt_info("EU total: %u\n", eu_total);
> +	}
> +}
> +
> +static void
> +exit_handler(int sig)
> +{
> +	deinit();
> +}
> +
> +igt_main
> +{
> +	igt_fixture {
> +		igt_install_exit_handler(exit_handler);
> +		init();
> +	}
> +
> +	igt_subtest("subslice-total")
> +		subslice_total();
> +
> +	igt_subtest("eu-total")
> +		eu_total();
> +}
> -- 
> 2.3.0
> 
> _______________________________________________
> dri-devel mailing list
> [e-mail]
> http://lists.freedesktop.org/mailman/listinfo/dri-devel

[e-mail] March 13, 2015, 9:09 a.m. UTC | #2

My only concern is about the following macros:

> +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL      33
> +#define LOCAL_I915_PARAM_EU_TOTAL    34

How about to just use the definitons in the kernel header file?
For an example:

  #include <drm/i915_drm.h>

  #ifdef LOCAL_I915_PARAM_SUBSLICE_TOTAL
  //Put all the code into this block.
  #endif

Then we can avoid put the same definitions in different files,
and we can avoid unecessary testing on an old kernel which doesn't
have this kernel interface.

For all the other part, it LGTM.

Reviewed-by: Zhigang Gong <[e-mail]>

Thanks,
Zhigang Gong.

On Thu, Mar 12, 2015 at 05:26:25PM -0700, [e-mail] wrote:
> From: Jeff McGee <[e-mail]>
> 
> New test core_getparams consists of 2 subtests, each one testing
> the ability of userspace to query the correct value of a GT config
> attribute: subslice total or EU total. drm/i915 implementation of
> these queries is required for Cherryview and Gen9+ devices (non-
> simulated).
> 
> v2: Duplicate small amount of new libdrm functionality to avoid
>     bumping libdrm version requirement (Daniel). Convert some
>     igt_asserts to the appropriate comparison variants. Add a
>     test description.
> v3: Actually use the LOCAL GETPARAM defines. Otherwise can't build
>     against older libdrm as intended by v2.
> 
> For: VIZ-4636
> Signed-off-by: Jeff McGee <[e-mail]>
> ---
>  tests/.gitignore       |   1 +
>  tests/Makefile.sources |   1 +
>  tests/core_getparams.c | 167 +++++++++++++++++++++++++++++++++++++++++++++++++
>  3 files changed, 169 insertions(+)
>  create mode 100644 tests/core_getparams.c
> 
> diff --git a/tests/.gitignore b/tests/.gitignore
> index 426cc67..c742308 100644
> --- a/tests/.gitignore
> +++ b/tests/.gitignore
> @@ -1,6 +1,7 @@
>  # Please keep sorted alphabetically
>  core_get_client_auth
>  core_getclient
> +core_getparams
>  core_getstats
>  core_getversion
>  drm_import_export
> diff --git a/tests/Makefile.sources b/tests/Makefile.sources
> index 51e8376..999c8f8 100644
> --- a/tests/Makefile.sources
> +++ b/tests/Makefile.sources
> @@ -15,6 +15,7 @@ NOUVEAU_TESTS_M = \
>  
>  TESTS_progs_M = \
>  	core_get_client_auth \
> +	core_getparams \
>  	drv_suspend \
>  	drv_hangman \
>  	gem_bad_reloc \
> diff --git a/tests/core_getparams.c b/tests/core_getparams.c
> new file mode 100644
> index 0000000..2855d06
> --- /dev/null
> +++ b/tests/core_getparams.c
> @@ -0,0 +1,167 @@
> +/*
> + * Copyright © 2014 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:
> + *    Jeff McGee <[e-mail]>
> + *
> + */
> +
> +#include <unistd.h>
> +#include <errno.h>
> +#include <xf86drm.h>
> +#include <i915_drm.h>
> +#include "drmtest.h"
> +#include "intel_chipset.h"
> +#include "intel_bufmgr.h"
> +
> +IGT_TEST_DESCRIPTION("Tests the export of parameters via DRM_IOCTL_I915_GETPARAM\n");
> +
> +int drm_fd;
> +int devid;
> +
> +static void
> +init(void)
> +{
> +	drm_fd = drm_open_any();
> +	devid = intel_get_drm_devid(drm_fd);
> +}
> +
> +static void
> +deinit(void)
> +{
> +	close(drm_fd);
> +}
> +
> +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL	33
> +#define LOCAL_I915_PARAM_EU_TOTAL	34
> +
> +static int
> +getparam(int param, int *value)
> +{
> +	drm_i915_getparam_t gp;
> +	int ret;
> +
> +	memset(&gp, 0, sizeof(gp));
> +	gp.value = value;
> +	gp.param = param;
> +	ret = drmIoctl(drm_fd, DRM_IOCTL_I915_GETPARAM, &gp);
> +	if (ret)
> +		return -errno;
> +
> +	return 0;
> +}
> +
> +static void
> +subslice_total(void)
> +{
> +	unsigned int subslice_total = 0;
> +	int ret;
> +
> +	ret = getparam(LOCAL_I915_PARAM_SUBSLICE_TOTAL, (int*)&subslice_total);
> +
> +	if (ret) {
> +		/*
> +		 * These devices are not required to implement the
> +		 * interface. If they do not, -ENODEV must be returned.
> +		*/
> +		if ((intel_gen(devid) < 8) ||
> +		    IS_BROADWELL(devid) ||
> +		    igt_run_in_simulation()) {
> +			igt_assert_eq(ret, -ENODEV);
> +			igt_info("subslice total: unknown\n");
> +		/*
> +		 * All other devices must implement the interface, so
> +		 * fail them if we are here.
> +		*/
> +		} else {
> +			igt_assert_neq(ret, EINVAL); /* request not recognized? */
> +			igt_assert_neq(ret, ENODEV); /* device not supported? */
> +			igt_assert_eq(ret, 0); /* other error? */
> +		}
> +	} else {
> +		/*
> +		 * On success, just make sure the returned count value is
> +		 * non-zero. The validity of the count value for the given
> +		 * device is not checked.
> +		*/
> +		igt_assert_neq(subslice_total, 0);
> +		igt_info("subslice total: %u\n", subslice_total);
> +	}
> +}
> +
> +static void
> +eu_total(void)
> +{
> +	unsigned int eu_total = 0;
> +	int ret;
> +
> +	ret = getparam(LOCAL_I915_PARAM_EU_TOTAL, (int*)&eu_total);
> +
> +	if (ret) {
> +		/*
> +		 * These devices are not required to implement the
> +		 * interface. If they do not, -ENODEV must be returned.
> +		*/
> +		if ((intel_gen(devid) < 8) ||
> +		    IS_BROADWELL(devid) ||
> +		    igt_run_in_simulation()) {
> +			igt_assert_eq(ret, -ENODEV);
> +			igt_info("EU total: unknown\n");
> +		/*
> +		 * All other devices must implement the interface, so
> +		 * fail them if we are here.
> +		*/
> +		} else {
> +			igt_assert_neq(ret, EINVAL); /* request not recognized? */
> +			igt_assert_neq(ret, ENODEV); /* device not supported? */
> +			igt_assert_eq(ret, 0); /* other error? */
> +		}
> +	} else {
> +		/*
> +		 * On success, just make sure the returned count value is
> +		 * non-zero. The validity of the count value for the given
> +		 * device is not checked.
> +		*/
> +		igt_assert_neq(eu_total, 0);
> +		igt_info("EU total: %u\n", eu_total);
> +	}
> +}
> +
> +static void
> +exit_handler(int sig)
> +{
> +	deinit();
> +}
> +
> +igt_main
> +{
> +	igt_fixture {
> +		igt_install_exit_handler(exit_handler);
> +		init();
> +	}
> +
> +	igt_subtest("subslice-total")
> +		subslice_total();
> +
> +	igt_subtest("eu-total")
> +		eu_total();
> +}
> -- 
> 2.3.0
> 
> _______________________________________________
> Intel-gfx mailing list
> [e-mail]
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Daniel Vetter March 13, 2015, 4:32 p.m. UTC | #3

On Fri, Mar 13, 2015 at 05:09:46PM +0800, Zhigang Gong wrote:
> My only concern is about the following macros:
> 
> > +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL      33
> > +#define LOCAL_I915_PARAM_EU_TOTAL    34
> 
> How about to just use the definitons in the kernel header file?
> For an example:
> 
>   #include <drm/i915_drm.h>
> 
>   #ifdef LOCAL_I915_PARAM_SUBSLICE_TOTAL
>   //Put all the code into this block.
>   #endif
> 
> Then we can avoid put the same definitions in different files,
> and we can avoid unecessary testing on an old kernel which doesn't
> have this kernel interface.
> 
> For all the other part, it LGTM.
> 
> Reviewed-by: Zhigang Gong <[e-mail]>

Once we update the libdrm requirements in igt we tend to go around and
replace all the now obsolete LOCAL_ defines. Imo not worth doing extra
work until then.

Patch applied, thanks.
-Daniel

> 
> Thanks,
> Zhigang Gong.
> 
> On Thu, Mar 12, 2015 at 05:26:25PM -0700, [e-mail] wrote:
> > From: Jeff McGee <[e-mail]>
> > 
> > New test core_getparams consists of 2 subtests, each one testing
> > the ability of userspace to query the correct value of a GT config
> > attribute: subslice total or EU total. drm/i915 implementation of
> > these queries is required for Cherryview and Gen9+ devices (non-
> > simulated).
> > 
> > v2: Duplicate small amount of new libdrm functionality to avoid
> >     bumping libdrm version requirement (Daniel). Convert some
> >     igt_asserts to the appropriate comparison variants. Add a
> >     test description.
> > v3: Actually use the LOCAL GETPARAM defines. Otherwise can't build
> >     against older libdrm as intended by v2.
> > 
> > For: VIZ-4636
> > Signed-off-by: Jeff McGee <[e-mail]>
> > ---
> >  tests/.gitignore       |   1 +
> >  tests/Makefile.sources |   1 +
> >  tests/core_getparams.c | 167 +++++++++++++++++++++++++++++++++++++++++++++++++
> >  3 files changed, 169 insertions(+)
> >  create mode 100644 tests/core_getparams.c
> > 
> > diff --git a/tests/.gitignore b/tests/.gitignore
> > index 426cc67..c742308 100644
> > --- a/tests/.gitignore
> > +++ b/tests/.gitignore
> > @@ -1,6 +1,7 @@
> >  # Please keep sorted alphabetically
> >  core_get_client_auth
> >  core_getclient
> > +core_getparams
> >  core_getstats
> >  core_getversion
> >  drm_import_export
> > diff --git a/tests/Makefile.sources b/tests/Makefile.sources
> > index 51e8376..999c8f8 100644
> > --- a/tests/Makefile.sources
> > +++ b/tests/Makefile.sources
> > @@ -15,6 +15,7 @@ NOUVEAU_TESTS_M = \
> >  
> >  TESTS_progs_M = \
> >  	core_get_client_auth \
> > +	core_getparams \
> >  	drv_suspend \
> >  	drv_hangman \
> >  	gem_bad_reloc \
> > diff --git a/tests/core_getparams.c b/tests/core_getparams.c
> > new file mode 100644
> > index 0000000..2855d06
> > --- /dev/null
> > +++ b/tests/core_getparams.c
> > @@ -0,0 +1,167 @@
> > +/*
> > + * Copyright © 2014 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:
> > + *    Jeff McGee <[e-mail]>
> > + *
> > + */
> > +
> > +#include <unistd.h>
> > +#include <errno.h>
> > +#include <xf86drm.h>
> > +#include <i915_drm.h>
> > +#include "drmtest.h"
> > +#include "intel_chipset.h"
> > +#include "intel_bufmgr.h"
> > +
> > +IGT_TEST_DESCRIPTION("Tests the export of parameters via DRM_IOCTL_I915_GETPARAM\n");
> > +
> > +int drm_fd;
> > +int devid;
> > +
> > +static void
> > +init(void)
> > +{
> > +	drm_fd = drm_open_any();
> > +	devid = intel_get_drm_devid(drm_fd);
> > +}
> > +
> > +static void
> > +deinit(void)
> > +{
> > +	close(drm_fd);
> > +}
> > +
> > +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL	33
> > +#define LOCAL_I915_PARAM_EU_TOTAL	34
> > +
> > +static int
> > +getparam(int param, int *value)
> > +{
> > +	drm_i915_getparam_t gp;
> > +	int ret;
> > +
> > +	memset(&gp, 0, sizeof(gp));
> > +	gp.value = value;
> > +	gp.param = param;
> > +	ret = drmIoctl(drm_fd, DRM_IOCTL_I915_GETPARAM, &gp);
> > +	if (ret)
> > +		return -errno;
> > +
> > +	return 0;
> > +}
> > +
> > +static void
> > +subslice_total(void)
> > +{
> > +	unsigned int subslice_total = 0;
> > +	int ret;
> > +
> > +	ret = getparam(LOCAL_I915_PARAM_SUBSLICE_TOTAL, (int*)&subslice_total);
> > +
> > +	if (ret) {
> > +		/*
> > +		 * These devices are not required to implement the
> > +		 * interface. If they do not, -ENODEV must be returned.
> > +		*/
> > +		if ((intel_gen(devid) < 8) ||
> > +		    IS_BROADWELL(devid) ||
> > +		    igt_run_in_simulation()) {
> > +			igt_assert_eq(ret, -ENODEV);
> > +			igt_info("subslice total: unknown\n");
> > +		/*
> > +		 * All other devices must implement the interface, so
> > +		 * fail them if we are here.
> > +		*/
> > +		} else {
> > +			igt_assert_neq(ret, EINVAL); /* request not recognized? */
> > +			igt_assert_neq(ret, ENODEV); /* device not supported? */
> > +			igt_assert_eq(ret, 0); /* other error? */
> > +		}
> > +	} else {
> > +		/*
> > +		 * On success, just make sure the returned count value is
> > +		 * non-zero. The validity of the count value for the given
> > +		 * device is not checked.
> > +		*/
> > +		igt_assert_neq(subslice_total, 0);
> > +		igt_info("subslice total: %u\n", subslice_total);
> > +	}
> > +}
> > +
> > +static void
> > +eu_total(void)
> > +{
> > +	unsigned int eu_total = 0;
> > +	int ret;
> > +
> > +	ret = getparam(LOCAL_I915_PARAM_EU_TOTAL, (int*)&eu_total);
> > +
> > +	if (ret) {
> > +		/*
> > +		 * These devices are not required to implement the
> > +		 * interface. If they do not, -ENODEV must be returned.
> > +		*/
> > +		if ((intel_gen(devid) < 8) ||
> > +		    IS_BROADWELL(devid) ||
> > +		    igt_run_in_simulation()) {
> > +			igt_assert_eq(ret, -ENODEV);
> > +			igt_info("EU total: unknown\n");
> > +		/*
> > +		 * All other devices must implement the interface, so
> > +		 * fail them if we are here.
> > +		*/
> > +		} else {
> > +			igt_assert_neq(ret, EINVAL); /* request not recognized? */
> > +			igt_assert_neq(ret, ENODEV); /* device not supported? */
> > +			igt_assert_eq(ret, 0); /* other error? */
> > +		}
> > +	} else {
> > +		/*
> > +		 * On success, just make sure the returned count value is
> > +		 * non-zero. The validity of the count value for the given
> > +		 * device is not checked.
> > +		*/
> > +		igt_assert_neq(eu_total, 0);
> > +		igt_info("EU total: %u\n", eu_total);
> > +	}
> > +}
> > +
> > +static void
> > +exit_handler(int sig)
> > +{
> > +	deinit();
> > +}
> > +
> > +igt_main
> > +{
> > +	igt_fixture {
> > +		igt_install_exit_handler(exit_handler);
> > +		init();
> > +	}
> > +
> > +	igt_subtest("subslice-total")
> > +		subslice_total();
> > +
> > +	igt_subtest("eu-total")
> > +		eu_total();
> > +}
> > -- 
> > 2.3.0
> > 
> > _______________________________________________
> > Intel-gfx mailing list
> > [e-mail]
> > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
> _______________________________________________
> Intel-gfx mailing list
> [e-mail]
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

[e-mail] March 13, 2015, 4:51 p.m. UTC | #4

On Fri, Mar 13, 2015 at 05:32:41PM +0100, Daniel Vetter wrote:
> On Fri, Mar 13, 2015 at 05:09:46PM +0800, Zhigang Gong wrote:
> > My only concern is about the following macros:
> > 
> > > +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL      33
> > > +#define LOCAL_I915_PARAM_EU_TOTAL    34
> > 
> > How about to just use the definitons in the kernel header file?
> > For an example:
> > 
> >   #include <drm/i915_drm.h>
> > 
> >   #ifdef LOCAL_I915_PARAM_SUBSLICE_TOTAL
> >   //Put all the code into this block.
> >   #endif
> > 
> > Then we can avoid put the same definitions in different files,
> > and we can avoid unecessary testing on an old kernel which doesn't
> > have this kernel interface.
> > 
> > For all the other part, it LGTM.
> > 
> > Reviewed-by: Zhigang Gong <[e-mail]>
> 
> Once we update the libdrm requirements in igt we tend to go around and
> replace all the now obsolete LOCAL_ defines. Imo not worth doing extra
> work until then.
> 
> Patch applied, thanks.
> -Daniel
> 

Patch applied? Do you want me to make the name change first? Should the
kernel part be reviewed and merged first?
-Jeff

> > 
> > Thanks,
> > Zhigang Gong.
> > 
> > On Thu, Mar 12, 2015 at 05:26:25PM -0700, [e-mail] wrote:
> > > From: Jeff McGee <[e-mail]>
> > > 
> > > New test core_getparams consists of 2 subtests, each one testing
> > > the ability of userspace to query the correct value of a GT config
> > > attribute: subslice total or EU total. drm/i915 implementation of
> > > these queries is required for Cherryview and Gen9+ devices (non-
> > > simulated).
> > > 
> > > v2: Duplicate small amount of new libdrm functionality to avoid
> > >     bumping libdrm version requirement (Daniel). Convert some
> > >     igt_asserts to the appropriate comparison variants. Add a
> > >     test description.
> > > v3: Actually use the LOCAL GETPARAM defines. Otherwise can't build
> > >     against older libdrm as intended by v2.
> > > 
> > > For: VIZ-4636
> > > Signed-off-by: Jeff McGee <[e-mail]>
> > > ---
> > >  tests/.gitignore       |   1 +
> > >  tests/Makefile.sources |   1 +
> > >  tests/core_getparams.c | 167 +++++++++++++++++++++++++++++++++++++++++++++++++
> > >  3 files changed, 169 insertions(+)
> > >  create mode 100644 tests/core_getparams.c
> > > 
> > > diff --git a/tests/.gitignore b/tests/.gitignore
> > > index 426cc67..c742308 100644
> > > --- a/tests/.gitignore
> > > +++ b/tests/.gitignore
> > > @@ -1,6 +1,7 @@
> > >  # Please keep sorted alphabetically
> > >  core_get_client_auth
> > >  core_getclient
> > > +core_getparams
> > >  core_getstats
> > >  core_getversion
> > >  drm_import_export
> > > diff --git a/tests/Makefile.sources b/tests/Makefile.sources
> > > index 51e8376..999c8f8 100644
> > > --- a/tests/Makefile.sources
> > > +++ b/tests/Makefile.sources
> > > @@ -15,6 +15,7 @@ NOUVEAU_TESTS_M = \
> > >  
> > >  TESTS_progs_M = \
> > >  	core_get_client_auth \
> > > +	core_getparams \
> > >  	drv_suspend \
> > >  	drv_hangman \
> > >  	gem_bad_reloc \
> > > diff --git a/tests/core_getparams.c b/tests/core_getparams.c
> > > new file mode 100644
> > > index 0000000..2855d06
> > > --- /dev/null
> > > +++ b/tests/core_getparams.c
> > > @@ -0,0 +1,167 @@
> > > +/*
> > > + * Copyright © 2014 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:
> > > + *    Jeff McGee <[e-mail]>
> > > + *
> > > + */
> > > +
> > > +#include <unistd.h>
> > > +#include <errno.h>
> > > +#include <xf86drm.h>
> > > +#include <i915_drm.h>
> > > +#include "drmtest.h"
> > > +#include "intel_chipset.h"
> > > +#include "intel_bufmgr.h"
> > > +
> > > +IGT_TEST_DESCRIPTION("Tests the export of parameters via DRM_IOCTL_I915_GETPARAM\n");
> > > +
> > > +int drm_fd;
> > > +int devid;
> > > +
> > > +static void
> > > +init(void)
> > > +{
> > > +	drm_fd = drm_open_any();
> > > +	devid = intel_get_drm_devid(drm_fd);
> > > +}
> > > +
> > > +static void
> > > +deinit(void)
> > > +{
> > > +	close(drm_fd);
> > > +}
> > > +
> > > +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL	33
> > > +#define LOCAL_I915_PARAM_EU_TOTAL	34
> > > +
> > > +static int
> > > +getparam(int param, int *value)
> > > +{
> > > +	drm_i915_getparam_t gp;
> > > +	int ret;
> > > +
> > > +	memset(&gp, 0, sizeof(gp));
> > > +	gp.value = value;
> > > +	gp.param = param;
> > > +	ret = drmIoctl(drm_fd, DRM_IOCTL_I915_GETPARAM, &gp);
> > > +	if (ret)
> > > +		return -errno;
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static void
> > > +subslice_total(void)
> > > +{
> > > +	unsigned int subslice_total = 0;
> > > +	int ret;
> > > +
> > > +	ret = getparam(LOCAL_I915_PARAM_SUBSLICE_TOTAL, (int*)&subslice_total);
> > > +
> > > +	if (ret) {
> > > +		/*
> > > +		 * These devices are not required to implement the
> > > +		 * interface. If they do not, -ENODEV must be returned.
> > > +		*/
> > > +		if ((intel_gen(devid) < 8) ||
> > > +		    IS_BROADWELL(devid) ||
> > > +		    igt_run_in_simulation()) {
> > > +			igt_assert_eq(ret, -ENODEV);
> > > +			igt_info("subslice total: unknown\n");
> > > +		/*
> > > +		 * All other devices must implement the interface, so
> > > +		 * fail them if we are here.
> > > +		*/
> > > +		} else {
> > > +			igt_assert_neq(ret, EINVAL); /* request not recognized? */
> > > +			igt_assert_neq(ret, ENODEV); /* device not supported? */
> > > +			igt_assert_eq(ret, 0); /* other error? */
> > > +		}
> > > +	} else {
> > > +		/*
> > > +		 * On success, just make sure the returned count value is
> > > +		 * non-zero. The validity of the count value for the given
> > > +		 * device is not checked.
> > > +		*/
> > > +		igt_assert_neq(subslice_total, 0);
> > > +		igt_info("subslice total: %u\n", subslice_total);
> > > +	}
> > > +}
> > > +
> > > +static void
> > > +eu_total(void)
> > > +{
> > > +	unsigned int eu_total = 0;
> > > +	int ret;
> > > +
> > > +	ret = getparam(LOCAL_I915_PARAM_EU_TOTAL, (int*)&eu_total);
> > > +
> > > +	if (ret) {
> > > +		/*
> > > +		 * These devices are not required to implement the
> > > +		 * interface. If they do not, -ENODEV must be returned.
> > > +		*/
> > > +		if ((intel_gen(devid) < 8) ||
> > > +		    IS_BROADWELL(devid) ||
> > > +		    igt_run_in_simulation()) {
> > > +			igt_assert_eq(ret, -ENODEV);
> > > +			igt_info("EU total: unknown\n");
> > > +		/*
> > > +		 * All other devices must implement the interface, so
> > > +		 * fail them if we are here.
> > > +		*/
> > > +		} else {
> > > +			igt_assert_neq(ret, EINVAL); /* request not recognized? */
> > > +			igt_assert_neq(ret, ENODEV); /* device not supported? */
> > > +			igt_assert_eq(ret, 0); /* other error? */
> > > +		}
> > > +	} else {
> > > +		/*
> > > +		 * On success, just make sure the returned count value is
> > > +		 * non-zero. The validity of the count value for the given
> > > +		 * device is not checked.
> > > +		*/
> > > +		igt_assert_neq(eu_total, 0);
> > > +		igt_info("EU total: %u\n", eu_total);
> > > +	}
> > > +}
> > > +
> > > +static void
> > > +exit_handler(int sig)
> > > +{
> > > +	deinit();
> > > +}
> > > +
> > > +igt_main
> > > +{
> > > +	igt_fixture {
> > > +		igt_install_exit_handler(exit_handler);
> > > +		init();
> > > +	}
> > > +
> > > +	igt_subtest("subslice-total")
> > > +		subslice_total();
> > > +
> > > +	igt_subtest("eu-total")
> > > +		eu_total();
> > > +}
> > > -- 
> > > 2.3.0
> > > 
> > > _______________________________________________
> > > Intel-gfx mailing list
> > > [e-mail]
> > > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
> > _______________________________________________
> > Intel-gfx mailing list
> > [e-mail]
> > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
> 
> -- 
> Daniel Vetter
> Software Engineer, Intel Corporation
> [phone] - http://blog.ffwll.ch
> _______________________________________________
> Beignet mailing list
> [e-mail]
> http://lists.freedesktop.org/mailman/listinfo/beignet

Daniel Vetter March 13, 2015, 4:59 p.m. UTC | #5

On Fri, Mar 13, 2015 at 09:51:57AM -0700, Jeff McGee wrote:
> On Fri, Mar 13, 2015 at 05:32:41PM +0100, Daniel Vetter wrote:
> > On Fri, Mar 13, 2015 at 05:09:46PM +0800, Zhigang Gong wrote:
> > > My only concern is about the following macros:
> > > 
> > > > +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL      33
> > > > +#define LOCAL_I915_PARAM_EU_TOTAL    34
> > > 
> > > How about to just use the definitons in the kernel header file?
> > > For an example:
> > > 
> > >   #include <drm/i915_drm.h>
> > > 
> > >   #ifdef LOCAL_I915_PARAM_SUBSLICE_TOTAL
> > >   //Put all the code into this block.
> > >   #endif
> > > 
> > > Then we can avoid put the same definitions in different files,
> > > and we can avoid unecessary testing on an old kernel which doesn't
> > > have this kernel interface.
> > > 
> > > For all the other part, it LGTM.
> > > 
> > > Reviewed-by: Zhigang Gong <[e-mail]>
> > 
> > Once we update the libdrm requirements in igt we tend to go around and
> > replace all the now obsolete LOCAL_ defines. Imo not worth doing extra
> > work until then.
> > 
> > Patch applied, thanks.
> > -Daniel
> > 
> 
> Patch applied? Do you want me to make the name change first? Should the
> kernel part be reviewed and merged first?

Forgot my own review comment already ;-) Fixed up with a follow-up patch.
And I'll pull the kernel part in now too.
-Daniel

[e-mail] March 13, 2015, 5:03 p.m. UTC | #6

On Fri, Mar 13, 2015 at 05:59:13PM +0100, Daniel Vetter wrote:
> On Fri, Mar 13, 2015 at 09:51:57AM -0700, Jeff McGee wrote:
> > On Fri, Mar 13, 2015 at 05:32:41PM +0100, Daniel Vetter wrote:
> > > On Fri, Mar 13, 2015 at 05:09:46PM +0800, Zhigang Gong wrote:
> > > > My only concern is about the following macros:
> > > > 
> > > > > +#define LOCAL_I915_PARAM_SUBSLICE_TOTAL      33
> > > > > +#define LOCAL_I915_PARAM_EU_TOTAL    34
> > > > 
> > > > How about to just use the definitons in the kernel header file?
> > > > For an example:
> > > > 
> > > >   #include <drm/i915_drm.h>
> > > > 
> > > >   #ifdef LOCAL_I915_PARAM_SUBSLICE_TOTAL
> > > >   //Put all the code into this block.
> > > >   #endif
> > > > 
> > > > Then we can avoid put the same definitions in different files,
> > > > and we can avoid unecessary testing on an old kernel which doesn't
> > > > have this kernel interface.
> > > > 
> > > > For all the other part, it LGTM.
> > > > 
> > > > Reviewed-by: Zhigang Gong <[e-mail]>
> > > 
> > > Once we update the libdrm requirements in igt we tend to go around and
> > > replace all the now obsolete LOCAL_ defines. Imo not worth doing extra
> > > work until then.
> > > 
> > > Patch applied, thanks.
> > > -Daniel
> > > 
> > 
> > Patch applied? Do you want me to make the name change first? Should the
> > kernel part be reviewed and merged first?
> 
> Forgot my own review comment already ;-) Fixed up with a follow-up patch.
> And I'll pull the kernel part in now too.
> -Daniel

Thanks. That leaves just the libdrm patch and version bump before Beignet
can merge their patches.
-Jeff

Patch

diff --git a/tests/.gitignore b/tests/.gitignore
index 426cc67..c742308 100644
--- a/tests/.gitignore
+++ b/tests/.gitignore
@@ -1,6 +1,7 @@ 
 # Please keep sorted alphabetically
 core_get_client_auth
 core_getclient
+core_getparams
 core_getstats
 core_getversion
 drm_import_export
diff --git a/tests/Makefile.sources b/tests/Makefile.sources
index 51e8376..999c8f8 100644
--- a/tests/Makefile.sources
+++ b/tests/Makefile.sources
@@ -15,6 +15,7 @@  NOUVEAU_TESTS_M = \
 
 TESTS_progs_M = \
 	core_get_client_auth \
+	core_getparams \
 	drv_suspend \
 	drv_hangman \
 	gem_bad_reloc \
diff --git a/tests/core_getparams.c b/tests/core_getparams.c
new file mode 100644
index 0000000..2855d06
--- /dev/null
+++ b/tests/core_getparams.c
@@ -0,0 +1,167 @@ 
+/*
+ * Copyright © 2014 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:
+ *    Jeff McGee <[e-mail]>
+ *
+ */
+
+#include <unistd.h>
+#include <errno.h>
+#include <xf86drm.h>
+#include <i915_drm.h>
+#include "drmtest.h"
+#include "intel_chipset.h"
+#include "intel_bufmgr.h"
+
+IGT_TEST_DESCRIPTION("Tests the export of parameters via DRM_IOCTL_I915_GETPARAM\n");
+
+int drm_fd;
+int devid;
+
+static void
+init(void)
+{
+	drm_fd = drm_open_any();
+	devid = intel_get_drm_devid(drm_fd);
+}
+
+static void
+deinit(void)
+{
+	close(drm_fd);
+}
+
+#define LOCAL_I915_PARAM_SUBSLICE_TOTAL	33
+#define LOCAL_I915_PARAM_EU_TOTAL	34
+
+static int
+getparam(int param, int *value)
+{
+	drm_i915_getparam_t gp;
+	int ret;
+
+	memset(&gp, 0, sizeof(gp));
+	gp.value = value;
+	gp.param = param;
+	ret = drmIoctl(drm_fd, DRM_IOCTL_I915_GETPARAM, &gp);
+	if (ret)
+		return -errno;
+
+	return 0;
+}
+
+static void
+subslice_total(void)
+{
+	unsigned int subslice_total = 0;
+	int ret;
+
+	ret = getparam(LOCAL_I915_PARAM_SUBSLICE_TOTAL, (int*)&subslice_total);
+
+	if (ret) {
+		/*
+		 * These devices are not required to implement the
+		 * interface. If they do not, -ENODEV must be returned.
+		*/
+		if ((intel_gen(devid) < 8) ||
+		    IS_BROADWELL(devid) ||
+		    igt_run_in_simulation()) {
+			igt_assert_eq(ret, -ENODEV);
+			igt_info("subslice total: unknown\n");
+		/*
+		 * All other devices must implement the interface, so
+		 * fail them if we are here.
+		*/
+		} else {
+			igt_assert_neq(ret, EINVAL); /* request not recognized? */
+			igt_assert_neq(ret, ENODEV); /* device not supported? */
+			igt_assert_eq(ret, 0); /* other error? */
+		}
+	} else {
+		/*
+		 * On success, just make sure the returned count value is
+		 * non-zero. The validity of the count value for the given
+		 * device is not checked.
+		*/
+		igt_assert_neq(subslice_total, 0);
+		igt_info("subslice total: %u\n", subslice_total);
+	}
+}
+
+static void
+eu_total(void)
+{
+	unsigned int eu_total = 0;
+	int ret;
+
+	ret = getparam(LOCAL_I915_PARAM_EU_TOTAL, (int*)&eu_total);
+
+	if (ret) {
+		/*
+		 * These devices are not required to implement the
+		 * interface. If they do not, -ENODEV must be returned.
+		*/
+		if ((intel_gen(devid) < 8) ||
+		    IS_BROADWELL(devid) ||
+		    igt_run_in_simulation()) {
+			igt_assert_eq(ret, -ENODEV);
+			igt_info("EU total: unknown\n");
+		/*
+		 * All other devices must implement the interface, so
+		 * fail them if we are here.
+		*/
+		} else {
+			igt_assert_neq(ret, EINVAL); /* request not recognized? */
+			igt_assert_neq(ret, ENODEV); /* device not supported? */
+			igt_assert_eq(ret, 0); /* other error? */
+		}
+	} else {
+		/*
+		 * On success, just make sure the returned count value is
+		 * non-zero. The validity of the count value for the given
+		 * device is not checked.
+		*/
+		igt_assert_neq(eu_total, 0);
+		igt_info("EU total: %u\n", eu_total);
+	}
+}
+
+static void
+exit_handler(int sig)
+{
+	deinit();
+}
+
+igt_main
+{
+	igt_fixture {
+		igt_install_exit_handler(exit_handler);
+		init();
+	}
+
+	igt_subtest("subslice-total")
+		subslice_total();
+
+	igt_subtest("eu-total")
+		eu_total();
+}