[PATCH v2] tests/intel: Enhance the wedged_mode sysfs support

Gurram, Pravalika pravalika.gurram at intel.com
Mon Jan 13 13:13:46 UTC 2025



> -----Original Message-----
> From: Kamil Konieczny <kamil.konieczny at linux.intel.com>
> Sent: Friday, January 10, 2025 5:03 PM
> To: Gurram, Pravalika <pravalika.gurram at intel.com>
> Cc: igt-dev at lists.freedesktop.org; Cavitt, Jonathan
> <jonathan.cavitt at intel.com>
> Subject: Re: [PATCH v2] tests/intel: Enhance the wedged_mode sysfs support
> 
> Hi Pravalika,
> On 2025-01-08 at 11:58:26 +0530, Pravalika Gurram wrote:
> > 1. Added support to read the wedged_mode sysfs 2. Handle the
> > -ETIMEDOUT return value during forcewake_all sysfs open
> 
> How is testing forcewake related to wedged? imho this should be separate
> patch, see also below.
> 
> >
> > Signed-off-by: Pravalika Gurram <pravalika.gurram at intel.com>
> > Reviewed-by: Jonathan Cavitt <jonathan.cavitt at intel.com>
> > ---
> >  tests/intel/xe_debugfs.c |  2 +-
> >  tests/intel/xe_wedged.c  | 12 ++++++++++++
> >  2 files changed, 13 insertions(+), 1 deletion(-)
> >
> > diff --git a/tests/intel/xe_debugfs.c b/tests/intel/xe_debugfs.c index
> > bcbb5036a..3a83868ba 100644
> > --- a/tests/intel/xe_debugfs.c
> > +++ b/tests/intel/xe_debugfs.c
> > @@ -268,7 +268,7 @@ test_forcewake(int fd)  {
> >  	int handle = igt_debugfs_open(fd, "forcewake_all", O_WRONLY);
> >
> > -	igt_assert_neq(handle, -1);
> > +	igt_assert_lte(0, handle);
> 
> Please drop this change, that function return -1 in case of error, if you have a
> report that it is not please send this in separate patch.
> 

This changes got pushed by mistake will remove from patch.
 
> >  	close(handle);
> >  }
> >
> > diff --git a/tests/intel/xe_wedged.c b/tests/intel/xe_wedged.c index
> > 613d571b4..844ab2809 100644
> > --- a/tests/intel/xe_wedged.c
> > +++ b/tests/intel/xe_wedged.c
> > @@ -203,11 +203,16 @@ simple_hang(int fd, struct drm_xe_sync *sync)
> >   * SUBTEST: wedged-mode-toggle
> >   * Description: Test wedged.mode=1 after testing wedged.mode=2
> >   */
> > +/**
> > + * SUBTEST: basic-wedged-read
> > + * Description: Read wedged debugfs
> > + */
> >  igt_main
> >  {
> >  	struct drm_xe_engine_class_instance *hwe;
> >  	int fd;
> >  	char pci_slot[NAME_MAX];
> > +	char str[150];
> >
> >  	igt_fixture {
> >  		fd = drm_open_driver(DRIVER_XE);
> > @@ -279,6 +284,13 @@ igt_main
> >  		igt_assert_eq(simple_ioctl(fd), 0);
> >  	}
> >
> > +	igt_subtest_f("basic-wedged-read") {
> > +		igt_require(igt_debugfs_exists(fd, "wedged_mode",
> O_RDONLY));
> > +
> > +		igt_debugfs_read(fd, "wedged_mode", str);
> > +		igt_assert_f(str[0] != '\0', "Failed to read wedged_mode from
> > +debugfs!\n");
> 
> Is this correct? Why not using
> 		ret = igt_debugfs_read(fd, "wedged_mode", str); and assert on
> ret <= 0
> 
> Regards,
> Kamil
> 
Since igt_debugfs_read is wraper for __igt_debugfs_read.
#define igt_debugfs_read(fd, filename, buf) \
                __igt_debugfs_read(fd, (filename), (buf), sizeof(buf))
> 
> > +	}
> > +
> >  	igt_fixture {
> >  		if (igt_debugfs_exists(fd, "fail_gt_reset/probability", O_RDWR))
> {
> >  			igt_debugfs_write(fd, "fail_gt_reset/probability", "0");
> > --
> > 2.34.1
> >


More information about the igt-dev mailing list