[PATCH v2] tests/intel/xe_vm: Always have at least 2 pages bound in hammer sections

Matthew Brost matthew.brost at intel.com
Fri Apr 12 01:58:55 UTC 2024


On Wed, Apr 10, 2024 at 09:42:37AM +0100, Matthew Auld wrote:
> On 10/04/2024 02:21, Matthew Brost wrote:
> > If the prefetch buffer is 2k or larger the hammer sections will prefetch
> > the next page, if the page is unbound a fault will occur. Account for
> > larger prefetchs by always having at least 2 pages bound for the hammer
> > thread.
> 
> Is this referring to bb_start pre-fetch? It should be + 4K at most for xe2.

Yes.

> Is it possible to make the test consider xe_cs_prefetch_size() somehow
> instead of hardcoding? Or perhaps that is tricky here? Otherwise maybe we

That's tricky.

> can add an assert that xe_cs_prefetch_size() remains <= 4K for this test,

Will add an assert to the test.

> with comment explaining what needs to change if we start triggering the
> assert? Thinking is that would be easier to debug and fix than seeing CAT
> errors, faults etc. if the pre-fetch changes again.

Will also add a comment.

Thanks,
Matt

> 
> > 
> > v2:
> >   - Missed table entry
> > 
> > Signed-off-by: Matthew Brost <matthew.brost at intel.com>
> > ---
> >   tests/intel/xe_vm.c | 16 ++++++++--------
> >   1 file changed, 8 insertions(+), 8 deletions(-)
> > 
> > diff --git a/tests/intel/xe_vm.c b/tests/intel/xe_vm.c
> > index bd4973cc38..9f9f61b7ef 100644
> > --- a/tests/intel/xe_vm.c
> > +++ b/tests/intel/xe_vm.c
> > @@ -1843,12 +1843,12 @@ igt_main
> >   		{ "all", 4, 2, 0, 4, 0 },
> >   		{ "one-partial", 4, 1, 1, 2, 0 },
> >   		{ "either-side-partial", 4, 2, 1, 2, 0 },
> > -		{ "either-side-partial-hammer", 4, 2, 1, 2,
> > +		{ "either-side-partial-hammer", 6, 2, 2, 2,
> >   			MAP_FLAG_HAMMER_FIRST_PAGE },
> > -		{ "either-side-partial-split-page-hammer", 4, 2, 1, 2,
> > +		{ "either-side-partial-split-page-hammer", 6, 2, 2, 2,
> >   			MAP_FLAG_HAMMER_FIRST_PAGE |
> >   			MAP_FLAG_LARGE_PAGE },
> > -		{ "either-side-partial-large-page-hammer", 4, 2, 1, 2,
> > +		{ "either-side-partial-large-page-hammer", 6, 2, 2, 2,
> >   			MAP_FLAG_HAMMER_FIRST_PAGE |
> >   			MAP_FLAG_LARGE_PAGE |
> >   			MAP_FLAG_LARGE_PAGE_NO_SPLIT },
> > @@ -1857,7 +1857,7 @@ igt_main
> >   		{ "front", 4, 2, 1, 3, 0 },
> >   		{ "many-all", 4 * 8, 2 * 8, 0 * 8, 4 * 8, 0 },
> >   		{ "many-either-side-partial", 4 * 8, 2 * 8, 1, 4 * 8 - 2, 0 },
> > -		{ "many-either-side-partial-hammer", 4 * 8, 2 * 8, 1, 4 * 8 - 2,
> > +		{ "many-either-side-partial-hammer", 4 * 8, 2 * 8, 2, 4 * 8 - 4,
> >   			MAP_FLAG_HAMMER_FIRST_PAGE },
> >   		{ "many-either-side-full", 4 * 8, 4 * 8, 1 * 8, 2 * 8, 0 },
> >   		{ "many-end", 4 * 8, 4, 0 * 8, 3 * 8 + 2, 0 },
> > @@ -1902,12 +1902,12 @@ igt_main
> >   		{ "one-partial", 4, 1, 1, 2, 0 },
> >   		{ "either-side-partial", 4, 2, 1, 2, 0 },
> >   		{ "either-side-full", 4, 4, 1, 2, 0 },
> > -		{ "either-side-partial-hammer", 4, 2, 1, 2,
> > +		{ "either-side-partial-hammer", 6, 2, 2, 2,
> >   			MAP_FLAG_HAMMER_FIRST_PAGE },
> > -		{ "either-side-partial-split-page-hammer", 4, 2, 1, 2,
> > +		{ "either-side-partial-split-page-hammer", 6, 2, 2, 2,
> >   			MAP_FLAG_HAMMER_FIRST_PAGE |
> >   			MAP_FLAG_LARGE_PAGE },
> > -		{ "either-side-partial-large-page-hammer", 4, 2, 1, 2,
> > +		{ "either-side-partial-large-page-hammer", 6, 2, 2, 2,
> >   			MAP_FLAG_HAMMER_FIRST_PAGE |
> >   			MAP_FLAG_LARGE_PAGE |
> >   			MAP_FLAG_LARGE_PAGE_NO_SPLIT },
> > @@ -1915,7 +1915,7 @@ igt_main
> >   		{ "front", 4, 2, 1, 3, 0 },
> >   		{ "many-all", 4 * 8, 2 * 8, 0 * 8, 4 * 8, 0 },
> >   		{ "many-either-side-partial", 4 * 8, 2 * 8, 1, 4 * 8 - 2, 0 },
> > -		{ "many-either-side-partial-hammer", 4 * 8, 2 * 8, 1, 4 * 8 - 2,
> > +		{ "many-either-side-partial-hammer", 4 * 8, 2 * 8, 2, 4 * 8 - 4,
> >   			MAP_FLAG_HAMMER_FIRST_PAGE },
> >   		{ "userptr-all", 4, 2, 0, 4, MAP_FLAG_USERPTR },
> >   		{ "userptr-one-partial", 4, 1, 1, 2, MAP_FLAG_USERPTR },


More information about the igt-dev mailing list