Re: [PATCH v3] tests/intel/xe_vm: Add test for 64k page corners

[Matthew Brost <matthew.brost@intel.com>]

On Fri, Jul 26, 2024 at 11:04:25AM +0200, Zbigniew Kempczyński wrote:
> On Thu, Jul 25, 2024 at 11:45:38AM -0700, Matthew Brost wrote:
> > Add sections which splits 64k pages in 4k pages or splits compact 64k in
> > 64k pages.
> > 
> > v2:
> >  - Use SZ_* (Zbigniew)
> >  - Remove clock block (Zbigniew)
> >  - Update commit message / comments
> > v3:
> >  - Send correct version
> > 
> > Cc: Zbigniew Kempczyński <zbigniew.kempczynski@intel.com>
> > Signed-off-by: Matthew Brost <matthew.brost@intel.com>
> > ---
> >  tests/intel/xe_vm.c | 125 ++++++++++++++++++++++++++++++++++++++++++++
> >  1 file changed, 125 insertions(+)
> > 
> > diff --git a/tests/intel/xe_vm.c b/tests/intel/xe_vm.c
> > index a4f6c7a0b4..f41c6aca53 100644
> > --- a/tests/intel/xe_vm.c
> > +++ b/tests/intel/xe_vm.c
> > @@ -366,6 +366,125 @@ static void userptr_invalid(int fd)
> >  	xe_vm_destroy(fd, vm);
> >  }
> >  
> > +/**
> > + * SUBTEST: compact-64k-pages
> > + * Description:
> > + *	Take corner cases related to compact and 64k pages
> > + * Functionality: bind
> > + * Test category: functionality test
> > + */
> > +static void compact_64k_pages(int fd, struct drm_xe_engine_class_instance *eci)
> > +{
> > +	size_t page_size = xe_get_default_alignment(fd);
> > +	uint64_t addr0 = 0x10000000ull, addr1;
> > +	uint32_t vm;
> > +	uint32_t bo0, bo1;
> > +	uint32_t exec_queue;
> > +	void *ptr0, *ptr1;
> > +	struct drm_xe_sync sync[2] = {
> > +		{ .type = DRM_XE_SYNC_TYPE_SYNCOBJ,
> > +			.flags = DRM_XE_SYNC_FLAG_SIGNAL, },
> > +		{ .type = DRM_XE_SYNC_TYPE_SYNCOBJ,
> > +			.flags = DRM_XE_SYNC_FLAG_SIGNAL, },
> > +	};
> > +	struct {
> > +		uint32_t batch[16];
> > +		uint64_t pad;
> > +		uint32_t data;
> > +	} *data = NULL;
> > +	struct drm_xe_exec exec = {
> > +		.num_batch_buffer = 1,
> > +		.num_syncs = 2,
> > +		.syncs = to_user_pointer(sync),
> > +	};
> > +	uint64_t batch_offset;
> > +	uint64_t batch_addr;
> > +	uint64_t sdi_offset;
> > +	uint64_t sdi_addr;
> > +	int b = 0;
> > +
> > +	vm = xe_vm_create(fd, 0, 0);
> > +	exec_queue = xe_exec_queue_create(fd, vm, eci, 0);
> > +
> > +	bo0 = xe_bo_create(fd, vm, SZ_8M,
> > +			   vram_if_possible(fd, eci->gt_id),
> > +			   DRM_XE_GEM_CREATE_FLAG_NEEDS_VISIBLE_VRAM);
> > +	ptr0 = xe_bo_map(fd, bo0, SZ_8M);
> > +
> > +	bo1 = xe_bo_create(fd, vm, SZ_8M / 2,
> > +			   vram_if_possible(fd, eci->gt_id),
> > +			   DRM_XE_GEM_CREATE_FLAG_NEEDS_VISIBLE_VRAM);
> > +	ptr1 = xe_bo_map(fd, bo1, SZ_8M / 2);
> > +
> > +	sync[0].handle = syncobj_create(fd, 0);
> > +	if (page_size == SZ_4K) {
> > +		/* Setup mapping to split a 64k PTE in cache */
> > +		xe_vm_bind_async(fd, vm, 0, bo0, 0, addr0, SZ_64K, 0, 0);
> > +
> > +		addr1 = addr0 + (SZ_64K / 2);
> > +		xe_vm_bind_async(fd, vm, 0, bo1, 0, addr1, SZ_64K / 2,
> > +				 sync, 1);
> > +	} else if (page_size == SZ_64K) {
> > +		addr0 += page_size;
> > +
> > +		/* Setup mapping to split compact 64k pages */
> > +		xe_vm_bind_async(fd, vm, 0, bo0, 0, addr0, SZ_8M, 0, 0);
> > +
> > +		addr1 = addr0 + (SZ_8M / 4);
> > +		xe_vm_bind_async(fd, vm, 0, bo1, 0, addr1, SZ_8M / 2,
> > +				 sync, 1);
> > +	}
> > +	igt_assert(syncobj_wait(fd, &sync[0].handle, 1, INT64_MAX, 0, NULL));
> > +
> > +	/* Verify 1st and 2nd mappings working */
> > +	batch_offset = (char *)&data[0].batch - (char *)data;
> > +	batch_addr = addr0 + batch_offset;
> > +	sdi_offset = (char *)&data[0].data - (char *)data;
> > +	sdi_addr = addr0 + sdi_offset;
> > +	data = ptr0;
> > +
> > +	data[0].batch[b++] = MI_STORE_DWORD_IMM_GEN4;
> > +	data[0].batch[b++] = sdi_addr;
> > +	data[0].batch[b++] = sdi_addr >> 32;
> > +	data[0].batch[b++] = 0xc0ffee;
> > +
> > +	sdi_addr = addr1 + sdi_offset;
> > +	data[0].batch[b++] = MI_STORE_DWORD_IMM_GEN4;
> > +	data[0].batch[b++] = sdi_addr;
> > +	data[0].batch[b++] = sdi_addr >> 32;
> > +	data[0].batch[b++] = 0xc0ffee;
> > +
> > +	data[0].batch[b++] = MI_BATCH_BUFFER_END;
> > +	igt_assert(b <= ARRAY_SIZE(data[0].batch));
> > +
> > +	sync[0].flags &= ~DRM_XE_SYNC_FLAG_SIGNAL;
> > +	sync[1].handle = syncobj_create(fd, 0);
> > +	exec.exec_queue_id = exec_queue;
> > +	exec.address = batch_addr;
> > +	xe_exec(fd, &exec);
> > +
> > +	igt_assert(syncobj_wait(fd, &sync[1].handle, 1, INT64_MAX, 0,
> > +				NULL));
> 
> This will fit in same line ase syncobj_wait() below.
> 

No, you need to wait here the exec to signal before checking the data
and unbinding.

> > +	igt_assert_eq(data[0].data, 0xc0ffee);
> > +	data = ptr1;
> > +	igt_assert_eq(data[0].data, 0xc0ffee);
> > +
> > +	sync[0].flags |= DRM_XE_SYNC_FLAG_SIGNAL;
> > +	syncobj_reset(fd, &sync[0].handle, 1);
> > +	xe_vm_unbind_all_async(fd, vm, 0, bo0, 0, 0);
> > +	xe_vm_unbind_all_async(fd, vm, 0, bo1, sync, 1);
> > +	igt_assert(syncobj_wait(fd, &sync[0].handle, 1, INT64_MAX, 0, NULL));

This waits on the unbinds.

> > +
> > +	xe_exec_queue_destroy(fd, exec_queue);
> > +	syncobj_destroy(fd, sync[0].handle);
> > +	syncobj_destroy(fd, sync[1].handle);
> > +	munmap(ptr0, SZ_8M);
> > +	munmap(ptr1, SZ_8M / 2);
> > +	gem_close(fd, bo0);
> > +	gem_close(fd, bo1);
> > +	xe_vm_destroy(fd, vm);
> > +}
> > +
> >  /**
> >   * SUBTEST: shared-%s-page
> >   * Description: Test shared arg[1] page
> > @@ -1973,6 +2092,12 @@ igt_main
> >  	igt_subtest("bind-flag-invalid")
> >  		bind_flag_invalid(fd);
> >  
> > +	igt_subtest("compact-64k-pages")
> > +		xe_for_each_engine(fd, hwe) {
> > +			compact_64k_pages(fd, hwe);
> > +			break;
> > +		}
> > +
> >  	igt_subtest("shared-pte-page")
> >  		xe_for_each_engine(fd, hwe)
> >  			shared_pte_page(fd, hwe, 4,
> > -- 
> > 2.34.1
> > 
> 
> Regardless addressing nit above:
> 
> Reviewed-by: Zbigniew Kempczyński <zbigniew.kempczynski@intel.com>
> 

Thanks.

Matt

> --
> Zbigniew