drm/i915: Flush pages on acquisition

When we return pages to the system, we ensure that they are marked as
being in the CPU domain since any external access is uncontrolled and we
must assume the worst. This means that we need to always flush the pages
on acquisition if we need to use them on the GPU, and from the beginning
have used set-domain. Set-domain is overkill for the purpose as it is a
general synchronisation barrier, but our intent is to only flush the
pages being swapped in. If we move that flush into the pages acquisition
phase, we know then that when we have obj->mm.pages, they are coherent
with the GPU and need only maintain that status without resorting to
heavy handed use of set-domain.

The principle knock-on effect for userspace is through mmap-gtt
pagefaulting. Our uAPI has always implied that the GTT mmap was async
(especially as when any pagefault occurs is unpredicatable to userspace)
and so userspace had to apply explicit domain control itself
(set-domain). However, swapping is transparent to the kernel, and so on
first fault we need to acquire the pages and make them coherent for
access through the GTT. Our use of set-domain here leaks into the uABI
that the first pagefault was synchronous. This is unintentional and
baring a few igt should be unoticed, nevertheless we bump the uABI
version for mmap-gtt to reflect the change in behaviour.

Another implication of the change is that gem_create() is presumed to
create an object that is coherent with the CPU and is in the CPU write
domain, so a set-domain(CPU) following a gem_create() would be a minor
operation that merely checked whether we could allocate all pages for
the object. On applying this change, a set-domain(CPU) causes a clflush
as we acquire the pages. This will have a small impact on mesa as we move
the clflush here on !llc from execbuf time to create, but that should
have minimal performance impact as the same clflush exists but is now
done early and because of the clflush issue, userspace recycles bo and
so should resist allocating fresh objects.

Internally, the presumption that objects are created in the CPU
write-domain and remain so through writes to obj->mm.mapping is more
prevalent than I expected; but easy enough to catch and apply a manual
flush.

For the future, we should push the page flush from the central
set_pages() into the callers so that we can more finely control when it
is applied, but for now doing it one location is easier to validate, at
the cost of sometimes flushing when there is no need.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.william.auld@gmail.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Antonio Argenziano <antonio.argenziano@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.william.auld@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190321161908.8007-1-chris@chris-wilson.co.uk

9 files changed, 20 insertions, 33 deletions

diff --git a/drivers/gpu/drm/i915/selftests/huge_pages.c b/drivers/gpu/drm/i915/selftests/huge_pages.c
index 1f419fb094f4..c5c8ba6c059f 100644
--- a/drivers/gpu/drm/i915/selftests/huge_pages.c
+++ b/drivers/gpu/drm/i915/selftests/huge_pages.c
@@ -908,10 +908,6 @@ gpu_write_dw(struct i915_vma *vma, u64 offset, u32 val)
 	if (IS_ERR(obj))
 		return ERR_CAST(obj);
 
-	err = i915_gem_object_set_to_wc_domain(obj, true);
-	if (err)
-		goto err;
-
 	cmd = i915_gem_object_pin_map(obj, I915_MAP_WC);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
@@ -1584,6 +1580,7 @@ static int igt_tmpfs_fallback(void *arg)
 	}
 	*vaddr = 0xdeadbeaf;
 
+	__i915_gem_object_flush_map(obj, 0, 64);
 	i915_gem_object_unpin_map(obj);
 
 	vma = i915_vma_instance(obj, vm, NULL);
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/selftests/i915_gem_context.c
index ab2aa32fd019..fb6dc54ce7ff 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_context.c
@@ -220,6 +220,7 @@ gpu_fill_dw(struct i915_vma *vma, u64 offset, unsigned long count, u32 value)
 		offset += PAGE_SIZE;
 	}
 	*cmd = MI_BATCH_BUFFER_END;
+	i915_gem_object_flush_map(obj);
 	i915_gem_object_unpin_map(obj);
 
 	err = i915_gem_object_set_to_gtt_domain(obj, false);
@@ -604,12 +605,9 @@ static struct i915_vma *rpcs_query_batch(struct i915_vma *vma)
 	*cmd++ = upper_32_bits(vma->node.start);
 	*cmd = MI_BATCH_BUFFER_END;
 
+	__i915_gem_object_flush_map(obj, 0, 64);
 	i915_gem_object_unpin_map(obj);
 
-	err = i915_gem_object_set_to_gtt_domain(obj, false);
-	if (err)
-		goto err;
-
 	vma = i915_vma_instance(obj, vma->vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
@@ -1202,12 +1200,9 @@ static int write_to_scratch(struct i915_gem_context *ctx,
 	}
 	*cmd++ = value;
 	*cmd = MI_BATCH_BUFFER_END;
+	__i915_gem_object_flush_map(obj, 0, 64);
 	i915_gem_object_unpin_map(obj);
 
-	err = i915_gem_object_set_to_gtt_domain(obj, false);
-	if (err)
-		goto err;
-
 	vma = i915_vma_instance(obj, &ctx->ppgtt->vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
@@ -1299,11 +1294,9 @@ static int read_from_scratch(struct i915_gem_context *ctx,
 		*cmd++ = result;
 	}
 	*cmd = MI_BATCH_BUFFER_END;
-	i915_gem_object_unpin_map(obj);
 
-	err = i915_gem_object_set_to_gtt_domain(obj, false);
-	if (err)
-		goto err;
+	i915_gem_object_flush_map(obj);
+	i915_gem_object_unpin_map(obj);
 
 	vma = i915_vma_instance(obj, &ctx->ppgtt->vm, NULL);
 	if (IS_ERR(vma)) {
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c
index a7055b12e53c..2b943ee246c9 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c
@@ -315,6 +315,7 @@ static int igt_dmabuf_export_kmap(void *arg)
 		goto err;
 	}
 	memset(ptr + PAGE_SIZE, 0xaa, PAGE_SIZE);
+	i915_gem_object_flush_map(obj);
 	i915_gem_object_unpin_map(obj);
 
 	ptr = dma_buf_kmap(dmabuf, 1);
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_evict.c b/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
index b270eab1cad1..9a9451846b33 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
@@ -274,7 +274,7 @@ static int igt_evict_for_cache_color(void *arg)
 		err = PTR_ERR(obj);
 		goto cleanup;
 	}
-	i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
+	i915_gem_object_set_cache_coherency(obj, I915_CACHE_LLC);
 	quirk_add(obj, &objects);
 
 	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0,
@@ -290,7 +290,7 @@ static int igt_evict_for_cache_color(void *arg)
 		err = PTR_ERR(obj);
 		goto cleanup;
 	}
-	i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
+	i915_gem_object_set_cache_coherency(obj, I915_CACHE_LLC);
 	quirk_add(obj, &objects);
 
 	/* Neighbouring; same colour - should fit */
diff --git a/drivers/gpu/drm/i915/selftests/i915_request.c b/drivers/gpu/drm/i915/selftests/i915_request.c
index 3eb6a6b075ab..e6ffe2240126 100644
--- a/drivers/gpu/drm/i915/selftests/i915_request.c
+++ b/drivers/gpu/drm/i915/selftests/i915_request.c
@@ -619,13 +619,11 @@ static struct i915_vma *empty_batch(struct drm_i915_private *i915)
 	}
 
 	*cmd = MI_BATCH_BUFFER_END;
-	i915_gem_chipset_flush(i915);
 
+	__i915_gem_object_flush_map(obj, 0, 64);
 	i915_gem_object_unpin_map(obj);
 
-	err = i915_gem_object_set_to_gtt_domain(obj, false);
-	if (err)
-		goto err;
+	i915_gem_chipset_flush(i915);
 
 	vma = i915_vma_instance(obj, &i915->ggtt.vm, NULL);
 	if (IS_ERR(vma)) {
@@ -777,10 +775,6 @@ static struct i915_vma *recursive_batch(struct drm_i915_private *i915)
 	if (err)
 		goto err;
 
-	err = i915_gem_object_set_to_wc_domain(obj, true);
-	if (err)
-		goto err;
-
 	cmd = i915_gem_object_pin_map(obj, I915_MAP_WC);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
@@ -799,10 +793,12 @@ static struct i915_vma *recursive_batch(struct drm_i915_private *i915)
 		*cmd++ = lower_32_bits(vma->node.start);
 	}
 	*cmd++ = MI_BATCH_BUFFER_END; /* terminate early in case of error */
-	i915_gem_chipset_flush(i915);
 
+	__i915_gem_object_flush_map(obj, 0, 64);
 	i915_gem_object_unpin_map(obj);
 
+	i915_gem_chipset_flush(i915);
+
 	return vma;
 
 err:
diff --git a/drivers/gpu/drm/i915/selftests/igt_spinner.c b/drivers/gpu/drm/i915/selftests/igt_spinner.c
index d0b93a3fbc54..16890dfe74c0 100644
--- a/drivers/gpu/drm/i915/selftests/igt_spinner.c
+++ b/drivers/gpu/drm/i915/selftests/igt_spinner.c
@@ -29,7 +29,7 @@ int igt_spinner_init(struct igt_spinner *spin, struct drm_i915_private *i915)
 		goto err_hws;
 	}
 
-	i915_gem_object_set_cache_level(spin->hws, I915_CACHE_LLC);
+	i915_gem_object_set_cache_coherency(spin->hws, I915_CACHE_LLC);
 	vaddr = i915_gem_object_pin_map(spin->hws, I915_MAP_WB);
 	if (IS_ERR(vaddr)) {
 		err = PTR_ERR(vaddr);
diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
index b5e35b2a925f..76b4fa150f2e 100644
--- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
@@ -70,7 +70,7 @@ static int hang_init(struct hang *h, struct drm_i915_private *i915)
 		goto err_hws;
 	}
 
-	i915_gem_object_set_cache_level(h->hws, I915_CACHE_LLC);
+	i915_gem_object_set_cache_coherency(h->hws, I915_CACHE_LLC);
 	vaddr = i915_gem_object_pin_map(h->hws, I915_MAP_WB);
 	if (IS_ERR(vaddr)) {
 		err = PTR_ERR(vaddr);
diff --git a/drivers/gpu/drm/i915/selftests/intel_lrc.c b/drivers/gpu/drm/i915/selftests/intel_lrc.c
index d61520ea03c1..9e871eb0bfb1 100644
--- a/drivers/gpu/drm/i915/selftests/intel_lrc.c
+++ b/drivers/gpu/drm/i915/selftests/intel_lrc.c
@@ -1018,12 +1018,9 @@ static int live_preempt_smoke(void *arg)
 	for (n = 0; n < PAGE_SIZE / sizeof(*cs) - 1; n++)
 		cs[n] = MI_ARB_CHECK;
 	cs[n] = MI_BATCH_BUFFER_END;
+	i915_gem_object_flush_map(smoke.batch);
 	i915_gem_object_unpin_map(smoke.batch);
 
-	err = i915_gem_object_set_to_gtt_domain(smoke.batch, false);
-	if (err)
-		goto err_batch;
-
 	for (n = 0; n < smoke.ncontext; n++) {
 		smoke.contexts[n] = kernel_context(smoke.i915);
 		if (!smoke.contexts[n])
diff --git a/drivers/gpu/drm/i915/selftests/intel_workarounds.c b/drivers/gpu/drm/i915/selftests/intel_workarounds.c
index f2a2b51a4662..3baed59008d7 100644
--- a/drivers/gpu/drm/i915/selftests/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/selftests/intel_workarounds.c
@@ -90,6 +90,7 @@ read_nonprivs(struct i915_gem_context *ctx, struct intel_engine_cs *engine)
 		goto err_obj;
 	}
 	memset(cs, 0xc5, PAGE_SIZE);
+	i915_gem_object_flush_map(result);
 	i915_gem_object_unpin_map(result);
 
 	vma = i915_vma_instance(result, &engine->i915->ggtt.vm, NULL);
@@ -358,6 +359,7 @@ static struct i915_vma *create_scratch(struct i915_gem_context *ctx)
 		goto err_obj;
 	}
 	memset(ptr, 0xc5, PAGE_SIZE);
+	i915_gem_object_flush_map(obj);
 	i915_gem_object_unpin_map(obj);
 
 	vma = i915_vma_instance(obj, &ctx->ppgtt->vm, NULL);
@@ -551,6 +553,7 @@ static int check_dirty_whitelist(struct i915_gem_context *ctx,
 
 		*cs++ = MI_BATCH_BUFFER_END;
 
+		i915_gem_object_flush_map(batch->obj);
 		i915_gem_object_unpin_map(batch->obj);
 		i915_gem_chipset_flush(ctx->i915);