[PATCH v10 1/5] drm/i915: Factor out setup_private_pat()

[PATCH v10 1/5] drm/i915: Factor out setup_private_pat()

[Zhi Wang]

Factor out setup_private_pat() for introducing the following patches.

Reviewed-by: Ben Widawsky <benjamin.widawsky@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ben Widawsky <benjamin.widawsky@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/i915_gem_gtt.c | 20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 933c4ea..e68cab4 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -2841,6 +2841,16 @@ static void gen6_gmch_remove(struct i915_address_space *vm)
 	cleanup_scratch_page(vm);
 }
 
+static void setup_private_pat(struct drm_i915_private *dev_priv)
+{
+	if (INTEL_GEN(dev_priv) >= 10)
+		cnl_setup_private_ppat(dev_priv);
+	else if (IS_CHERRYVIEW(dev_priv) || IS_GEN9_LP(dev_priv))
+		chv_setup_private_ppat(dev_priv);
+	else
+		bdw_setup_private_ppat(dev_priv);
+}
+
 static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 {
 	struct drm_i915_private *dev_priv = ggtt->base.i915;
@@ -2873,14 +2883,6 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 	}
 
 	ggtt->base.total = (size / sizeof(gen8_pte_t)) << PAGE_SHIFT;
-
-	if (INTEL_GEN(dev_priv) >= 10)
-		cnl_setup_private_ppat(dev_priv);
-	else if (IS_CHERRYVIEW(dev_priv) || IS_GEN9_LP(dev_priv))
-		chv_setup_private_ppat(dev_priv);
-	else
-		bdw_setup_private_ppat(dev_priv);
-
 	ggtt->base.cleanup = gen6_gmch_remove;
 	ggtt->base.bind_vma = ggtt_bind_vma;
 	ggtt->base.unbind_vma = ggtt_unbind_vma;
@@ -2901,6 +2903,8 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 
 	ggtt->invalidate = gen6_ggtt_invalidate;
 
+	setup_private_pat(dev_priv);
+
 	return ggtt_probe_common(ggtt, size);
 }
 
-- 
2.7.4

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH v10 2/5] drm/i915: Introduce private PAT management

[Zhi Wang]

The private PAT management is to support PPAT entry manipulation. Two
APIs are introduced for dynamically managing PPAT entries: intel_ppat_get
and intel_ppat_put.

intel_ppat_get will search for an existing PPAT entry which perfectly
matches the required PPAT value. If not, it will try to allocate a new
entry if there is any available PPAT indexs, or return a partially
matched PPAT entry if there is no available PPAT indexes.

intel_ppat_put will put back the PPAT entry which comes from
intel_ppat_get. If it's dynamically allocated, the reference count will
be decreased. If the reference count turns into zero, the PPAT index is
freed again.

Besides, another two callbacks are introduced to support the private PAT
management framework. One is ppat->update_hw(), which writes the PPAT
configurations in ppat->entries into HW. Another one is ppat->match, which
will return a score to show how two PPAT values match with each other.

v7:

- Keep all the register writes unchanged in this patch. (Joonas)

v6:

- Address all comments from Chris:
http://www.spinics.net/lists/intel-gfx/msg136850.html

- Address all comments from Joonas:
http://www.spinics.net/lists/intel-gfx/msg136845.html

v5:

- Add check and warnnings for those platforms which don't have PPAT.

v3:

- Introduce dirty bitmap for PPAT registers. (Chris)
- Change the name of the pointer "dev_priv" to "i915". (Chris)
- intel_ppat_{get, put} returns/takes a const intel_ppat_entry *. (Chris)

v2:

- API re-design. (Chris)

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ben Widawsky <benjamin.widawsky@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h     |   2 +
 drivers/gpu/drm/i915/i915_gem_gtt.c | 279 +++++++++++++++++++++++++++++-------
 drivers/gpu/drm/i915/i915_gem_gtt.h |  36 +++++
 3 files changed, 268 insertions(+), 49 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 571c4e2..77bdf55 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2307,6 +2307,8 @@ struct drm_i915_private {
 	DECLARE_HASHTABLE(mm_structs, 7);
 	struct mutex mm_lock;
 
+	struct intel_ppat ppat;
+
 	/* Kernel Modesetting */
 
 	struct intel_crtc *plane_to_crtc_mapping[I915_MAX_PIPES];
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index e68cab4..b48aac1 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -2742,41 +2742,203 @@ static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 	return 0;
 }
 
-static void cnl_setup_private_ppat(struct drm_i915_private *dev_priv)
+static struct intel_ppat_entry *
+__alloc_ppat_entry(struct intel_ppat *ppat, unsigned int index, u8 value)
 {
+	struct intel_ppat_entry *entry = &ppat->entries[index];
+
+	GEM_BUG_ON(index >= ppat->max_entries);
+	GEM_BUG_ON(test_bit(index, ppat->used));
+
+	entry->ppat = ppat;
+	entry->value = value;
+	kref_init(&entry->ref);
+	set_bit(index, ppat->used);
+	set_bit(index, ppat->dirty);
+
+	return entry;
+}
+
+static void __free_ppat_entry(struct intel_ppat_entry *entry)
+{
+	struct intel_ppat *ppat = entry->ppat;
+	unsigned int index = entry - ppat->entries;
+
+	GEM_BUG_ON(index >= ppat->max_entries);
+	GEM_BUG_ON(!test_bit(index, ppat->used));
+
+	entry->value = ppat->clear_value;
+	clear_bit(index, ppat->used);
+	set_bit(index, ppat->dirty);
+}
+
+/**
+ * intel_ppat_get - get a usable PPAT entry
+ * @i915: i915 device instance
+ * @value: the PPAT value required by the caller
+ *
+ * The function tries to search if there is an existing PPAT entry which
+ * matches with the required value. If perfectly matched, the existing PPAT
+ * entry will be used. If only partially matched, it will try to check if
+ * there is any available PPAT index. If yes, it will allocate a new PPAT
+ * index for the required entry and update the HW. If not, the partially
+ * matched entry will be used.
+ */
+const struct intel_ppat_entry *
+intel_ppat_get(struct drm_i915_private *i915, u8 value)
+{
+	struct intel_ppat *ppat = &i915->ppat;
+	struct intel_ppat_entry *entry;
+	unsigned int scanned, best_score;
+	int i;
+
+	GEM_BUG_ON(!ppat->max_entries);
+
+	scanned = best_score = 0;
+
+	for_each_set_bit(i, ppat->used, ppat->max_entries) {
+		unsigned int score;
+
+		entry = &ppat->entries[i];
+		score = ppat->match(entry->value, value);
+		if (score > best_score) {
+			if (score == INTEL_PPAT_PERFECT_MATCH) {
+				kref_get(&entry->ref);
+				return entry;
+			}
+			best_score = score;
+		}
+		scanned++;
+	}
+
+	if (scanned == ppat->max_entries) {
+		if (!best_score)
+			return ERR_PTR(-ENOSPC);
+
+		kref_get(&entry->ref);
+		return entry;
+	}
+
+	i = find_first_zero_bit(ppat->used, ppat->max_entries);
+	entry = __alloc_ppat_entry(ppat, i, value);
+	ppat->update_hw(i915);
+	return entry;
+}
+
+static void release_ppat(struct kref *kref)
+{
+	struct intel_ppat_entry *entry =
+		container_of(kref, struct intel_ppat_entry, ref);
+	struct drm_i915_private *i915 = entry->ppat->i915;
+
+	__free_ppat_entry(entry);
+	entry->ppat->update_hw(i915);
+}
+
+/**
+ * intel_ppat_put - put back the PPAT entry got from intel_ppat_get()
+ * @entry: an intel PPAT entry
+ *
+ * Put back the PPAT entry got from intel_ppat_get(). If the PPAT index of the
+ * entry is dynamically allocated, its reference count will be decreased. Once
+ * the reference count becomes into zero, the PPAT index becomes free again.
+ */
+void intel_ppat_put(const struct intel_ppat_entry *entry)
+{
+	struct intel_ppat *ppat = entry->ppat;
+	unsigned int index = entry - ppat->entries;
+
+	GEM_BUG_ON(!ppat->max_entries);
+
+	kref_put(&ppat->entries[index].ref, release_ppat);
+}
+
+static void cnl_private_pat_update_hw(struct drm_i915_private *dev_priv)
+{
+	struct intel_ppat *ppat = &dev_priv->ppat;
+	int i;
+
+	for_each_set_bit(i, ppat->dirty, ppat->max_entries) {
+		I915_WRITE(GEN10_PAT_INDEX(i), ppat->entries[i].value);
+		clear_bit(i, ppat->dirty);
+	}
+}
+
+static void bdw_private_pat_update_hw(struct drm_i915_private *dev_priv)
+{
+	struct intel_ppat *ppat = &dev_priv->ppat;
+	u64 pat = 0;
+	int i;
+
+	for (i = 0; i < ppat->max_entries; i++)
+		pat |= GEN8_PPAT(i, ppat->entries[i].value);
+
+	bitmap_clear(ppat->dirty, 0, ppat->max_entries);
+
+	I915_WRITE(GEN8_PRIVATE_PAT_LO, lower_32_bits(pat));
+	I915_WRITE(GEN8_PRIVATE_PAT_HI, upper_32_bits(pat));
+}
+
+static unsigned int bdw_private_pat_match(u8 src, u8 dst)
+{
+	unsigned int score = 0;
+
+	/* Cache attribute has to be matched. */
+	if (GEN8_PPAT_GET_CA(src) != GEN8_PPAT_GET_CA(dst))
+		return 0;
+
+	if (GEN8_PPAT_GET_TC(src) == GEN8_PPAT_GET_TC(dst))
+		score += 2;
+
+	if (GEN8_PPAT_GET_AGE(src) == GEN8_PPAT_GET_AGE(dst))
+		score += 1;
+
+	if (score == 3)
+		return INTEL_PPAT_PERFECT_MATCH;
+
+	return score;
+}
+
+static unsigned int chv_private_pat_match(u8 src, u8 dst)
+{
+	return (CHV_PPAT_GET_SNOOP(src) == CHV_PPAT_GET_SNOOP(dst)) ?
+		INTEL_PPAT_PERFECT_MATCH : 0;
+}
+
+static void cnl_setup_private_ppat(struct intel_ppat *ppat)
+{
+	ppat->max_entries = 8;
+	ppat->update_hw = cnl_private_pat_update_hw;
+	ppat->match = bdw_private_pat_match;
+	ppat->clear_value = GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3);
+
 	/* XXX: spec is unclear if this is still needed for CNL+ */
-	if (!USES_PPGTT(dev_priv)) {
-		I915_WRITE(GEN10_PAT_INDEX(0), GEN8_PPAT_UC);
+	if (!USES_PPGTT(ppat->i915)) {
+		__alloc_ppat_entry(ppat, 0, GEN8_PPAT_UC);
 		return;
 	}
 
-	I915_WRITE(GEN10_PAT_INDEX(0), GEN8_PPAT_WB | GEN8_PPAT_LLC);
-	I915_WRITE(GEN10_PAT_INDEX(1), GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);
-	I915_WRITE(GEN10_PAT_INDEX(2), GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
-	I915_WRITE(GEN10_PAT_INDEX(3), GEN8_PPAT_UC);
-	I915_WRITE(GEN10_PAT_INDEX(4), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
-	I915_WRITE(GEN10_PAT_INDEX(5), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
-	I915_WRITE(GEN10_PAT_INDEX(6), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
-	I915_WRITE(GEN10_PAT_INDEX(7), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
+	__alloc_ppat_entry(ppat, 0, GEN8_PPAT_WB | GEN8_PPAT_LLC);
+	__alloc_ppat_entry(ppat, 1, GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);
+	__alloc_ppat_entry(ppat, 2, GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
+	__alloc_ppat_entry(ppat, 3, GEN8_PPAT_UC);
+	__alloc_ppat_entry(ppat, 4, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
+	__alloc_ppat_entry(ppat, 5, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
+	__alloc_ppat_entry(ppat, 6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
+	__alloc_ppat_entry(ppat, 7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
 }
 
 /* The GGTT and PPGTT need a private PPAT setup in order to handle cacheability
  * bits. When using advanced contexts each context stores its own PAT, but
  * writing this data shouldn't be harmful even in those cases. */
-static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv)
+static void bdw_setup_private_ppat(struct intel_ppat *ppat)
 {
-	u64 pat;
+	ppat->max_entries = 8;
+	ppat->update_hw = bdw_private_pat_update_hw;
+	ppat->match = bdw_private_pat_match;
+	ppat->clear_value = GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3);
 
-	pat = GEN8_PPAT(0, GEN8_PPAT_WB | GEN8_PPAT_LLC)     | /* for normal objects, no eLLC */
-	      GEN8_PPAT(1, GEN8_PPAT_WC | GEN8_PPAT_LLCELLC) | /* for something pointing to ptes? */
-	      GEN8_PPAT(2, GEN8_PPAT_WT | GEN8_PPAT_LLCELLC) | /* for scanout with eLLC */
-	      GEN8_PPAT(3, GEN8_PPAT_UC)                     | /* Uncached objects, mostly for scanout */
-	      GEN8_PPAT(4, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0)) |
-	      GEN8_PPAT(5, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1)) |
-	      GEN8_PPAT(6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2)) |
-	      GEN8_PPAT(7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
-
-	if (!USES_PPGTT(dev_priv))
+	if (!USES_PPGTT(ppat->i915)) {
 		/* Spec: "For GGTT, there is NO pat_sel[2:0] from the entry,
 		 * so RTL will always use the value corresponding to
 		 * pat_sel = 000".
@@ -2790,17 +2952,26 @@ static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv)
 		 * So we can still hold onto all our assumptions wrt cpu
 		 * clflushing on LLC machines.
 		 */
-		pat = GEN8_PPAT(0, GEN8_PPAT_UC);
+		__alloc_ppat_entry(ppat, 0, GEN8_PPAT_UC);
+		return;
+	}
 
-	/* XXX: spec defines this as 2 distinct registers. It's unclear if a 64b
-	 * write would work. */
-	I915_WRITE(GEN8_PRIVATE_PAT_LO, pat);
-	I915_WRITE(GEN8_PRIVATE_PAT_HI, pat >> 32);
+	__alloc_ppat_entry(ppat, 0, GEN8_PPAT_WB | GEN8_PPAT_LLC);      /* for normal objects, no eLLC */
+	__alloc_ppat_entry(ppat, 1, GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);  /* for something pointing to ptes? */
+	__alloc_ppat_entry(ppat, 2, GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);  /* for scanout with eLLC */
+	__alloc_ppat_entry(ppat, 3, GEN8_PPAT_UC);                      /* Uncached objects, mostly for scanout */
+	__alloc_ppat_entry(ppat, 4, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
+	__alloc_ppat_entry(ppat, 5, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
+	__alloc_ppat_entry(ppat, 6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
+	__alloc_ppat_entry(ppat, 7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
 }
 
-static void chv_setup_private_ppat(struct drm_i915_private *dev_priv)
+static void chv_setup_private_ppat(struct intel_ppat *ppat)
 {
-	u64 pat;
+	ppat->max_entries = 8;
+	ppat->update_hw = bdw_private_pat_update_hw;
+	ppat->match = chv_private_pat_match;
+	ppat->clear_value = CHV_PPAT_SNOOP;
 
 	/*
 	 * Map WB on BDW to snooped on CHV.
@@ -2820,17 +2991,15 @@ static void chv_setup_private_ppat(struct drm_i915_private *dev_priv)
 	 * Which means we must set the snoop bit in PAT entry 0
 	 * in order to keep the global status page working.
 	 */
-	pat = GEN8_PPAT(0, CHV_PPAT_SNOOP) |
-	      GEN8_PPAT(1, 0) |
-	      GEN8_PPAT(2, 0) |
-	      GEN8_PPAT(3, 0) |
-	      GEN8_PPAT(4, CHV_PPAT_SNOOP) |
-	      GEN8_PPAT(5, CHV_PPAT_SNOOP) |
-	      GEN8_PPAT(6, CHV_PPAT_SNOOP) |
-	      GEN8_PPAT(7, CHV_PPAT_SNOOP);
 
-	I915_WRITE(GEN8_PRIVATE_PAT_LO, pat);
-	I915_WRITE(GEN8_PRIVATE_PAT_HI, pat >> 32);
+	__alloc_ppat_entry(ppat, 0, CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, 1, 0);
+	__alloc_ppat_entry(ppat, 2, 0);
+	__alloc_ppat_entry(ppat, 3, 0);
+	__alloc_ppat_entry(ppat, 4, CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, 5, CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, 6, CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, 7, CHV_PPAT_SNOOP);
 }
 
 static void gen6_gmch_remove(struct i915_address_space *vm)
@@ -2843,12 +3012,27 @@ static void gen6_gmch_remove(struct i915_address_space *vm)
 
 static void setup_private_pat(struct drm_i915_private *dev_priv)
 {
+	struct intel_ppat *ppat = &dev_priv->ppat;
+	int i;
+
+	ppat->i915 = dev_priv;
+
 	if (INTEL_GEN(dev_priv) >= 10)
-		cnl_setup_private_ppat(dev_priv);
+		cnl_setup_private_ppat(ppat);
 	else if (IS_CHERRYVIEW(dev_priv) || IS_GEN9_LP(dev_priv))
-		chv_setup_private_ppat(dev_priv);
+		chv_setup_private_ppat(ppat);
 	else
-		bdw_setup_private_ppat(dev_priv);
+		bdw_setup_private_ppat(ppat);
+
+	GEM_BUG_ON(ppat->max_entries > INTEL_MAX_PPAT_ENTRIES);
+
+	for_each_clear_bit(i, ppat->used, ppat->max_entries) {
+		ppat->entries[i].value = ppat->clear_value;
+		ppat->entries[i].ppat = ppat;
+		set_bit(i, ppat->dirty);
+	}
+
+	ppat->update_hw(dev_priv);
 }
 
 static int gen8_gmch_probe(struct i915_ggtt *ggtt)
@@ -3162,13 +3346,10 @@ void i915_gem_restore_gtt_mappings(struct drm_i915_private *dev_priv)
 	ggtt->base.closed = false;
 
 	if (INTEL_GEN(dev_priv) >= 8) {
-		if (INTEL_GEN(dev_priv) >= 10)
-			cnl_setup_private_ppat(dev_priv);
-		else if (IS_CHERRYVIEW(dev_priv) || IS_GEN9_LP(dev_priv))
-			chv_setup_private_ppat(dev_priv);
-		else
-			bdw_setup_private_ppat(dev_priv);
+		struct intel_ppat *ppat = &dev_priv->ppat;
 
+		bitmap_set(ppat->dirty, 0, ppat->max_entries);
+		dev_priv->ppat.update_hw(dev_priv);
 		return;
 	}
 
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index b4e3aa7..e10ca89 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -143,6 +143,11 @@ typedef u64 gen8_ppgtt_pml4e_t;
 #define GEN8_PPAT_ELLC_OVERRIDE		(0<<2)
 #define GEN8_PPAT(i, x)			((u64)(x) << ((i) * 8))
 
+#define GEN8_PPAT_GET_CA(x) ((x) & 3)
+#define GEN8_PPAT_GET_TC(x) ((x) & (3 << 2))
+#define GEN8_PPAT_GET_AGE(x) ((x) & (3 << 4))
+#define CHV_PPAT_GET_SNOOP(x) ((x) & (1 << 6))
+
 struct sg_table;
 
 struct intel_rotation_info {
@@ -536,6 +541,37 @@ i915_vm_to_ggtt(struct i915_address_space *vm)
 	return container_of(vm, struct i915_ggtt, base);
 }
 
+#define INTEL_MAX_PPAT_ENTRIES 8
+#define INTEL_PPAT_PERFECT_MATCH (~0U)
+
+struct intel_ppat;
+
+struct intel_ppat_entry {
+	struct intel_ppat *ppat;
+	struct kref ref;
+	u8 value;
+};
+
+struct intel_ppat {
+	struct intel_ppat_entry entries[INTEL_MAX_PPAT_ENTRIES];
+	DECLARE_BITMAP(used, INTEL_MAX_PPAT_ENTRIES);
+	DECLARE_BITMAP(dirty, INTEL_MAX_PPAT_ENTRIES);
+	unsigned int max_entries;
+	u8 clear_value;
+	/*
+	 * Return a score to show how two PPAT values match,
+	 * a INTEL_PPAT_PERFECT_MATCH indicates a perfect match
+	 */
+	unsigned int (*match)(u8 src, u8 dst);
+	void (*update_hw)(struct drm_i915_private *i915);
+
+	struct drm_i915_private *i915;
+};
+
+const struct intel_ppat_entry *
+intel_ppat_get(struct drm_i915_private *i915, u8 value);
+void intel_ppat_put(const struct intel_ppat_entry *entry);
+
 int i915_gem_init_aliasing_ppgtt(struct drm_i915_private *i915);
 void i915_gem_fini_aliasing_ppgtt(struct drm_i915_private *i915);
 
-- 
2.7.4

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH v10 3/5] drm/i915: Remove the "INDEX" suffix from PPAT marcos

[Zhi Wang]

Remove the "INDEX" suffix from PPAT marcos as they are bits actually, not
indexes.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ben Widawsky <benjamin.widawsky@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/gvt/gtt.c      |  2 +-
 drivers/gpu/drm/i915/i915_gem_gtt.c | 10 +++++-----
 drivers/gpu/drm/i915/i915_gem_gtt.h |  8 ++++----
 3 files changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c
index e6dfc33..7b5d9bf 100644
--- a/drivers/gpu/drm/i915/gvt/gtt.c
+++ b/drivers/gpu/drm/i915/gvt/gtt.c
@@ -1972,7 +1972,7 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu,
 		 */
 		se.val64 |= _PAGE_PRESENT | _PAGE_RW;
 		if (type == GTT_TYPE_PPGTT_PDE_PT)
-			se.val64 |= PPAT_CACHED_INDEX;
+			se.val64 |= PPAT_CACHED;
 
 		for (i = 0; i < page_entry_num; i++)
 			ops->set_entry(scratch_pt, &se, i, false, 0, vgpu);
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index b48aac1..e83063d 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -230,13 +230,13 @@ static gen8_pte_t gen8_pte_encode(dma_addr_t addr,
 
 	switch (level) {
 	case I915_CACHE_NONE:
-		pte |= PPAT_UNCACHED_INDEX;
+		pte |= PPAT_UNCACHED;
 		break;
 	case I915_CACHE_WT:
-		pte |= PPAT_DISPLAY_ELLC_INDEX;
+		pte |= PPAT_DISPLAY_ELLC;
 		break;
 	default:
-		pte |= PPAT_CACHED_INDEX;
+		pte |= PPAT_CACHED;
 		break;
 	}
 
@@ -249,9 +249,9 @@ static gen8_pde_t gen8_pde_encode(const dma_addr_t addr,
 	gen8_pde_t pde = _PAGE_PRESENT | _PAGE_RW;
 	pde |= addr;
 	if (level != I915_CACHE_NONE)
-		pde |= PPAT_CACHED_PDE_INDEX;
+		pde |= PPAT_CACHED_PDE;
 	else
-		pde |= PPAT_UNCACHED_INDEX;
+		pde |= PPAT_UNCACHED;
 	return pde;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index e10ca89..0178416 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -126,10 +126,10 @@ typedef u64 gen8_ppgtt_pml4e_t;
  * tables */
 #define GEN8_PDPE_MASK			0x1ff
 
-#define PPAT_UNCACHED_INDEX		(_PAGE_PWT | _PAGE_PCD)
-#define PPAT_CACHED_PDE_INDEX		0 /* WB LLC */
-#define PPAT_CACHED_INDEX		_PAGE_PAT /* WB LLCeLLC */
-#define PPAT_DISPLAY_ELLC_INDEX		_PAGE_PCD /* WT eLLC */
+#define PPAT_UNCACHED			(_PAGE_PWT | _PAGE_PCD)
+#define PPAT_CACHED_PDE			0 /* WB LLC */
+#define PPAT_CACHED			_PAGE_PAT /* WB LLCeLLC */
+#define PPAT_DISPLAY_ELLC		_PAGE_PCD /* WT eLLC */
 
 #define CHV_PPAT_SNOOP			(1<<6)
 #define GEN8_PPAT_AGE(x)		(x<<4)
-- 
2.7.4

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH v10 4/5] drm/i915: Do not allocate unused PPAT entries

[Zhi Wang]

Only PPAT entries 0/2/3/4 are using. Remove extra PPAT entry allocation
during initialization.

v8:

- Move ppat_index() into i915_gem_gtt.c. (Chris)
- Change the name of ppat_bits_to_index to ppat_index.

Cc: Ben Widawsky <benjamin.widawsky@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Suggested-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/i915_gem_gtt.c | 53 +++++++++++++++++++------------------
 1 file changed, 27 insertions(+), 26 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index e83063d..47feb93 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -2905,6 +2905,13 @@ static unsigned int chv_private_pat_match(u8 src, u8 dst)
 		INTEL_PPAT_PERFECT_MATCH : 0;
 }
 
+/* PPAT index = 4 * PAT + 2 * PCD + PWT */
+static inline unsigned int ppat_index(unsigned int bits)
+{
+	return (4 * !!(bits & _PAGE_PAT) + 2 * !!(bits & _PAGE_PCD)
+		+ !!(bits & _PAGE_PWT));
+}
+
 static void cnl_setup_private_ppat(struct intel_ppat *ppat)
 {
 	ppat->max_entries = 8;
@@ -2914,18 +2921,15 @@ static void cnl_setup_private_ppat(struct intel_ppat *ppat)
 
 	/* XXX: spec is unclear if this is still needed for CNL+ */
 	if (!USES_PPGTT(ppat->i915)) {
-		__alloc_ppat_entry(ppat, 0, GEN8_PPAT_UC);
+		__alloc_ppat_entry(ppat, ppat_index(PPAT_CACHED_PDE), GEN8_PPAT_UC);
 		return;
 	}
 
-	__alloc_ppat_entry(ppat, 0, GEN8_PPAT_WB | GEN8_PPAT_LLC);
-	__alloc_ppat_entry(ppat, 1, GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);
-	__alloc_ppat_entry(ppat, 2, GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
-	__alloc_ppat_entry(ppat, 3, GEN8_PPAT_UC);
-	__alloc_ppat_entry(ppat, 4, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
-	__alloc_ppat_entry(ppat, 5, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
-	__alloc_ppat_entry(ppat, 6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
-	__alloc_ppat_entry(ppat, 7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
+	/* See gen8_pte_encode() for the mapping from cache-level to PPAT */
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_CACHED_PDE), GEN8_PPAT_WB | GEN8_PPAT_LLC);
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_DISPLAY_ELLC), GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_UNCACHED), GEN8_PPAT_UC);
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_CACHED), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
 }
 
 /* The GGTT and PPGTT need a private PPAT setup in order to handle cacheability
@@ -2952,18 +2956,18 @@ static void bdw_setup_private_ppat(struct intel_ppat *ppat)
 		 * So we can still hold onto all our assumptions wrt cpu
 		 * clflushing on LLC machines.
 		 */
-		__alloc_ppat_entry(ppat, 0, GEN8_PPAT_UC);
+		__alloc_ppat_entry(ppat, ppat_index(PPAT_CACHED_PDE), GEN8_PPAT_UC);
 		return;
 	}
 
-	__alloc_ppat_entry(ppat, 0, GEN8_PPAT_WB | GEN8_PPAT_LLC);      /* for normal objects, no eLLC */
-	__alloc_ppat_entry(ppat, 1, GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);  /* for something pointing to ptes? */
-	__alloc_ppat_entry(ppat, 2, GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);  /* for scanout with eLLC */
-	__alloc_ppat_entry(ppat, 3, GEN8_PPAT_UC);                      /* Uncached objects, mostly for scanout */
-	__alloc_ppat_entry(ppat, 4, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
-	__alloc_ppat_entry(ppat, 5, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
-	__alloc_ppat_entry(ppat, 6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
-	__alloc_ppat_entry(ppat, 7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
+	/* See gen8_pte_encode() for the mapping from cache-level to PPAT */
+	/* for normal objects, no eLLC */
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_CACHED_PDE), GEN8_PPAT_WB | GEN8_PPAT_LLC);
+	/* for scanout with eLLC */
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_DISPLAY_ELLC), GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
+	/* Uncached objects, mostly for scanout */
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_UNCACHED), GEN8_PPAT_UC);
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_CACHED), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
 }
 
 static void chv_setup_private_ppat(struct intel_ppat *ppat)
@@ -2992,14 +2996,11 @@ static void chv_setup_private_ppat(struct intel_ppat *ppat)
 	 * in order to keep the global status page working.
 	 */
 
-	__alloc_ppat_entry(ppat, 0, CHV_PPAT_SNOOP);
-	__alloc_ppat_entry(ppat, 1, 0);
-	__alloc_ppat_entry(ppat, 2, 0);
-	__alloc_ppat_entry(ppat, 3, 0);
-	__alloc_ppat_entry(ppat, 4, CHV_PPAT_SNOOP);
-	__alloc_ppat_entry(ppat, 5, CHV_PPAT_SNOOP);
-	__alloc_ppat_entry(ppat, 6, CHV_PPAT_SNOOP);
-	__alloc_ppat_entry(ppat, 7, CHV_PPAT_SNOOP);
+	/* See gen8_pte_encode() for the mapping from cache-level to PPAT */
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_CACHED_PDE), CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_DISPLAY_ELLC), 0);
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_UNCACHED), 0);
+	__alloc_ppat_entry(ppat, ppat_index(PPAT_CACHED), CHV_PPAT_SNOOP);
 }
 
 static void gen6_gmch_remove(struct i915_address_space *vm)
-- 
2.7.4

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH v10 5/5] drm/i915/selftests: Introduce live tests of private PAT management

[Zhi Wang]

Introduce two live tests of private PAT management:

igt_ppat_init - This test is to check if all the PPAT configurations are
written into HW.

igt_ppat_get - This test performs several sub-tests on intel_ppat_get()
and intel_ppat_put().

The "perfect match" test case will try to get a PPAT entry with an existing
value, then check if the returned PPAT entry is the same one.

The "alloc entries" test case will run out of PPAT table, and check if all
the requested values are put into the newly allocated PPAT entries.

The negative test case will try to generate a new PPAT value, and get it
when PPAT table is full.

The "partial match" test case will generate a parital matched value from
the existing PPAT table and try to match it.

The "re-alloc" test case will try to free and then allocate a new entry
when the PPAT table is full.

The "put entries" test case will free all the PPAT entries that allocated
in "alloc entries" test case. It will check if the values of freed PPAT
entries turn into ppat->clear_value.

v10:

- Refine code structure.
- Introduce "re-alloc" test case. (Chris)

v9:

- Refine generate_new_value(). (Chris)
- Refine failure output. (Chris)
- Refine test flow of "perfect match". (Chris)
- Introduce another negative test case after "partial match". (Chris)

v8:

- Remove noisy output. (Chris)
- Add negative test case. (Chris)

Cc: Ben Widawsky <benjamin.widawsky@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/selftests/i915_gem_gtt.c | 378 ++++++++++++++++++++++++++
 1 file changed, 378 insertions(+)

diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
index 6b132ca..88c87ac 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
@@ -1094,6 +1094,382 @@ static int igt_ggtt_page(void *arg)
 	return err;
 }
 
+static int check_cnl_ppat(struct drm_i915_private *dev_priv)
+{
+	struct intel_ppat *ppat = &dev_priv->ppat;
+	int i;
+
+	for (i = 0; i < ppat->max_entries; i++) {
+		u32 value = I915_READ(GEN10_PAT_INDEX(i));
+
+		if (value != ppat->entries[i].value) {
+			pr_err("check PPAT failed: expected %x found %x\n",
+			       ppat->entries[i].value, value);
+			return -EINVAL;
+		}
+	}
+	return 0;
+}
+
+static int check_bdw_ppat(struct drm_i915_private *dev_priv)
+{
+	struct intel_ppat *ppat = &dev_priv->ppat;
+	u64 pat, hw_pat;
+	int i;
+
+	pat = hw_pat = 0;
+
+	for (i = 0; i < ppat->max_entries; i++)
+		pat |= GEN8_PPAT(i, ppat->entries[i].value);
+
+	hw_pat = I915_READ(GEN8_PRIVATE_PAT_HI);
+	hw_pat <<= 32;
+	hw_pat |= I915_READ(GEN8_PRIVATE_PAT_LO);
+
+	if (pat != hw_pat) {
+		pr_err("check PPAT failed: expected %llx found %llx\n",
+		        pat, hw_pat);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int igt_ppat_check(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	int ret;
+
+	if (!i915->ppat.max_entries)
+		return 0;
+
+	if (INTEL_GEN(i915) >= 10)
+		ret = check_cnl_ppat(i915);
+	else
+		ret = check_bdw_ppat(i915);
+
+	if (ret)
+		pr_err("check PPAT failed\n");
+
+	return ret;
+}
+
+static bool value_is_new(struct intel_ppat *ppat, u8 value)
+{
+	int i;
+
+	for_each_set_bit(i, ppat->used, ppat->max_entries) {
+		if (value != ppat->entries[i].value)
+			continue;
+
+		return false;
+	}
+	return true;
+}
+
+static bool value_for_partial_test(struct intel_ppat *ppat, u8 value)
+{
+	int i;
+
+	if (!value_is_new(ppat, value))
+		return false;
+
+	/*
+	 * At least, there should be one entry whose cache attribute is
+	 * same with the required value.
+	 */
+	for_each_set_bit(i, ppat->used, ppat->max_entries) {
+		if (GEN8_PPAT_GET_CA(value) !=
+		    GEN8_PPAT_GET_CA(ppat->entries[i].value))
+			continue;
+
+		return true;
+	}
+	return false;
+}
+
+static bool value_for_negative_test(struct intel_ppat *ppat, u8 value)
+{
+	int i;
+
+	if (!value_is_new(ppat, value))
+		return false;
+
+	/*
+	 * cache attribute has to be different, so i915_ppat_get() would
+	 * allocate a new entry.
+	 */
+	for_each_set_bit(i, ppat->used, ppat->max_entries) {
+		if (GEN8_PPAT_GET_CA(value) ==
+		    GEN8_PPAT_GET_CA(ppat->entries[i].value))
+			return false;
+	}
+	return true;
+}
+
+static u8 generate_new_value(struct intel_ppat *ppat,
+			     bool (*check_value)(struct intel_ppat *, u8))
+{
+	u8 ca[] = { GEN8_PPAT_WB, GEN8_PPAT_WT, GEN8_PPAT_UC, GEN8_PPAT_WC };
+	u8 tc[] = { GEN8_PPAT_LLC, GEN8_PPAT_LLCELLC, GEN8_PPAT_LLCeLLC };
+	u8 age[] = { GEN8_PPAT_AGE(3), GEN8_PPAT_AGE(2), GEN8_PPAT_AGE(1),
+		     GEN8_PPAT_AGE(0) };
+	int ca_index, tc_index, age_index;
+	u8 value;
+
+#define for_each_ppat_attr(ca_index, tc_index, age_index) \
+	for ((ca_index) = 0 ; (ca_index) < ARRAY_SIZE(ca); (ca_index)++) \
+	for ((tc_index) = 0; (tc_index) < ARRAY_SIZE(tc); (tc_index)++) \
+	for ((age_index) = 0; (age_index) < ARRAY_SIZE(age); (age_index)++)
+
+	for_each_ppat_attr(ca_index, tc_index, age_index) {
+		value = age[age_index] | ca[ca_index] | tc[tc_index];
+		if (check_value(ppat, value))
+			return value;
+	}
+#undef for_each_ppat_attr
+	return 0;
+}
+
+static const struct intel_ppat_entry *
+generate_and_check_new_value(struct intel_ppat *ppat)
+{
+	struct drm_i915_private *i915 = ppat->i915;
+	const struct intel_ppat_entry *entry;
+	u8 value;
+	DECLARE_BITMAP(used, INTEL_MAX_PPAT_ENTRIES);
+
+	value = generate_new_value(ppat, value_is_new);
+	if (!value) {
+		pr_err("fail to generate new value\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	bitmap_copy(used, ppat->used, ppat->max_entries);
+
+	entry = intel_ppat_get(i915, value);
+	if (IS_ERR(entry)) {
+		pr_err("fail to get new entry\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (entry->value != value) {
+		pr_err("value is not expected, expected %x found %x\n",
+				value, entry->value);
+		goto err;
+	}
+
+	if (bitmap_equal(used, ppat->used, ppat->max_entries)) {
+		pr_err("fail to alloc a new entry\n");
+		goto err;
+	}
+
+	return entry;
+err:
+	intel_ppat_put(entry);
+	return ERR_PTR(-EINVAL);
+}
+
+static int put_and_check_entry(const struct intel_ppat_entry *entry)
+{
+	intel_ppat_put(entry);
+
+	if (entry->value != entry->ppat->clear_value) {
+		pr_err("fail to put ppat value\n");
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int perform_perfect_match_test(struct intel_ppat *ppat)
+{
+	struct drm_i915_private *i915 = ppat->i915;
+	const struct intel_ppat_entry *entry;
+	int ret, i;
+
+	for_each_set_bit(i, ppat->used, ppat->max_entries) {
+		entry = intel_ppat_get(i915, ppat->entries[i].value);
+		if (IS_ERR(entry))
+			return PTR_ERR(entry);
+
+		if (entry != &ppat->entries[i]) {
+			pr_err("entry is not expected\n");
+			intel_ppat_put(entry);
+			return -EINVAL;
+		}
+		ret = put_and_check_entry(entry);
+		if (ret)
+			return ret;
+	}
+	return 0;
+}
+
+static int perform_negative_test(struct intel_ppat *ppat)
+{
+	struct drm_i915_private *i915 = ppat->i915;
+	const struct intel_ppat_entry *entry;
+	u8 value;
+
+	value = generate_new_value(ppat, value_for_negative_test);
+	if (!value) {
+		pr_err("fail to generate new value for negative test 2\n");
+		return -EINVAL;
+	}
+
+	entry = intel_ppat_get(i915, value);
+	if (!IS_ERR(entry) || PTR_ERR(entry) != -ENOSPC) {
+		pr_err("fail on negative test\n");
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int perform_partial_match_test(struct intel_ppat *ppat)
+{
+	struct drm_i915_private *i915 = ppat->i915;
+	const struct intel_ppat_entry *entry;
+	u8 value;
+	int ret;
+
+	value = generate_new_value(ppat, value_for_partial_test);
+	if (!value) {
+		pr_err("fail to generate new value for partial test\n");
+		return -EINVAL;
+	}
+
+	entry = intel_ppat_get(i915, value);
+	if (IS_ERR(entry)) {
+		pr_err("fail to get new entry\n");
+		return PTR_ERR(entry);
+	}
+
+	if (!(entry->value != value &&
+	    GEN8_PPAT_GET_CA(entry->value) == GEN8_PPAT_GET_CA(value))) {
+		pr_err("value is not expected, expected %x found %x\n",
+				value, entry->value);
+		intel_ppat_put(entry);
+		return -EINVAL;
+	}
+
+	ret = put_and_check_entry(entry);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+static int igt_ppat_get(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct intel_ppat *ppat = &i915->ppat;
+	const struct intel_ppat_entry **entries, **p;
+	const struct intel_ppat_entry *entry;
+	unsigned int size = 0;
+	int i, ret;
+
+	if (!ppat->max_entries)
+		return 0;
+
+	ret = igt_ppat_check(i915);
+	if (ret)
+		return ret;
+
+	/* case 1: perfect match */
+	ret = perform_perfect_match_test(ppat);
+	if (ret) {
+		pr_err("fail on perfect match test\n");
+		return ret;
+	}
+
+	/* case 2: alloc new entries */
+	entries = NULL;
+	ret = 0;
+
+	while (!bitmap_full(ppat->used, ppat->max_entries)) {
+		p = krealloc(entries, (size + 1) *
+				   sizeof(struct intel_ppat_entry *),
+				   GFP_KERNEL);
+		if (!p) {
+			ret = -ENOMEM;
+			goto ppat_put;
+		}
+
+		entries = p;
+
+		p = &entries[size++];
+		*p = NULL;
+
+		entry = generate_and_check_new_value(ppat);
+		if (IS_ERR(entry)) {
+			ret = PTR_ERR(entry);
+			pr_err("fail on alloc new entries test\n");
+			goto ppat_put;
+		}
+		*p = entry;
+	}
+
+	/* case 3: negative test 1, suppose PPAT table is full now */
+	ret = perform_negative_test(ppat);
+	if (ret) {
+		pr_err("fail on negative test 1\n");
+		goto ppat_put;
+	}
+
+	/* case 4: partial match */
+	ret = perform_partial_match_test(ppat);
+	if (ret) {
+		pr_err("fail on partial match test\n");
+		goto ppat_put;
+	}
+
+	/* case 3: negative test 2, suppose PPAT table is still full now */
+	ret = perform_negative_test(ppat);
+	if (ret) {
+		pr_err("fail on negative test 2\n");
+		goto ppat_put;
+	}
+
+	/* case 5: re-alloc test, make a hole and it should work again */
+	if (entries) {
+		for(i = 0; i < size; i++) {
+			entry = entries[i];
+
+			ret = put_and_check_entry(entry);
+			entries[i] = NULL;
+			if (ret) {
+				pr_err("fail on re-alloc test\n");
+				goto ppat_put;
+			}
+
+			entry = generate_and_check_new_value(ppat);
+			if (IS_ERR(entry)) {
+				ret = PTR_ERR(entry);
+				pr_err("fail on re-alloc test\n");
+				goto ppat_put;
+			}
+			entries[i] = entry;
+		}
+	}
+
+ppat_put:
+	if (entries) {
+		for (i = 0; i < size; i++) {
+			if (IS_ERR(entries[i]) || !entries[i])
+				continue;
+
+			if (ret)
+				intel_ppat_put(entry);
+			else
+				ret = put_and_check_entry(entries[i]);
+		}
+		kfree(entries);
+		entries = NULL;
+	}
+	if (ret)
+		return ret;
+
+	return igt_ppat_check(i915);
+}
+
 static void track_vma_bind(struct i915_vma *vma)
 {
 	struct drm_i915_gem_object *obj = vma->obj;
@@ -1560,6 +1936,8 @@ int i915_gem_gtt_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(igt_ggtt_pot),
 		SUBTEST(igt_ggtt_fill),
 		SUBTEST(igt_ggtt_page),
+		SUBTEST(igt_ppat_check),
+		SUBTEST(igt_ppat_get),
 	};
 
 	GEM_BUG_ON(offset_in_page(i915->ggtt.base.total));
-- 
2.7.4

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH v10 5/5] drm/i915/selftests: Introduce live tests of private PAT management

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 5062 bytes --]

Hi Zhi,

[auto build test ERROR on drm-intel/for-linux-next]
[also build test ERROR on next-20170913]
[cannot apply to v4.13]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Zhi-Wang/drm-i915-Factor-out-setup_private_pat/20170912-152713
base:   git://anongit.freedesktop.org/drm-intel for-linux-next
config: x86_64-randconfig-a0-09140023 (attached as .config)
compiler: gcc-4.4 (Debian 4.4.7-8) 4.4.7
reproduce:
        # save the attached .config to linux build tree
        make ARCH=x86_64 

Note: it may well be a FALSE warning. FWIW you are at least aware of it now.
http://gcc.gnu.org/wiki/Better_Uninitialized_Warnings

All errors (new ones prefixed by >>):

   cc1: warnings being treated as errors
   drivers/gpu/drm/i915/i915_gem_gtt.c: In function 'intel_ppat_get':
   drivers/gpu/drm/i915/i915_gem_gtt.c:2791: error: 'entry' may be used uninitialized in this function
   In file included from drivers/gpu/drm/i915/i915_gem_gtt.c:3767:
   drivers/gpu/drm/i915/selftests/i915_gem_gtt.c: In function 'igt_ppat_get':
>> drivers/gpu/drm/i915/selftests/i915_gem_gtt.c:1365: error: 'entry' may be used uninitialized in this function
   drivers/gpu/drm/i915/i915_gem_gtt.c: In function 'gen8_ppgtt_insert_4lvl':
   drivers/gpu/drm/i915/i915_gem_gtt.c:938: error: 'iter.sg' is used uninitialized in this function
   drivers/gpu/drm/i915/i915_gem_gtt.c:939: error: 'iter.dma' is used uninitialized in this function
   drivers/gpu/drm/i915/i915_gem_gtt.c: In function 'gen8_ppgtt_insert_3lvl':
   drivers/gpu/drm/i915/i915_gem_gtt.c:921: error: 'iter.sg' is used uninitialized in this function
   drivers/gpu/drm/i915/i915_gem_gtt.c:922: error: 'iter.dma' is used uninitialized in this function

vim +/entry +1365 drivers/gpu/drm/i915/selftests/i915_gem_gtt.c

  1359	
  1360	static int igt_ppat_get(void *arg)
  1361	{
  1362		struct drm_i915_private *i915 = arg;
  1363		struct intel_ppat *ppat = &i915->ppat;
  1364		const struct intel_ppat_entry **entries, **p;
> 1365		const struct intel_ppat_entry *entry;
  1366		unsigned int size = 0;
  1367		int i, ret;
  1368	
  1369		if (!ppat->max_entries)
  1370			return 0;
  1371	
  1372		ret = igt_ppat_check(i915);
  1373		if (ret)
  1374			return ret;
  1375	
  1376		/* case 1: perfect match */
  1377		ret = perform_perfect_match_test(ppat);
  1378		if (ret) {
  1379			pr_err("fail on perfect match test\n");
  1380			return ret;
  1381		}
  1382	
  1383		/* case 2: alloc new entries */
  1384		entries = NULL;
  1385		ret = 0;
  1386	
  1387		while (!bitmap_full(ppat->used, ppat->max_entries)) {
  1388			p = krealloc(entries, (size + 1) *
  1389					   sizeof(struct intel_ppat_entry *),
  1390					   GFP_KERNEL);
  1391			if (!p) {
  1392				ret = -ENOMEM;
  1393				goto ppat_put;
  1394			}
  1395	
  1396			entries = p;
  1397	
  1398			p = &entries[size++];
  1399			*p = NULL;
  1400	
  1401			entry = generate_and_check_new_value(ppat);
  1402			if (IS_ERR(entry)) {
  1403				ret = PTR_ERR(entry);
  1404				pr_err("fail on alloc new entries test\n");
  1405				goto ppat_put;
  1406			}
  1407			*p = entry;
  1408		}
  1409	
  1410		/* case 3: negative test 1, suppose PPAT table is full now */
  1411		ret = perform_negative_test(ppat);
  1412		if (ret) {
  1413			pr_err("fail on negative test 1\n");
  1414			goto ppat_put;
  1415		}
  1416	
  1417		/* case 4: partial match */
  1418		ret = perform_partial_match_test(ppat);
  1419		if (ret) {
  1420			pr_err("fail on partial match test\n");
  1421			goto ppat_put;
  1422		}
  1423	
  1424		/* case 3: negative test 2, suppose PPAT table is still full now */
  1425		ret = perform_negative_test(ppat);
  1426		if (ret) {
  1427			pr_err("fail on negative test 2\n");
  1428			goto ppat_put;
  1429		}
  1430	
  1431		/* case 5: re-alloc test, make a hole and it should work again */
  1432		if (entries) {
  1433			for(i = 0; i < size; i++) {
  1434				entry = entries[i];
  1435	
  1436				ret = put_and_check_entry(entry);
  1437				entries[i] = NULL;
  1438				if (ret) {
  1439					pr_err("fail on re-alloc test\n");
  1440					goto ppat_put;
  1441				}
  1442	
  1443				entry = generate_and_check_new_value(ppat);
  1444				if (IS_ERR(entry)) {
  1445					ret = PTR_ERR(entry);
  1446					pr_err("fail on re-alloc test\n");
  1447					goto ppat_put;
  1448				}
  1449				entries[i] = entry;
  1450			}
  1451		}
  1452	
  1453	ppat_put:
  1454		if (entries) {
  1455			for (i = 0; i < size; i++) {
  1456				if (IS_ERR(entries[i]) || !entries[i])
  1457					continue;
  1458	
  1459				if (ret)
  1460					intel_ppat_put(entry);
  1461				else
  1462					ret = put_and_check_entry(entries[i]);
  1463			}
  1464			kfree(entries);
  1465			entries = NULL;
  1466		}
  1467		if (ret)
  1468			return ret;
  1469	
  1470		return igt_ppat_check(i915);
  1471	}
  1472	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 28375 bytes --]

[-- Attachment #3: Type: text/plain, Size: 160 bytes --]

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✓ Fi.CI.BAT: success for series starting with [v10,1/5] drm/i915: Factor out setup_private_pat()

[Patchwork]

== Series Details ==

Series: series starting with [v10,1/5] drm/i915: Factor out setup_private_pat()
URL   : https://patchwork.freedesktop.org/series/30073/
State : success

== Summary ==

Series 30073v1 series starting with [v10,1/5] drm/i915: Factor out setup_private_pat()
https://patchwork.freedesktop.org/api/1.0/series/30073/revisions/1/mbox/

Test gem_exec_store:
        Subgroup basic-default:
                incomplete -> PASS       (fi-byt-j1900)
Test kms_cursor_legacy:
        Subgroup basic-busy-flip-before-cursor-legacy:
                fail       -> PASS       (fi-snb-2600) fdo#100215 +1
Test kms_pipe_crc_basic:
        Subgroup suspend-read-crc-pipe-b:
                pass       -> DMESG-WARN (fi-byt-n2820) fdo#101705

fdo#100215 https://bugs.freedesktop.org/show_bug.cgi?id=100215
fdo#101705 https://bugs.freedesktop.org/show_bug.cgi?id=101705

fi-bdw-5557u     total:289  pass:268  dwarn:0   dfail:0   fail:0   skip:21  time:443s
fi-bdw-gvtdvm    total:289  pass:265  dwarn:0   dfail:0   fail:0   skip:24  time:450s
fi-blb-e6850     total:289  pass:224  dwarn:1   dfail:0   fail:0   skip:64  time:379s
fi-bsw-n3050     total:289  pass:243  dwarn:0   dfail:0   fail:0   skip:46  time:533s
fi-bwr-2160      total:289  pass:184  dwarn:0   dfail:0   fail:0   skip:105 time:270s
fi-bxt-j4205     total:289  pass:260  dwarn:0   dfail:0   fail:0   skip:29  time:509s
fi-byt-j1900     total:289  pass:254  dwarn:1   dfail:0   fail:0   skip:34  time:499s
fi-byt-n2820     total:289  pass:250  dwarn:1   dfail:0   fail:0   skip:38  time:497s
fi-cfl-s         total:289  pass:250  dwarn:4   dfail:0   fail:0   skip:35  time:455s
fi-elk-e7500     total:289  pass:230  dwarn:0   dfail:0   fail:0   skip:59  time:449s
fi-glk-2a        total:289  pass:260  dwarn:0   dfail:0   fail:0   skip:29  time:595s
fi-hsw-4770      total:289  pass:263  dwarn:0   dfail:0   fail:0   skip:26  time:428s
fi-hsw-4770r     total:289  pass:263  dwarn:0   dfail:0   fail:0   skip:26  time:414s
fi-ilk-650       total:289  pass:229  dwarn:0   dfail:0   fail:0   skip:60  time:437s
fi-ivb-3520m     total:289  pass:261  dwarn:0   dfail:0   fail:0   skip:28  time:480s
fi-ivb-3770      total:289  pass:261  dwarn:0   dfail:0   fail:0   skip:28  time:459s
fi-kbl-7500u     total:289  pass:264  dwarn:1   dfail:0   fail:0   skip:24  time:485s
fi-kbl-7560u     total:289  pass:270  dwarn:0   dfail:0   fail:0   skip:19  time:580s
fi-kbl-r         total:289  pass:262  dwarn:0   dfail:0   fail:0   skip:27  time:597s
fi-pnv-d510      total:289  pass:223  dwarn:1   dfail:0   fail:0   skip:65  time:551s
fi-skl-6260u     total:289  pass:269  dwarn:0   dfail:0   fail:0   skip:20  time:460s
fi-skl-6700k     total:289  pass:265  dwarn:0   dfail:0   fail:0   skip:24  time:523s
fi-skl-6770hq    total:289  pass:269  dwarn:0   dfail:0   fail:0   skip:20  time:498s
fi-skl-gvtdvm    total:289  pass:266  dwarn:0   dfail:0   fail:0   skip:23  time:467s
fi-skl-x1585l    total:289  pass:268  dwarn:0   dfail:0   fail:0   skip:21  time:469s
fi-snb-2520m     total:289  pass:251  dwarn:0   dfail:0   fail:0   skip:38  time:567s
fi-snb-2600      total:289  pass:249  dwarn:0   dfail:0   fail:1   skip:39  time:429s

b7ca0db5dc374e1f58839cfdc3803d4f4fb47eaf drm-tip: 2017y-09m-08d-20h-12m-37s UTC integration manifest
67ff3d633019 drm/i915: Introduce private PAT management
54ae6d84469e drm/i915: Factor out setup_private_pat()

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_5626/
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✓ Fi.CI.IGT: success for series starting with [v10,1/5] drm/i915: Factor out setup_private_pat()

[Patchwork]

== Series Details ==

Series: series starting with [v10,1/5] drm/i915: Factor out setup_private_pat()
URL   : https://patchwork.freedesktop.org/series/30073/
State : success

== Summary ==

Test gem_eio:
        Subgroup in-flight:
                fail       -> PASS       (shard-hsw) fdo#102616
Test gem_pwrite_pread:
        Subgroup display-copy-correctness:
                fail       -> PASS       (shard-hsw)
Test gem_exec_reloc:
        Subgroup basic-write-gtt-active:
                skip       -> PASS       (shard-hsw)
Test kms_flip:
        Subgroup blt-wf_vblank-vs-modeset:
                skip       -> PASS       (shard-hsw)
Test perf:
        Subgroup blocking:
                fail       -> PASS       (shard-hsw) fdo#102252
Test gem_flink_race:
        Subgroup flink_close:
                fail       -> PASS       (shard-hsw)

fdo#102616 https://bugs.freedesktop.org/show_bug.cgi?id=102616
fdo#102252 https://bugs.freedesktop.org/show_bug.cgi?id=102252

shard-hsw        total:2302 pass:1237 dwarn:0   dfail:0   fail:13  skip:1052 time:9416s

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_5626/shards.html
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH v10 1/5] drm/i915: Factor out setup_private_pat()

[Joonas Lahtinen]

On Sat, 2017-09-09 at 16:53 +0800, Zhi Wang wrote:
> Factor out setup_private_pat() for introducing the following patches.
> 
> Reviewed-by: Ben Widawsky <benjamin.widawsky@intel.com>
> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Ben Widawsky <benjamin.widawsky@intel.com>
> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
> Cc: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>

The right order would be (please corrected other patches too);

(Suggested-by, if any)
Signed-off-by
Cc
Cc
Reviewed-by
Reviewed-by (in chronological order)

Regards, Joonas
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH v10 1/5] drm/i915: Factor out setup_private_pat()

[Wang, Zhi A]

Thanks! :) I'm going to refresh them again. 

-----Original Message-----
From: Joonas Lahtinen [mailto:joonas.lahtinen@linux.intel.com] 
Sent: Monday, September 11, 2017 11:28 AM
To: Wang, Zhi A <zhi.a.wang@intel.com>; intel-gfx@lists.freedesktop.org; intel-gvt-dev@lists.freedesktop.org
Cc: chris@chris-wilson.co.uk; zhenyuw@linux.intel.com; Widawsky, Benjamin <benjamin.widawsky@intel.com>; Vivi, Rodrigo <rodrigo.vivi@intel.com>
Subject: Re: [PATCH v10 1/5] drm/i915: Factor out setup_private_pat()

On Sat, 2017-09-09 at 16:53 +0800, Zhi Wang wrote:
> Factor out setup_private_pat() for introducing the following patches.
> 
> Reviewed-by: Ben Widawsky <benjamin.widawsky@intel.com>
> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Ben Widawsky <benjamin.widawsky@intel.com>
> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
> Cc: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>

The right order would be (please corrected other patches too);

(Suggested-by, if any)
Signed-off-by
Cc
Cc
Reviewed-by
Reviewed-by (in chronological order)

Regards, Joonas
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx