Skip to contentSkip to conversation

Classify the engines in class + instance (v4)

PATCH In review Oscar Mateo 4 participants 15 messages Apr 7, 2017

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

cover Classify the engines in class + instance (v4)

This refactoring helps simplify a few things here and there.

Daniele Ceraolo Spurio (2):
drm/i915: Classify the engines in class + instance drm/i915: Use the engine class to get the context size

Oscar Mateo (3):
drm/i915: Use the same vfunc for BSD2 ring init drm/i915: Generate the engine name based on the instance number drm/i915: Split the engine info table in two levels, using class +
instance

 drivers/gpu/drm/i915/i915_reg.h              |  8 +++
 drivers/gpu/drm/i915/intel_engine_cs.c       | 80 ++++++++++++++++++++--------
 drivers/gpu/drm/i915/intel_lrc.c             | 33 +++++++-----
 drivers/gpu/drm/i915/intel_lrc.h             |  6 ++-
 drivers/gpu/drm/i915/intel_ringbuffer.c      | 14 -----
 drivers/gpu/drm/i915/intel_ringbuffer.h      |  9 +++-
 drivers/gpu/drm/i915/selftests/mock_engine.c |  3 +-
 7 files changed, 100 insertions(+), 53 deletions(-)
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#1 lore raw

PATCH 1/5 drm/i915: Classify the engines in class + instance

From: Daniele Ceraolo Spurio <[email protected]>

In such a way that vcs and vcs2 are just two different instances (0 and 1) of the same engine class (VIDEO_DECODE_CLASS).

v2: Align the instance types (Tvrtko)

v3: Don't use enums for bspec-defined stuff (Michal)

---
 drivers/gpu/drm/i915/i915_reg.h         |  8 ++++++++
 drivers/gpu/drm/i915/intel_engine_cs.c  | 14 ++++++++++++++
 drivers/gpu/drm/i915/intel_ringbuffer.h |  4 ++++
 3 files changed, 26 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +8 -0
index 11b12f4..4c72ada 100644
@@ -85,6 +85,14 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define VECS_HW		3
 #define VCS2_HW		4
 
+/* Engine class */
+
+#define RENDER_CLASS		0
+#define VIDEO_DECODE_CLASS	1
+#define VIDEO_ENHANCEMENT_CLASS	2
+#define COPY_ENGINE_CLASS	3
+#define OTHER_CLASS		4
+
 /* PCI config space */
 
 #define MCHBAR_I915 0x44
drivers/gpu/drm/i915/intel_engine_cs.c +14 -0
index 854e8e0..2d3456d 100644
@@ -30,6 +30,8 @@
 	const char *name;
 	unsigned int exec_id;
 	unsigned int hw_id;
+	u8 class;
+	u8 instance;
 	u32 mmio_base;
 	unsigned irq_shift;
 	int (*init_legacy)(struct intel_engine_cs *engine);
@@ -39,6 +41,8 @@
 		.name = "rcs",
 		.hw_id = RCS_HW,
 		.exec_id = I915_EXEC_RENDER,
+		.class = RENDER_CLASS,
+		.instance = 0,
 		.mmio_base = RENDER_RING_BASE,
 		.irq_shift = GEN8_RCS_IRQ_SHIFT,
 		.init_execlists = logical_render_ring_init,
@@ -48,6 +52,8 @@
 		.name = "bcs",
 		.hw_id = BCS_HW,
 		.exec_id = I915_EXEC_BLT,
+		.class = COPY_ENGINE_CLASS,
+		.instance = 0,
 		.mmio_base = BLT_RING_BASE,
 		.irq_shift = GEN8_BCS_IRQ_SHIFT,
 		.init_execlists = logical_xcs_ring_init,
@@ -57,6 +63,8 @@
 		.name = "vcs",
 		.hw_id = VCS_HW,
 		.exec_id = I915_EXEC_BSD,
+		.class = VIDEO_DECODE_CLASS,
+		.instance = 0,
 		.mmio_base = GEN6_BSD_RING_BASE,
 		.irq_shift = GEN8_VCS1_IRQ_SHIFT,
 		.init_execlists = logical_xcs_ring_init,
@@ -66,6 +74,8 @@
 		.name = "vcs2",
 		.hw_id = VCS2_HW,
 		.exec_id = I915_EXEC_BSD,
+		.class = VIDEO_DECODE_CLASS,
+		.instance = 1,
 		.mmio_base = GEN8_BSD2_RING_BASE,
 		.irq_shift = GEN8_VCS2_IRQ_SHIFT,
 		.init_execlists = logical_xcs_ring_init,
@@ -75,6 +85,8 @@
 		.name = "vecs",
 		.hw_id = VECS_HW,
 		.exec_id = I915_EXEC_VEBOX,
+		.class = VIDEO_ENHANCEMENT_CLASS,
+		.instance = 0,
 		.mmio_base = VEBOX_RING_BASE,
 		.irq_shift = GEN8_VECS_IRQ_SHIFT,
 		.init_execlists = logical_xcs_ring_init,
@@ -101,6 +113,8 @@
 	engine->hw_id = engine->guc_id = info->hw_id;
 	engine->mmio_base = info->mmio_base;
 	engine->irq_shift = info->irq_shift;
+	engine->class = info->class;
+	engine->instance = info->instance;
 
 	/* Nothing to do here, execute in order of dependencies */
 	engine->schedule = NULL;
drivers/gpu/drm/i915/intel_ringbuffer.h +4 -0
index cbe61d3..f54fe7d 100644
@@ -193,6 +193,10 @@ struct intel_engine_cs {
 	enum intel_engine_id id;
 	unsigned int exec_id;
 	unsigned int hw_id;
+
+	u8 class;
+	u8 instance;
+
 	unsigned int guc_id;
 	u32		mmio_base;
 	unsigned int irq_shift;
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v1 1/5
#2 lore raw

Re: [PATCH 1/5] drm/i915: Classify the engines in class + instance

Quoted 9 lines · Oscar Mateo

From: Daniele Ceraolo Spurio <[email protected]>

In such a way that vcs and vcs2 are just two different instances (0 and 1) of the same engine class (VIDEO_DECODE_CLASS).

v2: Align the instance types (Tvrtko)

v3: Don't use enums for bspec-defined stuff (Michal)

Quoted 8 lines

---

-Michal
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#3 lore raw

PATCH 2/5 drm/i915: Use the same vfunc for BSD2 ring init

In reply to Oscar Mateo

If we needed to do something different for the init functions, we could always look at the engine instance to make the distinction. But, in any case, the two functions are virtually identical already (please notice that BSD2_RING is only used from gen8 onwards).

With this, the init functions depends excusively on the engine class (a fact that we will use soon).

v2: Commit message

Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
---
 drivers/gpu/drm/i915/intel_engine_cs.c  |  2 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c | 14 --------------
 drivers/gpu/drm/i915/intel_ringbuffer.h |  1 -
 3 files changed, 1 insertion(+), 16 deletions(-)
drivers/gpu/drm/i915/intel_engine_cs.c +1 -1
index 2d3456d..146d6b6 100644
@@ -79,7 +79,7 @@
 		.mmio_base = GEN8_BSD2_RING_BASE,
 		.irq_shift = GEN8_VCS2_IRQ_SHIFT,
 		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_bsd2_ring_buffer,
+		.init_legacy = intel_init_bsd_ring_buffer,
 	},
 	[VECS] = {
 		.name = "vecs",
drivers/gpu/drm/i915/intel_ringbuffer.c +0 -14
index c98acc2..81eee42 100644
@@ -2175,20 +2175,6 @@ int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine)
 	return intel_init_ring_buffer(engine);
 }
 
-/**
- * Initialize the second BSD ring (eg. Broadwell GT3, Skylake GT3)
- */
-int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine)
-{
-	struct drm_i915_private *dev_priv = engine->i915;
-
-	intel_ring_default_vfuncs(dev_priv, engine);
-
-	engine->emit_flush = gen6_bsd_ring_flush;
-
-	return intel_init_ring_buffer(engine);
-}
-
 int intel_init_blt_ring_buffer(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
drivers/gpu/drm/i915/intel_ringbuffer.h +0 -1
index f54fe7d..8b53ddb 100644
@@ -555,7 +555,6 @@ int intel_ring_pin(struct intel_ring *ring,
 
 int intel_init_render_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine);
-int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_blt_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine);
 
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#4 lore raw

PATCH 3/5 drm/i915: Generate the engine name based on the instance number

In reply to Oscar Mateo

Not really needed, but makes the next change a little bit more compact.

v2:
- Use zero-based numbering for engine names: xcs0, xcs1.. xcsN (Tvrtko, Chris)
- Make sure the mock engine name is null-terminated (Tvrtko, Chris)

v3: Because I'm stupid (Chris)

v4: Verify engine name wasn't truncated (Michal)

Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
Cc Michal Wajdeczko <[email protected]>
---
 drivers/gpu/drm/i915/intel_engine_cs.c       | 5 +++--
 drivers/gpu/drm/i915/intel_ringbuffer.h      | 4 +++-
 drivers/gpu/drm/i915/selftests/mock_engine.c | 3 ++-
 3 files changed, 8 insertions(+), 4 deletions(-)
drivers/gpu/drm/i915/intel_engine_cs.c +3 -2
index 146d6b6..e40355c 100644
@@ -71,7 +71,7 @@
 		.init_legacy = intel_init_bsd_ring_buffer,
 	},
 	[VCS2] = {
-		.name = "vcs2",
+		.name = "vcs",
 		.hw_id = VCS2_HW,
 		.exec_id = I915_EXEC_BSD,
 		.class = VIDEO_DECODE_CLASS,
@@ -108,7 +108,8 @@
 
 	engine->id = id;
 	engine->i915 = dev_priv;
-	engine->name = info->name;
+	WARN_ON(snprintf(engine->name, sizeof(engine->name), "%s%u",
+			 info->name, info->instance) >= sizeof(engine->name));
 	engine->exec_id = info->exec_id;
 	engine->hw_id = engine->guc_id = info->hw_id;
 	engine->mmio_base = info->mmio_base;
drivers/gpu/drm/i915/intel_ringbuffer.h +3 -1
index 8b53ddb..fd8994b 100644
@@ -187,9 +187,11 @@ enum intel_engine_id {
 	VECS
 };
 
+#define INTEL_ENGINE_CS_MAX_NAME 8
+
 struct intel_engine_cs {
 	struct drm_i915_private *i915;
-	const char	*name;
+	char name[INTEL_ENGINE_CS_MAX_NAME];
 	enum intel_engine_id id;
 	unsigned int exec_id;
 	unsigned int hw_id;
drivers/gpu/drm/i915/selftests/mock_engine.c +2 -1
index b89050e..9adacfb 100644
@@ -140,7 +140,8 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 
 	/* minimal engine setup for requests */
 	engine->base.i915 = i915;
-	engine->base.name = name;
+	WARN_ON(snprintf(engine->base.name, sizeof(engine->base.name), "%s",
+			 name) > sizeof(engine->base.name));
 	engine->base.id = id++;
 	engine->base.status_page.page_addr = (void *)(engine + 1);
 
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v1 3/5
#5 lore raw

Re: [PATCH 3/5] drm/i915: Generate the engine name based on the instance number

Quoted 10 lines · Oscar Mateo

Not really needed, but makes the next change a little bit more compact.

v2:
- Use zero-based numbering for engine names: xcs0, xcs1.. xcsN (Tvrtko, Chris)
- Make sure the mock engine name is null-terminated (Tvrtko, Chris)

v3: Because I'm stupid (Chris)

v4: Verify engine name wasn't truncated (Michal)

Quoted 8 lines
Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
Cc Michal Wajdeczko <[email protected]>

---

-Michal
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#6 lore raw

Re: [PATCH 3/5] drm/i915: Generate the engine name based on the instance number

Quoted 20 lines · Michal Wajdeczko
Quoted 10 lines · Oscar Mateo

Not really needed, but makes the next change a little bit more compact.

v2:
- Use zero-based numbering for engine names: xcs0, xcs1.. xcsN (Tvrtko, Chris)
- Make sure the mock engine name is null-terminated (Tvrtko, Chris)

v3: Because I'm stupid (Chris)

v4: Verify engine name wasn't truncated (Michal)

Quoted 8 lines
Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
Cc Michal Wajdeczko <[email protected]>

---

Quoted 1 line

s/WARN_ON/GEM_WARN_ON/ ?

The WARN_ON in mock_engine.c is off-by-one, but really we don't care too much there so just kill that warn.
-Chris

Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#7 lore raw

Re: [PATCH 3/5] drm/i915: Generate the engine name based on the instance number

Quoted 20 lines · Chris Wilson
Quoted 19 lines · Michal Wajdeczko
Quoted 18 lines · Oscar Mateo

Not really needed, but makes the next change a little bit more compact.

v2:
- Use zero-based numbering for engine names: xcs0, xcs1.. xcsN (Tvrtko, Chris)
- Make sure the mock engine name is null-terminated (Tvrtko, Chris)

v3: Because I'm stupid (Chris)

v4: Verify engine name wasn't truncated (Michal)

Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
Cc Michal Wajdeczko <[email protected]>

---

Quoted 2 lines
Quoted 1 line

s/WARN_ON/GEM_WARN_ON/ ?

As in:

    if (GEM_WARN_ON(snprintf(engine->name, sizeof(engine->name), "%s%u",
            class_info->name, info->instance) >= sizeof(engine->name)))
        ;

or do you want to return an error? (I don't really see the need, but...)

Quoted 4 lines

The WARN_ON in mock_engine.c is off-by-one, but really we don't care too much there so just kill that warn.
-Chris

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#8 lore raw

PATCH 4/5 drm/i915: Split the engine info table in two levels, using class + instance

In reply to Oscar Mateo

There are some properties that logically belong to the engine class, and some that belong to the engine instance. Make it explicit.

v2: Commit message (Tvrtko)

v3:
- Rebased
- Exec/uabi id should be per instance (Chris)

v4:
- Rebased
- Avoid re-ordering fields for smaller diff (Tvrtko)
- Bug on oob access to the class array (Michal)

Cc Tvrtko Ursulin <[email protected]>
Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
Cc Michal Wajdeczko <[email protected]>
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 65 ++++++++++++++++++++++------------
 1 file changed, 42 insertions(+), 23 deletions(-)
drivers/gpu/drm/i915/intel_engine_cs.c +42 -23
index e40355c..4e2b5da 100644
@@ -26,71 +26,84 @@
 #include "intel_ringbuffer.h"
 #include "intel_lrc.h"
 
-static const struct engine_info {
+struct engine_class_info {
 	const char *name;
-	unsigned int exec_id;
+	int (*init_legacy)(struct intel_engine_cs *engine);
+	int (*init_execlists)(struct intel_engine_cs *engine);
+};
+
+static const struct engine_class_info intel_engine_classes[] = {
+	[RENDER_CLASS] = {
+		.name = "rcs",
+		.init_execlists = logical_render_ring_init,
+		.init_legacy = intel_init_render_ring_buffer,
+	},
+	[COPY_ENGINE_CLASS] = {
+		.name = "bcs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_blt_ring_buffer,
+	},
+	[VIDEO_DECODE_CLASS] = {
+		.name = "vcs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_bsd_ring_buffer,
+	},
+	[VIDEO_ENHANCEMENT_CLASS] = {
+		.name = "vecs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_vebox_ring_buffer,
+	},
+};
+
+struct engine_info {
 	unsigned int hw_id;
+	unsigned int exec_id;
 	u8 class;
 	u8 instance;
 	u32 mmio_base;
 	unsigned irq_shift;
-	int (*init_legacy)(struct intel_engine_cs *engine);
-	int (*init_execlists)(struct intel_engine_cs *engine);
-} intel_engines[] = {
+};
+
+static const struct engine_info intel_engines[] = {
 	[RCS] = {
-		.name = "rcs",
 		.hw_id = RCS_HW,
 		.exec_id = I915_EXEC_RENDER,
 		.class = RENDER_CLASS,
 		.instance = 0,
 		.mmio_base = RENDER_RING_BASE,
 		.irq_shift = GEN8_RCS_IRQ_SHIFT,
-		.init_execlists = logical_render_ring_init,
-		.init_legacy = intel_init_render_ring_buffer,
 	},
 	[BCS] = {
-		.name = "bcs",
 		.hw_id = BCS_HW,
 		.exec_id = I915_EXEC_BLT,
 		.class = COPY_ENGINE_CLASS,
 		.instance = 0,
 		.mmio_base = BLT_RING_BASE,
 		.irq_shift = GEN8_BCS_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_blt_ring_buffer,
 	},
 	[VCS] = {
-		.name = "vcs",
 		.hw_id = VCS_HW,
 		.exec_id = I915_EXEC_BSD,
 		.class = VIDEO_DECODE_CLASS,
 		.instance = 0,
 		.mmio_base = GEN6_BSD_RING_BASE,
 		.irq_shift = GEN8_VCS1_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_bsd_ring_buffer,
 	},
 	[VCS2] = {
-		.name = "vcs",
 		.hw_id = VCS2_HW,
 		.exec_id = I915_EXEC_BSD,
 		.class = VIDEO_DECODE_CLASS,
 		.instance = 1,
 		.mmio_base = GEN8_BSD2_RING_BASE,
 		.irq_shift = GEN8_VCS2_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_bsd_ring_buffer,
 	},
 	[VECS] = {
-		.name = "vecs",
 		.hw_id = VECS_HW,
 		.exec_id = I915_EXEC_VEBOX,
 		.class = VIDEO_ENHANCEMENT_CLASS,
 		.instance = 0,
 		.mmio_base = VEBOX_RING_BASE,
 		.irq_shift = GEN8_VECS_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_vebox_ring_buffer,
 	},
 };
 
@@ -99,8 +112,12 @@
 		   enum intel_engine_id id)
 {
 	const struct engine_info *info = &intel_engines[id];
+	const struct engine_class_info *class_info;
 	struct intel_engine_cs *engine;
 
+	GEM_BUG_ON(info->class > ARRAY_SIZE(intel_engine_classes));
+	class_info = &intel_engine_classes[info->class];
+
 	GEM_BUG_ON(dev_priv->engine[id]);
 	engine = kzalloc(sizeof(*engine), GFP_KERNEL);
 	if (!engine)
@@ -109,7 +126,7 @@
 	engine->id = id;
 	engine->i915 = dev_priv;
 	WARN_ON(snprintf(engine->name, sizeof(engine->name), "%s%u",
-			 info->name, info->instance) >= sizeof(engine->name));
+			 class_info->name, info->instance) >= sizeof(engine->name));
 	engine->exec_id = info->exec_id;
 	engine->hw_id = engine->guc_id = info->hw_id;
 	engine->mmio_base = info->mmio_base;
@@ -190,12 +207,14 @@ int intel_engines_init(struct drm_i915_private *dev_priv)
 	int err = 0;
 
 	for_each_engine(engine, dev_priv, id) {
+		const struct engine_class_info *class_info =
+					&intel_engine_classes[engine->class];
 		int (*init)(struct intel_engine_cs *engine);
 
 		if (i915.enable_execlists)
-			init = intel_engines[id].init_execlists;
+			init = class_info->init_execlists;
 		else
-			init = intel_engines[id].init_legacy;
+			init = class_info->init_legacy;
 		if (!init) {
 			kfree(engine);
 			dev_priv->engine[id] = NULL;
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#9 lore raw

Re: [PATCH 4/5] drm/i915: Split the engine info table in two levels, using class + instance

Quoted 22 lines · Oscar Mateo

There are some properties that logically belong to the engine class, and some that belong to the engine instance. Make it explicit.

v2: Commit message (Tvrtko)

v3:
- Rebased
- Exec/uabi id should be per instance (Chris)

v4:
- Rebased
- Avoid re-ordering fields for smaller diff (Tvrtko)
- Bug on oob access to the class array (Michal)

Cc Tvrtko Ursulin <[email protected]>
Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
Cc Michal Wajdeczko <[email protected]>

---

<snip>

drivers/gpu/drm/i915/intel_engine_cs.c line 120

Quoted 10 lines
drivers/gpu/drm/i915/intel_engine_cs.c +4 -0
@@ -99,8 +112,12 @@
 		   enum intel_engine_id id)
 {
 	const struct engine_info *info = &intel_engines[id];
+	const struct engine_class_info *class_info;
 	struct intel_engine_cs *engine;
 
+	GEM_BUG_ON(info->class > ARRAY_SIZE(intel_engine_classes));
+	class_info = &intel_engine_classes[info->class];
+

Hmm, I think we should check against info->class >= ARRAY_SIZE, as index 4 can't be used on array[4]

-Michal

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#10 lore raw

PATCH v5 1/1 drm/i915: Split the engine info table in two levels, using class + instance

There are some properties that logically belong to the engine class, and some that belong to the engine instance. Make it explicit.

v2: Commit message (Tvrtko)

v3:
- Rebased
- Exec/uabi id should be per instance (Chris)

v4:
- Rebased
- Avoid re-ordering fields for smaller diff (Tvrtko)
- Bug on oob access to the class array (Michal)

v5: Bug on the right thing (Michal)

Cc Tvrtko Ursulin <[email protected]>
Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
Cc Michal Wajdeczko <[email protected]>
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 65 ++++++++++++++++++++++------------
 1 file changed, 42 insertions(+), 23 deletions(-)
drivers/gpu/drm/i915/intel_engine_cs.c +42 -23
index e40355c..37bd008 100644
@@ -26,71 +26,84 @@
 #include "intel_ringbuffer.h"
 #include "intel_lrc.h"
 
-static const struct engine_info {
+struct engine_class_info {
 	const char *name;
-	unsigned int exec_id;
+	int (*init_legacy)(struct intel_engine_cs *engine);
+	int (*init_execlists)(struct intel_engine_cs *engine);
+};
+
+static const struct engine_class_info intel_engine_classes[] = {
+	[RENDER_CLASS] = {
+		.name = "rcs",
+		.init_execlists = logical_render_ring_init,
+		.init_legacy = intel_init_render_ring_buffer,
+	},
+	[COPY_ENGINE_CLASS] = {
+		.name = "bcs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_blt_ring_buffer,
+	},
+	[VIDEO_DECODE_CLASS] = {
+		.name = "vcs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_bsd_ring_buffer,
+	},
+	[VIDEO_ENHANCEMENT_CLASS] = {
+		.name = "vecs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_vebox_ring_buffer,
+	},
+};
+
+struct engine_info {
 	unsigned int hw_id;
+	unsigned int exec_id;
 	u8 class;
 	u8 instance;
 	u32 mmio_base;
 	unsigned irq_shift;
-	int (*init_legacy)(struct intel_engine_cs *engine);
-	int (*init_execlists)(struct intel_engine_cs *engine);
-} intel_engines[] = {
+};
+
+static const struct engine_info intel_engines[] = {
 	[RCS] = {
-		.name = "rcs",
 		.hw_id = RCS_HW,
 		.exec_id = I915_EXEC_RENDER,
 		.class = RENDER_CLASS,
 		.instance = 0,
 		.mmio_base = RENDER_RING_BASE,
 		.irq_shift = GEN8_RCS_IRQ_SHIFT,
-		.init_execlists = logical_render_ring_init,
-		.init_legacy = intel_init_render_ring_buffer,
 	},
 	[BCS] = {
-		.name = "bcs",
 		.hw_id = BCS_HW,
 		.exec_id = I915_EXEC_BLT,
 		.class = COPY_ENGINE_CLASS,
 		.instance = 0,
 		.mmio_base = BLT_RING_BASE,
 		.irq_shift = GEN8_BCS_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_blt_ring_buffer,
 	},
 	[VCS] = {
-		.name = "vcs",
 		.hw_id = VCS_HW,
 		.exec_id = I915_EXEC_BSD,
 		.class = VIDEO_DECODE_CLASS,
 		.instance = 0,
 		.mmio_base = GEN6_BSD_RING_BASE,
 		.irq_shift = GEN8_VCS1_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_bsd_ring_buffer,
 	},
 	[VCS2] = {
-		.name = "vcs",
 		.hw_id = VCS2_HW,
 		.exec_id = I915_EXEC_BSD,
 		.class = VIDEO_DECODE_CLASS,
 		.instance = 1,
 		.mmio_base = GEN8_BSD2_RING_BASE,
 		.irq_shift = GEN8_VCS2_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_bsd_ring_buffer,
 	},
 	[VECS] = {
-		.name = "vecs",
 		.hw_id = VECS_HW,
 		.exec_id = I915_EXEC_VEBOX,
 		.class = VIDEO_ENHANCEMENT_CLASS,
 		.instance = 0,
 		.mmio_base = VEBOX_RING_BASE,
 		.irq_shift = GEN8_VECS_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_vebox_ring_buffer,
 	},
 };
 
@@ -99,8 +112,12 @@
 		   enum intel_engine_id id)
 {
 	const struct engine_info *info = &intel_engines[id];
+	const struct engine_class_info *class_info;
 	struct intel_engine_cs *engine;
 
+	GEM_BUG_ON(info->class >= ARRAY_SIZE(intel_engine_classes));
+	class_info = &intel_engine_classes[info->class];
+
 	GEM_BUG_ON(dev_priv->engine[id]);
 	engine = kzalloc(sizeof(*engine), GFP_KERNEL);
 	if (!engine)
@@ -109,7 +126,7 @@
 	engine->id = id;
 	engine->i915 = dev_priv;
 	WARN_ON(snprintf(engine->name, sizeof(engine->name), "%s%u",
-			 info->name, info->instance) >= sizeof(engine->name));
+			 class_info->name, info->instance) >= sizeof(engine->name));
 	engine->exec_id = info->exec_id;
 	engine->hw_id = engine->guc_id = info->hw_id;
 	engine->mmio_base = info->mmio_base;
@@ -190,12 +207,14 @@ int intel_engines_init(struct drm_i915_private *dev_priv)
 	int err = 0;
 
 	for_each_engine(engine, dev_priv, id) {
+		const struct engine_class_info *class_info =
+					&intel_engine_classes[engine->class];
 		int (*init)(struct intel_engine_cs *engine);
 
 		if (i915.enable_execlists)
-			init = intel_engines[id].init_execlists;
+			init = class_info->init_execlists;
 		else
-			init = intel_engines[id].init_legacy;
+			init = class_info->init_legacy;
 		if (!init) {
 			kfree(engine);
 			dev_priv->engine[id] = NULL;
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v5 1/1
#11 lore raw

Re: [PATCH v5] drm/i915: Split the engine info table in two levels, using class + instance

Quoted 16 lines · Oscar Mateo

There are some properties that logically belong to the engine class, and some that belong to the engine instance. Make it explicit.

v2: Commit message (Tvrtko)

v3:
- Rebased
- Exec/uabi id should be per instance (Chris)

v4:
- Rebased
- Avoid re-ordering fields for smaller diff (Tvrtko)
- Bug on oob access to the class array (Michal)

v5: Bug on the right thing (Michal)

Quoted 8 lines
Cc Tvrtko Ursulin <[email protected]>
Cc Paulo Zanoni <[email protected]>
Cc Rodrigo Vivi <[email protected]>
Cc Chris Wilson <[email protected]>
Cc Daniele Ceraolo Spurio <[email protected]>
Cc Michal Wajdeczko <[email protected]>

---

-Michal
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#12 lore raw

PATCH 5/5 drm/i915: Use the engine class to get the context size

In reply to Oscar Mateo

From: Daniele Ceraolo Spurio <[email protected]>

Technically speaking, the context size is per engine class, not per instance.

v2: Add MISSING_CASE (Tvrtko)

v3: Rebased

---
 drivers/gpu/drm/i915/intel_lrc.c | 33 +++++++++++++++++++++------------
 drivers/gpu/drm/i915/intel_lrc.h |  6 +++++-
 2 files changed, 26 insertions(+), 13 deletions(-)
drivers/gpu/drm/i915/intel_lrc.c +21 -12
index 0dc1cc4..1c6672c 100644
@@ -1908,8 +1908,10 @@ static void execlists_init_reg_state(u32 *regs,
 }
 
 /**
- * intel_lr_context_size() - return the size of the context for an engine
- * @engine: which engine to find the context size for
+ * intel_lr_class_context_size() - return the size of the context for a given
+ * engine class
+ * @dev_priv: i915 device private
+ * @class: which engine class to find the context size for
  *
  * Each engine may require a different amount of space for a context image,
  * so when allocating (or copying) an image, this function can be used to
@@ -1921,25 +1923,32 @@ static void execlists_init_reg_state(u32 *regs,
  * in LRC mode, but does not include the "shared data page" used with
  * GuC submission. The caller should account for this if using the GuC.
  */
-uint32_t intel_lr_context_size(struct intel_engine_cs *engine)
+uint32_t intel_lr_class_context_size(struct drm_i915_private *dev_priv, u8 class)
 {
 	int ret = 0;
 
-	WARN_ON(INTEL_GEN(engine->i915) < 8);
+	WARN_ON(INTEL_GEN(dev_priv) < 8);
 
-	switch (engine->id) {
-	case RCS:
-		if (INTEL_GEN(engine->i915) >= 9)
+	switch (class) {
+	case RENDER_CLASS:
+		switch (INTEL_GEN(dev_priv)) {
+		default:
+			MISSING_CASE(INTEL_GEN(dev_priv));
+		case 9:
 			ret = GEN9_LR_CONTEXT_RENDER_SIZE;
-		else
+			break;
+		case 8:
 			ret = GEN8_LR_CONTEXT_RENDER_SIZE;
+			break;
+		}
 		break;
-	case VCS:
-	case BCS:
-	case VECS:
-	case VCS2:
+	case VIDEO_DECODE_CLASS:
+	case VIDEO_ENHANCEMENT_CLASS:
+	case COPY_ENGINE_CLASS:
 		ret = GEN8_LR_CONTEXT_OTHER_SIZE;
 		break;
+	default:
+		MISSING_CASE(class);
 	}
 
 	return ret;
drivers/gpu/drm/i915/intel_lrc.h +5 -1
index e8015e7..bde2b6e 100644
@@ -78,7 +78,11 @@ enum {
 struct drm_i915_private;
 struct i915_gem_context;
 
-uint32_t intel_lr_context_size(struct intel_engine_cs *engine);
+uint32_t intel_lr_class_context_size(struct drm_i915_private *dev_priv, u8 class);
+static inline uint32_t intel_lr_context_size(struct intel_engine_cs *engine)
+{
+	return intel_lr_class_context_size(engine->i915, engine->class);
+}
 
 void intel_lr_context_resume(struct drm_i915_private *dev_priv);
 uint64_t intel_lr_context_descriptor(struct i915_gem_context *ctx,
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#13 lore raw

✓ Fi.CI.BAT: success for Classify the engines in class + instance (rev5)

In reply to Oscar Mateo
== Series Details ==

== Series Details ==

Series: Classify the engines in class + instance (rev5)
URL : https://patchwork.freedesktop.org/series/22535/
State : success

== Summary ==

Series 22535v5 Classify the engines in class + instance
https://patchwork.freedesktop.org/api/1.0/series/22535/revisions/5/mbox/

fi-bdw-5557u     total:278  pass:267  dwarn:0   dfail:0   fail:0   skip:11  time:430s
fi-bdw-gvtdvm    total:278  pass:256  dwarn:8   dfail:0   fail:0   skip:14  time:432s
fi-bsw-n3050     total:278  pass:242  dwarn:0   dfail:0   fail:0   skip:36  time:578s
fi-bxt-j4205     total:278  pass:259  dwarn:0   dfail:0   fail:0   skip:19  time:503s
fi-bxt-t5700     total:278  pass:258  dwarn:0   dfail:0   fail:0   skip:20  time:546s
fi-byt-j1900     total:278  pass:254  dwarn:0   dfail:0   fail:0   skip:24  time:486s
fi-byt-n2820     total:278  pass:250  dwarn:0   dfail:0   fail:0   skip:28  time:483s
fi-hsw-4770      total:278  pass:262  dwarn:0   dfail:0   fail:0   skip:16  time:407s
fi-hsw-4770r     total:278  pass:262  dwarn:0   dfail:0   fail:0   skip:16  time:410s
fi-ilk-650       total:278  pass:228  dwarn:0   dfail:0   fail:0   skip:50  time:428s
fi-ivb-3520m     total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:484s
fi-ivb-3770      total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:487s
fi-kbl-7500u     total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:457s
fi-kbl-7560u     total:278  pass:267  dwarn:1   dfail:0   fail:0   skip:10  time:569s
fi-skl-6260u     total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10  time:448s
fi-skl-6700hq    total:278  pass:261  dwarn:0   dfail:0   fail:0   skip:17  time:569s
fi-skl-6700k     total:278  pass:256  dwarn:4   dfail:0   fail:0   skip:18  time:468s
fi-skl-6770hq    total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10  time:498s
fi-skl-gvtdvm    total:278  pass:265  dwarn:0   dfail:0   fail:0   skip:13  time:435s
fi-snb-2520m     total:278  pass:250  dwarn:0   dfail:0   fail:0   skip:28  time:526s
fi-snb-2600      total:278  pass:249  dwarn:0   dfail:0   fail:0   skip:29  time:415s

5a74def419415233546c4e853a3f02bb31daee82 drm-tip: 2017y-04m-07d-15h-00m-28s UTC integration manifest b4acaec drm/i915: Use the engine class to get the context size
c1e7ada drm/i915: Split the engine info table in two levels, using class + instance e62b796 drm/i915: Generate the engine name based on the instance number
a70545a2 drm/i915: Use the same vfunc for BSD2 ring init
190b146 drm/i915: Classify the engines in class + instance

== Logs ==

For more details see: https://intel-gfx-ci.01.org/CI/Patchwork_4442/
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#14 lore raw

✓ Fi.CI.BAT: success for Classify the engines in class + instance (rev6)

In reply to Oscar Mateo
== Series Details ==

== Series Details ==

Series: Classify the engines in class + instance (rev6)
URL : https://patchwork.freedesktop.org/series/22535/
State : success

== Summary ==

Series 22535v6 Classify the engines in class + instance
https://patchwork.freedesktop.org/api/1.0/series/22535/revisions/6/mbox/

Test gem_exec_suspend:

        Subgroup basic-s4-devices:
                dmesg-warn -> PASS       (fi-kbl-7560u) fdo#100125

fdo#100125 https://bugs.freedesktop.org/show_bug.cgi?id=100125

fi-bdw-5557u     total:278  pass:267  dwarn:0   dfail:0   fail:0   skip:11  time:430s
fi-bdw-gvtdvm    total:278  pass:256  dwarn:8   dfail:0   fail:0   skip:14  time:424s
fi-bsw-n3050     total:278  pass:242  dwarn:0   dfail:0   fail:0   skip:36  time:571s
fi-bxt-j4205     total:278  pass:259  dwarn:0   dfail:0   fail:0   skip:19  time:506s
fi-bxt-t5700     total:278  pass:258  dwarn:0   dfail:0   fail:0   skip:20  time:538s
fi-byt-j1900     total:278  pass:254  dwarn:0   dfail:0   fail:0   skip:24  time:485s
fi-byt-n2820     total:278  pass:250  dwarn:0   dfail:0   fail:0   skip:28  time:476s
fi-hsw-4770      total:278  pass:262  dwarn:0   dfail:0   fail:0   skip:16  time:407s
fi-hsw-4770r     total:278  pass:262  dwarn:0   dfail:0   fail:0   skip:16  time:401s
fi-ilk-650       total:278  pass:228  dwarn:0   dfail:0   fail:0   skip:50  time:426s
fi-ivb-3520m     total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:484s
fi-ivb-3770      total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:466s
fi-kbl-7500u     total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:456s
fi-kbl-7560u     total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10  time:567s
fi-skl-6260u     total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10  time:447s
fi-skl-6700hq    total:278  pass:261  dwarn:0   dfail:0   fail:0   skip:17  time:565s
fi-skl-6700k     total:278  pass:256  dwarn:4   dfail:0   fail:0   skip:18  time:457s
fi-skl-6770hq    total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10  time:492s
fi-skl-gvtdvm    total:278  pass:265  dwarn:0   dfail:0   fail:0   skip:13  time:431s
fi-snb-2520m     total:278  pass:250  dwarn:0   dfail:0   fail:0   skip:28  time:534s
fi-snb-2600      total:278  pass:249  dwarn:0   dfail:0   fail:0   skip:29  time:407s

5a74def419415233546c4e853a3f02bb31daee82 drm-tip: 2017y-04m-07d-15h-00m-28s UTC integration manifest 2765f03 drm/i915: Use the engine class to get the context size
9ed6b00 drm/i915: Split the engine info table in two levels, using class + instance 3ca1bf6 drm/i915: Generate the engine name based on the instance number
e9ca4d4 drm/i915: Use the same vfunc for BSD2 ring init
d29d43e drm/i915: Classify the engines in class + instance

== Logs ==

For more details see: https://intel-gfx-ci.01.org/CI/Patchwork_4443/
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx