From e9a851ed634628489ca4a392740694d0ded78cb9 Mon Sep 17 00:00:00 2001
From: "Xu, Anhua" <anhua.xu@intel.com>
Date: Mon, 13 Aug 2012 03:08:33 +0000
Subject: [PATCH 01/41] drm/i915: fix wrong order of parameters in port
 checking functions

Wrong order of parameters passed-in when calling hdmi/adpa
/lvds_pipe_enabled(), 2nd and 3rd parameters are reversed.

This bug was indroduced by

commit 1519b9956eb4b4180fa3f47c73341463cdcfaa37
Author: Keith Packard <keithp@keithp.com>
Date:   Sat Aug 6 10:35:34 2011 -0700

    drm/i915: Fix PCH port pipe select in CPT disable paths

The reachable tag for this commit is v3.1-rc1-3-g1519b99

Signed-off-by: Anhua Xu <anhua.xu@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_display.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index a9ab1aff2c77..46a4c921a174 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -1384,7 +1384,7 @@ static void assert_pch_hdmi_disabled(struct drm_i915_private *dev_priv,
 				     enum pipe pipe, int reg)
 {
 	u32 val = I915_READ(reg);
-	WARN(hdmi_pipe_enabled(dev_priv, val, pipe),
+	WARN(hdmi_pipe_enabled(dev_priv, pipe, val),
 	     "PCH HDMI (0x%08x) enabled on transcoder %c, should be disabled\n",
 	     reg, pipe_name(pipe));
 
@@ -1404,13 +1404,13 @@ static void assert_pch_ports_disabled(struct drm_i915_private *dev_priv,
 
 	reg = PCH_ADPA;
 	val = I915_READ(reg);
-	WARN(adpa_pipe_enabled(dev_priv, val, pipe),
+	WARN(adpa_pipe_enabled(dev_priv, pipe, val),
 	     "PCH VGA enabled on transcoder %c, should be disabled\n",
 	     pipe_name(pipe));
 
 	reg = PCH_LVDS;
 	val = I915_READ(reg);
-	WARN(lvds_pipe_enabled(dev_priv, val, pipe),
+	WARN(lvds_pipe_enabled(dev_priv, pipe, val),
 	     "PCH LVDS enabled on transcoder %c, should be disabled\n",
 	     pipe_name(pipe));
 
@@ -1874,7 +1874,7 @@ static void disable_pch_hdmi(struct drm_i915_private *dev_priv,
 			     enum pipe pipe, int reg)
 {
 	u32 val = I915_READ(reg);
-	if (hdmi_pipe_enabled(dev_priv, val, pipe)) {
+	if (hdmi_pipe_enabled(dev_priv, pipe, val)) {
 		DRM_DEBUG_KMS("Disabling pch HDMI %x on pipe %d\n",
 			      reg, pipe);
 		I915_WRITE(reg, val & ~PORT_ENABLE);
@@ -1896,12 +1896,12 @@ static void intel_disable_pch_ports(struct drm_i915_private *dev_priv,
 
 	reg = PCH_ADPA;
 	val = I915_READ(reg);
-	if (adpa_pipe_enabled(dev_priv, val, pipe))
+	if (adpa_pipe_enabled(dev_priv, pipe, val))
 		I915_WRITE(reg, val & ~ADPA_DAC_ENABLE);
 
 	reg = PCH_LVDS;
 	val = I915_READ(reg);
-	if (lvds_pipe_enabled(dev_priv, val, pipe)) {
+	if (lvds_pipe_enabled(dev_priv, pipe, val)) {
 		DRM_DEBUG_KMS("disable lvds on pipe %d val 0x%08x\n", pipe, val);
 		I915_WRITE(reg, val & ~LVDS_PORT_EN);
 		POSTING_READ(reg);

From dcdaed6eae0d7ed0c68aaa1ebfceb242625b3bf0 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 21:19:34 +0200
Subject: [PATCH 02/41] drm/i915: add missing gen2 pipe A quirk entries

For some odd reason we've missed i830 and a i855 variant. Also
kill the two now redundant i830 entries.

v2: Don't add the missing 855 id to the pipe A quirk list, we seem to
lack justification for it.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_display.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 46a4c921a174..dd760f43f577 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -7017,21 +7017,16 @@ static struct intel_quirk intel_quirks[] = {
 	/* HP Mini needs pipe A force quirk (LP: #322104) */
 	{ 0x27ae, 0x103c, 0x361a, quirk_pipea_force },
 
-	/* Thinkpad R31 needs pipe A force quirk */
-	{ 0x3577, 0x1014, 0x0505, quirk_pipea_force },
 	/* Toshiba Protege R-205, S-209 needs pipe A force quirk */
 	{ 0x2592, 0x1179, 0x0001, quirk_pipea_force },
 
-	/* ThinkPad X30 needs pipe A force quirk (LP: #304614) */
-	{ 0x3577,  0x1014, 0x0513, quirk_pipea_force },
-	/* ThinkPad X40 needs pipe A force quirk */
-
 	/* ThinkPad T60 needs pipe A force quirk (bug #16494) */
 	{ 0x2782, 0x17aa, 0x201a, quirk_pipea_force },
 
 	/* 855 & before need to leave pipe A & dpll A up */
 	{ 0x3582, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
 	{ 0x2562, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
+	{ 0x3577, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
 
 	/* Lenovo U160 cannot use SSC on LVDS */
 	{ 0x0046, 0x17aa, 0x3920, quirk_ssc_force_disable },

From a37b9b349ef26cb44f0e8d59c3efbcd9485018b4 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 19:27:09 +0200
Subject: [PATCH 03/41] drm/i915/ns2501: kill pll A enabling hack

With the pipe A quirk properly fixed up for i830M, this shouldn't be
required any longer.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/dvo_ns2501.c    | 7 -------
 drivers/gpu/drm/i915/intel_display.c | 2 +-
 2 files changed, 1 insertion(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/i915/dvo_ns2501.c b/drivers/gpu/drm/i915/dvo_ns2501.c
index 1a0bad9a5fab..6bd383dfbb09 100644
--- a/drivers/gpu/drm/i915/dvo_ns2501.c
+++ b/drivers/gpu/drm/i915/dvo_ns2501.c
@@ -74,11 +74,6 @@ struct ns2501_priv {
 
 #define NSPTR(d) ((NS2501Ptr)(d->DriverPrivate.ptr))
 
-/*
- * Include the PLL launcher prototype
- */
-extern void intel_enable_pll(struct drm_i915_private *dev_priv, enum pipe pipe);
-
 /*
  * For reasons unclear to me, the ns2501 at least on the Fujitsu/Siemens
  * laptops does not react on the i2c bus unless
@@ -113,8 +108,6 @@ static void enable_dvo(struct intel_dvo_device *dvo)
 	I915_WRITE(DVOC_SRCDIM, 0x400300);	// 1024x768
 	I915_WRITE(FW_BLC, 0x1080304);
 
-	intel_enable_pll(dev_priv, 0);
-
 	I915_WRITE(DVOC, 0x90004084);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index dd760f43f577..a55f6642be20 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -1432,7 +1432,7 @@ static void assert_pch_ports_disabled(struct drm_i915_private *dev_priv,
  *
  * Unfortunately needed by dvo_ns2501 since the dvo depends on it running.
  */
-void intel_enable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
+static void intel_enable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
 	int reg;
 	u32 val;

From 6306cb4f809ccf92c1b7bf446635c571e33f5151 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 19:27:10 +0200
Subject: [PATCH 04/41] drm/i915: rip out the overlay pipe A workaround

Now that all affected i830M systems have the pipe A quirk set,
we don't need to do any special dances in the overlay code any
longer. And reading through the code I'm rather dubios that it
actually does what it claims to do ...

As a nice benefit this rips out a users of the crtc helper dpms
callback.

v2: As suggested by Chris Wilson, replace the code by an appropriate
WARN to ensure that the pipe A is indeed running.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_overlay.c | 58 +---------------------------
 1 file changed, 1 insertion(+), 57 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index 830d0dd610e1..c0f48580405a 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -235,54 +235,6 @@ static int intel_overlay_do_wait_request(struct intel_overlay *overlay,
 	return 0;
 }
 
-/* Workaround for i830 bug where pipe a must be enable to change control regs */
-static int
-i830_activate_pipe_a(struct drm_device *dev)
-{
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_crtc *crtc;
-	struct drm_crtc_helper_funcs *crtc_funcs;
-	struct drm_display_mode vesa_640x480 = {
-		DRM_MODE("640x480", DRM_MODE_TYPE_DRIVER, 25175, 640, 656,
-			 752, 800, 0, 480, 489, 492, 525, 0,
-			 DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC)
-	}, *mode;
-
-	crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[0]);
-	if (crtc->dpms_mode == DRM_MODE_DPMS_ON)
-		return 0;
-
-	/* most i8xx have pipe a forced on, so don't trust dpms mode */
-	if (I915_READ(_PIPEACONF) & PIPECONF_ENABLE)
-		return 0;
-
-	crtc_funcs = crtc->base.helper_private;
-	if (crtc_funcs->dpms == NULL)
-		return 0;
-
-	DRM_DEBUG_DRIVER("Enabling pipe A in order to enable overlay\n");
-
-	mode = drm_mode_duplicate(dev, &vesa_640x480);
-
-	if (!drm_crtc_helper_set_mode(&crtc->base, mode,
-				       crtc->base.x, crtc->base.y,
-				       crtc->base.fb))
-		return 0;
-
-	crtc_funcs->dpms(&crtc->base, DRM_MODE_DPMS_ON);
-	return 1;
-}
-
-static void
-i830_deactivate_pipe_a(struct drm_device *dev)
-{
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[0];
-	struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private;
-
-	crtc_funcs->dpms(crtc, DRM_MODE_DPMS_OFF);
-}
-
 /* overlay needs to be disable in OCMD reg */
 static int intel_overlay_on(struct intel_overlay *overlay)
 {
@@ -290,17 +242,12 @@ static int intel_overlay_on(struct intel_overlay *overlay)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
 	struct drm_i915_gem_request *request;
-	int pipe_a_quirk = 0;
 	int ret;
 
 	BUG_ON(overlay->active);
 	overlay->active = 1;
 
-	if (IS_I830(dev)) {
-		pipe_a_quirk = i830_activate_pipe_a(dev);
-		if (pipe_a_quirk < 0)
-			return pipe_a_quirk;
-	}
+	WARN_ON(IS_I830(dev) && !(dev_priv->quirks & QUIRK_PIPEA_FORCE));
 
 	request = kzalloc(sizeof(*request), GFP_KERNEL);
 	if (request == NULL) {
@@ -322,9 +269,6 @@ static int intel_overlay_on(struct intel_overlay *overlay)
 
 	ret = intel_overlay_do_wait_request(overlay, request, NULL);
 out:
-	if (pipe_a_quirk)
-		i830_deactivate_pipe_a(dev);
-
 	return ret;
 }
 

From 24218aacac65f5b731bbff8eeb09ea7847dda2ef Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 19:27:11 +0200
Subject: [PATCH 05/41] drm/i915: prepare load-detect pipe code for dpms
 changes

A few things need adjustement:
- Change the dpms state by calling the dpms connector function and
  not some crtc helper internal callbacks. Otherwise this will break
  once we switch to our own dpms handling.
- Instead of tracking and restoring intel_crtc->dpms_mode use the
  connector's dpms variable - the former relies on the dpms compuation
  rules used by the crtc helper. And it would break when the encoder
  is cloned and the other output has a different dpms state. But luckily
  no one is crazy enough for that.
- Properly clear the connector -> encoder -> crtc linking, even when
  failing (note that the crtc helper removes the encoder -> crtc link
  in disabled_unused_functions for us).

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_display.c | 38 ++++++++++------------------
 1 file changed, 14 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index a55f6642be20..f851474fe066 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -5615,21 +5615,12 @@ bool intel_get_load_detect_pipe(struct intel_encoder *intel_encoder,
 	if (encoder->crtc) {
 		crtc = encoder->crtc;
 
-		intel_crtc = to_intel_crtc(crtc);
-		old->dpms_mode = intel_crtc->dpms_mode;
+		old->dpms_mode = connector->dpms;
 		old->load_detect_temp = false;
 
 		/* Make sure the crtc and connector are running */
-		if (intel_crtc->dpms_mode != DRM_MODE_DPMS_ON) {
-			struct drm_encoder_helper_funcs *encoder_funcs;
-			struct drm_crtc_helper_funcs *crtc_funcs;
-
-			crtc_funcs = crtc->helper_private;
-			crtc_funcs->dpms(crtc, DRM_MODE_DPMS_ON);
-
-			encoder_funcs = encoder->helper_private;
-			encoder_funcs->dpms(encoder, DRM_MODE_DPMS_ON);
-		}
+		if (connector->dpms != DRM_MODE_DPMS_ON)
+			connector->funcs->dpms(connector, DRM_MODE_DPMS_ON);
 
 		return true;
 	}
@@ -5657,7 +5648,7 @@ bool intel_get_load_detect_pipe(struct intel_encoder *intel_encoder,
 	connector->encoder = encoder;
 
 	intel_crtc = to_intel_crtc(crtc);
-	old->dpms_mode = intel_crtc->dpms_mode;
+	old->dpms_mode = connector->dpms;
 	old->load_detect_temp = true;
 	old->release_fb = NULL;
 
@@ -5682,22 +5673,25 @@ bool intel_get_load_detect_pipe(struct intel_encoder *intel_encoder,
 		DRM_DEBUG_KMS("reusing fbdev for load-detection framebuffer\n");
 	if (IS_ERR(crtc->fb)) {
 		DRM_DEBUG_KMS("failed to allocate framebuffer for load-detection\n");
-		crtc->fb = old_fb;
-		return false;
+		goto fail;
 	}
 
 	if (!drm_crtc_helper_set_mode(crtc, mode, 0, 0, old_fb)) {
 		DRM_DEBUG_KMS("failed to set mode on load-detect pipe\n");
 		if (old->release_fb)
 			old->release_fb->funcs->destroy(old->release_fb);
-		crtc->fb = old_fb;
-		return false;
+		goto fail;
 	}
 
 	/* let the connector get through one full cycle before testing */
 	intel_wait_for_vblank(dev, intel_crtc->pipe);
 
 	return true;
+fail:
+	connector->encoder = NULL;
+	encoder->crtc = NULL;
+	crtc->fb = old_fb;
+	return false;
 }
 
 void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder,
@@ -5706,9 +5700,6 @@ void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder,
 {
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_device *dev = encoder->dev;
-	struct drm_crtc *crtc = encoder->crtc;
-	struct drm_encoder_helper_funcs *encoder_funcs = encoder->helper_private;
-	struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s], [ENCODER:%d:%s]\n",
 		      connector->base.id, drm_get_connector_name(connector),
@@ -5716,6 +5707,7 @@ void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder,
 
 	if (old->load_detect_temp) {
 		connector->encoder = NULL;
+		encoder->crtc = NULL;
 		drm_helper_disable_unused_functions(dev);
 
 		if (old->release_fb)
@@ -5725,10 +5717,8 @@ void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder,
 	}
 
 	/* Switch crtc and encoder back off if necessary */
-	if (old->dpms_mode != DRM_MODE_DPMS_ON) {
-		encoder_funcs->dpms(encoder, old->dpms_mode);
-		crtc_funcs->dpms(crtc, old->dpms_mode);
-	}
+	if (old->dpms_mode != DRM_MODE_DPMS_ON)
+		connector->funcs->dpms(connector, old->dpms_mode);
 }
 
 /* Returns the clock of the currently programmed mode of the given pipe. */

From d2434ab7fb085ac6848acd9d82366f96a642e471 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 21:20:10 +0200
Subject: [PATCH 06/41] drm/i915: drop intel_encoder argument to
 load_detect_pipe functions

Since it's redundant - we can get the attached encoder in the
functions themselves.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_crt.c     |  6 ++----
 drivers/gpu/drm/i915/intel_display.c | 10 ++++++----
 drivers/gpu/drm/i915/intel_drv.h     |  6 ++----
 drivers/gpu/drm/i915/intel_tv.c      |  7 ++-----
 4 files changed, 12 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index bc5e2c97db61..62b6555cfa27 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -517,14 +517,12 @@ intel_crt_detect(struct drm_connector *connector, bool force)
 		return connector->status;
 
 	/* for pre-945g platforms use load detect */
-	if (intel_get_load_detect_pipe(&crt->base, connector, NULL,
-				       &tmp)) {
+	if (intel_get_load_detect_pipe(connector, NULL, &tmp)) {
 		if (intel_crt_detect_ddc(connector))
 			status = connector_status_connected;
 		else
 			status = intel_crt_load_detect(crt);
-		intel_release_load_detect_pipe(&crt->base, connector,
-					       &tmp);
+		intel_release_load_detect_pipe(connector, &tmp);
 	} else
 		status = connector_status_unknown;
 
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index f851474fe066..62524b036bd1 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -5584,12 +5584,13 @@ mode_fits_in_fbdev(struct drm_device *dev,
 	return fb;
 }
 
-bool intel_get_load_detect_pipe(struct intel_encoder *intel_encoder,
-				struct drm_connector *connector,
+bool intel_get_load_detect_pipe(struct drm_connector *connector,
 				struct drm_display_mode *mode,
 				struct intel_load_detect_pipe *old)
 {
 	struct intel_crtc *intel_crtc;
+	struct intel_encoder *intel_encoder =
+		intel_attached_encoder(connector);
 	struct drm_crtc *possible_crtc;
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_crtc *crtc = NULL;
@@ -5694,10 +5695,11 @@ fail:
 	return false;
 }
 
-void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder,
-				    struct drm_connector *connector,
+void intel_release_load_detect_pipe(struct drm_connector *connector,
 				    struct intel_load_detect_pipe *old)
 {
+	struct intel_encoder *intel_encoder =
+		intel_attached_encoder(connector);
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_device *dev = encoder->dev;
 
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 6920bc26a2e9..95f635b70b5b 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -433,12 +433,10 @@ struct intel_load_detect_pipe {
 	bool load_detect_temp;
 	int dpms_mode;
 };
-extern bool intel_get_load_detect_pipe(struct intel_encoder *intel_encoder,
-				       struct drm_connector *connector,
+extern bool intel_get_load_detect_pipe(struct drm_connector *connector,
 				       struct drm_display_mode *mode,
 				       struct intel_load_detect_pipe *old);
-extern void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder,
-					   struct drm_connector *connector,
+extern void intel_release_load_detect_pipe(struct drm_connector *connector,
 					   struct intel_load_detect_pipe *old);
 
 extern void intelfb_restore(void);
diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c
index 1a0bab07699e..95653a508987 100644
--- a/drivers/gpu/drm/i915/intel_tv.c
+++ b/drivers/gpu/drm/i915/intel_tv.c
@@ -1303,12 +1303,9 @@ intel_tv_detect(struct drm_connector *connector, bool force)
 	if (force) {
 		struct intel_load_detect_pipe tmp;
 
-		if (intel_get_load_detect_pipe(&intel_tv->base, connector,
-					       &mode, &tmp)) {
+		if (intel_get_load_detect_pipe(connector, &mode, &tmp)) {
 			type = intel_tv_detect_type(intel_tv, connector);
-			intel_release_load_detect_pipe(&intel_tv->base,
-						       connector,
-						       &tmp);
+			intel_release_load_detect_pipe(connector, &tmp);
 		} else
 			return connector_status_unknown;
 	} else

From fac3274c4ee090140410b2f29d3fbd8f10eae186 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 19:27:12 +0200
Subject: [PATCH 07/41] drm/i915: simplify dvo dpms interface

All dvo drivers only support 2 dpms states, and our dvo driver
even switches of the dvo port for anything else than DPMS_ON. Hence
ditch this complexity and simply use bool enable.

While reading through this code I've noticed that the mode_set
function of ch7017 is a bit peculiar - it disable the lvds again, even
though the crtc helper code should have done that ... This might be to
work around an issue at driver load, we pretty much ignore the hw
state when taking over.

v2: Also do the conversion for the new ns2501 driver.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/dvo.h        |  9 ++++-----
 drivers/gpu/drm/i915/dvo_ch7017.c |  8 ++++----
 drivers/gpu/drm/i915/dvo_ch7xxx.c |  4 ++--
 drivers/gpu/drm/i915/dvo_ivch.c   |  8 ++++----
 drivers/gpu/drm/i915/dvo_ns2501.c | 14 ++++++--------
 drivers/gpu/drm/i915/dvo_sil164.c |  4 ++--
 drivers/gpu/drm/i915/dvo_tfp410.c |  4 ++--
 drivers/gpu/drm/i915/intel_dvo.c  |  4 ++--
 8 files changed, 26 insertions(+), 29 deletions(-)

diff --git a/drivers/gpu/drm/i915/dvo.h b/drivers/gpu/drm/i915/dvo.h
index 0c8ac4d92deb..0fa839e439b3 100644
--- a/drivers/gpu/drm/i915/dvo.h
+++ b/drivers/gpu/drm/i915/dvo.h
@@ -58,13 +58,12 @@ struct intel_dvo_dev_ops {
 	void (*create_resources)(struct intel_dvo_device *dvo);
 
 	/*
-	 * Turn on/off output or set intermediate power levels if available.
+	 * Turn on/off output.
 	 *
-	 * Unsupported intermediate modes drop to the lower power setting.
-	 * If the  mode is DPMSModeOff, the output must be disabled,
-	 * as the DPLL may be disabled afterwards.
+	 * Because none of our dvo drivers support an intermediate power levels,
+	 * we don't expose this in the interfac.
 	 */
-	void (*dpms)(struct intel_dvo_device *dvo, int mode);
+	void (*dpms)(struct intel_dvo_device *dvo, bool enable);
 
 	/*
 	 * Callback for testing a video mode for a given output.
diff --git a/drivers/gpu/drm/i915/dvo_ch7017.c b/drivers/gpu/drm/i915/dvo_ch7017.c
index 1ca799a1e1fc..71e7650a2994 100644
--- a/drivers/gpu/drm/i915/dvo_ch7017.c
+++ b/drivers/gpu/drm/i915/dvo_ch7017.c
@@ -163,7 +163,7 @@ struct ch7017_priv {
 };
 
 static void ch7017_dump_regs(struct intel_dvo_device *dvo);
-static void ch7017_dpms(struct intel_dvo_device *dvo, int mode);
+static void ch7017_dpms(struct intel_dvo_device *dvo, bool enable);
 
 static bool ch7017_read(struct intel_dvo_device *dvo, u8 addr, u8 *val)
 {
@@ -309,7 +309,7 @@ static void ch7017_mode_set(struct intel_dvo_device *dvo,
 	lvds_power_down = CH7017_LVDS_POWER_DOWN_DEFAULT_RESERVED |
 			  (mode->hdisplay & 0x0700) >> 8;
 
-	ch7017_dpms(dvo, DRM_MODE_DPMS_OFF);
+	ch7017_dpms(dvo, false);
 	ch7017_write(dvo, CH7017_HORIZONTAL_ACTIVE_PIXEL_INPUT,
 			horizontal_active_pixel_input);
 	ch7017_write(dvo, CH7017_HORIZONTAL_ACTIVE_PIXEL_OUTPUT,
@@ -331,7 +331,7 @@ static void ch7017_mode_set(struct intel_dvo_device *dvo,
 }
 
 /* set the CH7017 power state */
-static void ch7017_dpms(struct intel_dvo_device *dvo, int mode)
+static void ch7017_dpms(struct intel_dvo_device *dvo, bool enable)
 {
 	uint8_t val;
 
@@ -345,7 +345,7 @@ static void ch7017_dpms(struct intel_dvo_device *dvo, int mode)
 			CH7017_DAC3_POWER_DOWN |
 			CH7017_TV_POWER_DOWN_EN);
 
-	if (mode == DRM_MODE_DPMS_ON) {
+	if (enable) {
 		/* Turn on the LVDS */
 		ch7017_write(dvo, CH7017_LVDS_POWER_DOWN,
 			     val & ~CH7017_LVDS_POWER_DOWN_EN);
diff --git a/drivers/gpu/drm/i915/dvo_ch7xxx.c b/drivers/gpu/drm/i915/dvo_ch7xxx.c
index 4a036600e806..c1dea5b11f91 100644
--- a/drivers/gpu/drm/i915/dvo_ch7xxx.c
+++ b/drivers/gpu/drm/i915/dvo_ch7xxx.c
@@ -289,9 +289,9 @@ static void ch7xxx_mode_set(struct intel_dvo_device *dvo,
 }
 
 /* set the CH7xxx power state */
-static void ch7xxx_dpms(struct intel_dvo_device *dvo, int mode)
+static void ch7xxx_dpms(struct intel_dvo_device *dvo, bool enable)
 {
-	if (mode == DRM_MODE_DPMS_ON)
+	if (enable)
 		ch7xxx_writeb(dvo, CH7xxx_PM, CH7xxx_PM_DVIL | CH7xxx_PM_DVIP);
 	else
 		ch7xxx_writeb(dvo, CH7xxx_PM, CH7xxx_PM_FPD);
diff --git a/drivers/gpu/drm/i915/dvo_ivch.c b/drivers/gpu/drm/i915/dvo_ivch.c
index 04f2893d5e3c..fa8ff6b050fa 100644
--- a/drivers/gpu/drm/i915/dvo_ivch.c
+++ b/drivers/gpu/drm/i915/dvo_ivch.c
@@ -288,7 +288,7 @@ static enum drm_mode_status ivch_mode_valid(struct intel_dvo_device *dvo,
 }
 
 /** Sets the power state of the panel connected to the ivch */
-static void ivch_dpms(struct intel_dvo_device *dvo, int mode)
+static void ivch_dpms(struct intel_dvo_device *dvo, bool enable)
 {
 	int i;
 	uint16_t vr01, vr30, backlight;
@@ -297,13 +297,13 @@ static void ivch_dpms(struct intel_dvo_device *dvo, int mode)
 	if (!ivch_read(dvo, VR01, &vr01))
 		return;
 
-	if (mode == DRM_MODE_DPMS_ON)
+	if (enable)
 		backlight = 1;
 	else
 		backlight = 0;
 	ivch_write(dvo, VR80, backlight);
 
-	if (mode == DRM_MODE_DPMS_ON)
+	if (enable)
 		vr01 |= VR01_LCD_ENABLE | VR01_DVO_ENABLE;
 	else
 		vr01 &= ~(VR01_LCD_ENABLE | VR01_DVO_ENABLE);
@@ -315,7 +315,7 @@ static void ivch_dpms(struct intel_dvo_device *dvo, int mode)
 		if (!ivch_read(dvo, VR30, &vr30))
 			break;
 
-		if (((vr30 & VR30_PANEL_ON) != 0) == (mode == DRM_MODE_DPMS_ON))
+		if (((vr30 & VR30_PANEL_ON) != 0) == enable)
 			break;
 		udelay(1000);
 	}
diff --git a/drivers/gpu/drm/i915/dvo_ns2501.c b/drivers/gpu/drm/i915/dvo_ns2501.c
index 6bd383dfbb09..c4d9f2f395e6 100644
--- a/drivers/gpu/drm/i915/dvo_ns2501.c
+++ b/drivers/gpu/drm/i915/dvo_ns2501.c
@@ -493,19 +493,19 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 }
 
 /* set the NS2501 power state */
-static void ns2501_dpms(struct intel_dvo_device *dvo, int mode)
+static void ns2501_dpms(struct intel_dvo_device *dvo, bool enable)
 {
 	bool ok;
 	bool restore = false;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
 	unsigned char ch;
 
-	DRM_DEBUG_KMS("%s: Trying set the dpms of the DVO to %d\n",
-		      __FUNCTION__, mode);
+	DRM_DEBUG_KMS("%s: Trying set the dpms of the DVO to %i\n",
+		      __FUNCTION__, enable);
 
 	ch = ns->reg_8_shadow;
 
-	if (mode == DRM_MODE_DPMS_ON)
+	if (enable)
 		ch |= NS2501_8_PD;
 	else
 		ch &= ~NS2501_8_PD;
@@ -519,12 +519,10 @@ static void ns2501_dpms(struct intel_dvo_device *dvo, int mode)
 			ok &= ns2501_writeb(dvo, NS2501_REG8, ch);
 			ok &=
 			    ns2501_writeb(dvo, 0x34,
-					  (mode ==
-					   DRM_MODE_DPMS_ON) ? (0x03) : (0x00));
+					  enable ? 0x03 : 0x00);
 			ok &=
 			    ns2501_writeb(dvo, 0x35,
-					  (mode ==
-					   DRM_MODE_DPMS_ON) ? (0xff) : (0x00));
+					  enable ? 0xff : 0x00);
 			if (!ok) {
 				if (restore)
 					restore_dvo(dvo);
diff --git a/drivers/gpu/drm/i915/dvo_sil164.c b/drivers/gpu/drm/i915/dvo_sil164.c
index a0b13a6f619d..cc24c1cabecd 100644
--- a/drivers/gpu/drm/i915/dvo_sil164.c
+++ b/drivers/gpu/drm/i915/dvo_sil164.c
@@ -208,7 +208,7 @@ static void sil164_mode_set(struct intel_dvo_device *dvo,
 }
 
 /* set the SIL164 power state */
-static void sil164_dpms(struct intel_dvo_device *dvo, int mode)
+static void sil164_dpms(struct intel_dvo_device *dvo, bool enable)
 {
 	int ret;
 	unsigned char ch;
@@ -217,7 +217,7 @@ static void sil164_dpms(struct intel_dvo_device *dvo, int mode)
 	if (ret == false)
 		return;
 
-	if (mode == DRM_MODE_DPMS_ON)
+	if (enable)
 		ch |= SIL164_8_PD;
 	else
 		ch &= ~SIL164_8_PD;
diff --git a/drivers/gpu/drm/i915/dvo_tfp410.c b/drivers/gpu/drm/i915/dvo_tfp410.c
index aa2cd3ec54aa..097b3e82b00f 100644
--- a/drivers/gpu/drm/i915/dvo_tfp410.c
+++ b/drivers/gpu/drm/i915/dvo_tfp410.c
@@ -234,14 +234,14 @@ static void tfp410_mode_set(struct intel_dvo_device *dvo,
 }
 
 /* set the tfp410 power state */
-static void tfp410_dpms(struct intel_dvo_device *dvo, int mode)
+static void tfp410_dpms(struct intel_dvo_device *dvo, bool enable)
 {
 	uint8_t ctl1;
 
 	if (!tfp410_readb(dvo, TFP410_CTL_1, &ctl1))
 		return;
 
-	if (mode == DRM_MODE_DPMS_ON)
+	if (enable)
 		ctl1 |= TFP410_CTL_1_PD;
 	else
 		ctl1 &= ~TFP410_CTL_1_PD;
diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c
index 03dfdff8e003..227551f12d25 100644
--- a/drivers/gpu/drm/i915/intel_dvo.c
+++ b/drivers/gpu/drm/i915/intel_dvo.c
@@ -115,9 +115,9 @@ static void intel_dvo_dpms(struct drm_encoder *encoder, int mode)
 	if (mode == DRM_MODE_DPMS_ON) {
 		I915_WRITE(dvo_reg, temp | DVO_ENABLE);
 		I915_READ(dvo_reg);
-		intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, mode);
+		intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, true);
 	} else {
-		intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, mode);
+		intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, false);
 		I915_WRITE(dvo_reg, temp & ~DVO_ENABLE);
 		I915_READ(dvo_reg);
 	}

From 5d985ac81ab15631570cad83d187565c7fcfd00a Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 19:27:13 +0200
Subject: [PATCH 08/41] drm/i915: kill a few unused things in dev_priv

... and move a few others only used by i915_dma.c into the dri1
dungeon.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_dma.c | 22 +++++++++++-----------
 drivers/gpu/drm/i915/i915_drv.h | 13 ++++++-------
 2 files changed, 17 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 0a1b64f8d442..2285ae31717a 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -235,10 +235,10 @@ static int i915_initialize(struct drm_device * dev, drm_i915_init_t * init)
 		}
 	}
 
-	dev_priv->cpp = init->cpp;
-	dev_priv->back_offset = init->back_offset;
-	dev_priv->front_offset = init->front_offset;
-	dev_priv->current_page = 0;
+	dev_priv->dri1.cpp = init->cpp;
+	dev_priv->dri1.back_offset = init->back_offset;
+	dev_priv->dri1.front_offset = init->front_offset;
+	dev_priv->dri1.current_page = 0;
 	if (master_priv->sarea_priv)
 		master_priv->sarea_priv->pf_current_page = 0;
 
@@ -575,7 +575,7 @@ static int i915_dispatch_flip(struct drm_device * dev)
 
 	DRM_DEBUG_DRIVER("%s: page=%d pfCurrentPage=%d\n",
 			  __func__,
-			 dev_priv->current_page,
+			 dev_priv->dri1.current_page,
 			 master_priv->sarea_priv->pf_current_page);
 
 	i915_kernel_lost_context(dev);
@@ -589,12 +589,12 @@ static int i915_dispatch_flip(struct drm_device * dev)
 
 	OUT_RING(CMD_OP_DISPLAYBUFFER_INFO | ASYNC_FLIP);
 	OUT_RING(0);
-	if (dev_priv->current_page == 0) {
-		OUT_RING(dev_priv->back_offset);
-		dev_priv->current_page = 1;
+	if (dev_priv->dri1.current_page == 0) {
+		OUT_RING(dev_priv->dri1.back_offset);
+		dev_priv->dri1.current_page = 1;
 	} else {
-		OUT_RING(dev_priv->front_offset);
-		dev_priv->current_page = 0;
+		OUT_RING(dev_priv->dri1.front_offset);
+		dev_priv->dri1.current_page = 0;
 	}
 	OUT_RING(0);
 
@@ -613,7 +613,7 @@ static int i915_dispatch_flip(struct drm_device * dev)
 		ADVANCE_LP_RING();
 	}
 
-	master_priv->sarea_priv->pf_current_page = dev_priv->current_page;
+	master_priv->sarea_priv->pf_current_page = dev_priv->dri1.current_page;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 261fe2175afb..ed3ba70923ac 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -428,12 +428,6 @@ typedef struct drm_i915_private {
 
 	struct resource mch_res;
 
-	unsigned int cpp;
-	int back_offset;
-	int front_offset;
-	int current_page;
-	int page_flipping;
-
 	atomic_t irq_received;
 
 	/* protects the irq masks */
@@ -451,7 +445,6 @@ typedef struct drm_i915_private {
 	u32 hotplug_supported_mask;
 	struct work_struct hotplug_work;
 
-	unsigned int sr01, adpa, ppcr, dvob, dvoc, lvds;
 	int num_pipe;
 	int num_pch_pll;
 
@@ -790,6 +783,12 @@ typedef struct drm_i915_private {
 	struct {
 		unsigned allow_batchbuffer : 1;
 		u32 __iomem *gfx_hws_cpu_addr;
+
+		unsigned int cpp;
+		int back_offset;
+		int front_offset;
+		int current_page;
+		int page_flipping;
 	} dri1;
 
 	/* Kernel Modesetting */

From 88cefb6c60712982f739f6bbd9f82968aca46033 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 19:27:14 +0200
Subject: [PATCH 09/41] drm/i915: extract ironlake_fdi_pll_disable

Simply to make the ilk+ crtc disable path clearer and more symmetric
with the enable function.

Also switch to intel_crtc for the enable function.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_display.c | 57 ++++++++++++++++------------
 1 file changed, 33 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 62524b036bd1..b580d949d7d2 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -2709,11 +2709,10 @@ static void ivb_manual_fdi_link_train(struct drm_crtc *crtc)
 	DRM_DEBUG_KMS("FDI train done.\n");
 }
 
-static void ironlake_fdi_pll_enable(struct drm_crtc *crtc)
+static void ironlake_fdi_pll_enable(struct intel_crtc *intel_crtc)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = intel_crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_crtc->pipe;
 	u32 reg, temp;
 
@@ -2754,6 +2753,35 @@ static void ironlake_fdi_pll_enable(struct drm_crtc *crtc)
 	}
 }
 
+static void ironlake_fdi_pll_disable(struct intel_crtc *intel_crtc)
+{
+	struct drm_device *dev = intel_crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe = intel_crtc->pipe;
+	u32 reg, temp;
+
+	/* Switch from PCDclk to Rawclk */
+	reg = FDI_RX_CTL(pipe);
+	temp = I915_READ(reg);
+	I915_WRITE(reg, temp & ~FDI_PCDCLK);
+
+	/* Disable CPU FDI TX PLL */
+	reg = FDI_TX_CTL(pipe);
+	temp = I915_READ(reg);
+	I915_WRITE(reg, temp & ~FDI_TX_PLL_ENABLE);
+
+	POSTING_READ(reg);
+	udelay(100);
+
+	reg = FDI_RX_CTL(pipe);
+	temp = I915_READ(reg);
+	I915_WRITE(reg, temp & ~FDI_RX_PLL_ENABLE);
+
+	/* Wait for the clocks to turn off. */
+	POSTING_READ(reg);
+	udelay(100);
+}
+
 static void cpt_phase_pointer_disable(struct drm_device *dev, int pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -3201,7 +3229,7 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
 	is_pch_port = intel_crtc_driving_pch(crtc);
 
 	if (is_pch_port)
-		ironlake_fdi_pll_enable(crtc);
+		ironlake_fdi_pll_enable(intel_crtc);
 	else
 		ironlake_fdi_disable(crtc);
 
@@ -3304,26 +3332,7 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc)
 	/* disable PCH DPLL */
 	intel_disable_pch_pll(intel_crtc);
 
-	/* Switch from PCDclk to Rawclk */
-	reg = FDI_RX_CTL(pipe);
-	temp = I915_READ(reg);
-	I915_WRITE(reg, temp & ~FDI_PCDCLK);
-
-	/* Disable CPU FDI TX PLL */
-	reg = FDI_TX_CTL(pipe);
-	temp = I915_READ(reg);
-	I915_WRITE(reg, temp & ~FDI_TX_PLL_ENABLE);
-
-	POSTING_READ(reg);
-	udelay(100);
-
-	reg = FDI_RX_CTL(pipe);
-	temp = I915_READ(reg);
-	I915_WRITE(reg, temp & ~FDI_RX_PLL_ENABLE);
-
-	/* Wait for the clocks to turn off. */
-	POSTING_READ(reg);
-	udelay(100);
+	ironlake_fdi_pll_disable(intel_crtc);
 
 	intel_crtc->active = false;
 	intel_update_watermarks(dev);

From 9a78b6cce5ebb04ae1b2f99b6c9f5fa54a3428b0 Mon Sep 17 00:00:00 2001
From: Wang Xingchao <xingchao.wang@intel.com>
Date: Thu, 9 Aug 2012 16:52:15 +0800
Subject: [PATCH 10/41] drm/i915: HSW audio registers definition

Add hsw audio registers definition

Signed-off-by: Wang Xingchao <xingchao.wang@intel.com>
Reviewed-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_reg.h | 47 +++++++++++++++++++++++++++++++++
 1 file changed, 47 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 321ae7261320..cbddc7d8e169 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -4277,6 +4277,53 @@
 #define   AUD_CONFIG_PIXEL_CLOCK_HDMI		(0xf << 16)
 #define   AUD_CONFIG_DISABLE_NCTS		(1 << 3)
 
+/* HSW Audio */
+#define   HSW_AUD_CONFIG_A		0x65000 /* Audio Configuration Transcoder A */
+#define   HSW_AUD_CONFIG_B		0x65100 /* Audio Configuration Transcoder B */
+#define   HSW_AUD_CFG(pipe) _PIPE(pipe, \
+					HSW_AUD_CONFIG_A, \
+					HSW_AUD_CONFIG_B)
+
+#define   HSW_AUD_MISC_CTRL_A		0x65010 /* Audio Misc Control Convert 1 */
+#define   HSW_AUD_MISC_CTRL_B		0x65110 /* Audio Misc Control Convert 2 */
+#define   HSW_AUD_MISC_CTRL(pipe) _PIPE(pipe, \
+					HSW_AUD_MISC_CTRL_A, \
+					HSW_AUD_MISC_CTRL_B)
+
+#define   HSW_AUD_DIP_ELD_CTRL_ST_A	0x650b4 /* Audio DIP and ELD Control State Transcoder A */
+#define   HSW_AUD_DIP_ELD_CTRL_ST_B	0x651b4 /* Audio DIP and ELD Control State Transcoder B */
+#define   HSW_AUD_DIP_ELD_CTRL(pipe) _PIPE(pipe, \
+					HSW_AUD_DIP_ELD_CTRL_ST_A, \
+					HSW_AUD_DIP_ELD_CTRL_ST_B)
+
+/* Audio Digital Converter */
+#define   HSW_AUD_DIG_CNVT_1		0x65080 /* Audio Converter 1 */
+#define   HSW_AUD_DIG_CNVT_2		0x65180 /* Audio Converter 1 */
+#define   AUD_DIG_CNVT(pipe) _PIPE(pipe, \
+					HSW_AUD_DIG_CNVT_1, \
+					HSW_AUD_DIG_CNVT_2)
+
+#define   HSW_AUD_EDID_DATA_A		0x65050
+#define   HSW_AUD_EDID_DATA_B		0x65150
+#define   HSW_AUD_EDID_DATA(pipe) _PIPE(pipe, \
+					HSW_AUD_EDID_DATA_A, \
+					HSW_AUD_EDID_DATA_B)
+
+#define   HSW_AUD_PIPE_CONV_CFG		0x6507c /* Audio pipe and converter configs */
+#define   HSW_AUD_PIN_ELD_CP_VLD	0x650c0 /* Audio ELD and CP Ready Status */
+#define   AUDIO_INACTIVE_C		(1<<11)
+#define   AUDIO_INACTIVE_B		(1<<7)
+#define   AUDIO_INACTIVE_A		(1<<3)
+#define   AUDIO_OUTPUT_ENABLE_A		(1<<2)
+#define   AUDIO_OUTPUT_ENABLE_B		(1<<6)
+#define   AUDIO_OUTPUT_ENABLE_C		(1<<10)
+#define   AUDIO_ELD_VALID_A		(1<<0)
+#define   AUDIO_ELD_VALID_B		(1<<4)
+#define   AUDIO_ELD_VALID_C		(1<<8)
+#define   AUDIO_CP_READY_A		(1<<1)
+#define   AUDIO_CP_READY_B		(1<<5)
+#define   AUDIO_CP_READY_C		(1<<9)
+
 /* HSW Power Wells */
 #define HSW_PWR_WELL_CTL1			0x45400 /* BIOS */
 #define HSW_PWR_WELL_CTL2			0x45404 /* Driver */

From 4f07854dd7fc197d9ea2bd83c26736baabb0b981 Mon Sep 17 00:00:00 2001
From: Wang Xingchao <xingchao.wang@intel.com>
Date: Thu, 9 Aug 2012 16:52:16 +0800
Subject: [PATCH 11/41] drm/i915: write eld info for HDMI audio

HDMI audio related registers will be configured in write_eld callback.

Signed-off-by: Wang Xingchao <xingchao.wang@intel.com>
Reviewed-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_ddi.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 958422606bc7..170e3861aa4e 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -713,8 +713,12 @@ void intel_ddi_mode_set(struct drm_encoder *encoder,
 		/* Proper support for digital audio needs a new logic and a new set
 		 * of registers, so we leave it for future patch bombing.
 		 */
-		DRM_DEBUG_DRIVER("HDMI audio on pipe %c not yet supported on DDI\n",
+		DRM_DEBUG_DRIVER("HDMI audio on pipe %c on DDI\n",
 				 pipe_name(intel_crtc->pipe));
+
+		/* write eld */
+		DRM_DEBUG_DRIVER("HDMI audio: write eld information\n");
+		intel_write_eld(encoder, adjusted_mode);
 	}
 
 	/* Enable PIPE_DDI_FUNC_CTL for the pipe to work in HDMI mode */

From 9b138a83675364366388af8b6986b2e52be36bcc Mon Sep 17 00:00:00 2001
From: Wang Xingchao <xingchao.wang@intel.com>
Date: Thu, 9 Aug 2012 16:52:18 +0800
Subject: [PATCH 12/41] drm/i915: ironlake_write_eld code cleanup

Use _PIPE macro to get correct register definition for IBX/CPT, discard
old variable "i" way.

Signed-off-by: Wang Xingchao <xingchao.wang@intel.com>
Reviewed-by: Imre Deak <imre.deak@intel.com>
[danvet: Added the DIP_PORT_SEL #define from a preceeding patch in the
series that needs more work.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_reg.h      | 25 +++++++++++++++++++++++++
 drivers/gpu/drm/i915/intel_display.c | 22 +++++++++-------------
 2 files changed, 34 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index cbddc7d8e169..fd6a26a49b42 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -4247,7 +4247,15 @@
 #define G4X_HDMIW_HDMIEDID		0x6210C
 
 #define IBX_HDMIW_HDMIEDID_A		0xE2050
+#define IBX_HDMIW_HDMIEDID_B		0xE2150
+#define IBX_HDMIW_HDMIEDID(pipe) _PIPE(pipe, \
+					IBX_HDMIW_HDMIEDID_A, \
+					IBX_HDMIW_HDMIEDID_B)
 #define IBX_AUD_CNTL_ST_A		0xE20B4
+#define IBX_AUD_CNTL_ST_B		0xE21B4
+#define IBX_AUD_CNTL_ST(pipe) _PIPE(pipe, \
+					IBX_AUD_CNTL_ST_A, \
+					IBX_AUD_CNTL_ST_B)
 #define IBX_ELD_BUFFER_SIZE		(0x1f << 10)
 #define IBX_ELD_ADDRESS			(0x1f << 5)
 #define IBX_ELD_ACK			(1 << 4)
@@ -4256,7 +4264,15 @@
 #define IBX_CP_READYB			(1 << 1)
 
 #define CPT_HDMIW_HDMIEDID_A		0xE5050
+#define CPT_HDMIW_HDMIEDID_B		0xE5150
+#define CPT_HDMIW_HDMIEDID(pipe) _PIPE(pipe, \
+					CPT_HDMIW_HDMIEDID_A, \
+					CPT_HDMIW_HDMIEDID_B)
 #define CPT_AUD_CNTL_ST_A		0xE50B4
+#define CPT_AUD_CNTL_ST_B		0xE51B4
+#define CPT_AUD_CNTL_ST(pipe) _PIPE(pipe, \
+					CPT_AUD_CNTL_ST_A, \
+					CPT_AUD_CNTL_ST_B)
 #define CPT_AUD_CNTRL_ST2		0xE50C0
 
 /* These are the 4 32-bit write offset registers for each stream
@@ -4266,7 +4282,15 @@
 #define GEN7_SO_WRITE_OFFSET(n)		(0x5280 + (n) * 4)
 
 #define IBX_AUD_CONFIG_A			0xe2000
+#define IBX_AUD_CONFIG_B			0xe2100
+#define IBX_AUD_CFG(pipe) _PIPE(pipe, \
+					IBX_AUD_CONFIG_A, \
+					IBX_AUD_CONFIG_B)
 #define CPT_AUD_CONFIG_A			0xe5000
+#define CPT_AUD_CONFIG_B			0xe5100
+#define CPT_AUD_CFG(pipe) _PIPE(pipe, \
+					CPT_AUD_CONFIG_A, \
+					CPT_AUD_CONFIG_B)
 #define   AUD_CONFIG_N_VALUE_INDEX		(1 << 29)
 #define   AUD_CONFIG_N_PROG_ENABLE		(1 << 28)
 #define   AUD_CONFIG_UPPER_N_SHIFT		20
@@ -4302,6 +4326,7 @@
 #define   AUD_DIG_CNVT(pipe) _PIPE(pipe, \
 					HSW_AUD_DIG_CNVT_1, \
 					HSW_AUD_DIG_CNVT_2)
+#define   DIP_PORT_SEL_MASK		0x3
 
 #define   HSW_AUD_EDID_DATA_A		0x65050
 #define   HSW_AUD_EDID_DATA_B		0x65150
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index b580d949d7d2..f55e024c6df9 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -5078,28 +5078,24 @@ static void ironlake_write_eld(struct drm_connector *connector,
 	int aud_config;
 	int aud_cntl_st;
 	int aud_cntrl_st2;
+	int pipe = to_intel_crtc(crtc)->pipe;
 
 	if (HAS_PCH_IBX(connector->dev)) {
-		hdmiw_hdmiedid = IBX_HDMIW_HDMIEDID_A;
-		aud_config = IBX_AUD_CONFIG_A;
-		aud_cntl_st = IBX_AUD_CNTL_ST_A;
+		hdmiw_hdmiedid = IBX_HDMIW_HDMIEDID(pipe);
+		aud_config = IBX_AUD_CFG(pipe);
+		aud_cntl_st = IBX_AUD_CNTL_ST(pipe);
 		aud_cntrl_st2 = IBX_AUD_CNTL_ST2;
 	} else {
-		hdmiw_hdmiedid = CPT_HDMIW_HDMIEDID_A;
-		aud_config = CPT_AUD_CONFIG_A;
-		aud_cntl_st = CPT_AUD_CNTL_ST_A;
+		hdmiw_hdmiedid = CPT_HDMIW_HDMIEDID(pipe);
+		aud_config = CPT_AUD_CFG(pipe);
+		aud_cntl_st = CPT_AUD_CNTL_ST(pipe);
 		aud_cntrl_st2 = CPT_AUD_CNTRL_ST2;
 	}
 
-	i = to_intel_crtc(crtc)->pipe;
-	hdmiw_hdmiedid += i * 0x100;
-	aud_cntl_st += i * 0x100;
-	aud_config += i * 0x100;
-
-	DRM_DEBUG_DRIVER("ELD on pipe %c\n", pipe_name(i));
+	DRM_DEBUG_DRIVER("ELD on pipe %c\n", pipe_name(pipe));
 
 	i = I915_READ(aud_cntl_st);
-	i = (i >> 29) & 0x3;		/* DIP_Port_Select, 0x1 = PortB */
+	i = (i >> 29) & DIP_PORT_SEL_MASK;		/* DIP_Port_Select, 0x1 = PortB */
 	if (!i) {
 		DRM_DEBUG_DRIVER("Audio directed to unknown port\n");
 		/* operate blindly on all ports */

From 33faad195eec0c72fec84e638623ea25652eed72 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 12 Aug 2012 18:15:48 +0200
Subject: [PATCH 13/41] drm/fb-helper: don't clobber output routing in
 setup_crtcs

Yet again the too close relationship between the fb helper and the
crtc helper code strikes. This time around the fb helper resets all
encoder->crtc pointers to NULL before starting to set up it's own
mode. Which is total bullocks, because this will clobber the existing
output routing, which the new drm/i915 code depends upon to be
absolutely correct.

The crtc helper itself doesn't really care about that, since it
disables unused encoders in a rather roundabout way anyway.

Two places call drm_setup_crts:

- For the initial fb config. I've auditted all current drivers, and
  they all allocate their encoders with kzalloc. So there's no need to
  clear encoder->crtc once more.

- When processing hotplug events while showing the fb console. This
  one is a bit more tricky, but both the crtc helper code and the new
  drm/i915 modeset code disable encoders if their crtc is changed and
  that encoder isn't part of the new config. Also, both disable any
  disconnected outputs, too.

  Which only leaves encoders that are on, connected, but for some odd
  reason the fb helper code doesn't want to use. That would be a bug
  in the fb configuration selector, since it tries to light up as many
  outputs as possible.

v2: Kill the now unused encoders variable.

Acked-by: Dave Airlie <airlied@gmail.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/drm_fb_helper.c | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index f546d1e8af82..4ecc869fb5be 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -1230,7 +1230,6 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper)
 	struct drm_device *dev = fb_helper->dev;
 	struct drm_fb_helper_crtc **crtcs;
 	struct drm_display_mode **modes;
-	struct drm_encoder *encoder;
 	struct drm_mode_set *modeset;
 	bool *enabled;
 	int width, height;
@@ -1241,11 +1240,6 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper)
 	width = dev->mode_config.max_width;
 	height = dev->mode_config.max_height;
 
-	/* clean out all the encoder/crtc combos */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		encoder->crtc = NULL;
-	}
-
 	crtcs = kcalloc(dev->mode_config.num_connector,
 			sizeof(struct drm_fb_helper_crtc *), GFP_KERNEL);
 	modes = kcalloc(dev->mode_config.num_connector,

From 523313092abffb21bfe7cd8d37615402b2228850 Mon Sep 17 00:00:00 2001
From: Damien Lespiau <damien.lespiau@intel.com>
Date: Wed, 15 Aug 2012 19:23:25 +0100
Subject: [PATCH 14/41] drm/i915: Don't hardcode the number of pipes in the
 error state dump

New-ish devices have 3 pipes, so let's not just hardcode 2 but use the
for_each_pipe() macro and make struct intel_display_error_state is big
enough.

V2: Also add the number of pipes emitted (Chris Wilson)

Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_display.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index f55e024c6df9..7a1c19756884 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -7234,7 +7234,7 @@ struct intel_display_error_state {
 		u32 position;
 		u32 base;
 		u32 size;
-	} cursor[2];
+	} cursor[I915_MAX_PIPES];
 
 	struct intel_pipe_error_state {
 		u32 conf;
@@ -7246,7 +7246,7 @@ struct intel_display_error_state {
 		u32 vtotal;
 		u32 vblank;
 		u32 vsync;
-	} pipe[2];
+	} pipe[I915_MAX_PIPES];
 
 	struct intel_plane_error_state {
 		u32 control;
@@ -7256,7 +7256,7 @@ struct intel_display_error_state {
 		u32 addr;
 		u32 surface;
 		u32 tile_offset;
-	} plane[2];
+	} plane[I915_MAX_PIPES];
 };
 
 struct intel_display_error_state *
@@ -7270,7 +7270,7 @@ intel_display_capture_error_state(struct drm_device *dev)
 	if (error == NULL)
 		return NULL;
 
-	for (i = 0; i < 2; i++) {
+	for_each_pipe(i) {
 		error->cursor[i].control = I915_READ(CURCNTR(i));
 		error->cursor[i].position = I915_READ(CURPOS(i));
 		error->cursor[i].base = I915_READ(CURBASE(i));
@@ -7303,9 +7303,11 @@ intel_display_print_error_state(struct seq_file *m,
 				struct drm_device *dev,
 				struct intel_display_error_state *error)
 {
+	drm_i915_private_t *dev_priv = dev->dev_private;
 	int i;
 
-	for (i = 0; i < 2; i++) {
+	seq_printf(m, "Num Pipes: %d\n", dev_priv->num_pipe);
+	for_each_pipe(i) {
 		seq_printf(m, "Pipe [%d]:\n", i);
 		seq_printf(m, "  CONF: %08x\n", error->pipe[i].conf);
 		seq_printf(m, "  SRC: %08x\n", error->pipe[i].source);

From 0826874a664f9e9479d50fda23923b87a50cda0d Mon Sep 17 00:00:00 2001
From: Keith Packard <keithp@keithp.com>
Date: Mon, 13 Aug 2012 21:34:45 -0700
Subject: [PATCH 15/41] drm/i915: Allow VGA on CRTC 2

This is left over from the old PLL sharing code and isn't useful now
that PLLs are shared when possible.

Signed-off-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_crt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index 62b6555cfa27..1d05b2d5abed 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -662,7 +662,7 @@ void intel_crt_init(struct drm_device *dev)
 	if (IS_HASWELL(dev))
 		crt->base.crtc_mask = (1 << 0);
 	else
-		crt->base.crtc_mask = (1 << 0) | (1 << 1);
+		crt->base.crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
 
 	if (IS_GEN2(dev))
 		connector->interlace_allowed = 0;

From ec6f1bb90c9865d4a5af01b1c643bcf020d88706 Mon Sep 17 00:00:00 2001
From: Dave Airlie <airlied@redhat.com>
Date: Thu, 16 Aug 2012 10:15:34 +1000
Subject: [PATCH 16/41] drm/i915: implement dma buf begin_cpu_access (v2)

In order for udl vmap to work properly, we need to push the object
into the CPU domain before we start copying the data to the USB device.

This along with the udl change avoids userspace explicit mapping to
be used.

v2: add a flag for userspace to query to know if Intel kernel driver can
deal with the vmap flushing properly. In theory udl would need a flag also,
but I intend to push the patches very close to each other and other drivers
should do the right thing from the start.

I've added a test to my intel-gpu-tools prime branch, however testing
this is a bit messy since the only way to get udl to vmap is to rendering
something. I've tested this with real code as well to make sure it works.

Signed-off-by: Dave Airlie <airlied@redhat.com>
[danvet: resolved conflict, which required reallocating the PARAM
number to 21.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_dma.c        |  3 +++
 drivers/gpu/drm/i915/i915_gem_dmabuf.c | 17 +++++++++++++++++
 include/drm/i915_drm.h                 |  1 +
 3 files changed, 21 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 2285ae31717a..2cba7b4a04ed 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -1012,6 +1012,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 	case I915_PARAM_HAS_SEMAPHORES:
 		value = i915_semaphore_is_enabled(dev);
 		break;
+	case I915_PARAM_HAS_PRIME_VMAP_FLUSH:
+		value = 1;
+		break;
 	default:
 		DRM_DEBUG_DRIVER("Unknown parameter %d\n",
 				 param->param);
diff --git a/drivers/gpu/drm/i915/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
index aa308e1337db..ceaad5af01a4 100644
--- a/drivers/gpu/drm/i915/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
@@ -151,6 +151,22 @@ static int i915_gem_dmabuf_mmap(struct dma_buf *dma_buf, struct vm_area_struct *
 	return -EINVAL;
 }
 
+static int i915_gem_begin_cpu_access(struct dma_buf *dma_buf, size_t start, size_t length, enum dma_data_direction direction)
+{
+	struct drm_i915_gem_object *obj = dma_buf->priv;
+	struct drm_device *dev = obj->base.dev;
+	int ret;
+	bool write = (direction == DMA_BIDIRECTIONAL || direction == DMA_TO_DEVICE);
+
+	ret = i915_mutex_lock_interruptible(dev);
+	if (ret)
+		return ret;
+
+	ret = i915_gem_object_set_to_cpu_domain(obj, write);
+	mutex_unlock(&dev->struct_mutex);
+	return ret;
+}
+
 static const struct dma_buf_ops i915_dmabuf_ops =  {
 	.map_dma_buf = i915_gem_map_dma_buf,
 	.unmap_dma_buf = i915_gem_unmap_dma_buf,
@@ -162,6 +178,7 @@ static const struct dma_buf_ops i915_dmabuf_ops =  {
 	.mmap = i915_gem_dmabuf_mmap,
 	.vmap = i915_gem_dmabuf_vmap,
 	.vunmap = i915_gem_dmabuf_vunmap,
+	.begin_cpu_access = i915_gem_begin_cpu_access,
 };
 
 struct dma_buf *i915_gem_prime_export(struct drm_device *dev,
diff --git a/include/drm/i915_drm.h b/include/drm/i915_drm.h
index d8a79bf59ae7..05e24d31f329 100644
--- a/include/drm/i915_drm.h
+++ b/include/drm/i915_drm.h
@@ -312,6 +312,7 @@ typedef struct drm_i915_irq_wait {
 #define I915_PARAM_HAS_ALIASING_PPGTT	 18
 #define I915_PARAM_HAS_WAIT_TIMEOUT	 19
 #define I915_PARAM_HAS_SEMAPHORES	 20
+#define I915_PARAM_HAS_PRIME_VMAP_FLUSH	 21
 
 typedef struct drm_i915_getparam {
 	int param;

From 32ecd24264542328893e8649312c301150dbadc5 Mon Sep 17 00:00:00 2001
From: Dave Airlie <airlied@redhat.com>
Date: Mon, 30 Jul 2012 14:06:55 +1000
Subject: [PATCH 17/41] drm/udl: call begin/end cpu access at more appropriate
 time

We need to call these before we transfer the damaged areas to the device
not before/after we setup the long lived vmaps.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/udl/udl_fb.c  | 22 ++++++++++++++++++++--
 drivers/gpu/drm/udl/udl_gem.c |  7 -------
 2 files changed, 20 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c
index ce9a61179925..b8c00ed33051 100644
--- a/drivers/gpu/drm/udl/udl_fb.c
+++ b/drivers/gpu/drm/udl/udl_fb.c
@@ -13,6 +13,7 @@
 #include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/fb.h>
+#include <linux/dma-buf.h>
 
 #include "drmP.h"
 #include "drm.h"
@@ -377,16 +378,33 @@ static int udl_user_framebuffer_dirty(struct drm_framebuffer *fb,
 {
 	struct udl_framebuffer *ufb = to_udl_fb(fb);
 	int i;
+	int ret = 0;
 
 	if (!ufb->active_16)
 		return 0;
 
+	if (ufb->obj->base.import_attach) {
+		ret = dma_buf_begin_cpu_access(ufb->obj->base.import_attach->dmabuf,
+					       0, ufb->obj->base.size,
+					       DMA_FROM_DEVICE);
+		if (ret)
+			return ret;
+	}
+
 	for (i = 0; i < num_clips; i++) {
-		udl_handle_damage(ufb, clips[i].x1, clips[i].y1,
+		ret = udl_handle_damage(ufb, clips[i].x1, clips[i].y1,
 				  clips[i].x2 - clips[i].x1,
 				  clips[i].y2 - clips[i].y1);
+		if (ret)
+			break;
 	}
-	return 0;
+
+	if (ufb->obj->base.import_attach) {
+		dma_buf_end_cpu_access(ufb->obj->base.import_attach->dmabuf,
+				       0, ufb->obj->base.size,
+				       DMA_FROM_DEVICE);
+	}
+	return ret;
 }
 
 static void udl_user_framebuffer_destroy(struct drm_framebuffer *fb)
diff --git a/drivers/gpu/drm/udl/udl_gem.c b/drivers/gpu/drm/udl/udl_gem.c
index 291ecc145585..47b256375831 100644
--- a/drivers/gpu/drm/udl/udl_gem.c
+++ b/drivers/gpu/drm/udl/udl_gem.c
@@ -181,11 +181,6 @@ int udl_gem_vmap(struct udl_gem_object *obj)
 	int ret;
 
 	if (obj->base.import_attach) {
-		ret = dma_buf_begin_cpu_access(obj->base.import_attach->dmabuf,
-					       0, obj->base.size, DMA_BIDIRECTIONAL);
-		if (ret)
-			return -EINVAL;
-
 		obj->vmapping = dma_buf_vmap(obj->base.import_attach->dmabuf);
 		if (!obj->vmapping)
 			return -ENOMEM;
@@ -206,8 +201,6 @@ void udl_gem_vunmap(struct udl_gem_object *obj)
 {
 	if (obj->base.import_attach) {
 		dma_buf_vunmap(obj->base.import_attach->dmabuf, obj->vmapping);
-		dma_buf_end_cpu_access(obj->base.import_attach->dmabuf, 0,
-				       obj->base.size, DMA_BIDIRECTIONAL);
 		return;
 	}
 

From 83358c85866ebd2af1229fc9870b93e126690671 Mon Sep 17 00:00:00 2001
From: Wang Xingchao <xingchao.wang@intel.com>
Date: Thu, 16 Aug 2012 22:43:37 +0800
Subject: [PATCH 18/41] drm/i915: Haswell HDMI audio initialization

Added new haswell_write_eld() to initialize Haswell HDMI audio registers
to generate an unsolicited response to the audio controller driver to
indicate that the controller sequence should start.

Reviewed-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Wang Xingchao <xingchao.wang@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_display.c | 87 +++++++++++++++++++++++++++-
 1 file changed, 86 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 7a1c19756884..42c57581c53f 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -5066,6 +5066,91 @@ static void g4x_write_eld(struct drm_connector *connector,
 	I915_WRITE(G4X_AUD_CNTL_ST, i);
 }
 
+static void haswell_write_eld(struct drm_connector *connector,
+				     struct drm_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = connector->dev->dev_private;
+	uint8_t *eld = connector->eld;
+	struct drm_device *dev = crtc->dev;
+	uint32_t eldv;
+	uint32_t i;
+	int len;
+	int pipe = to_intel_crtc(crtc)->pipe;
+	int tmp;
+
+	int hdmiw_hdmiedid = HSW_AUD_EDID_DATA(pipe);
+	int aud_cntl_st = HSW_AUD_DIP_ELD_CTRL(pipe);
+	int aud_config = HSW_AUD_CFG(pipe);
+	int aud_cntrl_st2 = HSW_AUD_PIN_ELD_CP_VLD;
+
+
+	DRM_DEBUG_DRIVER("HDMI: Haswell Audio initialize....\n");
+
+	/* Audio output enable */
+	DRM_DEBUG_DRIVER("HDMI audio: enable codec\n");
+	tmp = I915_READ(aud_cntrl_st2);
+	tmp |= (AUDIO_OUTPUT_ENABLE_A << (pipe * 4));
+	I915_WRITE(aud_cntrl_st2, tmp);
+
+	/* Wait for 1 vertical blank */
+	intel_wait_for_vblank(dev, pipe);
+
+	/* Set ELD valid state */
+	tmp = I915_READ(aud_cntrl_st2);
+	DRM_DEBUG_DRIVER("HDMI audio: pin eld vld status=0x%8x\n", tmp);
+	tmp |= (AUDIO_ELD_VALID_A << (pipe * 4));
+	I915_WRITE(aud_cntrl_st2, tmp);
+	tmp = I915_READ(aud_cntrl_st2);
+	DRM_DEBUG_DRIVER("HDMI audio: eld vld status=0x%8x\n", tmp);
+
+	/* Enable HDMI mode */
+	tmp = I915_READ(aud_config);
+	DRM_DEBUG_DRIVER("HDMI audio: audio conf: 0x%8x\n", tmp);
+	/* clear N_programing_enable and N_value_index */
+	tmp &= ~(AUD_CONFIG_N_VALUE_INDEX | AUD_CONFIG_N_PROG_ENABLE);
+	I915_WRITE(aud_config, tmp);
+
+	DRM_DEBUG_DRIVER("ELD on pipe %c\n", pipe_name(pipe));
+
+	eldv = AUDIO_ELD_VALID_A << (pipe * 4);
+
+	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DISPLAYPORT)) {
+		DRM_DEBUG_DRIVER("ELD: DisplayPort detected\n");
+		eld[5] |= (1 << 2);	/* Conn_Type, 0x1 = DisplayPort */
+		I915_WRITE(aud_config, AUD_CONFIG_N_VALUE_INDEX); /* 0x1 = DP */
+	} else
+		I915_WRITE(aud_config, 0);
+
+	if (intel_eld_uptodate(connector,
+			       aud_cntrl_st2, eldv,
+			       aud_cntl_st, IBX_ELD_ADDRESS,
+			       hdmiw_hdmiedid))
+		return;
+
+	i = I915_READ(aud_cntrl_st2);
+	i &= ~eldv;
+	I915_WRITE(aud_cntrl_st2, i);
+
+	if (!eld[0])
+		return;
+
+	i = I915_READ(aud_cntl_st);
+	i &= ~IBX_ELD_ADDRESS;
+	I915_WRITE(aud_cntl_st, i);
+	i = (i >> 29) & DIP_PORT_SEL_MASK;		/* DIP_Port_Select, 0x1 = PortB */
+	DRM_DEBUG_DRIVER("port num:%d\n", i);
+
+	len = min_t(uint8_t, eld[2], 21);	/* 84 bytes of hw ELD buffer */
+	DRM_DEBUG_DRIVER("ELD size %d\n", len);
+	for (i = 0; i < len; i++)
+		I915_WRITE(hdmiw_hdmiedid, *((uint32_t *)eld + i));
+
+	i = I915_READ(aud_cntrl_st2);
+	i |= eldv;
+	I915_WRITE(aud_cntrl_st2, i);
+
+}
+
 static void ironlake_write_eld(struct drm_connector *connector,
 				     struct drm_crtc *crtc)
 {
@@ -6936,7 +7021,7 @@ static void intel_init_display(struct drm_device *dev)
 			dev_priv->display.write_eld = ironlake_write_eld;
 		} else if (IS_HASWELL(dev)) {
 			dev_priv->display.fdi_link_train = hsw_fdi_link_train;
-			dev_priv->display.write_eld = ironlake_write_eld;
+			dev_priv->display.write_eld = haswell_write_eld;
 		} else
 			dev_priv->display.update_wm = NULL;
 	} else if (IS_G4X(dev)) {

From 225067eedf1f4d4065940232c9069fcb255206ee Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Mon, 20 Aug 2012 10:23:20 +0200
Subject: [PATCH 19/41] drm/i915: move functions around

Prep work to make Chris Wilson's unbound tracking patch a bit easier
to read. Alas, I'd have preferred that moving the page allocation
retry loop from bind to get_pages would have been a separate patch,
too. But that looks like real work ;-)

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem.c | 116 ++++++++++++++++----------------
 1 file changed, 58 insertions(+), 58 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 051459324826..0f70c2acfefa 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1340,6 +1340,64 @@ i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data,
 	return i915_gem_mmap_gtt(file, dev, args->handle, &args->offset);
 }
 
+/* Immediately discard the backing storage */
+static void
+i915_gem_object_truncate(struct drm_i915_gem_object *obj)
+{
+	struct inode *inode;
+
+	/* Our goal here is to return as much of the memory as
+	 * is possible back to the system as we are called from OOM.
+	 * To do this we must instruct the shmfs to drop all of its
+	 * backing pages, *now*.
+	 */
+	inode = obj->base.filp->f_path.dentry->d_inode;
+	shmem_truncate_range(inode, 0, (loff_t)-1);
+
+	if (obj->base.map_list.map)
+		drm_gem_free_mmap_offset(&obj->base);
+
+	obj->madv = __I915_MADV_PURGED;
+}
+
+static inline int
+i915_gem_object_is_purgeable(struct drm_i915_gem_object *obj)
+{
+	return obj->madv == I915_MADV_DONTNEED;
+}
+
+static void
+i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj)
+{
+	int page_count = obj->base.size / PAGE_SIZE;
+	int i;
+
+	if (!obj->pages)
+		return;
+
+	BUG_ON(obj->madv == __I915_MADV_PURGED);
+
+	if (i915_gem_object_needs_bit17_swizzle(obj))
+		i915_gem_object_save_bit_17_swizzle(obj);
+
+	if (obj->madv == I915_MADV_DONTNEED)
+		obj->dirty = 0;
+
+	for (i = 0; i < page_count; i++) {
+		if (obj->dirty)
+			set_page_dirty(obj->pages[i]);
+
+		if (obj->madv == I915_MADV_WILLNEED)
+			mark_page_accessed(obj->pages[i]);
+
+		page_cache_release(obj->pages[i]);
+	}
+	obj->dirty = 0;
+
+	drm_free_large(obj->pages);
+	obj->pages = NULL;
+}
+
 int
 i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj,
 			      gfp_t gfpmask)
@@ -1387,38 +1445,6 @@ err_pages:
 	return PTR_ERR(page);
 }
 
-static void
-i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj)
-{
-	int page_count = obj->base.size / PAGE_SIZE;
-	int i;
-
-	if (!obj->pages)
-		return;
-
-	BUG_ON(obj->madv == __I915_MADV_PURGED);
-
-	if (i915_gem_object_needs_bit17_swizzle(obj))
-		i915_gem_object_save_bit_17_swizzle(obj);
-
-	if (obj->madv == I915_MADV_DONTNEED)
-		obj->dirty = 0;
-
-	for (i = 0; i < page_count; i++) {
-		if (obj->dirty)
-			set_page_dirty(obj->pages[i]);
-
-		if (obj->madv == I915_MADV_WILLNEED)
-			mark_page_accessed(obj->pages[i]);
-
-		page_cache_release(obj->pages[i]);
-	}
-	obj->dirty = 0;
-
-	drm_free_large(obj->pages);
-	obj->pages = NULL;
-}
-
 void
 i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
 			       struct intel_ring_buffer *ring,
@@ -1486,32 +1512,6 @@ i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
 	WARN_ON(i915_verify_lists(dev));
 }
 
-/* Immediately discard the backing storage */
-static void
-i915_gem_object_truncate(struct drm_i915_gem_object *obj)
-{
-	struct inode *inode;
-
-	/* Our goal here is to return as much of the memory as
-	 * is possible back to the system as we are called from OOM.
-	 * To do this we must instruct the shmfs to drop all of its
-	 * backing pages, *now*.
-	 */
-	inode = obj->base.filp->f_path.dentry->d_inode;
-	shmem_truncate_range(inode, 0, (loff_t)-1);
-
-	if (obj->base.map_list.map)
-		drm_gem_free_mmap_offset(&obj->base);
-
-	obj->madv = __I915_MADV_PURGED;
-}
-
-static inline int
-i915_gem_object_is_purgeable(struct drm_i915_gem_object *obj)
-{
-	return obj->madv == I915_MADV_DONTNEED;
-}
-
 static u32
 i915_gem_get_seqno(struct drm_device *dev)
 {

From 6c085a728cf000ac1865d66f8c9b52935558b328 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 20 Aug 2012 11:40:46 +0200
Subject: [PATCH 20/41] drm/i915: Track unbound pages

When dealing with a working set larger than the GATT, or even the
mappable aperture when touching through the GTT, we end up with evicting
objects only to rebind them at a new offset again later. Moving an
object into and out of the GTT requires clflushing the pages, thus
causing a double-clflush penalty for rebinding.

To avoid having to clflush on rebinding, we can track the pages as they
are evicted from the GTT and only relinquish those pages on memory
pressure.

As usual, if it were not for the handling of out-of-memory condition and
having to manually shrink our own bo caches, it would be a net reduction
of code. Alas.

Note: The patch also contains a few changes to the last-hope
evict_everything logic in i916_gem_execbuffer.c - we no longer try to
only evict the purgeable stuff in a first try (since that's superflous
and only helps in OOM corner-cases, not fragmented-gtt trashing
situations).

Also, the extraction of the get_pages retry loop from bind_to_gtt (and
other callsites) to get_pages should imo have been a separate patch.

v2: Ditch the newly added put_pages (for unbound objects only) in
i915_gem_reset. A quick irc discussion hasn't revealed any important
reason for this, so if we need this, I'd like to have a git blame'able
explanation for it.

v3: Undo the s/drm_malloc_ab/kmalloc/ in get_pages that Chris noticed.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Split out code movements and rant a bit in the commit message
with a few Notes. Done v2]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_debugfs.c        |  14 +-
 drivers/gpu/drm/i915/i915_drv.h            |  13 +-
 drivers/gpu/drm/i915/i915_gem.c            | 288 ++++++++++-----------
 drivers/gpu/drm/i915/i915_gem_dmabuf.c     |  20 +-
 drivers/gpu/drm/i915/i915_gem_evict.c      |  13 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |   9 +-
 drivers/gpu/drm/i915/i915_gem_gtt.c        |   2 +-
 drivers/gpu/drm/i915/i915_irq.c            |   4 +-
 drivers/gpu/drm/i915/i915_trace.h          |  10 +-
 9 files changed, 184 insertions(+), 189 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index a18e93687b8b..608d3ae2b5ba 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -211,7 +211,7 @@ static int i915_gem_object_info(struct seq_file *m, void* data)
 		   dev_priv->mm.object_memory);
 
 	size = count = mappable_size = mappable_count = 0;
-	count_objects(&dev_priv->mm.gtt_list, gtt_list);
+	count_objects(&dev_priv->mm.bound_list, gtt_list);
 	seq_printf(m, "%u [%u] objects, %zu [%zu] bytes in gtt\n",
 		   count, mappable_count, size, mappable_size);
 
@@ -225,8 +225,13 @@ static int i915_gem_object_info(struct seq_file *m, void* data)
 	seq_printf(m, "  %u [%u] inactive objects, %zu [%zu] bytes\n",
 		   count, mappable_count, size, mappable_size);
 
+	size = count = 0;
+	list_for_each_entry(obj, &dev_priv->mm.unbound_list, gtt_list)
+		size += obj->base.size, ++count;
+	seq_printf(m, "%u unbound objects, %zu bytes\n", count, size);
+
 	size = count = mappable_size = mappable_count = 0;
-	list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list) {
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, gtt_list) {
 		if (obj->fault_mappable) {
 			size += obj->gtt_space->size;
 			++count;
@@ -264,7 +269,7 @@ static int i915_gem_gtt_info(struct seq_file *m, void* data)
 		return ret;
 
 	total_obj_size = total_gtt_size = count = 0;
-	list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list) {
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, gtt_list) {
 		if (list == PINNED_LIST && obj->pin_count == 0)
 			continue;
 
@@ -526,7 +531,8 @@ static int i915_gem_fence_regs_info(struct seq_file *m, void *data)
 	for (i = 0; i < dev_priv->num_fence_regs; i++) {
 		struct drm_i915_gem_object *obj = dev_priv->fence_regs[i].obj;
 
-		seq_printf(m, "Fenced object[%2d] = ", i);
+		seq_printf(m, "Fence %d, pin count = %d, object = ",
+			   i, dev_priv->fence_regs[i].pin_count);
 		if (obj == NULL)
 			seq_printf(m, "unused");
 		else
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index ed3ba70923ac..a2382a175d92 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -685,7 +685,13 @@ typedef struct drm_i915_private {
 		struct drm_mm gtt_space;
 		/** List of all objects in gtt_space. Used to restore gtt
 		 * mappings on resume */
-		struct list_head gtt_list;
+		struct list_head bound_list;
+		/**
+		 * List of objects which are not bound to the GTT (thus
+		 * are idle and not used by the GPU) but still have
+		 * (presumably uncached) pages still attached.
+		 */
+		struct list_head unbound_list;
 
 		/** Usable portion of the GTT for GEM */
 		unsigned long gtt_start;
@@ -1306,8 +1312,7 @@ int __must_check i915_gem_object_unbind(struct drm_i915_gem_object *obj);
 void i915_gem_release_mmap(struct drm_i915_gem_object *obj);
 void i915_gem_lastclose(struct drm_device *dev);
 
-int i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj,
-				  gfp_t gfpmask);
+int __must_check i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj);
 int __must_check i915_mutex_lock_interruptible(struct drm_device *dev);
 int i915_gem_object_sync(struct drm_i915_gem_object *obj,
 			 struct intel_ring_buffer *to);
@@ -1449,7 +1454,7 @@ int __must_check i915_gem_evict_something(struct drm_device *dev, int min_size,
 					  unsigned alignment,
 					  unsigned cache_level,
 					  bool mappable);
-int i915_gem_evict_everything(struct drm_device *dev, bool purgeable_only);
+int i915_gem_evict_everything(struct drm_device *dev);
 
 /* i915_gem_stolen.c */
 int i915_gem_init_stolen(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 0f70c2acfefa..462a8f2ad3a2 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -55,6 +55,8 @@ static void i915_gem_object_update_fence(struct drm_i915_gem_object *obj,
 
 static int i915_gem_inactive_shrink(struct shrinker *shrinker,
 				    struct shrink_control *sc);
+static long i915_gem_purge(struct drm_i915_private *dev_priv, long target);
+static void i915_gem_shrink_all(struct drm_i915_private *dev_priv);
 static void i915_gem_object_truncate(struct drm_i915_gem_object *obj);
 
 static inline void i915_gem_object_fence_lost(struct drm_i915_gem_object *obj)
@@ -140,7 +142,7 @@ int i915_mutex_lock_interruptible(struct drm_device *dev)
 static inline bool
 i915_gem_object_is_inactive(struct drm_i915_gem_object *obj)
 {
-	return !obj->active;
+	return obj->gtt_space && !obj->active;
 }
 
 int
@@ -179,7 +181,7 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 
 	pinned = 0;
 	mutex_lock(&dev->struct_mutex);
-	list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list)
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, gtt_list)
 		if (obj->pin_count)
 			pinned += obj->gtt_space->size;
 	mutex_unlock(&dev->struct_mutex);
@@ -423,9 +425,11 @@ i915_gem_shmem_pread(struct drm_device *dev,
 		 * anyway again before the next pread happens. */
 		if (obj->cache_level == I915_CACHE_NONE)
 			needs_clflush = 1;
-		ret = i915_gem_object_set_to_gtt_domain(obj, false);
-		if (ret)
-			return ret;
+		if (obj->gtt_space) {
+			ret = i915_gem_object_set_to_gtt_domain(obj, false);
+			if (ret)
+				return ret;
+		}
 	}
 
 	offset = args->offset;
@@ -751,9 +755,11 @@ i915_gem_shmem_pwrite(struct drm_device *dev,
 		 * right away and we therefore have to clflush anyway. */
 		if (obj->cache_level == I915_CACHE_NONE)
 			needs_clflush_after = 1;
-		ret = i915_gem_object_set_to_gtt_domain(obj, true);
-		if (ret)
-			return ret;
+		if (obj->gtt_space) {
+			ret = i915_gem_object_set_to_gtt_domain(obj, true);
+			if (ret)
+				return ret;
+		}
 	}
 	/* Same trick applies for invalidate partially written cachelines before
 	 * writing.  */
@@ -1366,17 +1372,28 @@ i915_gem_object_is_purgeable(struct drm_i915_gem_object *obj)
 	return obj->madv == I915_MADV_DONTNEED;
 }
 
-static void
+static int
 i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj)
 {
 	int page_count = obj->base.size / PAGE_SIZE;
-	int i;
+	int ret, i;
 
-	if (!obj->pages)
-		return;
+	if (obj->pages == NULL)
+		return 0;
 
+	BUG_ON(obj->gtt_space);
 	BUG_ON(obj->madv == __I915_MADV_PURGED);
 
+	ret = i915_gem_object_set_to_cpu_domain(obj, true);
+	if (ret) {
+		/* In the event of a disaster, abandon all caches and
+		 * hope for the best.
+		 */
+		WARN_ON(ret != -EIO);
+		i915_gem_clflush_object(obj);
+		obj->base.read_domains = obj->base.write_domain = I915_GEM_DOMAIN_CPU;
+	}
+
 	if (i915_gem_object_needs_bit17_swizzle(obj))
 		i915_gem_object_save_bit_17_swizzle(obj);
 
@@ -1396,37 +1413,112 @@ i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj)
 
 	drm_free_large(obj->pages);
 	obj->pages = NULL;
+
+	list_del(&obj->gtt_list);
+
+	if (i915_gem_object_is_purgeable(obj))
+		i915_gem_object_truncate(obj);
+
+	return 0;
+}
+
+static long
+i915_gem_purge(struct drm_i915_private *dev_priv, long target)
+{
+	struct drm_i915_gem_object *obj, *next;
+	long count = 0;
+
+	list_for_each_entry_safe(obj, next,
+				 &dev_priv->mm.unbound_list,
+				 gtt_list) {
+		if (i915_gem_object_is_purgeable(obj) &&
+		    i915_gem_object_put_pages_gtt(obj) == 0) {
+			count += obj->base.size >> PAGE_SHIFT;
+			if (count >= target)
+				return count;
+		}
+	}
+
+	list_for_each_entry_safe(obj, next,
+				 &dev_priv->mm.inactive_list,
+				 mm_list) {
+		if (i915_gem_object_is_purgeable(obj) &&
+		    i915_gem_object_unbind(obj) == 0 &&
+		    i915_gem_object_put_pages_gtt(obj) == 0) {
+			count += obj->base.size >> PAGE_SHIFT;
+			if (count >= target)
+				return count;
+		}
+	}
+
+	return count;
+}
+
+static void
+i915_gem_shrink_all(struct drm_i915_private *dev_priv)
+{
+	struct drm_i915_gem_object *obj, *next;
+
+	i915_gem_evict_everything(dev_priv->dev);
+
+	list_for_each_entry_safe(obj, next, &dev_priv->mm.unbound_list, gtt_list)
+		i915_gem_object_put_pages_gtt(obj);
 }
 
 int
-i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj,
-			      gfp_t gfpmask)
+i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 {
+	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
 	int page_count, i;
 	struct address_space *mapping;
-	struct inode *inode;
 	struct page *page;
+	gfp_t gfp;
 
 	if (obj->pages || obj->sg_table)
 		return 0;
 
+	/* Assert that the object is not currently in any GPU domain. As it
+	 * wasn't in the GTT, there shouldn't be any way it could have been in
+	 * a GPU cache
+	 */
+	BUG_ON(obj->base.read_domains & I915_GEM_GPU_DOMAINS);
+	BUG_ON(obj->base.write_domain & I915_GEM_GPU_DOMAINS);
+
 	/* Get the list of pages out of our struct file.  They'll be pinned
 	 * at this point until we release them.
 	 */
 	page_count = obj->base.size / PAGE_SIZE;
-	BUG_ON(obj->pages != NULL);
 	obj->pages = drm_malloc_ab(page_count, sizeof(struct page *));
 	if (obj->pages == NULL)
 		return -ENOMEM;
 
-	inode = obj->base.filp->f_path.dentry->d_inode;
-	mapping = inode->i_mapping;
-	gfpmask |= mapping_gfp_mask(mapping);
-
+	/* Fail silently without starting the shrinker */
+	mapping = obj->base.filp->f_path.dentry->d_inode->i_mapping;
+	gfp = mapping_gfp_mask(mapping);
+	gfp |= __GFP_NORETRY | __GFP_NOWARN | __GFP_NO_KSWAPD;
+	gfp &= ~(__GFP_IO | __GFP_WAIT);
 	for (i = 0; i < page_count; i++) {
-		page = shmem_read_mapping_page_gfp(mapping, i, gfpmask);
-		if (IS_ERR(page))
-			goto err_pages;
+		page = shmem_read_mapping_page_gfp(mapping, i, gfp);
+		if (IS_ERR(page)) {
+			i915_gem_purge(dev_priv, page_count);
+			page = shmem_read_mapping_page_gfp(mapping, i, gfp);
+		}
+		if (IS_ERR(page)) {
+			/* We've tried hard to allocate the memory by reaping
+			 * our own buffer, now let the real VM do its job and
+			 * go down in flames if truly OOM.
+			 */
+			gfp &= ~(__GFP_NORETRY | __GFP_NOWARN | __GFP_NO_KSWAPD);
+			gfp |= __GFP_IO | __GFP_WAIT;
+
+			i915_gem_shrink_all(dev_priv);
+			page = shmem_read_mapping_page_gfp(mapping, i, gfp);
+			if (IS_ERR(page))
+				goto err_pages;
+
+			gfp |= __GFP_NORETRY | __GFP_NOWARN | __GFP_NO_KSWAPD;
+			gfp &= ~(__GFP_IO | __GFP_WAIT);
+		}
 
 		obj->pages[i] = page;
 	}
@@ -1434,6 +1526,7 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj,
 	if (i915_gem_object_needs_bit17_swizzle(obj))
 		i915_gem_object_do_bit_17_swizzle(obj);
 
+	list_add_tail(&obj->gtt_list, &dev_priv->mm.unbound_list);
 	return 0;
 
 err_pages:
@@ -1698,6 +1791,7 @@ void i915_gem_reset(struct drm_device *dev)
 		obj->base.read_domains &= ~I915_GEM_GPU_DOMAINS;
 	}
 
+
 	/* The fence registers are invalidated so clear them out */
 	i915_gem_reset_fences(dev);
 }
@@ -2209,22 +2303,6 @@ i915_gem_object_unbind(struct drm_i915_gem_object *obj)
 
 	i915_gem_object_finish_gtt(obj);
 
-	/* Move the object to the CPU domain to ensure that
-	 * any possible CPU writes while it's not in the GTT
-	 * are flushed when we go to remap it.
-	 */
-	if (ret == 0)
-		ret = i915_gem_object_set_to_cpu_domain(obj, 1);
-	if (ret == -ERESTARTSYS)
-		return ret;
-	if (ret) {
-		/* In the event of a disaster, abandon all caches and
-		 * hope for the best.
-		 */
-		i915_gem_clflush_object(obj);
-		obj->base.read_domains = obj->base.write_domain = I915_GEM_DOMAIN_CPU;
-	}
-
 	/* release the fence reg _after_ flushing */
 	ret = i915_gem_object_put_fence(obj);
 	if (ret)
@@ -2240,10 +2318,8 @@ i915_gem_object_unbind(struct drm_i915_gem_object *obj)
 	}
 	i915_gem_gtt_finish_object(obj);
 
-	i915_gem_object_put_pages_gtt(obj);
-
-	list_del_init(&obj->gtt_list);
-	list_del_init(&obj->mm_list);
+	list_del(&obj->mm_list);
+	list_move_tail(&obj->gtt_list, &dev_priv->mm.unbound_list);
 	/* Avoid an unnecessary call to unbind on rebind. */
 	obj->map_and_fenceable = true;
 
@@ -2251,10 +2327,7 @@ i915_gem_object_unbind(struct drm_i915_gem_object *obj)
 	obj->gtt_space = NULL;
 	obj->gtt_offset = 0;
 
-	if (i915_gem_object_is_purgeable(obj))
-		i915_gem_object_truncate(obj);
-
-	return ret;
+	return 0;
 }
 
 static int i915_ring_idle(struct intel_ring_buffer *ring)
@@ -2667,7 +2740,6 @@ i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj,
 	struct drm_device *dev = obj->base.dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct drm_mm_node *free_space;
-	gfp_t gfpmask = __GFP_NORETRY | __GFP_NOWARN;
 	u32 size, fence_size, fence_alignment, unfenced_alignment;
 	bool mappable, fenceable;
 	int ret;
@@ -2707,6 +2779,10 @@ i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj,
 		return -E2BIG;
 	}
 
+	ret = i915_gem_object_get_pages_gtt(obj);
+	if (ret)
+		return ret;
+
  search_free:
 	if (map_and_fenceable)
 		free_space =
@@ -2733,9 +2809,6 @@ i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj,
 							 false);
 	}
 	if (obj->gtt_space == NULL) {
-		/* If the gtt is empty and we're still having trouble
-		 * fitting our object in, we're out of memory.
-		 */
 		ret = i915_gem_evict_something(dev, size, alignment,
 					       obj->cache_level,
 					       map_and_fenceable);
@@ -2752,55 +2825,20 @@ i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj,
 		return -EINVAL;
 	}
 
-	ret = i915_gem_object_get_pages_gtt(obj, gfpmask);
-	if (ret) {
-		drm_mm_put_block(obj->gtt_space);
-		obj->gtt_space = NULL;
-
-		if (ret == -ENOMEM) {
-			/* first try to reclaim some memory by clearing the GTT */
-			ret = i915_gem_evict_everything(dev, false);
-			if (ret) {
-				/* now try to shrink everyone else */
-				if (gfpmask) {
-					gfpmask = 0;
-					goto search_free;
-				}
-
-				return -ENOMEM;
-			}
-
-			goto search_free;
-		}
-
-		return ret;
-	}
 
 	ret = i915_gem_gtt_prepare_object(obj);
 	if (ret) {
-		i915_gem_object_put_pages_gtt(obj);
 		drm_mm_put_block(obj->gtt_space);
 		obj->gtt_space = NULL;
-
-		if (i915_gem_evict_everything(dev, false))
-			return ret;
-
-		goto search_free;
+		return ret;
 	}
 
 	if (!dev_priv->mm.aliasing_ppgtt)
 		i915_gem_gtt_bind_object(obj, obj->cache_level);
 
-	list_add_tail(&obj->gtt_list, &dev_priv->mm.gtt_list);
+	list_move_tail(&obj->gtt_list, &dev_priv->mm.bound_list);
 	list_add_tail(&obj->mm_list, &dev_priv->mm.inactive_list);
 
-	/* Assert that the object is not currently in any GPU domain. As it
-	 * wasn't in the GTT, there shouldn't be any way it could have been in
-	 * a GPU cache
-	 */
-	BUG_ON(obj->base.read_domains & I915_GEM_GPU_DOMAINS);
-	BUG_ON(obj->base.write_domain & I915_GEM_GPU_DOMAINS);
-
 	obj->gtt_offset = obj->gtt_space->start;
 
 	fenceable =
@@ -3464,9 +3502,8 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void *data,
 	if (obj->madv != __I915_MADV_PURGED)
 		obj->madv = args->madv;
 
-	/* if the object is no longer bound, discard its backing storage */
-	if (i915_gem_object_is_purgeable(obj) &&
-	    obj->gtt_space == NULL)
+	/* if the object is no longer attached, discard its backing storage */
+	if (i915_gem_object_is_purgeable(obj) && obj->pages == NULL)
 		i915_gem_object_truncate(obj);
 
 	args->retained = obj->madv != __I915_MADV_PURGED;
@@ -3573,6 +3610,7 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 		dev_priv->mm.interruptible = was_interruptible;
 	}
 
+	i915_gem_object_put_pages_gtt(obj);
 	if (obj->base.map_list.map)
 		drm_gem_free_mmap_offset(&obj->base);
 
@@ -3605,7 +3643,7 @@ i915_gem_idle(struct drm_device *dev)
 
 	/* Under UMS, be paranoid and evict. */
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
-		i915_gem_evict_everything(dev, false);
+		i915_gem_evict_everything(dev);
 
 	i915_gem_reset_fences(dev);
 
@@ -3963,8 +4001,9 @@ i915_gem_load(struct drm_device *dev)
 
 	INIT_LIST_HEAD(&dev_priv->mm.active_list);
 	INIT_LIST_HEAD(&dev_priv->mm.inactive_list);
+	INIT_LIST_HEAD(&dev_priv->mm.unbound_list);
+	INIT_LIST_HEAD(&dev_priv->mm.bound_list);
 	INIT_LIST_HEAD(&dev_priv->mm.fence_list);
-	INIT_LIST_HEAD(&dev_priv->mm.gtt_list);
 	for (i = 0; i < I915_NUM_RINGS; i++)
 		init_ring_lists(&dev_priv->ring[i]);
 	for (i = 0; i < I915_MAX_NUM_FENCES; i++)
@@ -4208,13 +4247,6 @@ void i915_gem_release(struct drm_device *dev, struct drm_file *file)
 	spin_unlock(&file_priv->mm.lock);
 }
 
-static int
-i915_gpu_is_active(struct drm_device *dev)
-{
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	return !list_empty(&dev_priv->mm.active_list);
-}
-
 static int
 i915_gem_inactive_shrink(struct shrinker *shrinker, struct shrink_control *sc)
 {
@@ -4223,60 +4255,26 @@ i915_gem_inactive_shrink(struct shrinker *shrinker, struct shrink_control *sc)
 			     struct drm_i915_private,
 			     mm.inactive_shrinker);
 	struct drm_device *dev = dev_priv->dev;
-	struct drm_i915_gem_object *obj, *next;
+	struct drm_i915_gem_object *obj;
 	int nr_to_scan = sc->nr_to_scan;
 	int cnt;
 
 	if (!mutex_trylock(&dev->struct_mutex))
 		return 0;
 
-	/* "fast-path" to count number of available objects */
-	if (nr_to_scan == 0) {
-		cnt = 0;
-		list_for_each_entry(obj,
-				    &dev_priv->mm.inactive_list,
-				    mm_list)
-			cnt++;
-		mutex_unlock(&dev->struct_mutex);
-		return cnt / 100 * sysctl_vfs_cache_pressure;
+	if (nr_to_scan) {
+		nr_to_scan -= i915_gem_purge(dev_priv, nr_to_scan);
+		if (nr_to_scan > 0)
+			i915_gem_shrink_all(dev_priv);
 	}
 
-rescan:
-	/* first scan for clean buffers */
-	i915_gem_retire_requests(dev);
-
-	list_for_each_entry_safe(obj, next,
-				 &dev_priv->mm.inactive_list,
-				 mm_list) {
-		if (i915_gem_object_is_purgeable(obj)) {
-			if (i915_gem_object_unbind(obj) == 0 &&
-			    --nr_to_scan == 0)
-				break;
-		}
-	}
-
-	/* second pass, evict/count anything still on the inactive list */
 	cnt = 0;
-	list_for_each_entry_safe(obj, next,
-				 &dev_priv->mm.inactive_list,
-				 mm_list) {
-		if (nr_to_scan &&
-		    i915_gem_object_unbind(obj) == 0)
-			nr_to_scan--;
-		else
-			cnt++;
-	}
+	list_for_each_entry(obj, &dev_priv->mm.unbound_list, gtt_list)
+		cnt += obj->base.size >> PAGE_SHIFT;
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, gtt_list)
+		if (obj->pin_count == 0)
+			cnt += obj->base.size >> PAGE_SHIFT;
 
-	if (nr_to_scan && i915_gpu_is_active(dev)) {
-		/*
-		 * We are desperate for pages, so as a last resort, wait
-		 * for the GPU to finish and discard whatever we can.
-		 * This has a dramatic impact to reduce the number of
-		 * OOM-killer events whilst running the GPU aggressively.
-		 */
-		if (i915_gpu_idle(dev) == 0)
-			goto rescan;
-	}
 	mutex_unlock(&dev->struct_mutex);
-	return cnt / 100 * sysctl_vfs_cache_pressure;
+	return cnt;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
index ceaad5af01a4..43c95307f99e 100644
--- a/drivers/gpu/drm/i915/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
@@ -33,7 +33,7 @@ static struct sg_table *i915_gem_map_dma_buf(struct dma_buf_attachment *attachme
 	struct drm_i915_gem_object *obj = attachment->dmabuf->priv;
 	struct drm_device *dev = obj->base.dev;
 	int npages = obj->base.size / PAGE_SIZE;
-	struct sg_table *sg = NULL;
+	struct sg_table *sg;
 	int ret;
 	int nents;
 
@@ -41,10 +41,10 @@ static struct sg_table *i915_gem_map_dma_buf(struct dma_buf_attachment *attachme
 	if (ret)
 		return ERR_PTR(ret);
 
-	if (!obj->pages) {
-		ret = i915_gem_object_get_pages_gtt(obj, __GFP_NORETRY | __GFP_NOWARN);
-		if (ret)
-			goto out;
+	ret = i915_gem_object_get_pages_gtt(obj);
+	if (ret) {
+		sg = ERR_PTR(ret);
+		goto out;
 	}
 
 	/* link the pages into an SG then map the sg */
@@ -89,12 +89,10 @@ static void *i915_gem_dmabuf_vmap(struct dma_buf *dma_buf)
 		goto out_unlock;
 	}
 
-	if (!obj->pages) {
-		ret = i915_gem_object_get_pages_gtt(obj, __GFP_NORETRY | __GFP_NOWARN);
-		if (ret) {
-			mutex_unlock(&dev->struct_mutex);
-			return ERR_PTR(ret);
-		}
+	ret = i915_gem_object_get_pages_gtt(obj);
+	if (ret) {
+		mutex_unlock(&dev->struct_mutex);
+		return ERR_PTR(ret);
 	}
 
 	obj->dma_buf_vmapping = vmap(obj->pages, obj->base.size / PAGE_SIZE, 0, PAGE_KERNEL);
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 7279c31d4a9a..74635da7c498 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -148,7 +148,7 @@ found:
 }
 
 int
-i915_gem_evict_everything(struct drm_device *dev, bool purgeable_only)
+i915_gem_evict_everything(struct drm_device *dev)
 {
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj, *next;
@@ -160,7 +160,7 @@ i915_gem_evict_everything(struct drm_device *dev, bool purgeable_only)
 	if (lists_empty)
 		return -ENOSPC;
 
-	trace_i915_gem_evict_everything(dev, purgeable_only);
+	trace_i915_gem_evict_everything(dev);
 
 	/* The gpu_idle will flush everything in the write domain to the
 	 * active list. Then we must move everything off the active list
@@ -174,12 +174,9 @@ i915_gem_evict_everything(struct drm_device *dev, bool purgeable_only)
 
 	/* Having flushed everything, unbind() should never raise an error */
 	list_for_each_entry_safe(obj, next,
-				 &dev_priv->mm.inactive_list, mm_list) {
-		if (!purgeable_only || obj->madv != I915_MADV_WILLNEED) {
-			if (obj->pin_count == 0)
-				WARN_ON(i915_gem_object_unbind(obj));
-		}
-	}
+				 &dev_priv->mm.inactive_list, mm_list)
+		if (obj->pin_count == 0)
+			WARN_ON(i915_gem_object_unbind(obj));
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index afb312ee050c..834a636b44f0 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -502,17 +502,12 @@ i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
 			}
 		}
 
-		if (ret != -ENOSPC || retry > 1)
+		if (ret != -ENOSPC || retry++)
 			return ret;
 
-		/* First attempt, just clear anything that is purgeable.
-		 * Second attempt, clear the entire GTT.
-		 */
-		ret = i915_gem_evict_everything(ring->dev, retry == 0);
+		ret = i915_gem_evict_everything(ring->dev);
 		if (ret)
 			return ret;
-
-		retry++;
 	} while (1);
 
 err:
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 3b3b731a17c3..8329a14862ed 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -348,7 +348,7 @@ void i915_gem_restore_gtt_mappings(struct drm_device *dev)
 	intel_gtt_clear_range(dev_priv->mm.gtt_start / PAGE_SIZE,
 			      (dev_priv->mm.gtt_end - dev_priv->mm.gtt_start) / PAGE_SIZE);
 
-	list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list) {
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, gtt_list) {
 		i915_gem_clflush_object(obj);
 		i915_gem_gtt_bind_object(obj, obj->cache_level);
 	}
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index a61b41a8c607..002dceebbeab 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -1221,7 +1221,7 @@ static void i915_capture_error_state(struct drm_device *dev)
 	list_for_each_entry(obj, &dev_priv->mm.active_list, mm_list)
 		i++;
 	error->active_bo_count = i;
-	list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list)
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, gtt_list)
 		if (obj->pin_count)
 			i++;
 	error->pinned_bo_count = i - error->active_bo_count;
@@ -1246,7 +1246,7 @@ static void i915_capture_error_state(struct drm_device *dev)
 		error->pinned_bo_count =
 			capture_pinned_bo(error->pinned_bo,
 					  error->pinned_bo_count,
-					  &dev_priv->mm.gtt_list);
+					  &dev_priv->mm.bound_list);
 
 	do_gettimeofday(&error->time);
 
diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index fe90b3a84a6d..3c4093d91f60 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -214,22 +214,18 @@ TRACE_EVENT(i915_gem_evict,
 );
 
 TRACE_EVENT(i915_gem_evict_everything,
-	    TP_PROTO(struct drm_device *dev, bool purgeable),
-	    TP_ARGS(dev, purgeable),
+	    TP_PROTO(struct drm_device *dev),
+	    TP_ARGS(dev),
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
-			     __field(bool, purgeable)
 			    ),
 
 	    TP_fast_assign(
 			   __entry->dev = dev->primary->index;
-			   __entry->purgeable = purgeable;
 			  ),
 
-	    TP_printk("dev=%d%s",
-		      __entry->dev,
-		      __entry->purgeable ? ", purgeable only" : "")
+	    TP_printk("dev=%d", __entry->dev)
 );
 
 TRACE_EVENT(i915_gem_ring_dispatch,

From c4670ad080a1fe60d49f850e11f64a41db525496 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 20 Aug 2012 10:23:27 +0100
Subject: [PATCH 21/41] drm/i915: Add some sanity checks to unbound tracking

A pair of universally true checks that just need to be put in the right
place depending on where in the patch sequence you go. Note that
i915_gem_object_put_pages_gtt() already gains the
BUG_ON(obj->gtt_space), but on reflection that needed to migrate to
put_pages().

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 462a8f2ad3a2..a549be5f535b 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1378,6 +1378,8 @@ i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj)
 	int page_count = obj->base.size / PAGE_SIZE;
 	int ret, i;
 
+	BUG_ON(obj->gtt_space);
+
 	if (obj->pages == NULL)
 		return 0;
 
@@ -2293,6 +2295,8 @@ i915_gem_object_unbind(struct drm_i915_gem_object *obj)
 	if (obj->pin_count)
 		return -EBUSY;
 
+	BUG_ON(obj->pages == NULL);
+
 	ret = i915_gem_object_finish_gpu(obj);
 	if (ret)
 		return ret;

From b7abb7144356536c41a81eda2bbfd46027e529f6 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 20 Aug 2012 11:33:30 +0200
Subject: [PATCH 22/41] drm/i915: Show (count, size) of purgeable objects in
 i915_gem_objects

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_debugfs.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 608d3ae2b5ba..5548b02f40bb 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -197,8 +197,8 @@ static int i915_gem_object_info(struct seq_file *m, void* data)
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 count, mappable_count;
-	size_t size, mappable_size;
+	u32 count, mappable_count, purgeable_count;
+	size_t size, mappable_size, purgeable_size;
 	struct drm_i915_gem_object *obj;
 	int ret;
 
@@ -225,9 +225,12 @@ static int i915_gem_object_info(struct seq_file *m, void* data)
 	seq_printf(m, "  %u [%u] inactive objects, %zu [%zu] bytes\n",
 		   count, mappable_count, size, mappable_size);
 
-	size = count = 0;
-	list_for_each_entry(obj, &dev_priv->mm.unbound_list, gtt_list)
+	size = count = purgeable_size = purgeable_count = 0;
+	list_for_each_entry(obj, &dev_priv->mm.unbound_list, gtt_list) {
 		size += obj->base.size, ++count;
+		if (obj->madv == I915_MADV_DONTNEED)
+			purgeable_size += obj->base.size, ++purgeable_count;
+	}
 	seq_printf(m, "%u unbound objects, %zu bytes\n", count, size);
 
 	size = count = mappable_size = mappable_count = 0;
@@ -240,7 +243,13 @@ static int i915_gem_object_info(struct seq_file *m, void* data)
 			mappable_size += obj->gtt_space->size;
 			++mappable_count;
 		}
+		if (obj->madv == I915_MADV_DONTNEED) {
+			purgeable_size += obj->base.size;
+			++purgeable_count;
+		}
 	}
+	seq_printf(m, "%u purgeable objects, %zu bytes\n",
+		   purgeable_count, purgeable_size);
 	seq_printf(m, "%u pinned mappable objects, %zu bytes\n",
 		   mappable_count, mappable_size);
 	seq_printf(m, "%u fault mappable objects, %zu bytes\n",

From c110a6d728d0243a755973d34e363952b74af984 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 11 Aug 2012 15:41:02 +0100
Subject: [PATCH 23/41] drm/i915: Show pin count in debugfs

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_debugfs.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 5548b02f40bb..204c19a49969 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -118,6 +118,8 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 		   obj->madv == I915_MADV_DONTNEED ? " purgeable" : "");
 	if (obj->base.name)
 		seq_printf(m, " (name: %d)", obj->base.name);
+	if (obj->pin_count)
+		seq_printf(m, " (pinned x %d)", obj->pin_count);
 	if (obj->fence_reg != I915_FENCE_REG_NONE)
 		seq_printf(m, " (fence: %d)", obj->fence_reg);
 	if (obj->gtt_space != NULL)

From d8cb5086695dcdd076e911fc298a5a6701497371 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 11 Aug 2012 15:41:03 +0100
Subject: [PATCH 24/41] drm/i915: Try harder to allocate an mmap_offset

Given the persistence of an offset for the lifetime of an object, itis
easy to contemplate how the mmap space becomes badly fragmented to the
point that further allocations fail with ENOSPC. Our only recourse at
this point is to try to purge the objects to release some space and
reattempt the allocation.

References: https://bugs.freedesktop.org/show_bug.cgi?id=39552
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem.c | 50 +++++++++++++++++++++++++++------
 1 file changed, 41 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index a549be5f535b..e1ec587c42c8 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1276,6 +1276,42 @@ i915_gem_get_unfenced_gtt_alignment(struct drm_device *dev,
 	return i915_gem_get_gtt_size(dev, size, tiling_mode);
 }
 
+static int i915_gem_object_create_mmap_offset(struct drm_i915_gem_object *obj)
+{
+	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
+	int ret;
+
+	if (obj->base.map_list.map)
+		return 0;
+
+	ret = drm_gem_create_mmap_offset(&obj->base);
+	if (ret != -ENOSPC)
+		return ret;
+
+	/* Badly fragmented mmap space? The only way we can recover
+	 * space is by destroying unwanted objects. We can't randomly release
+	 * mmap_offsets as userspace expects them to be persistent for the
+	 * lifetime of the objects. The closest we can is to release the
+	 * offsets on purgeable objects by truncating it and marking it purged,
+	 * which prevents userspace from ever using that object again.
+	 */
+	i915_gem_purge(dev_priv, obj->base.size >> PAGE_SHIFT);
+	ret = drm_gem_create_mmap_offset(&obj->base);
+	if (ret != -ENOSPC)
+		return ret;
+
+	i915_gem_shrink_all(dev_priv);
+	return drm_gem_create_mmap_offset(&obj->base);
+}
+
+static void i915_gem_object_free_mmap_offset(struct drm_i915_gem_object *obj)
+{
+	if (!obj->base.map_list.map)
+		return;
+
+	drm_gem_free_mmap_offset(&obj->base);
+}
+
 int
 i915_gem_mmap_gtt(struct drm_file *file,
 		  struct drm_device *dev,
@@ -1307,11 +1343,9 @@ i915_gem_mmap_gtt(struct drm_file *file,
 		goto out;
 	}
 
-	if (!obj->base.map_list.map) {
-		ret = drm_gem_create_mmap_offset(&obj->base);
-		if (ret)
-			goto out;
-	}
+	ret = i915_gem_object_create_mmap_offset(obj);
+	if (ret)
+		goto out;
 
 	*offset = (u64)obj->base.map_list.hash.key << PAGE_SHIFT;
 
@@ -1360,8 +1394,7 @@ i915_gem_object_truncate(struct drm_i915_gem_object *obj)
 	inode = obj->base.filp->f_path.dentry->d_inode;
 	shmem_truncate_range(inode, 0, (loff_t)-1);
 
-	if (obj->base.map_list.map)
-		drm_gem_free_mmap_offset(&obj->base);
+	i915_gem_object_free_mmap_offset(obj);
 
 	obj->madv = __I915_MADV_PURGED;
 }
@@ -3615,8 +3648,7 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 	}
 
 	i915_gem_object_put_pages_gtt(obj);
-	if (obj->base.map_list.map)
-		drm_gem_free_mmap_offset(&obj->base);
+	i915_gem_object_free_mmap_offset(obj);
 
 	drm_gem_object_release(&obj->base);
 	i915_gem_info_remove_obj(dev_priv, obj->base.size);

From de9932d13c95c4125d2ed8703f5574fcc108bbcf Mon Sep 17 00:00:00 2001
From: "Xu, Anhua" <anhua.xu@intel.com>
Date: Tue, 21 Aug 2012 03:06:02 +0000
Subject: [PATCH 25/41] drm/i915: fix reassignment of variable "intel_dp->DP"

In intel_dp_mode_set we OR in the exact same bits twice at the same
spot. Kill one of the redundant assignments.

This little regression was introduced by:
commit 417e822deee1d2bcd8a8a60660c40a0903713f2b
Author: Keith Packard <keithp@keithp.com>
Date:   Tue Nov 1 19:54:11 2011 -0700

    drm/i915: Treat PCH eDP like DP in most places

	PCH eDP has many of the same needs as regular PCH DP connections,
	including the DP_CTl bit settings, the TRANS_DP_CTL register.

The reachable tag for this commit is: v3.1-5461-g417e822

Signed-off-by: Anhua Xu <anhua.xu@intel.com>
[danvet: Improved the commit message somewhat and ensured the diff is
clearer.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_dp.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 977d9d216c73..1ab371b8be01 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -850,10 +850,8 @@ intel_dp_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode,
 	 * supposed to be read-only.
 	 */
 	intel_dp->DP = I915_READ(intel_dp->output_reg) & DP_DETECTED;
-	intel_dp->DP |=  DP_VOLTAGE_0_4 | DP_PRE_EMPHASIS_0;
 
 	/* Handle DP bits in common between all three register formats */
-
 	intel_dp->DP |= DP_VOLTAGE_0_4 | DP_PRE_EMPHASIS_0;
 
 	switch (intel_dp->lane_count) {

From 44f46b4222a9ba5b21dcdf13f93dc53f48768d2b Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 21 Jun 2012 13:19:59 +0300
Subject: [PATCH 26/41] drm/i915: Cantiga+ cannot handle a hsync front porch of
 0

This addresses WaPruneModeWithIncorrectHsyncOffset.

Bugzilla: http://bugs.freedesktop.org/show_bug.cgi?id=50236
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_display.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 42c57581c53f..778cbb88bddc 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -3602,6 +3602,13 @@ static bool intel_crtc_mode_fixup(struct drm_crtc *crtc,
 	if (!(adjusted_mode->private_flags & INTEL_MODE_CRTC_TIMINGS_SET))
 		drm_mode_set_crtcinfo(adjusted_mode, 0);
 
+	/* WaPruneModeWithIncorrectHsyncOffset: Cantiga+ cannot handle modes
+	 * with a hsync front porch of 0.
+	 */
+	if ((INTEL_INFO(dev)->gen > 4 || IS_G4X(dev)) &&
+		adjusted_mode->hsync_start == adjusted_mode->hdisplay)
+		return false;
+
 	return true;
 }
 

From 71e172e8d141fc167b1cca28271fd39c8cf63ed3 Mon Sep 17 00:00:00 2001
From: Ben Widawsky <ben@bwidawsk.net>
Date: Mon, 20 Aug 2012 16:15:13 -0700
Subject: [PATCH 27/41] drm/i915: Add ERR_INT to gen7 error state

ERR_INT can generate interrupts. However since most of the conditions seem
quite fatal the patch opts to simply report it in error state instead of
adding more complexity to the interrupt handler for little gain (the
bits are sticky anyway).

Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Tested-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Antti Koskipaa <antti.koskipaa@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_debugfs.c | 3 +++
 drivers/gpu/drm/i915/i915_drv.h     | 1 +
 drivers/gpu/drm/i915/i915_irq.c     | 3 +++
 drivers/gpu/drm/i915/i915_reg.h     | 1 +
 4 files changed, 8 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 204c19a49969..d3003937ddf5 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -719,6 +719,9 @@ static int i915_error_state(struct seq_file *m, void *unused)
 		seq_printf(m, "DONE_REG: 0x%08x\n", error->done_reg);
 	}
 
+	if (INTEL_INFO(dev)->gen == 7)
+		seq_printf(m, "ERR_INT: 0x%08x\n", error->err_int);
+
 	for_each_ring(ring, dev_priv, i)
 		i915_ring_error_state(m, dev, error, i);
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index a2382a175d92..cbd3cd0747b9 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -196,6 +196,7 @@ struct drm_i915_error_state {
 	u32 cpu_ring_head[I915_NUM_RINGS];
 	u32 cpu_ring_tail[I915_NUM_RINGS];
 	u32 error; /* gen6+ */
+	u32 err_int; /* gen7 */
 	u32 instpm[I915_NUM_RINGS];
 	u32 instps[I915_NUM_RINGS];
 	u32 instdone1;
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 002dceebbeab..36330296932a 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -1210,6 +1210,9 @@ static void i915_capture_error_state(struct drm_device *dev)
 		error->done_reg = I915_READ(DONE_REG);
 	}
 
+	if (INTEL_INFO(dev)->gen == 7)
+		error->err_int = I915_READ(GEN7_ERR_INT);
+
 	i915_gem_record_fences(dev, error);
 	i915_gem_record_rings(dev, error);
 
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index fd6a26a49b42..a20885c0f35f 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -500,6 +500,7 @@
 #define DMA_FADD_I8XX	0x020d0
 
 #define ERROR_GEN6	0x040a0
+#define GEN7_ERR_INT	0x44040
 
 /* GM45+ chicken bits -- debug workaround bits that may be required
  * for various sorts of correct behavior.  The top 16 bits of each are

From b4c145c1d245c2cc19754dbe4b718f5a48755993 Mon Sep 17 00:00:00 2001
From: Ben Widawsky <ben@bwidawsk.net>
Date: Mon, 20 Aug 2012 16:15:14 -0700
Subject: [PATCH 28/41] drm/i915: Find unclaimed MMIO writes.

ERR_INT on HSW will display unclaimed MMIO accesses. This can be either
the result of a driver bug writing to an invalid addresses, or the
result of RC6.

Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Tested-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Antti Koskipaa <antti.koskipaa@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_drv.c | 4 ++++
 drivers/gpu/drm/i915/i915_reg.h | 1 +
 2 files changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 7ebb13b65133..7f3863fb138c 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -1174,6 +1174,10 @@ void i915_write##x(struct drm_i915_private *dev_priv, u32 reg, u##x val) { \
 	if (unlikely(__fifo_ret)) { \
 		gen6_gt_check_fifodbg(dev_priv); \
 	} \
+	if (IS_HASWELL(dev_priv->dev) && (I915_READ_NOTRACE(GEN7_ERR_INT) & ERR_INT_MMIO_UNCLAIMED)) { \
+		DRM_ERROR("Unclaimed write to %x\n", reg); \
+		writel(ERR_INT_MMIO_UNCLAIMED, dev_priv->regs + GEN7_ERR_INT);	\
+	} \
 }
 __i915_write(8, b)
 __i915_write(16, w)
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index a20885c0f35f..f27e9982bed0 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -501,6 +501,7 @@
 
 #define ERROR_GEN6	0x040a0
 #define GEN7_ERR_INT	0x44040
+#define   ERR_INT_MMIO_UNCLAIMED (1<<13)
 
 /* GM45+ chicken bits -- debug workaround bits that may be required
  * for various sorts of correct behavior.  The top 16 bits of each are

From 53621860c38caff37fff99ba5f0b817511907bc4 Mon Sep 17 00:00:00 2001
From: Tejun Heo <tj@kernel.org>
Date: Wed, 22 Aug 2012 16:40:57 -0700
Subject: [PATCH 29/41] i915: use alloc_ordered_workqueue() instead of explicit
 UNBOUND w/ max_active = 1

This is an equivalent conversion and will ease scheduled removal of
WQ_NON_REENTRANT.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_dma.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 2cba7b4a04ed..2c09900e3267 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -1558,11 +1558,9 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	 *
 	 * All tasks on the workqueue are expected to acquire the dev mutex
 	 * so there is no point in running more than one instance of the
-	 * workqueue at any time: max_active = 1 and NON_REENTRANT.
+	 * workqueue at any time.  Use an ordered one.
 	 */
-	dev_priv->wq = alloc_workqueue("i915",
-				       WQ_UNBOUND | WQ_NON_REENTRANT,
-				       1);
+	dev_priv->wq = alloc_ordered_workqueue("i915", 0);
 	if (dev_priv->wq == NULL) {
 		DRM_ERROR("Failed to create our workqueue.\n");
 		ret = -ENOMEM;

From 86a1ee26bb60e1ab8984e92f0e9186c354670aed Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 11 Aug 2012 15:41:04 +0100
Subject: [PATCH 30/41] drm/i915: Only pwrite through the GTT if there is space
 in the aperture

Avoid stalling and waiting for the GPU by checking to see if there is
sufficient inactive space in the aperture for us to bind the buffer
prior to writing through the GTT. If there is inadequate space we will
have to stall waiting for the GPU, and incur overheads moving objects
about. Instead, only incur the clflush overhead on the target object by
writing through shmem.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_drv.h            |  6 +++--
 drivers/gpu/drm/i915/i915_gem.c            | 29 ++++++++++++----------
 drivers/gpu/drm/i915/i915_gem_context.c    |  4 +--
 drivers/gpu/drm/i915/i915_gem_evict.c      |  6 ++++-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |  2 +-
 drivers/gpu/drm/i915/intel_overlay.c       |  2 +-
 drivers/gpu/drm/i915/intel_pm.c            |  2 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c    |  6 ++---
 8 files changed, 33 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index cbd3cd0747b9..06a88be067f1 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1307,7 +1307,8 @@ struct drm_i915_gem_object *i915_gem_alloc_object(struct drm_device *dev,
 void i915_gem_free_object(struct drm_gem_object *obj);
 int __must_check i915_gem_object_pin(struct drm_i915_gem_object *obj,
 				     uint32_t alignment,
-				     bool map_and_fenceable);
+				     bool map_and_fenceable,
+				     bool nonblocking);
 void i915_gem_object_unpin(struct drm_i915_gem_object *obj);
 int __must_check i915_gem_object_unbind(struct drm_i915_gem_object *obj);
 void i915_gem_release_mmap(struct drm_i915_gem_object *obj);
@@ -1454,7 +1455,8 @@ void i915_gem_init_global_gtt(struct drm_device *dev,
 int __must_check i915_gem_evict_something(struct drm_device *dev, int min_size,
 					  unsigned alignment,
 					  unsigned cache_level,
-					  bool mappable);
+					  bool mappable,
+					  bool nonblock);
 int i915_gem_evict_everything(struct drm_device *dev);
 
 /* i915_gem_stolen.c */
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index e1ec587c42c8..1ec5f7ef2cc2 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -41,7 +41,8 @@ static void i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *o
 static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj);
 static __must_check int i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj,
 						    unsigned alignment,
-						    bool map_and_fenceable);
+						    bool map_and_fenceable,
+						    bool nonblocking);
 static int i915_gem_phys_pwrite(struct drm_device *dev,
 				struct drm_i915_gem_object *obj,
 				struct drm_i915_gem_pwrite *args,
@@ -609,7 +610,7 @@ i915_gem_gtt_pwrite_fast(struct drm_device *dev,
 	char __user *user_data;
 	int page_offset, page_length, ret;
 
-	ret = i915_gem_object_pin(obj, 0, true);
+	ret = i915_gem_object_pin(obj, 0, true, true);
 	if (ret)
 		goto out;
 
@@ -925,10 +926,8 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 		goto out;
 	}
 
-	if (obj->gtt_space &&
-	    obj->cache_level == I915_CACHE_NONE &&
+	if (obj->cache_level == I915_CACHE_NONE &&
 	    obj->tiling_mode == I915_TILING_NONE &&
-	    obj->map_and_fenceable &&
 	    obj->base.write_domain != I915_GEM_DOMAIN_CPU) {
 		ret = i915_gem_gtt_pwrite_fast(dev, obj, args, file);
 		/* Note that the gtt paths might fail with non-page-backed user
@@ -936,7 +935,7 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 		 * textures). Fallback to the shmem path in that case. */
 	}
 
-	if (ret == -EFAULT)
+	if (ret == -EFAULT || ret == -ENOSPC)
 		ret = i915_gem_shmem_pwrite(dev, obj, args, file);
 
 out:
@@ -1115,7 +1114,7 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 			goto unlock;
 	}
 	if (!obj->gtt_space) {
-		ret = i915_gem_object_bind_to_gtt(obj, 0, true);
+		ret = i915_gem_object_bind_to_gtt(obj, 0, true, false);
 		if (ret)
 			goto unlock;
 
@@ -2772,7 +2771,8 @@ static void i915_gem_verify_gtt(struct drm_device *dev)
 static int
 i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj,
 			    unsigned alignment,
-			    bool map_and_fenceable)
+			    bool map_and_fenceable,
+			    bool nonblocking)
 {
 	struct drm_device *dev = obj->base.dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
@@ -2848,7 +2848,8 @@ i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj,
 	if (obj->gtt_space == NULL) {
 		ret = i915_gem_evict_something(dev, size, alignment,
 					       obj->cache_level,
-					       map_and_fenceable);
+					       map_and_fenceable,
+					       nonblocking);
 		if (ret)
 			return ret;
 
@@ -3188,7 +3189,7 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 	 * (e.g. libkms for the bootup splash), we have to ensure that we
 	 * always use map_and_fenceable for all scanout buffers.
 	 */
-	ret = i915_gem_object_pin(obj, alignment, true);
+	ret = i915_gem_object_pin(obj, alignment, true, false);
 	if (ret)
 		return ret;
 
@@ -3325,7 +3326,8 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
 int
 i915_gem_object_pin(struct drm_i915_gem_object *obj,
 		    uint32_t alignment,
-		    bool map_and_fenceable)
+		    bool map_and_fenceable,
+		    bool nonblocking)
 {
 	int ret;
 
@@ -3349,7 +3351,8 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 
 	if (obj->gtt_space == NULL) {
 		ret = i915_gem_object_bind_to_gtt(obj, alignment,
-						  map_and_fenceable);
+						  map_and_fenceable,
+						  nonblocking);
 		if (ret)
 			return ret;
 	}
@@ -3407,7 +3410,7 @@ i915_gem_pin_ioctl(struct drm_device *dev, void *data,
 	obj->user_pin_count++;
 	obj->pin_filp = file;
 	if (obj->user_pin_count == 1) {
-		ret = i915_gem_object_pin(obj, args->alignment, true);
+		ret = i915_gem_object_pin(obj, args->alignment, true, false);
 		if (ret)
 			goto out;
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 5c2d354cebbd..4aa7ecf77ede 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -221,7 +221,7 @@ static int create_default_context(struct drm_i915_private *dev_priv)
 	 * default context.
 	 */
 	dev_priv->ring[RCS].default_context = ctx;
-	ret = i915_gem_object_pin(ctx->obj, CONTEXT_ALIGN, false);
+	ret = i915_gem_object_pin(ctx->obj, CONTEXT_ALIGN, false, false);
 	if (ret)
 		goto err_destroy;
 
@@ -374,7 +374,7 @@ static int do_switch(struct i915_hw_context *to)
 	if (from_obj == to->obj)
 		return 0;
 
-	ret = i915_gem_object_pin(to->obj, CONTEXT_ALIGN, false);
+	ret = i915_gem_object_pin(to->obj, CONTEXT_ALIGN, false, false);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 74635da7c498..a2d8acde8550 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -45,7 +45,7 @@ mark_free(struct drm_i915_gem_object *obj, struct list_head *unwind)
 int
 i915_gem_evict_something(struct drm_device *dev, int min_size,
 			 unsigned alignment, unsigned cache_level,
-			 bool mappable)
+			 bool mappable, bool nonblocking)
 {
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct list_head eviction_list, unwind_list;
@@ -92,12 +92,16 @@ i915_gem_evict_something(struct drm_device *dev, int min_size,
 			goto found;
 	}
 
+	if (nonblocking)
+		goto none;
+
 	/* Now merge in the soon-to-be-expired objects... */
 	list_for_each_entry(obj, &dev_priv->mm.active_list, mm_list) {
 		if (mark_free(obj, &unwind_list))
 			goto found;
 	}
 
+none:
 	/* Nothing found, clean up and bail out! */
 	while (!list_empty(&unwind_list)) {
 		obj = list_first_entry(&unwind_list,
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 834a636b44f0..f7346d876551 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -354,7 +354,7 @@ pin_and_fence_object(struct drm_i915_gem_object *obj,
 		obj->tiling_mode != I915_TILING_NONE;
 	need_mappable = need_fence || need_reloc_mappable(obj);
 
-	ret = i915_gem_object_pin(obj, entry->alignment, need_mappable);
+	ret = i915_gem_object_pin(obj, entry->alignment, need_mappable, false);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index c0f48580405a..afd0f30ab882 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -1383,7 +1383,7 @@ void intel_setup_overlay(struct drm_device *dev)
 		}
 		overlay->flip_addr = reg_bo->phys_obj->handle->busaddr;
 	} else {
-		ret = i915_gem_object_pin(reg_bo, PAGE_SIZE, true);
+		ret = i915_gem_object_pin(reg_bo, PAGE_SIZE, true, false);
 		if (ret) {
 			DRM_ERROR("failed to pin overlay register bo\n");
 			goto out_free_bo;
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index c0407aa5baac..c0721ffb56a2 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -2138,7 +2138,7 @@ intel_alloc_context_page(struct drm_device *dev)
 		return NULL;
 	}
 
-	ret = i915_gem_object_pin(ctx, 4096, true);
+	ret = i915_gem_object_pin(ctx, 4096, true, false);
 	if (ret) {
 		DRM_ERROR("failed to pin power context: %d\n", ret);
 		goto err_unref;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index c828169c73ae..ac93643731aa 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -391,7 +391,7 @@ init_pipe_control(struct intel_ring_buffer *ring)
 
 	i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
 
-	ret = i915_gem_object_pin(obj, 4096, true);
+	ret = i915_gem_object_pin(obj, 4096, true, false);
 	if (ret)
 		goto err_unref;
 
@@ -979,7 +979,7 @@ static int init_status_page(struct intel_ring_buffer *ring)
 
 	i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
 
-	ret = i915_gem_object_pin(obj, 4096, true);
+	ret = i915_gem_object_pin(obj, 4096, true, false);
 	if (ret != 0) {
 		goto err_unref;
 	}
@@ -1036,7 +1036,7 @@ static int intel_init_ring_buffer(struct drm_device *dev,
 
 	ring->obj = obj;
 
-	ret = i915_gem_object_pin(obj, PAGE_SIZE, true);
+	ret = i915_gem_object_pin(obj, PAGE_SIZE, true, false);
 	if (ret)
 		goto err_unref;
 

From 4d6294bf77834276d41b51154a18663b0a8f2063 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 11 Aug 2012 15:41:05 +0100
Subject: [PATCH 31/41] drm/i915: Protect private gem objects from truncate
 (such as imported dmabuf)

If the object has no backing shmemfs filp, then we obviously cannot
perform a truncation operation upon it.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 1ec5f7ef2cc2..eb54a6d81517 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1385,6 +1385,11 @@ i915_gem_object_truncate(struct drm_i915_gem_object *obj)
 {
 	struct inode *inode;
 
+	i915_gem_object_free_mmap_offset(obj);
+
+	if (obj->base.filp == NULL)
+		return;
+
 	/* Our goal here is to return as much of the memory as
 	 * is possible back to the system as we are called from OOM.
 	 * To do this we must instruct the shmfs to drop all of its
@@ -1393,8 +1398,6 @@ i915_gem_object_truncate(struct drm_i915_gem_object *obj)
 	inode = obj->base.filp->f_path.dentry->d_inode;
 	shmem_truncate_range(inode, 0, (loff_t)-1);
 
-	i915_gem_object_free_mmap_offset(obj);
-
 	obj->madv = __I915_MADV_PURGED;
 }
 

From 0327d6ba998ca181013a5a1709701a6532a41972 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 11 Aug 2012 15:41:06 +0100
Subject: [PATCH 32/41] drm/i915: Extract general object init routine

As we wish to create specialised object constructions in the near
future that share the same basic GEM object struct, export the default
initializer.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_drv.h |  1 +
 drivers/gpu/drm/i915/i915_gem.c | 30 ++++++++++++++++++------------
 2 files changed, 19 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 06a88be067f1..9f4c27eab046 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1302,6 +1302,7 @@ int i915_gem_wait_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 void i915_gem_load(struct drm_device *dev);
 int i915_gem_init_object(struct drm_gem_object *obj);
+void i915_gem_object_init(struct drm_i915_gem_object *obj);
 struct drm_i915_gem_object *i915_gem_alloc_object(struct drm_device *dev,
 						  size_t size);
 void i915_gem_free_object(struct drm_gem_object *obj);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index eb54a6d81517..553175856e81 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3558,10 +3558,26 @@ unlock:
 	return ret;
 }
 
+void i915_gem_object_init(struct drm_i915_gem_object *obj)
+{
+	obj->base.driver_private = NULL;
+
+	INIT_LIST_HEAD(&obj->mm_list);
+	INIT_LIST_HEAD(&obj->gtt_list);
+	INIT_LIST_HEAD(&obj->ring_list);
+	INIT_LIST_HEAD(&obj->exec_list);
+
+	obj->fence_reg = I915_FENCE_REG_NONE;
+	obj->madv = I915_MADV_WILLNEED;
+	/* Avoid an unnecessary call to unbind on the first bind. */
+	obj->map_and_fenceable = true;
+
+	i915_gem_info_add_obj(obj->base.dev->dev_private, obj->base.size);
+}
+
 struct drm_i915_gem_object *i915_gem_alloc_object(struct drm_device *dev,
 						  size_t size)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj;
 	struct address_space *mapping;
 	u32 mask;
@@ -3585,7 +3601,7 @@ struct drm_i915_gem_object *i915_gem_alloc_object(struct drm_device *dev,
 	mapping = obj->base.filp->f_path.dentry->d_inode->i_mapping;
 	mapping_set_gfp_mask(mapping, mask);
 
-	i915_gem_info_add_obj(dev_priv, size);
+	i915_gem_object_init(obj);
 
 	obj->base.write_domain = I915_GEM_DOMAIN_CPU;
 	obj->base.read_domains = I915_GEM_DOMAIN_CPU;
@@ -3607,16 +3623,6 @@ struct drm_i915_gem_object *i915_gem_alloc_object(struct drm_device *dev,
 	} else
 		obj->cache_level = I915_CACHE_NONE;
 
-	obj->base.driver_private = NULL;
-	obj->fence_reg = I915_FENCE_REG_NONE;
-	INIT_LIST_HEAD(&obj->mm_list);
-	INIT_LIST_HEAD(&obj->gtt_list);
-	INIT_LIST_HEAD(&obj->ring_list);
-	INIT_LIST_HEAD(&obj->exec_list);
-	obj->madv = I915_MADV_WILLNEED;
-	/* Avoid an unnecessary call to unbind on the first bind. */
-	obj->map_and_fenceable = true;
-
 	return obj;
 }
 

From 504c7267a1e84b157cbd7e9c1b805e1bc0c2c846 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 23 Aug 2012 13:12:52 +0100
Subject: [PATCH 33/41] drm/i915: Use cpu relocations if the object is in the
 GTT but not mappable

This prevents the case of unbinding the object in order to process the
relocations through the GTT and then rebinding it only to then proceed
to use cpu relocations as the object is now in the CPU write domain. By
choosing to use cpu relocations up front, we can therefore avoid the
rebind penalty.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem_execbuffer.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index f7346d876551..dc87563440f9 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -95,6 +95,7 @@ eb_destroy(struct eb_objects *eb)
 static inline int use_cpu_reloc(struct drm_i915_gem_object *obj)
 {
 	return (obj->base.write_domain == I915_GEM_DOMAIN_CPU ||
+		!obj->map_and_fenceable ||
 		obj->cache_level != I915_CACHE_NONE);
 }
 

From b361237bcc7cea1d99f770490120d8bc2aed7777 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 24 Aug 2012 09:35:08 +0100
Subject: [PATCH 34/41] drm/i915: Juggle code order to ease flow of the next
 patch

Move the wait-for-rendering logic around in the file so that we can
group it together with the subsequent variations. The general goal is to
have the lower level routines clustered together and then the higher
level logic building upon those low level routines that came before.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem.c | 379 ++++++++++++++++----------------
 1 file changed, 188 insertions(+), 191 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 553175856e81..0b236ea0996c 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -945,6 +945,194 @@ unlock:
 	return ret;
 }
 
+int
+i915_gem_check_wedge(struct drm_i915_private *dev_priv,
+		     bool interruptible)
+{
+	if (atomic_read(&dev_priv->mm.wedged)) {
+		struct completion *x = &dev_priv->error_completion;
+		bool recovery_complete;
+		unsigned long flags;
+
+		/* Give the error handler a chance to run. */
+		spin_lock_irqsave(&x->wait.lock, flags);
+		recovery_complete = x->done > 0;
+		spin_unlock_irqrestore(&x->wait.lock, flags);
+
+		/* Non-interruptible callers can't handle -EAGAIN, hence return
+		 * -EIO unconditionally for these. */
+		if (!interruptible)
+			return -EIO;
+
+		/* Recovery complete, but still wedged means reset failure. */
+		if (recovery_complete)
+			return -EIO;
+
+		return -EAGAIN;
+	}
+
+	return 0;
+}
+
+/*
+ * Compare seqno against outstanding lazy request. Emit a request if they are
+ * equal.
+ */
+static int
+i915_gem_check_olr(struct intel_ring_buffer *ring, u32 seqno)
+{
+	int ret;
+
+	BUG_ON(!mutex_is_locked(&ring->dev->struct_mutex));
+
+	ret = 0;
+	if (seqno == ring->outstanding_lazy_request)
+		ret = i915_add_request(ring, NULL, NULL);
+
+	return ret;
+}
+
+/**
+ * __wait_seqno - wait until execution of seqno has finished
+ * @ring: the ring expected to report seqno
+ * @seqno: duh!
+ * @interruptible: do an interruptible wait (normally yes)
+ * @timeout: in - how long to wait (NULL forever); out - how much time remaining
+ *
+ * Returns 0 if the seqno was found within the alloted time. Else returns the
+ * errno with remaining time filled in timeout argument.
+ */
+static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
+			bool interruptible, struct timespec *timeout)
+{
+	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	struct timespec before, now, wait_time={1,0};
+	unsigned long timeout_jiffies;
+	long end;
+	bool wait_forever = true;
+	int ret;
+
+	if (i915_seqno_passed(ring->get_seqno(ring, true), seqno))
+		return 0;
+
+	trace_i915_gem_request_wait_begin(ring, seqno);
+
+	if (timeout != NULL) {
+		wait_time = *timeout;
+		wait_forever = false;
+	}
+
+	timeout_jiffies = timespec_to_jiffies(&wait_time);
+
+	if (WARN_ON(!ring->irq_get(ring)))
+		return -ENODEV;
+
+	/* Record current time in case interrupted by signal, or wedged * */
+	getrawmonotonic(&before);
+
+#define EXIT_COND \
+	(i915_seqno_passed(ring->get_seqno(ring, false), seqno) || \
+	atomic_read(&dev_priv->mm.wedged))
+	do {
+		if (interruptible)
+			end = wait_event_interruptible_timeout(ring->irq_queue,
+							       EXIT_COND,
+							       timeout_jiffies);
+		else
+			end = wait_event_timeout(ring->irq_queue, EXIT_COND,
+						 timeout_jiffies);
+
+		ret = i915_gem_check_wedge(dev_priv, interruptible);
+		if (ret)
+			end = ret;
+	} while (end == 0 && wait_forever);
+
+	getrawmonotonic(&now);
+
+	ring->irq_put(ring);
+	trace_i915_gem_request_wait_end(ring, seqno);
+#undef EXIT_COND
+
+	if (timeout) {
+		struct timespec sleep_time = timespec_sub(now, before);
+		*timeout = timespec_sub(*timeout, sleep_time);
+	}
+
+	switch (end) {
+	case -EIO:
+	case -EAGAIN: /* Wedged */
+	case -ERESTARTSYS: /* Signal */
+		return (int)end;
+	case 0: /* Timeout */
+		if (timeout)
+			set_normalized_timespec(timeout, 0, 0);
+		return -ETIME;
+	default: /* Completed */
+		WARN_ON(end < 0); /* We're not aware of other errors */
+		return 0;
+	}
+}
+
+/**
+ * Waits for a sequence number to be signaled, and cleans up the
+ * request and object lists appropriately for that event.
+ */
+int
+i915_wait_seqno(struct intel_ring_buffer *ring, uint32_t seqno)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	bool interruptible = dev_priv->mm.interruptible;
+	int ret;
+
+	BUG_ON(!mutex_is_locked(&dev->struct_mutex));
+	BUG_ON(seqno == 0);
+
+	ret = i915_gem_check_wedge(dev_priv, interruptible);
+	if (ret)
+		return ret;
+
+	ret = i915_gem_check_olr(ring, seqno);
+	if (ret)
+		return ret;
+
+	return __wait_seqno(ring, seqno, interruptible, NULL);
+}
+
+/**
+ * Ensures that all rendering to the object has completed and the object is
+ * safe to unbind from the GTT or access from the CPU.
+ */
+static __must_check int
+i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
+			       bool readonly)
+{
+	struct intel_ring_buffer *ring = obj->ring;
+	u32 seqno;
+	int ret;
+
+	seqno = readonly ? obj->last_write_seqno : obj->last_read_seqno;
+	if (seqno == 0)
+		return 0;
+
+	ret = i915_wait_seqno(ring, seqno);
+	if (ret)
+		return ret;
+
+	i915_gem_retire_requests_ring(ring);
+
+	/* Manually manage the write flush as we may have not yet
+	 * retired the buffer.
+	 */
+	if (obj->last_write_seqno &&
+	    i915_seqno_passed(seqno, obj->last_write_seqno)) {
+		obj->last_write_seqno = 0;
+		obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
+	}
+
+	return 0;
+}
+
 /**
  * Called when user space prepares to use an object with the CPU, either
  * through the mmap ioctl's mapping or a GTT mapping.
@@ -1952,197 +2140,6 @@ i915_gem_retire_work_handler(struct work_struct *work)
 	mutex_unlock(&dev->struct_mutex);
 }
 
-int
-i915_gem_check_wedge(struct drm_i915_private *dev_priv,
-		     bool interruptible)
-{
-	if (atomic_read(&dev_priv->mm.wedged)) {
-		struct completion *x = &dev_priv->error_completion;
-		bool recovery_complete;
-		unsigned long flags;
-
-		/* Give the error handler a chance to run. */
-		spin_lock_irqsave(&x->wait.lock, flags);
-		recovery_complete = x->done > 0;
-		spin_unlock_irqrestore(&x->wait.lock, flags);
-
-		/* Non-interruptible callers can't handle -EAGAIN, hence return
-		 * -EIO unconditionally for these. */
-		if (!interruptible)
-			return -EIO;
-
-		/* Recovery complete, but still wedged means reset failure. */
-		if (recovery_complete)
-			return -EIO;
-
-		return -EAGAIN;
-	}
-
-	return 0;
-}
-
-/*
- * Compare seqno against outstanding lazy request. Emit a request if they are
- * equal.
- */
-static int
-i915_gem_check_olr(struct intel_ring_buffer *ring, u32 seqno)
-{
-	int ret;
-
-	BUG_ON(!mutex_is_locked(&ring->dev->struct_mutex));
-
-	ret = 0;
-	if (seqno == ring->outstanding_lazy_request)
-		ret = i915_add_request(ring, NULL, NULL);
-
-	return ret;
-}
-
-/**
- * __wait_seqno - wait until execution of seqno has finished
- * @ring: the ring expected to report seqno
- * @seqno: duh!
- * @interruptible: do an interruptible wait (normally yes)
- * @timeout: in - how long to wait (NULL forever); out - how much time remaining
- *
- * Returns 0 if the seqno was found within the alloted time. Else returns the
- * errno with remaining time filled in timeout argument.
- */
-static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
-			bool interruptible, struct timespec *timeout)
-{
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
-	struct timespec before, now, wait_time={1,0};
-	unsigned long timeout_jiffies;
-	long end;
-	bool wait_forever = true;
-	int ret;
-
-	if (i915_seqno_passed(ring->get_seqno(ring, true), seqno))
-		return 0;
-
-	trace_i915_gem_request_wait_begin(ring, seqno);
-
-	if (timeout != NULL) {
-		wait_time = *timeout;
-		wait_forever = false;
-	}
-
-	timeout_jiffies = timespec_to_jiffies(&wait_time);
-
-	if (WARN_ON(!ring->irq_get(ring)))
-		return -ENODEV;
-
-	/* Record current time in case interrupted by signal, or wedged * */
-	getrawmonotonic(&before);
-
-#define EXIT_COND \
-	(i915_seqno_passed(ring->get_seqno(ring, false), seqno) || \
-	atomic_read(&dev_priv->mm.wedged))
-	do {
-		if (interruptible)
-			end = wait_event_interruptible_timeout(ring->irq_queue,
-							       EXIT_COND,
-							       timeout_jiffies);
-		else
-			end = wait_event_timeout(ring->irq_queue, EXIT_COND,
-						 timeout_jiffies);
-
-		ret = i915_gem_check_wedge(dev_priv, interruptible);
-		if (ret)
-			end = ret;
-	} while (end == 0 && wait_forever);
-
-	getrawmonotonic(&now);
-
-	ring->irq_put(ring);
-	trace_i915_gem_request_wait_end(ring, seqno);
-#undef EXIT_COND
-
-	if (timeout) {
-		struct timespec sleep_time = timespec_sub(now, before);
-		*timeout = timespec_sub(*timeout, sleep_time);
-	}
-
-	switch (end) {
-	case -EIO:
-	case -EAGAIN: /* Wedged */
-	case -ERESTARTSYS: /* Signal */
-		return (int)end;
-	case 0: /* Timeout */
-		if (timeout)
-			set_normalized_timespec(timeout, 0, 0);
-		return -ETIME;
-	default: /* Completed */
-		WARN_ON(end < 0); /* We're not aware of other errors */
-		return 0;
-	}
-}
-
-/**
- * Waits for a sequence number to be signaled, and cleans up the
- * request and object lists appropriately for that event.
- */
-int
-i915_wait_seqno(struct intel_ring_buffer *ring, uint32_t seqno)
-{
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
-	int ret = 0;
-
-	BUG_ON(seqno == 0);
-
-	ret = i915_gem_check_wedge(dev_priv, dev_priv->mm.interruptible);
-	if (ret)
-		return ret;
-
-	ret = i915_gem_check_olr(ring, seqno);
-	if (ret)
-		return ret;
-
-	ret = __wait_seqno(ring, seqno, dev_priv->mm.interruptible, NULL);
-
-	return ret;
-}
-
-/**
- * Ensures that all rendering to the object has completed and the object is
- * safe to unbind from the GTT or access from the CPU.
- */
-static __must_check int
-i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
-			       bool readonly)
-{
-	u32 seqno;
-	int ret;
-
-	/* If there is rendering queued on the buffer being evicted, wait for
-	 * it.
-	 */
-	if (readonly)
-		seqno = obj->last_write_seqno;
-	else
-		seqno = obj->last_read_seqno;
-	if (seqno == 0)
-		return 0;
-
-	ret = i915_wait_seqno(obj->ring, seqno);
-	if (ret)
-		return ret;
-
-	/* Manually manage the write flush as we may have not yet retired
-	 * the buffer.
-	 */
-	if (obj->last_write_seqno &&
-	    i915_seqno_passed(seqno, obj->last_write_seqno)) {
-		obj->last_write_seqno = 0;
-		obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
-	}
-
-	i915_gem_retire_requests_ring(obj->ring);
-	return 0;
-}
-
 /**
  * Ensures that an object will eventually get non-busy by flushing any required
  * write domains, emitting any outstanding lazy request and retiring and

From 3236f57a0162391f84b93f39fc1882c49a8998c7 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 24 Aug 2012 09:35:09 +0100
Subject: [PATCH 35/41] drm/i915: Use a non-blocking wait for set-to-domain
 ioctl

The principal use for set-to-domain is for userspace to serialise
operations with a particular buffer, for example to maintain coherency
with a CPU map or to ratelimit its rendering by waiting on all previous
operations before continuing. As such we tend to hold the struct_mutex
for long periods during the synchronisation and so cause contention
issues with other users of the graphics device, even for independent
operations as memory management. An example is the contention between
compiz and X which causes jitter in the display and a drop in peak
throughput.

The ultimate solution would be a set of fine grained locks and lockless
operations, but an intermediate step is to first attempt the
synchronisation for set-to-domain without holding the mutex. This
introduces a number of race conditions, so we limit it use to the ioctl
periphery where we have no dependent state and can safely complete with
a locked synchronisation afterwards.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem.c | 55 +++++++++++++++++++++++++++++++++
 1 file changed, 55 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 0b236ea0996c..719a933c5756 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1133,6 +1133,52 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
 	return 0;
 }
 
+/* A nonblocking variant of the above wait. This is a highly dangerous routine
+ * as the object state may change during this call.
+ */
+static __must_check int
+i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
+					    bool readonly)
+{
+	struct drm_device *dev = obj->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ring_buffer *ring = obj->ring;
+	u32 seqno;
+	int ret;
+
+	BUG_ON(!mutex_is_locked(&dev->struct_mutex));
+	BUG_ON(!dev_priv->mm.interruptible);
+
+	seqno = readonly ? obj->last_write_seqno : obj->last_read_seqno;
+	if (seqno == 0)
+		return 0;
+
+	ret = i915_gem_check_wedge(dev_priv, true);
+	if (ret)
+		return ret;
+
+	ret = i915_gem_check_olr(ring, seqno);
+	if (ret)
+		return ret;
+
+	mutex_unlock(&dev->struct_mutex);
+	ret = __wait_seqno(ring, seqno, true, NULL);
+	mutex_lock(&dev->struct_mutex);
+
+	i915_gem_retire_requests_ring(ring);
+
+	/* Manually manage the write flush as we may have not yet
+	 * retired the buffer.
+	 */
+	if (obj->last_write_seqno &&
+	    i915_seqno_passed(seqno, obj->last_write_seqno)) {
+		obj->last_write_seqno = 0;
+		obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
+	}
+
+	return ret;
+}
+
 /**
  * Called when user space prepares to use an object with the CPU, either
  * through the mmap ioctl's mapping or a GTT mapping.
@@ -1170,6 +1216,14 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 		goto unlock;
 	}
 
+	/* Try to flush the object off the GPU without holding the lock.
+	 * We will repeat the flush holding the lock in the normal manner
+	 * to catch cases where we are gazumped.
+	 */
+	ret = i915_gem_object_wait_rendering__nonblocking(obj, !write_domain);
+	if (ret)
+		goto unref;
+
 	if (read_domains & I915_GEM_DOMAIN_GTT) {
 		ret = i915_gem_object_set_to_gtt_domain(obj, write_domain != 0);
 
@@ -1183,6 +1237,7 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 		ret = i915_gem_object_set_to_cpu_domain(obj, write_domain != 0);
 	}
 
+unref:
 	drm_gem_object_unreference(&obj->base);
 unlock:
 	mutex_unlock(&dev->struct_mutex);

From bd9854f995f56f4bc3f0dc32dc92be02303c6f19 Mon Sep 17 00:00:00 2001
From: Ben Widawsky <ben@bwidawsk.net>
Date: Thu, 23 Aug 2012 15:18:09 -0700
Subject: [PATCH 36/41] drm/i915: Extract reading INSTDONE

INSTDONE is used in many places, and it varies from generation to
generation. This provides a good reason for us to extract the logic to
read the relevant information.

The patch has no functional change. It's prep for some new stuff.

v2: move the memset inside of i915_get_extra_instdone (Jani)
v3,4: bugs caught by (Jani)

Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_irq.c | 50 ++++++++++++++++++++-------------
 drivers/gpu/drm/i915/i915_reg.h |  1 +
 2 files changed, 32 insertions(+), 19 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 36330296932a..b7ec34ed5c30 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -1070,6 +1070,23 @@ i915_error_first_batchbuffer(struct drm_i915_private *dev_priv,
 	return NULL;
 }
 
+/* NB: please notice the memset */
+static void i915_get_extra_instdone(struct drm_device *dev,
+				    uint32_t *instdone)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	memset(instdone, 0, sizeof(*instdone) * I915_NUM_INSTDONE_REG);
+
+	if (INTEL_INFO(dev)->gen < 4) {
+		instdone[0] = I915_READ(INSTDONE);
+		instdone[1] = 0;
+	} else {
+		instdone[0] = I915_READ(INSTDONE_I965);
+		instdone[1] = I915_READ(INSTDONE1);
+	}
+}
+
+
 static void i915_record_ring_state(struct drm_device *dev,
 				   struct drm_i915_error_state *error,
 				   struct intel_ring_buffer *ring)
@@ -1288,6 +1305,7 @@ void i915_destroy_error_state(struct drm_device *dev)
 static void i915_report_and_clear_eir(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t instdone[I915_NUM_INSTDONE_REG];
 	u32 eir = I915_READ(EIR);
 	int pipe;
 
@@ -1296,16 +1314,17 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 
 	pr_err("render error detected, EIR: 0x%08x\n", eir);
 
+	i915_get_extra_instdone(dev, instdone);
+
 	if (IS_G4X(dev)) {
 		if (eir & (GM45_ERROR_MEM_PRIV | GM45_ERROR_CP_PRIV)) {
 			u32 ipeir = I915_READ(IPEIR_I965);
 
 			pr_err("  IPEIR: 0x%08x\n", I915_READ(IPEIR_I965));
 			pr_err("  IPEHR: 0x%08x\n", I915_READ(IPEHR_I965));
-			pr_err("  INSTDONE: 0x%08x\n",
-			       I915_READ(INSTDONE_I965));
+			pr_err("  INSTDONE: 0x%08x\n", instdone[0]);
 			pr_err("  INSTPS: 0x%08x\n", I915_READ(INSTPS));
-			pr_err("  INSTDONE1: 0x%08x\n", I915_READ(INSTDONE1));
+			pr_err("  INSTDONE1: 0x%08x\n", instdone[1]);
 			pr_err("  ACTHD: 0x%08x\n", I915_READ(ACTHD_I965));
 			I915_WRITE(IPEIR_I965, ipeir);
 			POSTING_READ(IPEIR_I965);
@@ -1344,7 +1363,7 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 
 			pr_err("  IPEIR: 0x%08x\n", I915_READ(IPEIR));
 			pr_err("  IPEHR: 0x%08x\n", I915_READ(IPEHR));
-			pr_err("  INSTDONE: 0x%08x\n", I915_READ(INSTDONE));
+			pr_err("  INSTDONE: 0x%08x\n", instdone[0]);
 			pr_err("  ACTHD: 0x%08x\n", I915_READ(ACTHD));
 			I915_WRITE(IPEIR, ipeir);
 			POSTING_READ(IPEIR);
@@ -1353,10 +1372,9 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 
 			pr_err("  IPEIR: 0x%08x\n", I915_READ(IPEIR_I965));
 			pr_err("  IPEHR: 0x%08x\n", I915_READ(IPEHR_I965));
-			pr_err("  INSTDONE: 0x%08x\n",
-			       I915_READ(INSTDONE_I965));
+			pr_err("  INSTDONE: 0x%08x\n", instdone[0]);
 			pr_err("  INSTPS: 0x%08x\n", I915_READ(INSTPS));
-			pr_err("  INSTDONE1: 0x%08x\n", I915_READ(INSTDONE1));
+			pr_err("  INSTDONE1: 0x%08x\n", instdone[1]);
 			pr_err("  ACTHD: 0x%08x\n", I915_READ(ACTHD_I965));
 			I915_WRITE(IPEIR_I965, ipeir);
 			POSTING_READ(IPEIR_I965);
@@ -1671,7 +1689,7 @@ void i915_hangcheck_elapsed(unsigned long data)
 {
 	struct drm_device *dev = (struct drm_device *)data;
 	drm_i915_private_t *dev_priv = dev->dev_private;
-	uint32_t acthd[I915_NUM_RINGS], instdone, instdone1;
+	uint32_t acthd[I915_NUM_RINGS], instdone[I915_NUM_INSTDONE_REG];
 	struct intel_ring_buffer *ring;
 	bool err = false, idle;
 	int i;
@@ -1699,25 +1717,19 @@ void i915_hangcheck_elapsed(unsigned long data)
 		return;
 	}
 
-	if (INTEL_INFO(dev)->gen < 4) {
-		instdone = I915_READ(INSTDONE);
-		instdone1 = 0;
-	} else {
-		instdone = I915_READ(INSTDONE_I965);
-		instdone1 = I915_READ(INSTDONE1);
-	}
+	i915_get_extra_instdone(dev, instdone);
 
 	if (memcmp(dev_priv->last_acthd, acthd, sizeof(acthd)) == 0 &&
-	    dev_priv->last_instdone == instdone &&
-	    dev_priv->last_instdone1 == instdone1) {
+	    dev_priv->last_instdone == instdone[0] &&
+	    dev_priv->last_instdone1 == instdone[1]) {
 		if (i915_hangcheck_hung(dev))
 			return;
 	} else {
 		dev_priv->hangcheck_count = 0;
 
 		memcpy(dev_priv->last_acthd, acthd, sizeof(acthd));
-		dev_priv->last_instdone = instdone;
-		dev_priv->last_instdone1 = instdone1;
+		dev_priv->last_instdone = instdone[0];
+		dev_priv->last_instdone1 = instdone[1];
 	}
 
 repeat:
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index f27e9982bed0..1f97b3fd2105 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -478,6 +478,7 @@
 #define IPEIR_I965	0x02064
 #define IPEHR_I965	0x02068
 #define INSTDONE_I965	0x0206c
+#define I915_NUM_INSTDONE_REG	2
 #define RING_IPEIR(base)	((base)+0x64)
 #define RING_IPEHR(base)	((base)+0x68)
 #define RING_INSTDONE(base)	((base)+0x6c)

From d53bd48459a554503e07bc3a4b8bba2b57c2ff48 Mon Sep 17 00:00:00 2001
From: Ben Widawsky <ben@bwidawsk.net>
Date: Wed, 22 Aug 2012 11:32:14 -0700
Subject: [PATCH 37/41] drm/i915: Add new INSTDONE registers

Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_reg.h | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 1f97b3fd2105..d0b60f2f1e50 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -478,7 +478,11 @@
 #define IPEIR_I965	0x02064
 #define IPEHR_I965	0x02068
 #define INSTDONE_I965	0x0206c
-#define I915_NUM_INSTDONE_REG	2
+#define GEN7_INSTDONE_1		0x0206c
+#define GEN7_SC_INSTDONE	0x07100
+#define GEN7_SAMPLER_INSTDONE	0x0e160
+#define GEN7_ROW_INSTDONE	0x0e164
+#define I915_NUM_INSTDONE_REG	4
 #define RING_IPEIR(base)	((base)+0x64)
 #define RING_IPEHR(base)	((base)+0x68)
 #define RING_INSTDONE(base)	((base)+0x6c)

From 050ee91f128bd767b1413383fea6c973aa464c54 Mon Sep 17 00:00:00 2001
From: Ben Widawsky <ben@bwidawsk.net>
Date: Wed, 22 Aug 2012 11:32:15 -0700
Subject: [PATCH 38/41] drm/i915: Use new INSTDONE registers (Gen7+)

Using the extracted INSTDONE reading, and our new register definitions,
update our hangcheck detection and error collection to use it. This
primarily means changing == to memcmp, and changing = to memcpy.
Hopefully this will give more info on error dump, and provide more
accurate hangcheck detection (both are actually TBD).

Also, remove the reading of instdone1 from the ring error collection
function, and just crap everything in capture_error_state (that could be
split into a separate patch if it wasn't so trivial).

v2: Now assuming i915_get_extra_instdone does the memset we can clean up the
code a bit (Jani)

v3: use ARRAY_SIZE as requested earlier by Jani (didn't change sizeof)
Updated commit msg

Cc: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_debugfs.c |  8 ++++--
 drivers/gpu/drm/i915/i915_drv.h     |  5 ++--
 drivers/gpu/drm/i915/i915_irq.c     | 43 +++++++++++++++++------------
 3 files changed, 33 insertions(+), 23 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index d3003937ddf5..3d886af2aae6 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -662,10 +662,9 @@ static void i915_ring_error_state(struct seq_file *m,
 	seq_printf(m, "  IPEIR: 0x%08x\n", error->ipeir[ring]);
 	seq_printf(m, "  IPEHR: 0x%08x\n", error->ipehr[ring]);
 	seq_printf(m, "  INSTDONE: 0x%08x\n", error->instdone[ring]);
-	if (ring == RCS && INTEL_INFO(dev)->gen >= 4) {
-		seq_printf(m, "  INSTDONE1: 0x%08x\n", error->instdone1);
+	if (ring == RCS && INTEL_INFO(dev)->gen >= 4)
 		seq_printf(m, "  BBADDR: 0x%08llx\n", error->bbaddr);
-	}
+
 	if (INTEL_INFO(dev)->gen >= 4)
 		seq_printf(m, "  INSTPS: 0x%08x\n", error->instps[ring]);
 	seq_printf(m, "  INSTPM: 0x%08x\n", error->instpm[ring]);
@@ -714,6 +713,9 @@ static int i915_error_state(struct seq_file *m, void *unused)
 	for (i = 0; i < dev_priv->num_fence_regs; i++)
 		seq_printf(m, "  fence[%d] = %08llx\n", i, error->fence[i]);
 
+	for (i = 0; i < ARRAY_SIZE(error->extra_instdone); i++)
+		seq_printf(m, "  INSTDONE_%d: 0x%08x\n", i, error->extra_instdone[i]);
+
 	if (INTEL_INFO(dev)->gen >= 6) {
 		seq_printf(m, "ERROR: 0x%08x\n", error->error);
 		seq_printf(m, "DONE_REG: 0x%08x\n", error->done_reg);
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 9f4c27eab046..58b43db0a134 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -199,7 +199,7 @@ struct drm_i915_error_state {
 	u32 err_int; /* gen7 */
 	u32 instpm[I915_NUM_RINGS];
 	u32 instps[I915_NUM_RINGS];
-	u32 instdone1;
+	u32 extra_instdone[I915_NUM_INSTDONE_REG];
 	u32 seqno[I915_NUM_RINGS];
 	u64 bbaddr;
 	u32 fault_reg[I915_NUM_RINGS];
@@ -454,8 +454,7 @@ typedef struct drm_i915_private {
 	struct timer_list hangcheck_timer;
 	int hangcheck_count;
 	uint32_t last_acthd[I915_NUM_RINGS];
-	uint32_t last_instdone;
-	uint32_t last_instdone1;
+	uint32_t prev_instdone[I915_NUM_INSTDONE_REG];
 
 	unsigned int stop_rings;
 
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index b7ec34ed5c30..34dc7158f03c 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -1077,12 +1077,25 @@ static void i915_get_extra_instdone(struct drm_device *dev,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	memset(instdone, 0, sizeof(*instdone) * I915_NUM_INSTDONE_REG);
 
-	if (INTEL_INFO(dev)->gen < 4) {
+	switch(INTEL_INFO(dev)->gen) {
+	case 2:
+	case 3:
 		instdone[0] = I915_READ(INSTDONE);
-		instdone[1] = 0;
-	} else {
+		break;
+	case 4:
+	case 5:
+	case 6:
 		instdone[0] = I915_READ(INSTDONE_I965);
 		instdone[1] = I915_READ(INSTDONE1);
+		break;
+	default:
+		WARN_ONCE(1, "Unsupported platform\n");
+	case 7:
+		instdone[0] = I915_READ(GEN7_INSTDONE_1);
+		instdone[1] = I915_READ(GEN7_SC_INSTDONE);
+		instdone[2] = I915_READ(GEN7_SAMPLER_INSTDONE);
+		instdone[3] = I915_READ(GEN7_ROW_INSTDONE);
+		break;
 	}
 }
 
@@ -1108,10 +1121,8 @@ static void i915_record_ring_state(struct drm_device *dev,
 		error->ipehr[ring->id] = I915_READ(RING_IPEHR(ring->mmio_base));
 		error->instdone[ring->id] = I915_READ(RING_INSTDONE(ring->mmio_base));
 		error->instps[ring->id] = I915_READ(RING_INSTPS(ring->mmio_base));
-		if (ring->id == RCS) {
-			error->instdone1 = I915_READ(INSTDONE1);
+		if (ring->id == RCS)
 			error->bbaddr = I915_READ64(BB_ADDR);
-		}
 	} else {
 		error->faddr[ring->id] = I915_READ(DMA_FADD_I8XX);
 		error->ipeir[ring->id] = I915_READ(IPEIR);
@@ -1230,6 +1241,8 @@ static void i915_capture_error_state(struct drm_device *dev)
 	if (INTEL_INFO(dev)->gen == 7)
 		error->err_int = I915_READ(GEN7_ERR_INT);
 
+	i915_get_extra_instdone(dev, error->extra_instdone);
+
 	i915_gem_record_fences(dev, error);
 	i915_gem_record_rings(dev, error);
 
@@ -1307,7 +1320,7 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t instdone[I915_NUM_INSTDONE_REG];
 	u32 eir = I915_READ(EIR);
-	int pipe;
+	int pipe, i;
 
 	if (!eir)
 		return;
@@ -1322,9 +1335,9 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 
 			pr_err("  IPEIR: 0x%08x\n", I915_READ(IPEIR_I965));
 			pr_err("  IPEHR: 0x%08x\n", I915_READ(IPEHR_I965));
-			pr_err("  INSTDONE: 0x%08x\n", instdone[0]);
+			for (i = 0; i < ARRAY_SIZE(instdone); i++)
+				pr_err("  INSTDONE_%d: 0x%08x\n", i, instdone[i]);
 			pr_err("  INSTPS: 0x%08x\n", I915_READ(INSTPS));
-			pr_err("  INSTDONE1: 0x%08x\n", instdone[1]);
 			pr_err("  ACTHD: 0x%08x\n", I915_READ(ACTHD_I965));
 			I915_WRITE(IPEIR_I965, ipeir);
 			POSTING_READ(IPEIR_I965);
@@ -1358,12 +1371,13 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 	if (eir & I915_ERROR_INSTRUCTION) {
 		pr_err("instruction error\n");
 		pr_err("  INSTPM: 0x%08x\n", I915_READ(INSTPM));
+		for (i = 0; i < ARRAY_SIZE(instdone); i++)
+			pr_err("  INSTDONE_%d: 0x%08x\n", i, instdone[i]);
 		if (INTEL_INFO(dev)->gen < 4) {
 			u32 ipeir = I915_READ(IPEIR);
 
 			pr_err("  IPEIR: 0x%08x\n", I915_READ(IPEIR));
 			pr_err("  IPEHR: 0x%08x\n", I915_READ(IPEHR));
-			pr_err("  INSTDONE: 0x%08x\n", instdone[0]);
 			pr_err("  ACTHD: 0x%08x\n", I915_READ(ACTHD));
 			I915_WRITE(IPEIR, ipeir);
 			POSTING_READ(IPEIR);
@@ -1372,9 +1386,7 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 
 			pr_err("  IPEIR: 0x%08x\n", I915_READ(IPEIR_I965));
 			pr_err("  IPEHR: 0x%08x\n", I915_READ(IPEHR_I965));
-			pr_err("  INSTDONE: 0x%08x\n", instdone[0]);
 			pr_err("  INSTPS: 0x%08x\n", I915_READ(INSTPS));
-			pr_err("  INSTDONE1: 0x%08x\n", instdone[1]);
 			pr_err("  ACTHD: 0x%08x\n", I915_READ(ACTHD_I965));
 			I915_WRITE(IPEIR_I965, ipeir);
 			POSTING_READ(IPEIR_I965);
@@ -1718,18 +1730,15 @@ void i915_hangcheck_elapsed(unsigned long data)
 	}
 
 	i915_get_extra_instdone(dev, instdone);
-
 	if (memcmp(dev_priv->last_acthd, acthd, sizeof(acthd)) == 0 &&
-	    dev_priv->last_instdone == instdone[0] &&
-	    dev_priv->last_instdone1 == instdone[1]) {
+	    memcmp(dev_priv->prev_instdone, instdone, sizeof(instdone)) == 0) {
 		if (i915_hangcheck_hung(dev))
 			return;
 	} else {
 		dev_priv->hangcheck_count = 0;
 
 		memcpy(dev_priv->last_acthd, acthd, sizeof(acthd));
-		dev_priv->last_instdone = instdone[0];
-		dev_priv->last_instdone1 = instdone[1];
+		memcpy(dev_priv->prev_instdone, instdone, sizeof(instdone));
 	}
 
 repeat:

From 7788a765205f63abcb8645c16c85a968bd578f4f Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 24 Aug 2012 19:18:18 +0100
Subject: [PATCH 39/41] drm/i915: Avoid unbinding due to an interrupted
 pin_and_fence during execbuffer

If we need to stall in order to complete the pin_and_fence operation
during execbuffer reservation, there is a high likelihood that the
operation will be interrupted by a signal (thanks X!). In order to
simplify the cleanup along that error path, the object was
unconditionally unbound and the error propagated. However, being
interrupted here is far more common than I would like and so we can
strive to avoid the extra work by eliminating the forced unbind.

v2: In discussion over the indecent colour of the new functions and
unwind path, we realised that we can use the new unreserve function to
clean up the code even further.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem_execbuffer.c | 114 ++++++++-------------
 1 file changed, 45 insertions(+), 69 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index dc87563440f9..e6b2205ecf6d 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -331,7 +331,8 @@ i915_gem_execbuffer_relocate(struct drm_device *dev,
 	return ret;
 }
 
-#define  __EXEC_OBJECT_HAS_FENCE (1<<31)
+#define  __EXEC_OBJECT_HAS_PIN (1<<31)
+#define  __EXEC_OBJECT_HAS_FENCE (1<<30)
 
 static int
 need_reloc_mappable(struct drm_i915_gem_object *obj)
@@ -341,9 +342,10 @@ need_reloc_mappable(struct drm_i915_gem_object *obj)
 }
 
 static int
-pin_and_fence_object(struct drm_i915_gem_object *obj,
-		     struct intel_ring_buffer *ring)
+i915_gem_execbuffer_reserve_object(struct drm_i915_gem_object *obj,
+				   struct intel_ring_buffer *ring)
 {
+	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
 	struct drm_i915_gem_exec_object2 *entry = obj->exec_entry;
 	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
 	bool need_fence, need_mappable;
@@ -359,11 +361,13 @@ pin_and_fence_object(struct drm_i915_gem_object *obj,
 	if (ret)
 		return ret;
 
+	entry->flags |= __EXEC_OBJECT_HAS_PIN;
+
 	if (has_fenced_gpu_access) {
 		if (entry->flags & EXEC_OBJECT_NEEDS_FENCE) {
 			ret = i915_gem_object_get_fence(obj);
 			if (ret)
-				goto err_unpin;
+				return ret;
 
 			if (i915_gem_object_pin_fence(obj))
 				entry->flags |= __EXEC_OBJECT_HAS_FENCE;
@@ -372,12 +376,35 @@ pin_and_fence_object(struct drm_i915_gem_object *obj,
 		}
 	}
 
+	/* Ensure ppgtt mapping exists if needed */
+	if (dev_priv->mm.aliasing_ppgtt && !obj->has_aliasing_ppgtt_mapping) {
+		i915_ppgtt_bind_object(dev_priv->mm.aliasing_ppgtt,
+				       obj, obj->cache_level);
+
+		obj->has_aliasing_ppgtt_mapping = 1;
+	}
+
 	entry->offset = obj->gtt_offset;
 	return 0;
+}
 
-err_unpin:
-	i915_gem_object_unpin(obj);
-	return ret;
+static void
+i915_gem_execbuffer_unreserve_object(struct drm_i915_gem_object *obj)
+{
+	struct drm_i915_gem_exec_object2 *entry;
+
+	if (!obj->gtt_space)
+		return;
+
+	entry = obj->exec_entry;
+
+	if (entry->flags & __EXEC_OBJECT_HAS_FENCE)
+		i915_gem_object_unpin_fence(obj);
+
+	if (entry->flags & __EXEC_OBJECT_HAS_PIN)
+		i915_gem_object_unpin(obj);
+
+	entry->flags &= ~(__EXEC_OBJECT_HAS_FENCE | __EXEC_OBJECT_HAS_PIN);
 }
 
 static int
@@ -385,11 +412,10 @@ i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
 			    struct drm_file *file,
 			    struct list_head *objects)
 {
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
 	struct drm_i915_gem_object *obj;
-	int ret, retry;
-	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
 	struct list_head ordered_objects;
+	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
+	int retry;
 
 	INIT_LIST_HEAD(&ordered_objects);
 	while (!list_empty(objects)) {
@@ -427,12 +453,12 @@ i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
 	 * 2.  Bind new objects.
 	 * 3.  Decrement pin count.
 	 *
-	 * This avoid unnecessary unbinding of later objects in order to makr
+	 * This avoid unnecessary unbinding of later objects in order to make
 	 * room for the earlier objects *unless* we need to defragment.
 	 */
 	retry = 0;
 	do {
-		ret = 0;
+		int ret = 0;
 
 		/* Unbind any ill-fitting objects or pin. */
 		list_for_each_entry(obj, objects, exec_list) {
@@ -452,7 +478,7 @@ i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
 			    (need_mappable && !obj->map_and_fenceable))
 				ret = i915_gem_object_unbind(obj);
 			else
-				ret = pin_and_fence_object(obj, ring);
+				ret = i915_gem_execbuffer_reserve_object(obj, ring);
 			if (ret)
 				goto err;
 		}
@@ -462,46 +488,14 @@ i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
 			if (obj->gtt_space)
 				continue;
 
-			ret = pin_and_fence_object(obj, ring);
-			if (ret) {
-				int ret_ignore;
-
-				/* This can potentially raise a harmless
-				 * -EINVAL if we failed to bind in the above
-				 * call. It cannot raise -EINTR since we know
-				 * that the bo is freshly bound and so will
-				 * not need to be flushed or waited upon.
-				 */
-				ret_ignore = i915_gem_object_unbind(obj);
-				(void)ret_ignore;
-				WARN_ON(obj->gtt_space);
-				break;
-			}
+			ret = i915_gem_execbuffer_reserve_object(obj, ring);
+			if (ret)
+				goto err;
 		}
 
-		/* Decrement pin count for bound objects */
-		list_for_each_entry(obj, objects, exec_list) {
-			struct drm_i915_gem_exec_object2 *entry;
-
-			if (!obj->gtt_space)
-				continue;
-
-			entry = obj->exec_entry;
-			if (entry->flags & __EXEC_OBJECT_HAS_FENCE) {
-				i915_gem_object_unpin_fence(obj);
-				entry->flags &= ~__EXEC_OBJECT_HAS_FENCE;
-			}
-
-			i915_gem_object_unpin(obj);
-
-			/* ... and ensure ppgtt mapping exist if needed. */
-			if (dev_priv->mm.aliasing_ppgtt && !obj->has_aliasing_ppgtt_mapping) {
-				i915_ppgtt_bind_object(dev_priv->mm.aliasing_ppgtt,
-						       obj, obj->cache_level);
-
-				obj->has_aliasing_ppgtt_mapping = 1;
-			}
-		}
+err:		/* Decrement pin count for bound objects */
+		list_for_each_entry(obj, objects, exec_list)
+			i915_gem_execbuffer_unreserve_object(obj);
 
 		if (ret != -ENOSPC || retry++)
 			return ret;
@@ -510,24 +504,6 @@ i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
 		if (ret)
 			return ret;
 	} while (1);
-
-err:
-	list_for_each_entry_continue_reverse(obj, objects, exec_list) {
-		struct drm_i915_gem_exec_object2 *entry;
-
-		if (!obj->gtt_space)
-			continue;
-
-		entry = obj->exec_entry;
-		if (entry->flags & __EXEC_OBJECT_HAS_FENCE) {
-			i915_gem_object_unpin_fence(obj);
-			entry->flags &= ~__EXEC_OBJECT_HAS_FENCE;
-		}
-
-		i915_gem_object_unpin(obj);
-	}
-
-	return ret;
 }
 
 static int

From cd7988eea561a70a4f98e431c1395f913672d626 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sun, 26 Aug 2012 20:33:18 +0200
Subject: [PATCH 40/41] drm/i915: disable rc6 on ilk when vt-d is enabled

It blows up. And hopefully this is the root-cause of the mysterious
rc6 related hang on ilk. For reference, the commit that enabled rc6 on
ilk again is:

commit 456470eb583f063ee84c6818251e638598be0fb8
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Wed Aug 8 23:35:40 2012 +0200

    drm/i915: enable rc6 on ilk again

Reported-by: Dave Airlie <airlied@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_pm.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index c0721ffb56a2..ebe3498bc9c0 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -2372,6 +2372,11 @@ int intel_enable_rc6(const struct drm_device *dev)
 		return i915_enable_rc6;
 
 	if (INTEL_INFO(dev)->gen == 5) {
+#ifdef CONFIG_INTEL_IOMMU
+		/* Disable rc6 on ilk if VT-d is on. */
+		if (intel_iommu_gfx_mapped)
+			return false;
+#endif
 		DRM_DEBUG_DRIVER("Ironlake: only RC6 available\n");
 		return INTEL_RC6_ENABLE;
 	}

From d7c3b937bdf45f0b844400b7bf6fd3ed50bac604 Mon Sep 17 00:00:00 2001
From: Sedat Dilek <sedat.dilek@gmail.com>
Date: Mon, 27 Aug 2012 14:02:37 +0200
Subject: [PATCH 41/41] drm/i915: Remove __GFP_NO_KSWAPD

When I pulled-in today's drm-intel-next into linux-next (next-20120824)
I saw this build-breakage:

drivers/gpu/drm/i915/i915_gem.c: In function 'i915_gem_object_get_pages_gtt':
drivers/gpu/drm/i915/i915_gem.c:1778:40: error: '__GFP_NO_KSWAPD' undeclared (first use in this function)
drivers/gpu/drm/i915/i915_gem.c:1778:40: note: each undeclared identifier is reported only once for each function it appears in

This is caused by commit ba099ef165f8 ("mm: remove __GFP_NO_KSWAPD")
and commit b6beae2c2014 ("mm: remove __GFP_NO_KSWAPD fixes") in
linux-next (next-20120824).

Fix this by removing __GFP_NO_KSWAPD from drm/i915 driver.

Signed-off-by: Sedat Dilek <sedat.dilek@gmail.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/i915_gem.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 719a933c5756..ac050e6425b6 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1775,7 +1775,7 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 	/* Fail silently without starting the shrinker */
 	mapping = obj->base.filp->f_path.dentry->d_inode->i_mapping;
 	gfp = mapping_gfp_mask(mapping);
-	gfp |= __GFP_NORETRY | __GFP_NOWARN | __GFP_NO_KSWAPD;
+	gfp |= __GFP_NORETRY | __GFP_NOWARN;
 	gfp &= ~(__GFP_IO | __GFP_WAIT);
 	for (i = 0; i < page_count; i++) {
 		page = shmem_read_mapping_page_gfp(mapping, i, gfp);
@@ -1788,7 +1788,7 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 			 * our own buffer, now let the real VM do its job and
 			 * go down in flames if truly OOM.
 			 */
-			gfp &= ~(__GFP_NORETRY | __GFP_NOWARN | __GFP_NO_KSWAPD);
+			gfp &= ~(__GFP_NORETRY | __GFP_NOWARN);
 			gfp |= __GFP_IO | __GFP_WAIT;
 
 			i915_gem_shrink_all(dev_priv);
@@ -1796,7 +1796,7 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 			if (IS_ERR(page))
 				goto err_pages;
 
-			gfp |= __GFP_NORETRY | __GFP_NOWARN | __GFP_NO_KSWAPD;
+			gfp |= __GFP_NORETRY | __GFP_NOWARN;
 			gfp &= ~(__GFP_IO | __GFP_WAIT);
 		}