mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Ben Hutchings <ben@decadent.org.uk>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: torvalds@linux-foundation.org, akpm@linux-foundation.org,
	alan@lxorguk.ukuu.org.uk, Daniel Vetter <daniel.vetter@ffwll.ch>,
	Eugeni Dodonov <eugeni.dodonov@intel.com>
Subject: [ 02/69] drm/i915: hold forcewake around ring hw init
Date: Sun, 17 Jun 2012 18:59:43 +0100	[thread overview]
Message-ID: <20120617175941.644246078@decadent.org.uk> (raw)
In-Reply-To: <20120617175941.281333656@decadent.org.uk>

3.2-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Daniel Vetter <daniel.vetter@ffwll.ch>

commit b7884eb45ec98c0d34c7f49005ae9d4b4b4e38f6 upstream.

Empirical evidence suggests that we need to: On at least one ivb
machine when running the hangman i-g-t test, the rings don't properly
initialize properly - the RING_START registers seems to be stuck at
all zeros.

Holding forcewake around this register init sequences makes chip reset
reliable again. Note that this is not the first such issue:

commit f01db988ef6f6c70a6cc36ee71e4a98a68901229
Author: Sean Paul <seanpaul@chromium.org>
Date:   Fri Mar 16 12:43:22 2012 -0400

    drm/i915: Add wait_for in init_ring_common

added delay loops to make RING_START and RING_CTL initialization
reliable on the blt ring at boot-up. So I guess it won't hurt if we do
this unconditionally for all force_wake needing gpus.

To avoid copy&pasting of the HAS_FORCE_WAKE check I've added a new
intel_info bit for that.

v2: Fixup missing commas in static struct and properly handling the
error case in init_ring_common, both noticed by Jani Nikula.

Reported-and-tested-by: Yang Guang <guang.a.yang@intel.com>
Reviewed-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=50522
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
[bwh: Backported to 3.2:
 - drop changes to Haswell device information
 - NEEDS_FORCE_WAKE didn't refer to Valley View anyway]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
---
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -214,6 +214,7 @@
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
+	.has_force_wake = 1,
 };
 
 static const struct intel_device_info intel_sandybridge_m_info = {
@@ -222,6 +223,7 @@
 	.has_fbc = 1,
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
+	.has_force_wake = 1,
 };
 
 static const struct intel_device_info intel_ivybridge_d_info = {
@@ -229,6 +231,7 @@
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
+	.has_force_wake = 1,
 };
 
 static const struct intel_device_info intel_ivybridge_m_info = {
@@ -237,6 +240,7 @@
 	.has_fbc = 0,	/* FBC is not enabled on Ivybridge mobile yet */
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
+	.has_force_wake = 1,
 };
 
 static const struct pci_device_id pciidlist[] = {		/* aka */
@@ -939,7 +943,7 @@
 
 /* We give fast paths for the really cool registers */
 #define NEEDS_FORCE_WAKE(dev_priv, reg) \
-	(((dev_priv)->info->gen >= 6) && \
+	((HAS_FORCE_WAKE((dev_priv)->dev)) && \
 	 ((reg) < 0x40000) &&		 \
 	 ((reg) != FORCEWAKE) &&	 \
 	 ((reg) != ECOBUS))
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -244,6 +244,7 @@
 	u8 is_broadwater:1;
 	u8 is_crestline:1;
 	u8 is_ivybridge:1;
+	u8 has_force_wake:1;
 	u8 has_fbc:1;
 	u8 has_pipe_cxsr:1;
 	u8 has_hotplug:1;
@@ -1001,6 +1002,8 @@
 #define HAS_PCH_CPT(dev) (INTEL_PCH_TYPE(dev) == PCH_CPT)
 #define HAS_PCH_IBX(dev) (INTEL_PCH_TYPE(dev) == PCH_IBX)
 
+#define HAS_FORCE_WAKE(dev) (INTEL_INFO(dev)->has_force_wake)
+
 #include "i915_trace.h"
 
 extern struct drm_ioctl_desc i915_ioctls[];
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -263,10 +263,15 @@
 
 static int init_ring_common(struct intel_ring_buffer *ring)
 {
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	struct drm_device *dev = ring->dev;
+	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj = ring->obj;
+	int ret = 0;
 	u32 head;
 
+	if (HAS_FORCE_WAKE(dev))
+		gen6_gt_force_wake_get(dev_priv);
+
 	/* Stop the ring if it's running. */
 	I915_WRITE_CTL(ring, 0);
 	I915_WRITE_HEAD(ring, 0);
@@ -314,7 +319,8 @@
 				I915_READ_HEAD(ring),
 				I915_READ_TAIL(ring),
 				I915_READ_START(ring));
-		return -EIO;
+		ret = -EIO;
+		goto out;
 	}
 
 	if (!drm_core_check_feature(ring->dev, DRIVER_MODESET))
@@ -325,7 +331,11 @@
 		ring->space = ring_space(ring);
 	}
 
-	return 0;
+out:
+	if (HAS_FORCE_WAKE(dev))
+		gen6_gt_force_wake_put(dev_priv);
+
+	return ret;
 }
 
 static int



  parent reply	other threads:[~2012-06-17 18:40 UTC|newest]

Thread overview: 86+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-06-17 17:59 [ 00/69] 3.2.21-stable review Ben Hutchings
2012-06-17 17:59 ` [ 01/69] drm/i915: Mark the ringbuffers as being in the GTT domain Ben Hutchings
2012-06-17 17:59 ` Ben Hutchings [this message]
2012-06-17 17:59 ` [ 03/69] drm/i915: Unpin the flip target if we fail to queue the flip Ben Hutchings
2012-06-17 17:59 ` [ 04/69] drm/i915: fix up ivb plane 3 pageflips Ben Hutchings
2012-06-18 21:55   ` Herton Ronaldo Krzesinski
2012-06-18 22:10     ` Eugeni Dodonov
2012-06-19  3:59       ` Ben Hutchings
2012-06-19  8:23         ` Daniel Vetter
2012-06-19 11:59           ` Ben Hutchings
2012-06-19 11:45         ` Eugeni Dodonov
2012-06-17 17:59 ` [ 05/69] char/agp: add another Ironlake host bridge Ben Hutchings
2012-06-17 17:59 ` [ 06/69] x86/uv: Fix UV2 BAU legacy mode Ben Hutchings
2012-06-17 17:59 ` [ 07/69] powerpc: Fix kernel panic during kernel module load Ben Hutchings
2012-06-17 17:59 ` [ 08/69] drm/nouveau: determine a value for display_info.bpc if edid doesnt Ben Hutchings
2012-06-17 17:59 ` [ 09/69] drm/nouveau: default to 8bpc for non-LVDS panels if EDID isnt useful Ben Hutchings
2012-06-17 17:59 ` [ 10/69] drm/nouveau/disp: fix dithering not being enabled on some eDP macbooks Ben Hutchings
2012-06-17 17:59 ` [ 11/69] fuse: fix stat call on 32 bit platforms Ben Hutchings
2012-06-17 17:59 ` [ 12/69] x86, MCE, AMD: Make APIC LVT thresholding interrupt optional Ben Hutchings
2012-06-17 17:59 ` [ 13/69] x86/amd: Re-enable CPU topology extensions in case BIOS has disabled it Ben Hutchings
2012-06-17 17:59 ` [ 14/69] hwmon: (fam15h_power) Increase output resolution Ben Hutchings
2012-06-17 17:59 ` [ 15/69] sched: Fix the relax_domain_level boot parameter Ben Hutchings
2012-06-17 17:59 ` [ 16/69] iwlwifi: dont mess up the SCD when removing a key Ben Hutchings
2012-06-17 17:59 ` [ 17/69] hwrng: atmel-rng - fix race condition leading to repeated bits Ben Hutchings
2012-06-17 17:59 ` [ 18/69] crypto: aesni-intel - fix unaligned cbc decrypt for x86-32 Ben Hutchings
2012-06-17 18:00 ` [ 19/69] xen/setup: filter APERFMPERF cpuid feature out Ben Hutchings
2012-06-17 18:00 ` [ 20/69] NFSv4.1: Fix a request leak on the back channel Ben Hutchings
2012-06-17 18:00 ` [ 21/69] can: c_can: fix "BUG! echo_skb is occupied!" during transmit Ben Hutchings
2012-06-17 18:00 ` [ 22/69] can: c_can: fix an interrupt thrash issue with c_can driver Ben Hutchings
2012-06-17 18:00 ` [ 23/69] can: c_can: fix race condition in c_can_open() Ben Hutchings
2012-06-17 18:00 ` [ 24/69] ARM i.MX53: Fix PLL4 base address Ben Hutchings
2012-06-17 18:00 ` [ 25/69] usb: musb_gadget: fix crash caused by dangling pointer Ben Hutchings
2012-06-17 18:00 ` [ 26/69] mac80211: clean up remain-on-channel on interface stop Ben Hutchings
2012-06-17 18:00 ` [ 27/69] rt2x00: use atomic variable for seqno Ben Hutchings
2012-06-17 18:00 ` [ 28/69] iwlwifi: disable WoWLAN if !CONFIG_PM_SLEEP Ben Hutchings
2012-06-17 18:00 ` [ 29/69] ASoC: wm8994: Ensure all AIFnCLK events are run from the _late variants Ben Hutchings
2012-06-17 18:00 ` [ 30/69] ASoC: wm8994: Apply volume updates with clocks enabled Ben Hutchings
2012-06-17 18:00 ` [ 31/69] iwlwifi: unregister LEDs if mac80211 registration fails Ben Hutchings
2012-06-17 18:00 ` [ 32/69] cfg80211: fix interface combinations check Ben Hutchings
2012-06-17 18:00 ` [ 33/69] [SCSI] mpt2sas: Fix unsafe using smp_processor_id() in preemptible Ben Hutchings
2012-06-17 18:00 ` [ 34/69] net: sierra_net: device IDs for Aircard 320U++ Ben Hutchings
2012-06-17 18:00 ` [ 35/69] ARM: imx6: exit coherency when shutting down a cpu Ben Hutchings
2012-06-17 18:00 ` [ 36/69] ARM i.MX imx21ads: Fix overlapping static i/o mappings Ben Hutchings
2012-06-17 18:00 ` [ 37/69] NFSv4: Fix unnecessary delegation returns in nfs4_do_open Ben Hutchings
2012-06-17 18:00 ` [ 38/69] iwlwifi: use correct supported firmware for 6035 and 6000g2 Ben Hutchings
2012-06-17 18:00 ` [ 39/69] iwlwifi: disable the buggy chain extension feature in HW Ben Hutchings
2012-06-18 19:18   ` Herton Ronaldo Krzesinski
2012-06-18 19:23     ` Grumbach, Emmanuel
2012-06-19  3:57       ` Ben Hutchings
2012-06-17 18:00 ` [ 40/69] ALSA: hda - Add codec->no_jack_detect flag Ben Hutchings
2012-06-17 18:00 ` [ 41/69] ALSA: hda - add support for Uniwill ECS M31EI notebook Ben Hutchings
2012-06-17 18:00 ` [ 42/69] ALSA: hda - Suppress auto-mute feature on some machines with ALC861 Ben Hutchings
2012-06-17 18:00 ` [ 43/69] ALSA: hda - Add another jack-detection suppression for ASUS ALC892 Ben Hutchings
2012-06-17 18:00 ` [ 44/69] ALSA: HDA: Pin fixup for Zotac Z68 motherboard Ben Hutchings
2012-06-17 18:00 ` [ 45/69] usb: cdc-wdm: Add device-id for Huawei 3G/LTE modems Ben Hutchings
2012-06-17 20:51   ` Bjørn Mork
2012-06-17 22:51     ` Ben Hutchings
2012-06-17 18:00 ` [ 46/69] USB: cdc-wdm: Add Vodafone/Huawei K5005 support Ben Hutchings
2012-06-17 20:50   ` Bjørn Mork
2012-06-17 18:00 ` [ 47/69] USB: option: " Ben Hutchings
2012-06-17 18:00 ` [ 48/69] USB: qcserial: Add Sierra Wireless device IDs Ben Hutchings
2012-06-17 18:00 ` [ 49/69] USB: ftdi-sio: Add support for RT Systems USB-RTS01 serial adapter Ben Hutchings
2012-06-17 18:00 ` [ 50/69] USB: serial: cp210x: add Optris MS Pro usb id Ben Hutchings
2012-06-17 18:00 ` [ 51/69] USB: serial: sierra: Add support for Sierra Wireless AirCard 320U modem Ben Hutchings
2012-06-17 18:00 ` [ 52/69] USB: option: Updated Huawei K4605 has better id Ben Hutchings
2012-06-17 18:00 ` [ 53/69] USB: mct_u232: Fix incorrect TIOCMSET return Ben Hutchings
2012-06-17 18:00 ` [ 54/69] USB: option: fix port-data abuse Ben Hutchings
2012-06-17 18:00 ` [ 55/69] USB: option: fix memory leak Ben Hutchings
2012-06-17 18:00 ` [ 56/69] USB: option: add more YUGA device ids Ben Hutchings
2012-06-17 18:00 ` [ 57/69] target: Return error to initiator if SET TARGET PORT GROUPS emulation fails Ben Hutchings
2012-06-17 18:00 ` [ 58/69] USB: add NO_D3_DURING_SLEEP flag and revert 151b61284776be2 Ben Hutchings
2012-06-17 18:00 ` [ 59/69] USB: serial: Enforce USB driver and USB serial driver match Ben Hutchings
2012-06-17 18:00 ` [ 60/69] usb-storage: Add 090c:1000 to unusal-devs Ben Hutchings
2012-06-18  8:39   ` Hans de Goede
2012-06-18  9:14     ` Simon Raffeiner
2012-06-18 13:38       ` Ben Hutchings
2012-06-17 18:00 ` [ 61/69] xhci: Fix invalid loop check in xhci_free_tt_info() Ben Hutchings
2012-06-17 18:00 ` [ 62/69] xhci: Dont free endpoints in xhci_mem_cleanup() Ben Hutchings
2012-06-17 18:00 ` [ 63/69] xHCI: Increase the timeout for controller save/restore state operation Ben Hutchings
2012-06-17 18:00 ` [ 64/69] usb: PS3 EHCI QH read work-around Ben Hutchings
2012-06-17 18:00 ` [ 65/69] USB: fix PS3 EHCI systems Ben Hutchings
2012-06-17 18:00 ` [ 66/69] usb: cdc-acm: fix devices not unthrottled on open Ben Hutchings
2012-06-17 18:00 ` [ 67/69] USB: fix gathering of interface associations Ben Hutchings
2012-06-17 18:00 ` [ 68/69] swap: fix shmem swapping when more than 8 areas Ben Hutchings
2012-06-17 18:00 ` [ 69/69] drm/radeon: add some additional 6xx/7xx/EG register init Ben Hutchings
2012-06-17 18:48 ` [ 00/69] 3.2.21-stable review Ben Hutchings

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20120617175941.644246078@decadent.org.uk \
    --to=ben@decadent.org.uk \
    --cc=akpm@linux-foundation.org \
    --cc=alan@lxorguk.ukuu.org.uk \
    --cc=daniel.vetter@ffwll.ch \
    --cc=eugeni.dodonov@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=torvalds@linux-foundation.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®