Once the machine gets to a certain point in the suspend process, we
expect the GPU to be idle. If it is not, we might corrupt memory.
Empirically (with an early version of this patch) we have seen this is
not the case. We cannot currently explain why the latent GPU writes
occur.

In the technical sense, this patch is a workaround in that we have an
issue we can't explain, and the patch indirectly solves the issue.
However, it's really better than a workaround because we understand why
it works, and it really should be a safe thing to do in all cases.

The noticeable effect other than the debug messages would be an increase
in the suspend time. I have not measure how expensive it actually is.

I think it would be good to spend further time to root cause why we're
seeing these latent writes, but it shouldn't preclude preventing the
fallout.

NOTE: It should be safe (and makes some sense IMO) to also keep the
VALID bit unset on resume when we clear_range(). I've opted not to do
this as properly clearing those bits at some later point would be extra
work.

v2: Fix bugzilla link

Bugzilla: http://bugs.freedesktop.org/show_bug.cgi?id=65496
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=59321Tested-by: NTakashi Iwai <tiwai@suse.de>
Tested-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Tested-By: NTodd Previte <tprevite@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

We need this to work around a corruption when the boot kernel image
loads the hibernated kernel image from swap on Haswell systems -
somehow not everything is properly shut off.

This is just the prep work, the next patch will implement the actual
workaround.
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
[danvet: Add a commit message suitable for -fixes and add cc: stable]
Cc: stable@vger.kernel.org
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Needed to prevent display corruption in high res panels.

v2: use correct unit names (Rodrigo)
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Tested-by: NUlrich Drepper <drepper@gmail.com>
Reviewed-by: NRodrigo Vivi <rodrigo.vivi@gmail.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This reverts commit 81b5c7bc.

Adding drm/i915 into the vga arbiter chain means that X (in a piece of
well-meant paranoia) will do a get/put on the vga decoding around
_every_ accel call down into the ddx. Which results in some nice
performance disasters [1]. This really breaks userspace, by disabling
DRI for everyone, and stops OpenGL from working, this isn't limited
to just the i915 but both the integrated and discrete GPUs on
multi-gpu systems, in other words this causes untold worlds of pain,

Ville tried to come up with a Great Hack to fiddle the required VGA
I/O ops behind everyone's back using stop_machine, but that didn't
really work out [2]. Given that we're fairly late in the -rc stage for
such games let's just revert this all.

One thing we might want to keep is to delay the disabling of the vga
decoding until the fbdev emulation and the fbcon screen is set up. If
we kill vga mem decoding beforehand fbcon can end up with a white
square in the top-left corner it tried to save from the vga memory for
a seamless transition. And we have bug reports on older platforms
which seem to match these symptoms.

But again that's something to play around with in -next.

References: [1] http://lists.x.org/archives/xorg-devel/2013-September/037763.html
References: [2] http://www.spinics.net/lists/intel-gfx/msg34062.html
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NDave Airlie <airlied@redhat.com>

This reverts commit 6e1b4fda.

This is part of a revert due to a userspace breakage, better explained in the revert of 1a1a4cbf4906a13c0c377f708df5d94168e7b582.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NDave Airlie <airlied@redhat.com>

HW ACR support may have issues on some older chips, so
use SW ACR for now until we've tested further.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
CC: Rafał Miłecki <zajec5@gmail.com>

Causes hangs on certain boards.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=70053Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Disable CG/PG before resetting.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Disable CG/PG and stop the rlc before resetting.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

These checks should be ">=" instead of ">".  j is used as an offset into
the table->mc_reg_address[] array and that has
SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE (16) elements.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

It should be ">=" instead of ">" here.  The table->mc_reg_address[]
array has SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE (16) elements.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

The error path does this:

	for (--i; i >= 0; --i) {

which is a forever loop because "i" is unsigned.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

They have 4 rather than 8.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=63599Signed-off-by: Nwojciech kapuscinski <wojtask9@wp.pl>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

Wrong bit offset for SRC endian swapping.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

Always use the regular UVD state for now.  This fixes
a performance regression with UVD playback on certain APUs.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Use the hw generated values rather than calculating
them in the driver.  There may be some older r6xx
asics where this doesn't work correctly.  This remains
to be seen.

See bug:
https://bugs.freedesktop.org/show_bug.cgi?id=69675Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The drm code that calculates the 1001 clocks rounds up
rather than truncating.  This allows the table to match
properly on those modes.

See bug:
https://bugs.freedesktop.org/show_bug.cgi?id=69675Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Avoid losing precision.  See bug:
https://bugs.freedesktop.org/show_bug.cgi?id=69675

v2: fix math as per Anssi's comments.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Return -ENOMEM if the allocation fails.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Reviewed-by: NJani Nikula <jani.nikula@intel.com>

This reverts commit 928c2f0c.

This patch double applied, two checks for the price of one.
Signed-off-by: NDave Airlie <airlied@redhat.com>

Commit 8b9ba7a3 'drm/gma500: use gem get/put page helpers' was missing a
line, which resulted in garbled screen.
Signed-off-by: NRob Clark <robdclark@gmail.com>
Tested-by: NGuillaume CLÉMENT <gclement@baobob.org>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Signed-off-by: NBen Skeggs <bskeggs@redhat.com>

The current test for an attached enabled encoder fails if we have
multiple connectors aliased to the same encoder - both connectors
believe they own the enabled encoder and so we attempt to both enable
and disable DPMS on the encoder, leading to hilarity and an OOPs:

[  354.803064] WARNING: CPU: 0 PID: 482 at
/usr/src/linux/dist/3.11.2/drivers/gpu/drm/i915/intel_display.c:3869 intel_modeset_check_state+0x764/0x770 [i915]()
[  354.803064] wrong connector dpms state
[  354.803084] Modules linked in: nfsd auth_rpcgss oid_registry exportfs nfs lockd sunrpc xt_nat iptable_nat nf_nat_ipv4 nf_nat xt_limit xt_LOG xt_tcpudp nf_conntrack_ipv4 nf_defrag_ipv4 ipt_REJECT ipv6 xt_recent xt_conntrack nf_conntrack iptable_filter ip_tables x_tables snd_hda_codec_realtek snd_hda_codec_hdmi x86_pkg_temp_thermal snd_hda_intel coretemp kvm_intel snd_hda_codec i915 kvm snd_hwdep snd_pcm_oss snd_mixer_oss crc32_pclmul snd_pcm crc32c_intel e1000e intel_agp igb ghash_clmulni_intel intel_gtt aesni_intel cfbfillrect aes_x86_64 cfbimgblt lrw cfbcopyarea drm_kms_helper ptp video thermal processor gf128mul snd_page_alloc drm snd_timer glue_helper 8250_pci snd pps_core ablk_helper agpgart cryptd sg soundcore fan i2c_algo_bit sr_mod thermal_sys 8250 i2c_i801 serial_core
hwmon cdrom i2c_core evdev button
[  354.803086] CPU: 0 PID: 482 Comm: kworker/0:1 Not tainted 3.11.2 #1
[  354.803087] Hardware name: Supermicro X10SAE/X10SAE, BIOS 1.00 05/03/2013 [  354.803091] Workqueue: events console_callback
[  354.803092]  0000000000000009 ffff88023611db48 ffffffff814048ac ffff88023611db90
[  354.803093]  ffff88023611db80 ffffffff8103d4e3 ffff880230d82800 ffff880230f9b800
[  354.803094]  ffff880230f99000 ffff880230f99448 ffff8802351c0e00 ffff88023611dbe0
[  354.803094] Call Trace:
[  354.803098]  [<ffffffff814048ac>] dump_stack+0x54/0x8d
[  354.803101]  [<ffffffff8103d4e3>] warn_slowpath_common+0x73/0x90
[  354.803103]  [<ffffffff8103d547>] warn_slowpath_fmt+0x47/0x50
[  354.803109]  [<ffffffffa089f1be>] ? intel_ddi_connector_get_hw_state+0x5e/0x110 [i915]
[  354.803114]  [<ffffffffa0896974>] intel_modeset_check_state+0x764/0x770 [i915]
[  354.803117]  [<ffffffffa08969bb>] intel_connector_dpms+0x3b/0x60 [i915]
[  354.803120]  [<ffffffffa037e1d0>] drm_fb_helper_dpms.isra.11+0x120/0x160 [drm_kms_helper]
[  354.803122]  [<ffffffffa037e24e>] drm_fb_helper_blank+0x3e/0x80 [drm_kms_helper]
[  354.803123]  [<ffffffff812116c2>] fb_blank+0x52/0xc0
[  354.803125]  [<ffffffff8121e04b>] fbcon_blank+0x21b/0x2d0
[  354.803127]  [<ffffffff81062243>] ? update_rq_clock.part.74+0x13/0x30
[  354.803129]  [<ffffffff81047486>] ? lock_timer_base.isra.30+0x26/0x50
[  354.803130]  [<ffffffff810472b2>] ? internal_add_timer+0x12/0x40
[  354.803131]  [<ffffffff81047f48>] ? mod_timer+0xf8/0x1c0
[  354.803133]  [<ffffffff81266d61>] do_unblank_screen+0xa1/0x1c0
[  354.803134]  [<ffffffff81268087>] poke_blanked_console+0xc7/0xd0
[  354.803136]  [<ffffffff812681cf>] console_callback+0x13f/0x160
[  354.803137]  [<ffffffff81053258>] process_one_work+0x148/0x3d0
[  354.803138]  [<ffffffff81053f19>] worker_thread+0x119/0x3a0
[  354.803140]  [<ffffffff81053e00>] ? manage_workers.isra.30+0x2a0/0x2a0
[  354.803141]  [<ffffffff8105994b>] kthread+0xbb/0xc0
[  354.803142]  [<ffffffff81059890>] ? kthread_create_on_node+0x120/0x120
[  354.803144]  [<ffffffff8140b32c>] ret_from_fork+0x7c/0xb0
[  354.803145]  [<ffffffff81059890>] ? kthread_create_on_node+0x120/0x120

This regression goes back to the big modeset rework and the conversion
to the new dpms helpers which started with:

commit 5ab432ef
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Sat Jun 30 08:59:56 2012 +0200

    drm/i915/hdmi: convert to encoder->disable/enable

Fixes: igt/kms_flip/dpms-off-confusion
Reported-and-tested-by: NWakko Warner <wakko@animx.eu.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=68030
Link:  http://lkml.kernel.org/r/20130928185023.GA21672@animx.eu.orgSigned-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@vger.kernel.org
[danvet: Add regression citation, mention the igt testcase this fixes
and slap a cc: stable on the patch.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Power Well in use forces constantly PSR to exit.
On recent Kernel I noticed that PSR Performance Counter was always 0
indicating that PSR was never really achieved.
By masking LPSP, PSR can work normally and save power on Haswell.

Two bugs had been raised with PSR flag enabled:
- "Screen flickers when booted by enabling PSR in the kernel (i915.enable_psr=1) , the system is booting to a gray screen."
- "When booting the DUT with PSR feature enabled in the kernel (i915.enable_psr=1) , the system is booting to a gray screen."

Both bugs has been fixed by this patch.

v2: proper comment for -fixes
Tested-by: NSelvaraj, Elavarasan <elavarasanx.selvaraj@intel.com>
Signed-off-by: NRodrigo Vivi <rodrigo.vivi@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Otherwise using any atomic memory operation will lock up the GPU due
to a Haswell hardware bug.

v2: Use the _MASKED_BIT_ENABLE macro.  Drop drm parameter definition.
Signed-off-by: NFrancisco Jerez <currojerez@riseup.net>
Reviewed-by: NBen Widawsky <ben@bwidawsk.net>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: <stable@vger.kernel.org>
[danvet: Fix checkpatch fail.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

During driver loading we are initializing rps.vlv_work in
valleyview_enable_rps() via the rps.delayed_resume_work delayed work.
This is too late since we are using vlv_work already via
i915_driver_load()->intel_uncore_sanitize()->
intel_disable_gt_powersave(). This at least leads to the following
kernel warning:

 INFO: trying to register non-static key.
 the code is fine but needs lockdep annotation.
 turning off the locking correctness validator.

Fix this by initialzing vlv_work before we call intel_uncore_sanitize().

The regression was introduced in

commit 7dcd2677
Author: Konstantin Khlebnikov <khlebnikov@openvz.org>
Date:   Wed Jul 17 10:22:58 2013 +0400

    drm/i915: fix long-standing SNB regression in power consumption
    after resume

though there was no good reason to initialize the static vlv_work from
another delayed work to begin with (especially since this will happen
multiple times).

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=69397Tested-by: Nshui yangwei <yangweix.shui@intel.com>
Signed-off-by: NImre Deak <imre.deak@intel.com>
Cc: stable@vger.kernel.org
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Signed-off-by: NRob Clark <robdclark@gmail.com>

We still need an API exported by msm iommu driver (but not visible in
any public header anymore).  For now, just declare the prototype
ourselves, but when msm iommu driver provides a better option, use that
instead.
Signed-off-by: NRob Clark <robdclark@gmail.com>

The include file has been removed and the file does not
need it anyway, so remove it. Fixes a compile error.
Signed-off-by: NJoerg Roedel <joro@8bytes.org>
Signed-off-by: NRob Clark <robdclark@gmail.com>

The variable priv->kms is not initialized yet.

Found by "scripts/coccinelle/tests/odd_ptr_err.cocci".
PTR_ERR should access the value just tested by IS_ERR.
Signed-off-by: NThomas Meyer <thomas@m3y3r.de>

In

commit 81e49f81
Author: Glauber Costa <glommer@openvz.org>
Date:   Wed Aug 28 10:18:13 2013 +1000

    i915: bail out earlier when shrinker cannot acquire mutex

SHRINK_STOP was added to tell the core shrinker code to bail out and
go to the next shrinker since the i915 shrinker couldn't acquire
required locks. But the SHRINK_STOP return code was added to the
->count_objects callback and not the ->scan_objects callback as it
should have been, resulting in tons of dmesg noise like

shrink_slab: i915_gem_inactive_scan+0x0/0x9c negative objects to delete nr=-xxxxxxxxx

Fix discusssed with Dave Chinner.

References: http://www.spinics.net/lists/intel-gfx/msg33597.htmlReported-by: NKnut Petersen <Knut_Petersen@t-online.de>
Cc: Knut Petersen <Knut_Petersen@t-online.de>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Glauber Costa <glommer@openvz.org>
Cc: Glauber Costa <glommer@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Acked-by: NDave Chinner <dchinner@redhat.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

These asics seem to use a mix of the DCE2.x and
DCE3.2 audio interfaces despite what the register spec
says.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=69729
https://bugs.freedesktop.org/show_bug.cgi?id=69671Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

This regression has been introduced in

commit 9f11a9e4
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Thu Jun 13 00:54:58 2013 +0200

    drm/i915: set up PIPECONF explicitly for i9xx/vlv platforms

Ville brough up the idea that this is just the pipe A quirk gone
wrong.

Note that after resume the bios might or might not have enabled pipe A
already.  We have a bit of magic to make sure that on resume we set up
a decent mode for pipe A, but I fear if I just smash pipe A to always
on we'd enable it in a bogus state and hang the hw. Hence the
readback.

v2: Clarify the logic a bit as suggested by Chris. Also amend the
commit message to clarify why we don't unconditionally enable the
pipe.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=66462
References: https://lkml.org/lkml/2013/8/26/238
Cc: Meelis Roos <mroos@ut.ee>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
[danvet: Use |= instead of = as suggested by Chris.]
Cc: stable@vger.kernel.org
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The native TV encoder has it's own flags to adjust sync modes and
enabled interlaced modes which are totally irrelevant for the adjusted
mode. This worked out nicely since the input modes used by both the
load detect code and reported in the ->get_modes callbacks all have no
flags set, and we also don't fill out any of them in the ->get_config
callback.

This changed with the additional sanitation done with

commit 2960bc9c
Author: Imre Deak <imre.deak@intel.com>
Date:   Tue Jul 30 13:36:32 2013 +0300

    drm/i915: make user mode sync polarity setting explicit

sinc now the "no flags at all" state wouldn't fit through core code
any more. So fix this up again by explicitly clearing the flags in the
->compute_config callback.

Aside: We have zero checking in place to make sure that the requested
mode is indeed the right input mode we want for the selected TV mode.
So we'll happily fall over if userspace tries to pull us.  But that's
definitely work for a different patch series. So just add a FIXME
comment for now.
Reported-by: NKnut Petersen <Knut_Petersen@t-online.de>
Cc: Knut Petersen <Knut_Petersen@t-online.de>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: NKnut Petersen <Knut_Petersen@t-online.de>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

There is no clear cut rules or specs for the retry interval, as there
are many factors that affect overall response time. Increase the
interval, and even more so on branch devices which may have limited i2c
bit rates.
Signed-off-by: NJani Nikula <jani.nikula@intel.com>
Reference: https://bugs.freedesktop.org/show_bug.cgi?id=60263Tested-by: NNicolas Suzor <nic@suzor.com>
Reviewed-by: NTodd Previte <tprevite@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Fix a bug that was introduced in commit c4c11dd1 ("drm/i2c: tda998x:
add video and audio input configuration") when Sebastian cleaned up my
original patch.  Without this being fixed, audio is muted when the
display is turned off, never to be re-enabled.
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>
Cc: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Cc: Darren Etheridge <detheridge@ti.com>
Cc: Dave Airlie <airlied@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Missing ULL when calculating the amount of vram
leads to an overflow when the amount of vram is >= 4G.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

When dpm was merged, I added a new asic struct for
rv6xx, but it never got properly updated when the
hdmi callbacks were added due to the two patch sets
being developed in parallel.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=69729Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Reviewed-by: NChristian König <christian.koenig@amd.com>
Cc: stable@vger.kernel.org

In

commit edc3d884
Author: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Date:   Thu May 23 13:55:35 2013 +0300

    drm/i915: avoid big kmallocs on reading error state

we introduce a two-pass mechanism for splitting long strings being
formatted into the error-state. The first pass finds the length, and the
second pass emits the right portion of the string into the accumulation
buffer. Unfortunately we use the same va_list for both passes, resulting
in the second pass reading garbage off the end of the argument list. As
the two passes are only used for boundaries between read() calls, the
corruption is only rarely seen.

This fixes the root cause behind

commit baf27f9b
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Sat Jun 29 23:26:50 2013 +0100

    drm/i915: Break up the large vsnprintf() in print_error_buffers()
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: stable@vger.kernel.org
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Starting with UVD3 message and feedback buffers have their
own 256MB segment, so no need to force them into VRAM any more.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>