The radeon driver uses placement range restrictions for several reasons,
in particular to make sure BOs in VRAM can be accessed by the CPU, e.g.
during a page fault.

Without this change, TTM could evict other BOs while trying to satisfy
the requested placement, even if the evicted BOs were outside of the
requested placement range. Doing so didn't free up any space in the
requested placement range, so the (potentially high) eviction cost was
incurred for no benefit.

Nominating for stable because radeon driver changes in 3.17 made this
much more noticeable than before.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84662
Cc: stable@vger.kernel.org
Signed-off-by: NMichel Dänzer <michel.daenzer@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NMichel Dänzer <michel.daenzer@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Avoids HDP cache flush issues when using vram which can
cause ring test failures on certain boards.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: Alexander Fyodorov <halcy@yandex.ru>
Cc: stable@vger.kernel.org

If the sad_count is 0, set the hw to stereo and change
the error message to a warn.  A lot of monitors don't
set the speaker allocation block.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

Fixes kfree of the sadb buffer when it's NULL.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

This reverts commit 186b1b2b.

There are still some stability problems on some SI boards so bring
this back.

This reverts commit fc9dfeb1.

There are still some stability problems on some SI boards so bring
this back.

The old code has problems with the Dell MST monitors due to some
assumptions I made that weren't true.

I initially thought the Virtual Channel Payload IDs had to be in
the DPCD table in ascending order, however it appears that assumption
is bogus.

The old code also assumed it was possible to insert a member
into the table and it would move other members up, like it does
when you remove table entries, however reality has shown this
isn't true.

So the new code allocates VCPIs separate from entries in the payload
tracking table, and when we remove an entry from the DPCD table,
I shuffle the tracking payload entries around in the struct.

This appears to make VT switch more robust (still not perfect)
with an MST enabled Dell monitor.
Signed-off-by: NDave Airlie <airlied@redhat.com>

The translation from the X driver to the KMS one typo'ed a couple
of array indices, causing the HW cursor to look weird (blocky with
leaking edge colors). This fixes it.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: stable@vger.kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

Track whether UVD or VCE are enabled in debugfs.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Track whether UVD or VCE are enabled in debugfs.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Adds an extra argument to radeon_bo_create, which is only used in radeon_prime.c.
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Not the whole world is a radeon! :-)
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The patch replaces direct access to driver_features field
by calls to helper function.
Signed-off-by: NAndrzej Hajda <a.hajda@samsung.com>
Reviewed-by: NDavid Herrmann <dh.herrmann@gmail.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Helpful for debugging as the version shows up in a
register dump.

Cc: Jay Cornwall <jay.cornwall@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

We should never have a packet 0 in the command stream.
Dump the full command stream to help debug.

bug:
https://bugs.freedesktop.org/show_bug.cgi?id=84500Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The requirement for these was removed in:
0a168933Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

legacy_connector_convert is used in radeon_combios.c only, so declare it as static.
Signed-off-by: NMichele Curti <michele.curti@gmail.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The tables:
 * supported_devices_connector_convert
 * supported_devices_connector_object_id_convert
 * object_connector_convert
are used in redeon_atombios.c only, so declare them as static.
Signed-off-by: NMichele Curti <michele.curti@gmail.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

It's no longer used now that the underlying bugs are fixed.
Reviewed-by: NAlexandre Demers <alexandre.f.demers@gmail.com>
Tested-by: NAlexandre Demers <alexandre.f.demers@gmail.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

No longer needed now that the underlying bug was fixed in
e0792981
(drm/radeon/dpm: fix typo in vddci setup for eg/btc).

bug:
https://bugs.freedesktop.org/show_bug.cgi?id=69721Reviewed-by: NAlexandre Demers <alexandre.f.demers@gmail.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Not sure this was ever necessary for CI, was just done
to be on the safe side.

bug:
https://bugs.freedesktop.org/show_bug.cgi?id=69721Reviewed-by: NAlexandre Demers <alexandre.f.demers@gmail.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Not sure this was ever necessary for SI, was just done
to be on the safe side.

bug:
https://bugs.freedesktop.org/show_bug.cgi?id=69721Reviewed-by: NAlexandre Demers <alexandre.f.demers@gmail.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

No longer needed now that the underlying bug was fixed in
b0880e87
(drm/radeon/dpm: fix vddci setup typo on cayman).

bug:
https://bugs.freedesktop.org/show_bug.cgi?id=69721Reviewed-by: NAlexandre Demers <alexandre.f.demers@gmail.com>
Tested-by: NAlexandre Demers <alexandre.f.demers@gmail.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

This should allow the audio driver to get a better
idea of whether the sink is connected or not.

v2: fix copy/paste typo noticed by David Henningsson
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Clean up the enable sequence as well.

V2: clean up duplicate defines
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Most of that functionality is only used by r600_hdmi.c
and I'm planning to change that further.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Only need one copy.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

no functional change.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

CEA VICs 44 and 45 were missing DRM_MODE_FLAG_INTERLACE.
Signed-off-by: NClint Taylor <clinton.a.taylor@intel.com>
Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Adds an extra argument to nouveau_bo_new, which is only used in nouveau_prime.c.
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>

Haswell and later silicon has added a new pixel replication register
to the pipe timings for each transcoder. Now in addition to the
DPLL_A_MD register for the pixel clock double, we also need to write
to the TRANS_MULT_n (0x6002c) register to double the pixel data. Writing
to the DPLL only double the pixel clock.

ver2: Macro name change from MULTIPLY to PIPE_MULTI. (Daniel)
ver3: Do not set pixel multiplier if transcoder is eDP (Ville)
ver4: Macro name change to PIPE_MULT and default else pixel_multiplier

Cc: Ville =?iso-8859-1?Q?Syrj=E4l=E4?= <ville.syrjala@linux.intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: NClint Taylor <clinton.a.taylor@intel.com>
Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
[danvet: Appease checkpatch and move one hunk back into the right
place that git am misplace!?]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This allows importing reservation objects from dma-bufs.
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>

Allows importing dma_reservation_objects from a dma-buf.
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>

Make nouveau_fence_chan refcounted, to make trace_fence_destroy
always return the correct name without a race condition.
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>

Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>

Allows userspace to detect shared fences are supported.
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>

This fixes a regression introduced by "drm/nouveau: rework to new fence interface"
(commit 29ba89b2).

The fence sequence should not be reset after creation, the old value is used instead.
On destruction the final value is written, to prevent another source of accidental
wraparound in case of a channel being destroyed after a hang, and unblocking any other
channel that may wait on the about-to-be-deleted channel to signal.

I'm nothing if not optimistic about any hope of recovery from that. ;-)
Reported-by: NTed Percival <ted@tedp.id.au>
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>
Tested-by: NTed Percival <ted@tedp.id.au>
Acked-by: NBen Skeggs <bskeggs@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

If the GPU frequency isn't going to change don't spam dmesg with
debug messages about it.
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NJani Nikula <jani.nikula@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@intel.com>