linux-toradex.git/drivers/gpu, branch tegra-11.2.12

drm/i915, intel_ips: When i915 loads after IPS, make IPS relink to i915.

2011-01-07T21:58:45+00:00

commit 63ee41d794d9c555f84205517a68509848988760 upstream.

The IPS driver is designed to be able to run detached from i915 and
just not enable GPU turbo in that case, in order to avoid module
dependencies between the two drivers.  This means that we don't know
what the load order between the two is going to be, and we had
previously only supported IPS after (optionally) i915, but not i915
after IPS.  If the wrong order was chosen, you'd get no GPU turbo, and
something like half the possible graphics performance.

Signed-off-by: Eric Anholt 
Signed-off-by: Chris Wilson 
Signed-off-by: Greg Kroah-Hartman

drm/i915/dp: Fix I2C/EDID handling with active DisplayPort to DVI converter

2011-01-07T21:58:45+00:00

commit 8316f33766a82907c694267ff911e45e256f09f9 upstream.

The DisplayPort standard (1.1a) states that:
  The I2C-over-AUX Reply field is valid only when Native AUX CH Reply
  field is AUX_ACK (00). When Native AUX CH Reply field is not 00, then,
  I2C-over-AUX Reply field must be 00 and be ignored.

This fixes broken EDID reading when using an active DisplayPort to
duallink DVI converter.  If the AUX CH replier chooses to defer the
transaction, a short read occurs and erroneous data is returned as
the i2c reply due to a lack of length checking and failure to check
for AUX ACK.

As a result, broken EDIDs can look like:
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f    0123456789abcdef
00: bc bc bc ff bc bc bc ff bc bc bc ac bc bc bc 45    ???.???.???????E
10: bc bc bc 10 bc bc bc 34 bc bc bc ee bc bc bc 4c    ???????4???????L
20: bc bc bc 50 bc bc bc 00 bc bc bc 40 bc bc bc 00    ???P???.???@???.
30: bc bc bc 01 bc bc bc 01 bc bc bc a0 bc bc bc 40    ???????????????@
40: bc bc bc 00 bc bc bc 00 bc bc bc 00 bc bc bc 55    ???.???.???.???U
50: bc bc bc 35 bc bc bc 31 bc bc bc 20 bc bc bc fc    ???5???1??? ????
60: bc bc bc 4c bc bc bc 34 bc bc bc 46 bc bc bc 00    ???L???4???F???.
70: bc bc bc 38 bc bc bc 11 bc bc bc 20 bc bc bc 20    ???8??????? ???
80: bc bc bc ff bc bc bc ff bc bc bc ff bc bc bc ff    ???.???.???.???.
...

which can lead to:
[drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder
[drm:drm_edid_block_valid] *ERROR* Raw EDID:
<3>30 30 30 30 30 30 30 32 38 32 30 32 63 63 31 61  000000028202cc1a
<3>28 00 02 8c 00 00 00 00 18 00 00 00 00 00 00 00  (...............
<3>20 4c 61 73 74 20 62 65 61 63 6f 6e 3a 20 33 32   Last beacon: 32
<3>32 30 6d 73 20 61 67 6f 46 00 05 8c 00 00 00 00  20ms agoF.......
<3>36 00 00 00 00 00 00 00 00 0c 57 69 2d 46 69 20  6.........Wi-Fi
<3>52 6f 75 74 65 72 01 08 82 84 8b 96 24 30 48 6c  Router......$0Hl
<3>03 01 01 06 02 00 00 2a 01 00 2f 01 00 32 04 0c  .......*../..2..
<3>12 18 60 dd 09 00 10 18 02 00 00 01 00 00 18 00  ..`.............

Signed-off-by: David Flynn 
[ickle: fix up some surrounding checkpatch warnings]
Signed-off-by: Chris Wilson 
Signed-off-by: Greg Kroah-Hartman

drm/radeon/kms: reorder display resume to avoid problems

2011-01-07T21:58:45+00:00

commit a93f344d3c04e4b84490c65f2a574387c593be40 upstream.

On resume, we were attemping to unblank the displays before the
timing and plls had be reprogrammed which led to atom timeouts
waiting for things that are not yet programmed.  Re-program
the mode first, then reset the dpms state.

This fixes the infamous atombios timeouts on resume.

Signed-off-by: Alex Deucher 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

drm/radeon/kms: fix evergreen asic reset

2011-01-07T21:58:44+00:00

commit 9f0c4f9c2f835eee1bbb93f96bf9483d56f1892b upstream.

Only reset the grbm blocks, srbm tends to lock the GPU
if not done properly and in most cases is not necessary.
Also, no need to call asic init after reset the grbm blocks.

Signed-off-by: Alex Deucher 
Reviewed-by: Jerome Glisse 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

drm/radeon/kms/evergreen: reset the grbm blocks at resume and init

2011-01-07T21:58:44+00:00

commit 86f5c9edbb3bac37cc8cee6528a929005ba72aad upstream.

This fixes module reloading and resume as the gfx block seems to
be left in a bad state in some cases.

Signed-off-by: Alex Deucher 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

nouveau: Acknowledge HPD irq in handler, not bottom half

2011-01-07T21:58:30+00:00

commit ab838338a2a9e0cb8346eb0cab9977be13e8dce5 upstream.

The old code generated an interrupt storm bad enough to completely
take down my system.

Signed-off-by: Andy Lutomirski 
Signed-off-by: Ben Skeggs 
Signed-off-by: Greg Kroah-Hartman

drm/i915: Always set the DP transcoder config to 8BPC.

2011-01-07T21:58:15+00:00

commit 220cad3cbf553f893432919b458da36489373fc6 upstream.

The pipe is always set to 8BPC, but here we were leaving whatever
previous bits were set by the BIOS in place.

Signed-off-by: Eric Anholt 
Tested-by: Keith Packard 
Signed-off-by: Chris Wilson 
Signed-off-by: Greg Kroah-Hartman

drm/radeon/kms: fix vram base calculation on rs780/rs880

2011-01-07T21:58:14+00:00

commit 8961d52d4cc52edf5672f8f2712c57162b736793 upstream.

Avoid overflowing a 32 bit value.

Signed-off-by: Alex Deucher 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

drm/kms: remove spaces from connector names (v2)

2011-01-07T21:58:14+00:00

commit e76116ca9671e2e5239054a40303b94feab585ad upstream.

Grub doesn't parse spaces in parameters correctly, so
this makes it impossible to force video= parameters
for kms on the grub kernel command line.

v2: shorten the names to make them easier to type.

Reported-by: Sergej Pupykin 

Cc: Sergej Pupykin 
Signed-off-by: Alex Deucher 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

drm/radeon/kms: don't apply 7xx HDP flush workaround on AGP

2011-01-07T21:58:14+00:00

commit f3886f85cfde578f1d0ba6e40ac5f9d70043923b upstream.

It should be required for all 7xx asics, but seems to cause
problems on some AGP 7xx chips.

Fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=19002

Signed-off-by: Alex Deucher 
Reported-and-Tested-by: Duncan <1i5t5.duncan@cox.net>
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman