nouveau
15 years agoRemove all object with mmaps in CloseScreen, so that drmClose actually calls the...
Stuart Bennett [Wed, 22 Oct 2008 01:47:45 +0000 (02:47 +0100)] 
Remove all object with mmaps in CloseScreen, so that drmClose actually calls the drm release method

(and so subsequently, we can be DRM master on the next drmOpen)

15 years agoNo reason to call AccelCommonInit twice in ScreenInit, but calling it in EnterVT...
Stuart Bennett [Wed, 22 Oct 2008 02:32:51 +0000 (03:32 +0100)] 
No reason to call AccelCommonInit twice in ScreenInit, but calling it in EnterVT helps resume...

15 years agonouveau_channel_free should remove mmaps made in nouveau_channel_alloc
Stuart Bennett [Wed, 22 Oct 2008 01:41:17 +0000 (02:41 +0100)] 
nouveau_channel_free should remove mmaps made in nouveau_channel_alloc

15 years agoDo frag prog allocation and shader upload in TCL init
Stuart Bennett [Wed, 22 Oct 2008 01:15:06 +0000 (02:15 +0100)] 
Do frag prog allocation and shader upload in TCL init

Similar in spirit to the nv50 inplementation.
Removes many dependencies on state stored in static variables, which in turn
makes reinitializing shaders on resume / server regeneration tractable.

15 years agoOnly hackup shaders once
Stuart Bennett [Wed, 22 Oct 2008 00:44:10 +0000 (01:44 +0100)] 
Only hackup shaders once

15 years agocr26 is just another view of 0x3c0
Stuart Bennett [Sat, 18 Oct 2008 01:23:18 +0000 (02:23 +0100)] 
cr26 is just another view of 0x3c0

remove superfluous crtc unlock too

15 years agoUse symbolic define values where known
Stuart Bennett [Wed, 15 Oct 2008 02:59:06 +0000 (03:59 +0100)] 
Use symbolic define values where known

15 years agoConvert CIO, DIO and VIO use to use defines from nvreg for index and data reg offsets
Stuart Bennett [Tue, 14 Oct 2008 23:25:26 +0000 (00:25 +0100)] 
Convert CIO, DIO and VIO use to use defines from nvreg for index and data reg offsets

15 years agoRename relevant functions, sizes and offsets to PRM.IO from P.IO, in keeping with...
Stuart Bennett [Tue, 14 Oct 2008 23:12:11 +0000 (00:12 +0100)] 
Rename relevant functions, sizes and offsets to PRM.IO from P.IO, in keeping with the nvidia scheme

15 years agoName crtc index regs according to rules.xml
Stuart Bennett [Tue, 14 Oct 2008 22:58:11 +0000 (23:58 +0100)] 
Name crtc index regs according to rules.xml

15 years agoSaveGeneration isn't doing a lot
Stuart Bennett [Tue, 21 Oct 2008 16:46:37 +0000 (17:46 +0100)] 
SaveGeneration isn't doing a lot

add another close screen thing too

15 years agoReplace a few memsets with initializers
Stuart Bennett [Tue, 21 Oct 2008 15:27:00 +0000 (16:27 +0100)] 
Replace a few memsets with initializers

15 years agoEliminate separate NVRec ctor and dtor
Stuart Bennett [Sun, 12 Oct 2008 14:52:43 +0000 (15:52 +0100)] 
Eliminate separate NVRec ctor and dtor

15 years agorandr12: remove pointless debug in nv_output and nv_crtc
Stuart Bennett [Sat, 11 Oct 2008 20:05:57 +0000 (21:05 +0100)] 
randr12: remove pointless debug in nv_output and nv_crtc

15 years agorandr12: deBoolification and eliminate `override' arg to nv_crtc_load_state_ext
Stuart Bennett [Sat, 11 Oct 2008 17:40:35 +0000 (18:40 +0100)] 
randr12: deBoolification and eliminate `override' arg to nv_crtc_load_state_ext

15 years agoMake all CR 57/58 access use the proper functions and defines
Stuart Bennett [Sat, 11 Oct 2008 17:29:40 +0000 (18:29 +0100)] 
Make all CR 57/58 access use the proper functions and defines

Use a define for usleep in bios too

15 years agoRedo bios logging
Stuart Bennett [Sat, 11 Oct 2008 16:56:23 +0000 (17:56 +0100)] 
Redo bios logging

15 years agoImprove some nv_bios messages, remove others
Stuart Bennett [Sat, 11 Oct 2008 17:04:58 +0000 (18:04 +0100)] 
Improve some nv_bios messages, remove others

15 years agoBios register list is easier to read when sorted numerically
Stuart Bennett [Sat, 11 Oct 2008 16:59:02 +0000 (17:59 +0100)] 
Bios register list is easier to read when sorted numerically

and remove NULL initialization for calloc'd struct

15 years agorandr12: panels with edid have only a constant mode
Stuart Bennett [Sat, 11 Oct 2008 01:35:40 +0000 (02:35 +0100)] 
randr12: panels with edid have only a constant mode

prove me wrong.

15 years agoRemove unnecessary "Setting owner" message
Stuart Bennett [Sat, 11 Oct 2008 01:16:50 +0000 (02:16 +0100)] 
Remove unnecessary "Setting owner" message

15 years agoThe fifo channel should be removed along with closing of the fd, so only clear the...
Maarten Maathuis [Fri, 10 Oct 2008 23:22:35 +0000 (01:22 +0200)] 
The fifo channel should be removed along with closing of the fd, so only clear the client memory.

15 years agogit-log is no longer valid for git 1.6, switch to git log.
Maarten Maathuis [Fri, 10 Oct 2008 21:37:55 +0000 (23:37 +0200)] 
git-log is no longer valid for git 1.6, switch to git log.

15 years agoSome fixes that bring me closer to surviving to the 2nd X server generation.
Maarten Maathuis [Fri, 10 Oct 2008 21:14:57 +0000 (23:14 +0200)] 
Some fixes that bring me closer to surviving to the 2nd X server generation.

- Not quite done, i get a fifo dump when something tries to render.
- Something is keeping an fd, which is preventing a full takedown, but this is not a valid reason for failing (kernel modesetting
for example will never have full takedown).

15 years agorandr12: separate cursor pixmap location setting function
Stuart Bennett [Thu, 9 Oct 2008 16:09:58 +0000 (17:09 +0100)] 
randr12: separate cursor pixmap location setting function

15 years agoMove cursor show/hide funcs to nv_hw, document nv40 bug, and set curctl2 before apply...
Stuart Bennett [Thu, 9 Oct 2008 16:07:30 +0000 (17:07 +0100)] 
Move cursor show/hide funcs to nv_hw, document nv40 bug, and set curctl2 before applying nv40 fix

15 years agoUse correct refclk for nv_get_clock, and use bool type as appropriate in arbitration
Stuart Bennett [Thu, 9 Oct 2008 16:24:23 +0000 (17:24 +0100)] 
Use correct refclk for nv_get_clock, and use bool type as appropriate in arbitration

15 years agoTidy arbitration prototypes and split randr12 cursor pixmap setting from arbitration
Stuart Bennett [Thu, 9 Oct 2008 14:25:16 +0000 (15:25 +0100)] 
Tidy arbitration prototypes and split randr12 cursor pixmap setting from arbitration

Also fix ages old warning in nv_hw

15 years agoMerge nv4, nv10 and nForce UpdateArbitrationSettings functions
Stuart Bennett [Fri, 10 Oct 2008 00:11:46 +0000 (01:11 +0100)] 
Merge nv4, nv10 and nForce UpdateArbitrationSettings functions

15 years agoAdd PCI slot reading defines, and use them for nForce code
Stuart Bennett [Thu, 9 Oct 2008 23:57:45 +0000 (00:57 +0100)] 
Add PCI slot reading defines, and use them for nForce code

Also move nForce dimm checking code to nv_setup, so it is only tested once

15 years agoSome tidying of arbitration functions
Stuart Bennett [Thu, 9 Oct 2008 19:22:34 +0000 (20:22 +0100)] 
Some tidying of arbitration functions

15 years agoReindent arbitration stuff
Stuart Bennett [Thu, 9 Oct 2008 18:05:01 +0000 (19:05 +0100)] 
Reindent arbitration stuff

15 years agoWrap usleep when called in bios code
Stuart Bennett [Wed, 8 Oct 2008 22:45:13 +0000 (23:45 +0100)] 
Wrap usleep when called in bios code

15 years agorandr12: don't use in use (by another vga output) crtc for load detect (avoids flicker)
Stuart Bennett [Fri, 10 Oct 2008 15:41:29 +0000 (16:41 +0100)] 
randr12: don't use in use (by another vga output) crtc for load detect (avoids flicker)

15 years agoBetter behaviour on allocator fail
Stuart Bennett [Tue, 7 Oct 2008 20:39:10 +0000 (21:39 +0100)] 
Better behaviour on allocator fail

also fix randr12 code which doesn't know how to call calloc
and be sure to use the number of actually inited crtcs

15 years agorandr12: reorder nv_output functions
Stuart Bennett [Tue, 7 Oct 2008 01:35:44 +0000 (02:35 +0100)] 
randr12: reorder nv_output functions

roughly into order of use

15 years agorandr12: distinguish detected encoder from active encoder
Stuart Bennett [Thu, 18 Sep 2008 17:32:36 +0000 (18:32 +0100)] 
randr12: distinguish detected encoder from active encoder

don't swap output props until a mode is set on new encoder either

15 years agoReduce includes, everything necessary is in nv_include.h
Stuart Bennett [Wed, 8 Oct 2008 00:22:17 +0000 (01:22 +0100)] 
Reduce includes, everything necessary is in nv_include.h

15 years agoOops.
Stuart Bennett [Sat, 20 Sep 2008 23:52:56 +0000 (00:52 +0100)] 
Oops.

15 years agoOptimizing memory usage for unaccelerated case is not interesting
Stuart Bennett [Sat, 20 Sep 2008 23:44:50 +0000 (00:44 +0100)] 
Optimizing memory usage for unaccelerated case is not interesting

15 years agorandr12: rename mon -> edid
Stuart Bennett [Sat, 20 Sep 2008 23:43:57 +0000 (00:43 +0100)] 
randr12: rename mon -> edid

15 years agorandr12: crtc destroy
Stuart Bennett [Sat, 20 Sep 2008 23:31:40 +0000 (00:31 +0100)] 
randr12: crtc destroy

also alleviate world shortage of curly braces

15 years agoNeed to increment dcb entry index when fabricating additional entries
Stuart Bennett [Sat, 20 Sep 2008 22:57:06 +0000 (23:57 +0100)] 
Need to increment dcb entry index when fabricating additional entries

15 years agoAlso remove man page entry.
Maarten Maathuis [Fri, 19 Sep 2008 19:23:04 +0000 (21:23 +0200)] 
Also remove man page entry.

15 years agoNV50: remove {Prepare,Finish}Access since i realise now it's not so useful
Maarten Maathuis [Fri, 19 Sep 2008 19:14:09 +0000 (21:14 +0200)] 
NV50: remove {Prepare,Finish}Access since i realise now it's not so useful

- despite that the xserver code suggests that pixmaps can be offscreen on this call, that is not done as far as i can tell
- not to mention that there is no way to know for sure if a pixmap is offscreen (exaPixmapIsOffscreen is not exported), making the
offset useless
- this will (obviously) become relevant as soon as driver managed pixmaps come into play, but that is for another day

15 years agoNV50: The TMDS dual link threshold is meaningless for LVDS
Maarten Maathuis [Fri, 19 Sep 2008 18:41:59 +0000 (20:41 +0200)] 
NV50: The TMDS dual link threshold is meaningless for LVDS

15 years agorandr12: multiple encoders per connector (DVI-I)
Stuart Bennett [Mon, 25 Aug 2008 00:47:45 +0000 (01:47 +0100)] 
randr12: multiple encoders per connector (DVI-I)

that this doesn't break things in subtle and unexpected ways is too much to
hope for. one not-subtle and expected way things will break is that xorg.confs
with specific setups for DVI-A-n and DVI-D-n connections will need renaming (to
DVI-I-n).

as the last patch in the series, this also fixes an issue introduced with
144b04a2 (randr12: fix dpms, detect, destroy, ...), that for a screen connected
to a dvi connector, both the analogue and digital variants were detected as
connected

15 years agonv50: 0xa0 family uses class 0x8397 for 3D
Ben Skeggs [Wed, 17 Sep 2008 12:12:07 +0000 (22:12 +1000)] 
nv50: 0xa0 family uses class 0x8397 for 3D

15 years agorandr12: avoid weirdness when tv-out happens to share i2c_index with another output
Stuart Bennett [Sun, 14 Sep 2008 22:01:49 +0000 (23:01 +0100)] 
randr12: avoid weirdness when tv-out happens to share i2c_index with another output

15 years agorandr12: fix dpms, detect, destroy, save and restore for multiple encoders per connector
Stuart Bennett [Sun, 24 Aug 2008 03:15:20 +0000 (04:15 +0100)] 
randr12: fix dpms, detect, destroy, save and restore for multiple encoders per connector

15 years agorandr12: re-do detection of dvi-a vs vga
Stuart Bennett [Sun, 24 Aug 2008 03:26:19 +0000 (04:26 +0100)] 
randr12: re-do detection of dvi-a vs vga

15 years agorandr12: unify output funcs for analogue and tmds
Stuart Bennett [Sun, 24 Aug 2008 02:56:16 +0000 (03:56 +0100)] 
randr12: unify output funcs for analogue and tmds

15 years agorandr12: unified entry path for dpms
Stuart Bennett [Sun, 24 Aug 2008 02:12:58 +0000 (03:12 +0100)] 
randr12: unified entry path for dpms

15 years agorandr12: nouveau_output -> nouveau_connector
Stuart Bennett [Sun, 24 Aug 2008 03:45:48 +0000 (04:45 +0100)] 
randr12: nouveau_output -> nouveau_connector

15 years agorandr12: separate encoder struct
Stuart Bennett [Sun, 24 Aug 2008 01:45:41 +0000 (02:45 +0100)] 
randr12: separate encoder struct

15 years agoAnother DCB 1.4/1.5 TV (#17471)
Stuart Bennett [Sun, 14 Sep 2008 15:53:12 +0000 (16:53 +0100)] 
Another DCB 1.4/1.5 TV (#17471)

I think we've got enough for TVs to no longer be fatal

15 years agoOnly use OF bios image on PPC
Stuart Bennett [Sun, 14 Sep 2008 15:44:32 +0000 (16:44 +0100)] 
Only use OF bios image on PPC

15 years agonv50/exa: work around corruption issues (see detailed commit msg)
Ben Skeggs [Wed, 10 Sep 2008 17:29:46 +0000 (03:29 +1000)] 
nv50/exa: work around corruption issues (see detailed commit msg)

It's not possible for us to do this properly without using the exa
pixmap hooks, they require a real memory manager, and is being
worked on in the ng branches, which doesn't currently support nv5x
at all.

For a better solution until then:
  http://nouveau.freedesktop.org/wiki/NV50Support

15 years agoPartly revert ea152819f45b6cf92f6742ed3f9f639cdbbd53f0, alignment is needed.
Stephane Marchesin [Fri, 5 Sep 2008 21:08:14 +0000 (23:08 +0200)] 
Partly revert ea152819f45b6cf92f6742ed3f9f639cdbbd53f0, alignment is needed.

15 years agoexa: align offscreenBase to something sensible
Maarten Maathuis [Fri, 5 Sep 2008 20:51:27 +0000 (22:51 +0200)] 
exa: align offscreenBase to something sensible
- Aligning virtualY makes no sense whatsoever.
- displayWidth is the "pitch-width" of the framebuffer.

15 years agoAdd 8200 detection.
Stephane Marchesin [Thu, 4 Sep 2008 05:51:03 +0000 (07:51 +0200)] 
Add 8200 detection.

15 years agonv50: solid fill shouldn't use a pattern rop + minor changes
Maarten Maathuis [Sat, 30 Aug 2008 15:51:05 +0000 (17:51 +0200)] 
nv50: solid fill shouldn't use a pattern rop + minor changes

15 years agorandr12: merge output detection functions
Stuart Bennett [Sat, 23 Aug 2008 23:18:40 +0000 (00:18 +0100)] 
randr12: merge output detection functions

15 years agorandr12: remember the edid from detection for getting modes later
Stuart Bennett [Sat, 23 Aug 2008 23:08:41 +0000 (00:08 +0100)] 
randr12: remember the edid from detection for getting modes later

also split native mode finding into separate function

15 years agorandr12: detypedef private structs and use a define to get to the privates
Stuart Bennett [Sun, 24 Aug 2008 18:56:24 +0000 (19:56 +0100)] 
randr12: detypedef private structs and use a define to get to the privates

15 years agoRemove unused code
Stuart Bennett [Sat, 23 Aug 2008 23:31:35 +0000 (00:31 +0100)] 
Remove unused code

15 years agorandr12: make the nouveau experience less green
Stuart Bennett [Mon, 25 Aug 2008 19:23:05 +0000 (20:23 +0100)] 
randr12: make the nouveau experience less green

15 years agonv50: micro cleanup
Maarten Maathuis [Sat, 23 Aug 2008 01:44:28 +0000 (03:44 +0200)] 
nv50: micro cleanup

15 years agonv50: do ROPs properly this time
Maarten Maathuis [Sat, 23 Aug 2008 01:39:39 +0000 (03:39 +0200)] 
nv50: do ROPs properly this time
- Table should work for pre-nv50 as well, i just don't have a card plugged in to be sure.

15 years agonv10: composite is an async operation by default, no need to call exaMarkSync().
Maarten Maathuis [Fri, 22 Aug 2008 15:56:31 +0000 (17:56 +0200)] 
nv10: composite is an async operation by default, no need to call exaMarkSync().

15 years agoexa: FIRE_RING at the end of composite
Maarten Maathuis [Fri, 22 Aug 2008 15:53:11 +0000 (17:53 +0200)] 
exa: FIRE_RING at the end of composite

15 years agonv50: move VERTEX_{START,END} to {Prepare,Done}Composite
Maarten Maathuis [Thu, 21 Aug 2008 21:17:06 +0000 (23:17 +0200)] 
nv50: move VERTEX_{START,END} to {Prepare,Done}Composite
- If PrepareComposite returns TRUE, then everything upto DoneComposite is required.

15 years agonv50: minor tweaks to composite
Maarten Maathuis [Thu, 21 Aug 2008 20:50:49 +0000 (22:50 +0200)] 
nv50: minor tweaks to composite
- Make the blend stuff more consistent with other composite implementations.
- Verified with "rendercheck -f a8r8g8b8,x8r8g8b8,r5g6b5,a8 -t {blend,composite,cacomposite}" for all accelerated ops.

15 years agonv50: handle ROPs better
Maarten Maathuis [Thu, 21 Aug 2008 18:01:49 +0000 (20:01 +0200)] 
nv50: handle ROPs better
- Handle ROP's more like pre-NV50 code.
- Special case for 32bpp which seems to work.
- This does NOT fix all rendering issues, just an improvement.

15 years agonv50: clip SIFC and add {Prepare,Finish}Access hooks.
Maarten Maathuis [Wed, 20 Aug 2008 17:14:54 +0000 (19:14 +0200)] 
nv50: clip SIFC and add {Prepare,Finish}Access hooks.

- when w*cpp is not aligned to a dword, then it's possible to end up with crap at the end of a line, so clip it.
- {Prepare,Finish}Access hooks force tiled memory to be migrated upon cpu access.
- Added an option to disable these hooks in case you hit a xserver bug that existed up until very recently.

15 years agorandr12: don't cache pll values (fixes newrestore removal regression)
Stuart Bennett [Mon, 18 Aug 2008 23:02:00 +0000 (00:02 +0100)] 
randr12: don't cache pll values (fixes newrestore removal regression)

for 5ms, it's really not worth it

15 years agoRemove some of the inaccuracies in the manpage
Stuart Bennett [Mon, 18 Aug 2008 20:50:37 +0000 (21:50 +0100)] 
Remove some of the inaccuracies in the manpage

15 years agoNV_ARCH_04 can have rotated output, it's just not accelerated
Stuart Bennett [Mon, 18 Aug 2008 20:40:36 +0000 (21:40 +0100)] 
NV_ARCH_04 can have rotated output, it's just not accelerated

15 years agoDelete ShadowFB rotation
Stuart Bennett [Mon, 18 Aug 2008 20:39:18 +0000 (21:39 +0100)] 
Delete ShadowFB rotation

And nothing of value was lost.

15 years agoBye bye 8 bit depth
Stuart Bennett [Sun, 17 Aug 2008 23:59:32 +0000 (00:59 +0100)] 
Bye bye 8 bit depth

15 years agoCrtcNumber option is long gone
Stuart Bennett [Sun, 17 Aug 2008 22:52:28 +0000 (23:52 +0100)] 
CrtcNumber option is long gone

15 years agoMicro-tidyups to nv_driver and nouveau_xv
Stuart Bennett [Sun, 17 Aug 2008 22:17:01 +0000 (23:17 +0100)] 
Micro-tidyups to nv_driver and nouveau_xv

15 years agorandr12: a few line length improvements
Stuart Bennett [Sun, 17 Aug 2008 15:57:17 +0000 (16:57 +0100)] 
randr12: a few line length improvements

15 years agoKill some 8 bit code.
Stephane Marchesin [Sun, 17 Aug 2008 20:44:11 +0000 (22:44 +0200)] 
Kill some 8 bit code.

15 years agoFix IFC for 16bpp. Also fix the color accuracy of 16bpp solid fills.
Stephane Marchesin [Sun, 17 Aug 2008 20:25:43 +0000 (22:25 +0200)] 
Fix IFC for 16bpp. Also fix the color accuracy of 16bpp solid fills.

15 years ago0x48 exception does not apply on BIT cards
Stuart Bennett [Sun, 17 Aug 2008 02:51:39 +0000 (03:51 +0100)] 
0x48 exception does not apply on BIT cards

15 years agorandr12: some code simplification, rearrangement and tidying
Stuart Bennett [Sun, 17 Aug 2008 01:48:33 +0000 (02:48 +0100)] 
randr12: some code simplification, rearrangement and tidying

We don't claim to support a bpp of 15 or 24, so some old calculations can be simplified

15 years agoNewRestore: remove, due to lack of interest
Stuart Bennett [Sun, 17 Aug 2008 01:27:42 +0000 (02:27 +0100)] 
NewRestore: remove, due to lack of interest

nobody wants to do text modes manually. for any future masochists, it's here in the history

15 years agorandr12: remove another field member
Stuart Bennett [Sun, 17 Aug 2008 00:37:41 +0000 (01:37 +0100)] 
randr12: remove another field member

15 years agorandr12: fpWidth and fpHeight no more
Stuart Bennett [Sun, 17 Aug 2008 00:26:20 +0000 (01:26 +0100)] 
randr12: fpWidth and fpHeight no more

15 years agorandr12: improve aspect scaling code
Stuart Bennett [Sat, 16 Aug 2008 23:38:23 +0000 (00:38 +0100)] 
randr12: improve aspect scaling code

15 years agorandr12: for native scaling, a clock check isn't enough
Stuart Bennett [Sat, 16 Aug 2008 22:15:29 +0000 (23:15 +0100)] 
randr12: for native scaling, a clock check isn't enough

1280x1024 and 1280x960 can have the same clock

15 years agoApply minimum front porch only when necessary (#15949)
Stuart Bennett [Fri, 15 Aug 2008 22:02:48 +0000 (23:02 +0100)] 
Apply minimum front porch only when necessary (#15949)

Also, read minimum front porch from bios table

15 years agoNV50: disallow doublescan modes on TMDS/LVDS
Maarten Maathuis [Sat, 16 Aug 2008 15:18:21 +0000 (17:18 +0200)] 
NV50: disallow doublescan modes on TMDS/LVDS

15 years agoNV50: extra warning message
Maarten Maathuis [Sat, 16 Aug 2008 13:33:13 +0000 (15:33 +0200)] 
NV50: extra warning message

15 years agoNV50: a half decent attempt at doing something when LVDS has no DDC
Maarten Maathuis [Sat, 16 Aug 2008 13:28:43 +0000 (15:28 +0200)] 
NV50: a half decent attempt at doing something when LVDS has no DDC

15 years agoForgot something.
Maarten Maathuis [Sat, 16 Aug 2008 12:35:05 +0000 (14:35 +0200)] 
Forgot something.

15 years agoNV50: support extended i2c ports (4 and 5)
Maarten Maathuis [Sat, 16 Aug 2008 12:24:18 +0000 (14:24 +0200)] 
NV50: support extended i2c ports (4 and 5)

15 years agonv50: fix some of the rendering bugs
Maarten Maathuis [Sat, 16 Aug 2008 12:02:51 +0000 (14:02 +0200)] 
nv50: fix some of the rendering bugs

- It seems that 32bpp solid fill only works well for for the non-rop case.
- The SRCCOPY_PREMULT comes from nv and seems to work well, in combination with the above fallback.
- This does not affect the strange stripes with font rendering issue.

15 years agoRedo parsing for panels with EDID, and fix nv3x in the process (#17138)
Stuart Bennett [Fri, 15 Aug 2008 00:40:01 +0000 (01:40 +0100)] 
Redo parsing for panels with EDID, and fix nv3x in the process (#17138)

breakage a possibility :)