GitHub/exynos8895/android_kernel_samsung_universal8895.git
14 years agodrm/radeon/kms: use active device to pick connector for encoder
Dave Airlie [Fri, 29 Jan 2010 05:55:30 +0000 (15:55 +1000)]
drm/radeon/kms: use active device to pick connector for encoder

On the W500 we have UNIPHY routed to both DVI and DP, this seems
to always pick the DVI connector which means link training fails.

Switch to using active device to pick the connector, this seems
like it should be safe from a code review, and it fixes things
a bit more here.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: fix incorrect logic in DP vs eDP connector checking.
Dave Airlie [Fri, 29 Jan 2010 05:31:47 +0000 (15:31 +1000)]
drm/radeon/kms: fix incorrect logic in DP vs eDP connector checking.

This makes displayport work again here.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: preface warning printk with driver name
Dave Airlie [Mon, 25 Jan 2010 06:13:55 +0000 (16:13 +1000)]
drm/radeon/kms: preface warning printk with driver name

This just adds a little more info to the warning for old -ati/mesa
userspaces.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: drop unnecessary printks.
Dave Airlie [Mon, 25 Jan 2010 06:13:12 +0000 (16:13 +1000)]
drm/radeon/kms: drop unnecessary printks.

These printks aren't required anymore.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm: fix regression in fb blank handling
Zhenyu Wang [Mon, 18 Jan 2010 08:47:04 +0000 (16:47 +0800)]
drm: fix regression in fb blank handling

commit 731b5a15a3b1474a41c2ca29b4c32b0f21bc852e
Author: James Simmons <jsimmons@infradead.org>
Date:   Thu Oct 29 20:39:07 2009 +0000

    drm/kms: properly handle fbdev blanking

uses DRM_MODE_DPMS_ON for FB_BLANK_NORMAL, but DRM_MODE_DPMS_ON
is actually for turning output on instead of blank.

This makes fb blank broken on my T61, it put LVDS on but leave
pipe disabled which made screen totally white or caused some
'burning' effect.

[airlied: James objects to this but at this point in 2.6.33,
I can't see a patch that will fix this properly like he wants coming
in time and otherwise this is a regression - proper fix for 2.6.34
hopefully.]

Cc: James Simmons <jsimmons@infradead.org>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: make hibernate work on IGPs
Dave Airlie [Mon, 25 Jan 2010 03:08:08 +0000 (13:08 +1000)]
drm/radeon/kms: make hibernate work on IGPs

This is the least invasive fix without migrating the radeon driver
to pm_ops from what I can see. We just always migrate VRAM objects
on IGPs for now and we can fix it up later to migrate depending
on STR vs STD.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/vmwgfx: Optimize memory footprint for DMA buffers.
Thomas Hellstrom [Sat, 16 Jan 2010 15:05:05 +0000 (16:05 +0100)]
drm/vmwgfx: Optimize memory footprint for DMA buffers.

Use VRAM whenever there is free space for DMA buffers,
but use system GMR memory if using VRAM would cause an eviction.

This significantly reduces the guest system memory usage for
VMs with a large amount of VRAM allocated.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/ttm: Allow system memory as a busy placement.
Thomas Hellstrom [Sat, 16 Jan 2010 15:05:04 +0000 (16:05 +0100)]
drm/ttm: Allow system memory as a busy placement.

This is needed to fix a vmwgfx memory usage bug.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agoMerge remote branch 'korg/drm-radeon-next' into drm-linus
Dave Airlie [Mon, 25 Jan 2010 06:04:21 +0000 (16:04 +1000)]
Merge remote branch 'korg/drm-radeon-next' into drm-linus

* korg/drm-radeon-next:
  drm/radeon/kms: fix legacy get_engine/memory clock
  drm/radeon/kms/atom: atom parser fixes
  drm/radeon/kms: clean up atombios pll code
  drm/radeon/kms: clean up pll struct
  drm/radeon/kms/atom: fix crtc lock ordering
  drm/radeon: r6xx/r7xx possible security issue, system ram access
  drm/radeon/kms: r600/r700 don't test ib if ib initialization fails
  drm/radeon/kms: Forbid creation of framebuffer with no valid GEM object
  drm/radeon/kms: r600 handle irq vector ring overflow
  drm/radeon/kms: r600/r700 don't process IRQ if not initialized
  drm/radeon/kms: r600/r700 disable irq at suspend
  drm/radeon/kms/r4xx: cleanup atom path
  drm/radeon/kms: fix atombios_crtc_set_base
  drm/radeon/kms/atom: upstream parser updates
  drm/radeon/kms/atom: fix some parser bugs
  drm/radeon/kms: fix hardcoded mmio size in register functions
  drm/radeon/kms/r100: fix bug in CS parser
  drm/radeon/kms/r200: fix bug in CS parser
  drm/radeon/kms/r200: fix bug in CS parser

14 years agoMerge remote branch 'nouveau/for-airlied' of ../drm-nouveau-next into drm-linus
Dave Airlie [Mon, 25 Jan 2010 06:04:11 +0000 (16:04 +1000)]
Merge remote branch 'nouveau/for-airlied' of ../drm-nouveau-next into drm-linus

* 'nouveau/for-airlied' of ../drm-nouveau-next:
  drm/nv50: prevent switching off SOR when in use for DVI-over-DP
  drm/nv50: fail auxch transaction if reply count not what we expect
  drm/nouveau: fix failure path if userspace specifies no valid memtypes
  drm/nouveau: report LVDS as disconnected if lid closed
  drm/nv50: prevent accidently turning off encoders we're actually using
  drm/nv50: fix alignment of per-channel fifo cache
  drm/nouveau: Evict buffers in VRAM before freeing sgdma
  drm/nouveau: Acknowledge DMA_VTX_PROTECTION PGRAPH interrupts
  drm/nouveau: fix thinko in nv04_instmem.c
  drm/nouveau: fix a race condition in nouveau_dma_wait()

14 years agodrm/ttm: Fix race condition in ttm_bo_delayed_delete (v3, final)
Luca Barbieri [Wed, 20 Jan 2010 19:01:30 +0000 (20:01 +0100)]
drm/ttm: Fix race condition in ttm_bo_delayed_delete (v3, final)

Resending this with Thomas Hellstrom's signoff for merging into 2.6.33

ttm_bo_delayed_delete has a race condition, because after we do:
kref_put(&nentry->list_kref, ttm_bo_release_list);

we are not holding the list lock and not holding any reference to
objects, and thus every bo in the list can be removed and freed at
this point.

However, we then use the next pointer we stored, which is not guaranteed
to be valid.

This was apparently the cause of some Nouveau oopses I experienced.

This patch rewrites the function so that it keeps the reference to nentry
until nentry itself is freed and we already got a reference to nentry->next.

v2 updated by me according to Thomas Hellstrom's feedback.
v3 proposed by Thomas Hellstrom. Commit comment updated by me.

Both updates fixed minor efficiency/style issues only and all three versions
should be correct.

Signed-off-by: Luca Barbieri <luca@luca-barbieri.com>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/nv50: prevent switching off SOR when in use for DVI-over-DP
Ben Skeggs [Fri, 22 Jan 2010 00:57:01 +0000 (10:57 +1000)]
drm/nv50: prevent switching off SOR when in use for DVI-over-DP

Another hack because of us exposing each encoder block's function as
an encoder rather than exposing a single encoder that deals with them
all.

A proper fix will come, it's just rather invasive so this hack will
do until then.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nv50: fail auxch transaction if reply count not what we expect
Ben Skeggs [Thu, 21 Jan 2010 23:10:05 +0000 (09:10 +1000)]
drm/nv50: fail auxch transaction if reply count not what we expect

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: fix failure path if userspace specifies no valid memtypes
Ben Skeggs [Thu, 21 Jan 2010 05:03:23 +0000 (15:03 +1000)]
drm/nouveau: fix failure path if userspace specifies no valid memtypes

We need to add the buffer to the list even if we fail, otherwise the
validate_fini() call won't unreserve + unreference the GEM object,
making TTM very unhappy.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: report LVDS as disconnected if lid closed
Ben Skeggs [Mon, 18 Jan 2010 01:42:37 +0000 (11:42 +1000)]
drm/nouveau: report LVDS as disconnected if lid closed

Also adds a module option to ignore the status reported via ACPI, in case
we hit systems with broken ACPI.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/radeon/kms: fix legacy get_engine/memory clock
Alex Deucher [Fri, 22 Jan 2010 00:28:18 +0000 (19:28 -0500)]
drm/radeon/kms: fix legacy get_engine/memory clock

Fix a bad shift in the post div.

Should fix fdo bug 26145

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms/atom: atom parser fixes
Alex Deucher [Thu, 21 Jan 2010 22:14:49 +0000 (17:14 -0500)]
drm/radeon/kms/atom: atom parser fixes

Only reset the reg block on the initial execute
table call; nested calls require the reg block not be
reset on each call.  Also reset the fb window and
io mode.  This matches the upstream parser behavior.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: clean up atombios pll code
Alex Deucher [Tue, 19 Jan 2010 22:32:27 +0000 (17:32 -0500)]
drm/radeon/kms: clean up atombios pll code

- split pll adjust into a separate function
- use a union for SetPixelClock params

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: clean up pll struct
Alex Deucher [Tue, 19 Jan 2010 22:16:10 +0000 (17:16 -0500)]
drm/radeon/kms: clean up pll struct

- add a new flag for fixed post div
- pull the pll flags into the struct

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms/atom: fix crtc lock ordering
Alex Deucher [Thu, 21 Jan 2010 21:50:30 +0000 (16:50 -0500)]
drm/radeon/kms/atom: fix crtc lock ordering

This makes crtc_prepare and crtc_commit match.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon: r6xx/r7xx possible security issue, system ram access
Jerome Glisse [Mon, 18 Jan 2010 12:01:36 +0000 (13:01 +0100)]
drm/radeon: r6xx/r7xx possible security issue, system ram access

This patch workaround a possible security issue which can allow
user to abuse drm on r6xx/r7xx hw to access any system ram memory.
This patch doesn't break userspace, it detect "valid" old use of
CB_COLOR[0-7]_FRAG & CB_COLOR[0-7]_TILE registers and overwritte
the address these registers are pointing to with the one of the
last color buffer. This workaround will work for old mesa &
xf86-video-ati and any old user which did use similar register
programming pattern as those (we expect that there is no others
user of those ioctl except possibly a malicious one). This patch
add a warning if it detects such usage, warning encourage people
to update their mesa & xf86-video-ati. New userspace will submit
proper relocation.

Fix for xf86-video-ati / mesa (this kernel patch is enough to
prevent abuse, fix for userspace are to set proper cs stream and
avoid kernel warning) :
http://cgit.freedesktop.org/xorg/driver/xf86-video-ati/commit/?id=95d63e408cc88b6934bec84a0b1ef94dfe8bee7b
http://cgit.freedesktop.org/mesa/mesa/commit/?id=46dc6fd3ed5ef96cda53641a97bc68c3bc104a9f

Abusing this register to perform system ram memory is not easy,
here is outline on how it could be achieve. First attacker must
have access to the drm device and be able to submit command stream
throught cs ioctl. Then attacker must build a proper command stream
for r6xx/r7xx hw which will abuse the FRAG or TILE buffer to
overwrite the GPU GART which is in VRAM. To achieve so attacker
as to setup CB_COLOR[0-7]_FRAG or CB_COLOR[0-7]_TILE to point
to the GPU GART, then it has to find a way to write predictable
value into those buffer (with little cleverness i believe this
can be done but this is an hard task). Once attacker have such
program it can overwritte GPU GART to program GPU gart to point
anywhere in system memory. It then can reusse same method as he
used to reprogram GART to overwritte the system ram through the
GART mapping. In the process the attacker has to be carefull to
not overwritte any sensitive area of the GART table, like ring
or IB gart entry as it will more then likely lead to GPU lockup.
Bottom line is that i think it's very hard to use this flaw
to get system ram access but in theory one can achieve so.

Side note: I am not aware of anyone ever using the GPU as an
attack vector, nevertheless we take great care in the opensource
driver to try to detect and forbid malicious use of GPU. I don't
think the closed source driver are as cautious as we are.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms: r600/r700 don't test ib if ib initialization fails
Jerome Glisse [Sun, 17 Jan 2010 20:21:56 +0000 (21:21 +0100)]
drm/radeon/kms: r600/r700 don't test ib if ib initialization fails

If ib initialization failed don't try to test ib as it will result
in an oops (accessing NULL ib buffer ptr).

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms: Forbid creation of framebuffer with no valid GEM object
Jerome Glisse [Sun, 17 Jan 2010 20:21:41 +0000 (21:21 +0100)]
drm/radeon/kms: Forbid creation of framebuffer with no valid GEM object

This will avoid oops if at later point the fb is use. Trying to create
a framebuffer with no valid GEM object is bogus and should be forbidden
as this patch does.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms: r600 handle irq vector ring overflow
Jerome Glisse [Fri, 15 Jan 2010 13:44:39 +0000 (14:44 +0100)]
drm/radeon/kms: r600 handle irq vector ring overflow

In some rare case i faced an irq overflow quickly followed by
a GPU lockup (hard hang) this patch try to deal with irq vector
ring overflow, so far haven't been able to reproduce it with
the patch.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms: r600/r700 don't process IRQ if not initialized
Jerome Glisse [Fri, 15 Jan 2010 13:44:38 +0000 (14:44 +0100)]
drm/radeon/kms: r600/r700 don't process IRQ if not initialized

In some rare case the wptr returned from the hw wasn't 0 and leaded
to trick r600_process_irq that their were irq to process. Add a
check to bail out if irq hasn't been initialized this will avoid
oops provoqued by the rare wptr != 0 on initialization.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms: r600/r700 disable irq at suspend
Jerome Glisse [Fri, 15 Jan 2010 13:44:37 +0000 (14:44 +0100)]
drm/radeon/kms: r600/r700 disable irq at suspend

To avoid hw doing anythings after we disabled PCIE GART, fully
disable IRQ at suspend. Also cleanup a bit the ih structure
and process function.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms/r4xx: cleanup atom path
Alex Deucher [Wed, 20 Jan 2010 21:22:53 +0000 (16:22 -0500)]
drm/radeon/kms/r4xx: cleanup atom path

most of radeon_legacy_atom_set_surface() is taken care
of in atombios_set_base(), so remove the duplicate
setup and move the remaining bits (DISP_MERGE setup and
FP2 sync) to atombios_crtc.c where they are used.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms: fix atombios_crtc_set_base
Alex Deucher [Tue, 19 Jan 2010 21:34:01 +0000 (16:34 -0500)]
drm/radeon/kms: fix atombios_crtc_set_base

Make it call the proper backend depending on the
GPU family.  Right now r4xx cards with atombios modesetting
enabled were using the avivo crtc base code.  This also
allows us to add support for new asics more easily.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms/atom: upstream parser updates
Alex Deucher [Tue, 19 Jan 2010 17:45:29 +0000 (12:45 -0500)]
drm/radeon/kms/atom: upstream parser updates

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms/atom: fix some parser bugs
Alex Deucher [Tue, 19 Jan 2010 17:38:48 +0000 (12:38 -0500)]
drm/radeon/kms/atom: fix some parser bugs

- add support for inline src params
- fix shift_left/shift_right and shl/shr ops
  shift_* ops use inline src params, shl/r use full params
- fix mask op (uses inline params)

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms: fix hardcoded mmio size in register functions
Alex Deucher [Thu, 14 Jan 2010 00:09:12 +0000 (19:09 -0500)]
drm/radeon/kms: fix hardcoded mmio size in register functions

newer asics have large mmio apertures

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms/r100: fix bug in CS parser
Alex Deucher [Tue, 19 Jan 2010 01:20:07 +0000 (20:20 -0500)]
drm/radeon/kms/r100: fix bug in CS parser

The first dword of PACKET3_3D_DRAW_IMMD maps to
SE_VTX_FMT so the vertex size is part of the draw
packet.

This patch fixes a possible case where you have a
command buffer that does not contain SE_VTX_FMT
register write, but does contain PACKET3_3D_DRAW_IMMD.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms/r200: fix bug in CS parser
Andrew Randrianasulu [Wed, 20 Jan 2010 16:56:07 +0000 (11:56 -0500)]
drm/radeon/kms/r200: fix bug in CS parser

Add missing vertex shader regs for r200.

fixed fdo bug 26061

agd5f: use official reg names

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/radeon/kms/r200: fix bug in CS parser
Andrew Randrianasulu [Wed, 20 Jan 2010 16:36:30 +0000 (11:36 -0500)]
drm/radeon/kms/r200: fix bug in CS parser

The checks for CUBE and 3D textures were inverted.

fixes fdo bug 24159

agd5f: added comments for clarity.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
14 years agodrm/nv50: prevent accidently turning off encoders we're actually using
Ben Skeggs [Sun, 17 Jan 2010 22:52:35 +0000 (08:52 +1000)]
drm/nv50: prevent accidently turning off encoders we're actually using

On most cards the DisplayPort connector is created with 2 encoders sharing
a single SOR (for native DP, and for DVI-over-DP).  The previous logic
for turning off unused encoders didn't take into account that we could
have multiple drm_encoders on a single hw encoder and ended up turning off
encoders that were actually being used still.

This patch fixes that issue.  We probably want to look at something a bit
better later on, and only expose one drm_encoder per hw encoder block.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nv50: fix alignment of per-channel fifo cache
Ben Skeggs [Sun, 17 Jan 2010 22:33:04 +0000 (08:33 +1000)]
drm/nv50: fix alignment of per-channel fifo cache

GPU pointer to the structure is shifted right by 10 bits, so we need to
align to 1024 bytes, not 256.

Reported-by: Maarten Maathuis <madman2003@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: Evict buffers in VRAM before freeing sgdma
Luca Barbieri [Sat, 16 Jan 2010 14:30:15 +0000 (15:30 +0100)]
drm/nouveau: Evict buffers in VRAM before freeing sgdma

Currently, we take down the sgdma engine without evicting all buffers
from VRAM.

The TTM device release will try to evict anything in VRAM to GART
memory, but this will fail since sgdma has already been taken down.

This causes an infinite loop in kernel mode on module unload.
It usually doesn't happen because there aren't any buffer on close.
However, if the GPU is locked up, this condition is easily triggered.

This patch fixes it in the simplest way possible by cleaning VRAM
right before cleaning SGDMA memory.

Signed-off-by: Luca Barbieri <luca@luca-barbieri.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: Acknowledge DMA_VTX_PROTECTION PGRAPH interrupts
Luca Barbieri [Sat, 16 Jan 2010 14:27:51 +0000 (15:27 +0100)]
drm/nouveau: Acknowledge DMA_VTX_PROTECTION PGRAPH interrupts

Currently Nouveau is unable to dismiss DMA_VTX_PROTECTION errors,
which results in an infinite loop in the interrupt handler.

These errors are caused both by bugs in the Gallium driver and by
user-specified index buffers with out of bounds indices.

By mmio-tracing the nVidia drivers, I found out how this is done.
On DMA_VTX_PROTECTION, The nVidia driver reads the register 0x402000,
always getting the value 4, and then writes 4 back to 0x402000.

This patch adds that logic by reading 0x402000 and writing the same
value back.
It's unclear what should happen if the value read is not 4, and
the current approach might not be the correct one.

To test this, modify mesa/progs/trivial/vbo-drawrange.c, defining
ELTOBJ to 1 and replacing indices with huge out of bounds integers.

Without this patch, the GPU and/or kernel should lock up.
With this patch, it should misrender as expected but not lock up.

The errors are still logged since they are useful for development.

This has been tested on NV49 and may not work on other cards.

To find out how things work on other cards, run the aforementioned
test using the blob with mmiotrace and grep for a read of the PGRAPH
source register.

Signed-off-by: Luca Barbieri <luca@luca-barbieri.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: fix thinko in nv04_instmem.c
Ben Skeggs [Fri, 15 Jan 2010 02:21:37 +0000 (12:21 +1000)]
drm/nouveau: fix thinko in nv04_instmem.c

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: fix a race condition in nouveau_dma_wait()
Ben Skeggs [Fri, 15 Jan 2010 02:08:57 +0000 (12:08 +1000)]
drm/nouveau: fix a race condition in nouveau_dma_wait()

Can be triggered easily on certain cards (NV46 and NV50 of mine) by
running "dmesg", the DRM's channel will lockup.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agoMerge remote branch 'korg/drm-radeon-testing' into drm-linus
Dave Airlie [Fri, 15 Jan 2010 05:40:39 +0000 (15:40 +1000)]
Merge remote branch 'korg/drm-radeon-testing' into drm-linus

* korg/drm-radeon-testing:
  drm/radeon/kms: Use radeon_agp_disable when disabling AGP
  drm/radeon/kms: Disable AGP is aperture size < 32M
  drm/radeon/kms: Fix r600 blit cleanup path
  drm/radeon/kms: Do not unpin buffer in fb destruction
  drm/radeon/kms: fix displayport->dvi connector DDC.

14 years agodrm: EDID accept separate sync video mode
Jerome Glisse [Thu, 14 Jan 2010 18:02:20 +0000 (19:02 +0100)]
drm: EDID accept separate sync video mode

X is accepting such video mode, do the same. Pointed out by Joshua Roys
on IRC. Fix https://bugzilla.redhat.com/show_bug.cgi?id=540024

[fix printf to use composite not integrated :- airlied]

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agoMerge remote branch 'nouveau/for-airlied' of ../drm-nouveau-next into drm-linus
Dave Airlie [Fri, 15 Jan 2010 03:22:59 +0000 (13:22 +1000)]
Merge remote branch 'nouveau/for-airlied' of ../drm-nouveau-next into drm-linus

* 'nouveau/for-airlied' of ../drm-nouveau-next: (44 commits)
  drm/nouveau: check pushbuffer bounds in ioctl
  drm/nouveau: reserve VGA area for the moment
  drm/nouveau: Unset the EDID connector property when the EDID block goes away.
  drm/nouveau: Fallback to analog load detection when the EDID block is invalid.
  drm/nouveau: fix edid memleak in nouveau_connector
  drm/nouveau: Break some long lines.
  drm/nouveau: add NV18 device id to call_lvds_manufacturer_script
  drm/nv50: Fix typo in PGRAPH initialisation.
  drm/nouveau: less magic DCB 1.5 parsing
  drm/nouveau: assume no nv04 board has a DCB table
  drm/nouveau: remove PRIV0 check in nouveau_mem_close()
  drm/nouveau: wait on fence after bo move if validating for another channel
  drm/nouveau: trust init table registers are safe
  drm/nv50: wait for pgraph to idle before unloading the context

14 years agodrm/nouveau: check pushbuffer bounds in ioctl
Luca Barbieri [Sun, 10 Jan 2010 19:10:53 +0000 (20:10 +0100)]
drm/nouveau: check pushbuffer bounds in ioctl

Currently there is no check that the pushbuffer request bounds are inside
the TTM BO.

This allows to instruct the kernel to do relocations on user-selected
addresses, since the relocation bounds checking relies on the request
bounds.

This can oops the kernel accidentally and is easily exploitable.

This patch adds bound checking and alignment checking for ->offset and
->nr_dwords.

It also makes some variables unsigned, which should have no effect,
but prevents possible bounds checking problems.

Signed-off-by: Luca Barbieri <luca@luca-barbieri.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: reserve VGA area for the moment
Ben Skeggs [Thu, 14 Jan 2010 23:24:20 +0000 (09:24 +1000)]
drm/nouveau: reserve VGA area for the moment

This is to prevent things such as GART tables and other important GPU
structures being allocated there before we take over fbcon ourselves.

This is more of a workaround for the moment, a better solution will
require some more invasive changes, but it'll be done at some point.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: Unset the EDID connector property when the EDID block goes away.
Francisco Jerez [Thu, 14 Jan 2010 14:54:23 +0000 (15:54 +0100)]
drm/nouveau: Unset the EDID connector property when the EDID block goes away.

Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: Fallback to analog load detection when the EDID block is invalid.
Francisco Jerez [Thu, 14 Jan 2010 14:47:03 +0000 (15:47 +0100)]
drm/nouveau: Fallback to analog load detection when the EDID block is invalid.

Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: fix edid memleak in nouveau_connector
Xavier Chantry [Mon, 11 Jan 2010 21:42:21 +0000 (22:42 +0100)]
drm/nouveau: fix edid memleak in nouveau_connector

This was spotted by kmemleak.

Signed-off-by: Xavier Chantry <shiningxc@gmail.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: Break some long lines.
Francisco Jerez [Thu, 14 Jan 2010 13:56:54 +0000 (14:56 +0100)]
drm/nouveau: Break some long lines.

Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: add NV18 device id to call_lvds_manufacturer_script
Andrea Tacconi [Thu, 14 Jan 2010 13:52:37 +0000 (14:52 +0100)]
drm/nouveau: add NV18 device id to call_lvds_manufacturer_script

This fixes imac black screen (NV18 card)

Signed-off-by: Andrea Tacconi <tacconet@libero.it>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nv50: Fix typo in PGRAPH initialisation.
Marcin Kościelnicki [Tue, 12 Jan 2010 18:21:56 +0000 (18:21 +0000)]
drm/nv50: Fix typo in PGRAPH initialisation.

This enables streamout functionality.

Signed-off-by: Marcin Kościelnicki <koriakin@0x04.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: less magic DCB 1.5 parsing
Ben Skeggs [Thu, 14 Jan 2010 07:53:04 +0000 (17:53 +1000)]
drm/nouveau: less magic DCB 1.5 parsing

This in the very least matches the parsing of all the previously known
entries, and hopefully (at least closer to) correct for any we haven't
seen yet.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: assume no nv04 board has a DCB table
Ben Skeggs [Thu, 14 Jan 2010 05:58:10 +0000 (15:58 +1000)]
drm/nouveau: assume no nv04 board has a DCB table

There's a report of a TNT2 where the DCB table pointer is *not* NULL
(it contains a part of a VBIOS data string), and we assume this means
a DCB table is present, causing all kinds of hilarity.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: remove PRIV0 check in nouveau_mem_close()
Ben Skeggs [Wed, 13 Jan 2010 01:22:33 +0000 (11:22 +1000)]
drm/nouveau: remove PRIV0 check in nouveau_mem_close()

We don't setup PRIV0 anymore, so this is unnecessary.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: wait on fence after bo move if validating for another channel
Ben Skeggs [Tue, 12 Jan 2010 05:28:19 +0000 (15:28 +1000)]
drm/nouveau: wait on fence after bo move if validating for another channel

Not an ideal solution, but it'll do for the moment for correctness.  We
need to come up with a nicer way to manage inter-channel sync, the hw
is unfortunately a little lacking in this area.

Should fix some resume corruption, as well as corruption that may be seen
while under memory pressure.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: trust init table registers are safe
Ben Skeggs [Tue, 12 Jan 2010 03:02:19 +0000 (13:02 +1000)]
drm/nouveau: trust init table registers are safe

Apparently the original reason for checking this was there were known
register accesses that caused hangs on some chipsets.  This was more
than likely because of incorrect parsing of previous opcodes, and I
hardly think aborting a script half way through is going to be any
better (in fact, we have had bug reports where this has been the cause
of s/r failures among other things).

This patch (which has been in Fedora 12 for a long time now) removes
all checking for known register ranges, and just leaves the check to
ensure the access is within the mapped aperture to avoid an oops.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nv50: wait for pgraph to idle before unloading the context
Maarten Maathuis [Mon, 11 Jan 2010 20:18:53 +0000 (21:18 +0100)]
drm/nv50: wait for pgraph to idle before unloading the context

This should fix the problem with gpu hangs people have had when closing
channels.

Signed-off-by: Maarten Maathuis <madman2003@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/ttm: Make sure system buffer objects has offset == 0.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:45 +0000 (22:28 +0100)]
drm/ttm: Make sure system buffer objects has offset == 0.

This is a convention that the vmwgfx driver has come to rely on.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/vmwgfx: Implement basic pm operations.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:43 +0000 (22:28 +0100)]
drm/vmwgfx: Implement basic pm operations.

Currently we really only support S3, since the device doesn't support
saving of the 3D state.

On S3/S4, move all buffer objects to swappable memory and take down
GMR bindings. We need to do that from a PM notifier since we can't
do persistant memory allocations from the standard PM callbacks.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/ttm: Export symbols needed for vmwgfx suspend / resume operations.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:42 +0000 (22:28 +0100)]
drm/ttm: Export symbols needed for vmwgfx suspend / resume operations.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/vmwgfx: Implement a swap_notify callback.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:41 +0000 (22:28 +0100)]
drm/vmwgfx: Implement a swap_notify callback.

Unbind GMR bindings on the buffer about to be swapped out.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/ttm: Add a swap_notify callback.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:40 +0000 (22:28 +0100)]
drm/ttm: Add a swap_notify callback.

This is needed for a bugfix in the vmwgfx driver.
Drivers may have GPU bindings on buffers that core TTM is not aware of,
and TTM may view those buffers as ordinary system memory buffers.
Add a notifier to such drivers when TTM is about to move the buffer
contents out to swappable memory. The driver must then release any
private GPU bindings on those buffers.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/vmwgfx: Use bo_driver::move_notify to unbind GMRs.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:39 +0000 (22:28 +0100)]
drm/vmwgfx: Use bo_driver::move_notify to unbind GMRs.

This was previously done explicitly for overlay- and fb buffers.
Now it's done for any buffer leaving the SYSTEM memory region.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/vmwgfx: Make fence sequences continous across a VT switch.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:38 +0000 (22:28 +0100)]
drm/vmwgfx: Make fence sequences continous across a VT switch.

A vt switch in stealth mode would take down the FIFO, and re-
initialize fence sequence numbers. This patch
saves the current state of the fence sequence when the FIFO is
disabled.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/vmwgfx: Fix an error path causing an oops.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:37 +0000 (22:28 +0100)]
drm/vmwgfx: Fix an error path causing an oops.

An error happening before the snooper.image member had been set up
would cause a kfree of an arbitrary pointer. Set up the snooper.image
member early.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/vmwgfx: Don't promote updates from GMR-backed scanouts to fullscreen.
Thomas Hellstrom [Wed, 13 Jan 2010 21:28:36 +0000 (22:28 +0100)]
drm/vmwgfx: Don't promote updates from GMR-backed scanouts to fullscreen.

That's unnecessary since partial screen updates from GMRs are fast.
Also fix cliprect pointer dereferencing

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: Use radeon_agp_disable when disabling AGP
Jerome Glisse [Wed, 13 Jan 2010 14:16:38 +0000 (15:16 +0100)]
drm/radeon/kms: Use radeon_agp_disable when disabling AGP

Use same common function to disable agp so we replace the GART
callback by the proper one when we do so. This fix oops if
radeon_agp_init report failure.

This patch also move radeon_agp_init out of *_mc_init for r600
& rv770 so that we can have a similar behavior than for previous
hw, ie if agp_init fails it will fallback to GPU GART and disable
AGP.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: Disable AGP is aperture size < 32M
Jerome Glisse [Wed, 13 Jan 2010 14:16:39 +0000 (15:16 +0100)]
drm/radeon/kms: Disable AGP is aperture size < 32M

radeon KMS need a GART of at least 32M to properly work. This patch
check the AGP aperture size and disable if it's less than 32M. Note
than unlike non KMS path we don't staticaly allocate AGP memory so
we are not wasting memory not used by graphic processing.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: Fix r600 blit cleanup path
Jerome Glisse [Wed, 13 Jan 2010 09:29:27 +0000 (10:29 +0100)]
drm/radeon/kms: Fix r600 blit cleanup path

r600 blit cleanup path need to check if a bo was allocated before
trying to free or unpin it. This patch add this check and avoid
oops when the initialization on r6xx or r7xx hw fails.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: Do not unpin buffer in fb destruction
Jerome Glisse [Wed, 13 Jan 2010 11:32:36 +0000 (12:32 +0100)]
drm/radeon/kms: Do not unpin buffer in fb destruction

It's not necessary to unpin buffer in fb destruction. pin/unpin
need to be balanced and we don't pin in fb creation. We pin when
an fb is associated to a crtc and unpin when the fb is disassociated
from the crtc.

Note:
Maybe we should take reference on fb in set_base callback so fb
doesn't disappear until it's unbind from ctrc.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: fix displayport->dvi connector DDC.
Dave Airlie [Thu, 14 Jan 2010 01:42:17 +0000 (11:42 +1000)]
drm/radeon/kms: fix displayport->dvi connector DDC.

It appears that attempting AUXCH DDC breaks the subsequent attempt
to do DDC over the i2c lines, so use the sink type to determine
if we should be doing AUXCH or i2c DDC.

This fixes my DVI monitor plugged into DP->DVI convertor.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agoMerge remote branch 'korg/drm-radeon-next' into drm-linus
Dave Airlie [Wed, 13 Jan 2010 06:17:38 +0000 (16:17 +1000)]
Merge remote branch 'korg/drm-radeon-next' into drm-linus

* korg/drm-radeon-next
  drm/radeon/kms: only evict to GTT if CP is ready
  drm/radeon/kms: Fix crash getting TV info with no BIOS.
  drm/radeon/kms/rv100: reject modes > 135 Mhz on DVI (v2)
  drm/radeon/kms/r6xx+: make irq handler less verbose
  drm/radeon/kms: fix up LVDS handling on macs (v2)

14 years agodrm: change drm set mode messages as DRM_DEBUG
Dave Young [Wed, 13 Jan 2010 05:38:59 +0000 (13:38 +0800)]
drm: change drm set mode messages as DRM_DEBUG

Following drm info repeat 207 times during one hour, it's quite annoying
[ 1266.286747] [drm] TV-19: set mode NTSC 480i 0

Change from DRM_INFO to DRM_DEBUG

Signed-off-by: Dave Young <hidave.darkstar@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm: fix crtc no modes printf + typo
Dave Airlie [Wed, 13 Jan 2010 06:15:11 +0000 (16:15 +1000)]
drm: fix crtc no modes printf + typo

Toralf Förster pointed out the typo, the fact I forget the if
statement is purely personal fail.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: only evict to GTT if CP is ready
Dave Airlie [Tue, 12 Jan 2010 23:21:49 +0000 (09:21 +1000)]
drm/radeon/kms: only evict to GTT if CP is ready

Testing GTT ready might be more correct but cp.ready
works fine and has been tested on irc by 2-3 ppl.

fixes bug k.org 15035 and fd.o 25733

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: Fix crash getting TV info with no BIOS.
Michel Dänzer [Mon, 11 Jan 2010 07:58:38 +0000 (08:58 +0100)]
drm/radeon/kms: Fix crash getting TV info with no BIOS.

Signed-off-by: Michel Dänzer <daenzer@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms/rv100: reject modes > 135 Mhz on DVI (v2)
Alex Deucher [Mon, 11 Jan 2010 20:02:31 +0000 (15:02 -0500)]
drm/radeon/kms/rv100: reject modes > 135 Mhz on DVI (v2)

Due to heat issues.  Fixes fdo bug 25992

v2: fix typo noticed by Maarten Maathuis

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms/r6xx+: make irq handler less verbose
Alex Deucher [Tue, 12 Jan 2010 00:47:38 +0000 (19:47 -0500)]
drm/radeon/kms/r6xx+: make irq handler less verbose

Unhandled vectors can be safely ignored, no need
to spam the kernel log by default.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodrm/radeon/kms: fix up LVDS handling on macs (v2)
Alex Deucher [Tue, 12 Jan 2010 16:16:57 +0000 (11:16 -0500)]
drm/radeon/kms: fix up LVDS handling on macs (v2)

Based on radeonfb code and recent ddx fix.

v2: minor formatting fix from Michel Dänzer

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Reviewed-by: Michel Dänzer <michel@daenzer.net>
Tested-by: Michel Dänzer <michel@daenzer.net>
Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agoMerge branch 'for-airlied' of /ssd/git/drm-nouveau-next into drm-linus
Dave Airlie [Mon, 11 Jan 2010 04:43:16 +0000 (14:43 +1000)]
Merge branch 'for-airlied' of /ssd/git/drm-nouveau-next into drm-linus

* 'for-airlied' of /ssd/git/drm-nouveau-next: (28 commits)
  drm/nv04: Fix set_operation software method.
  drm/nouveau: initialise DMA tracking parameters earlier
  drm/nouveau: use dma.max rather than pushbuf size for checking GET validity
  drm/nv04: differentiate between nv04/nv05
  drm/nouveau: Fix null deref in nouveau_fence_emit due to deleted fence
  drm/nv50: prevent a possible ctxprog hang
  drm/nouveau: have ttm's fault handler called directly
  drm/nv50: restore correct cache1 get/put address on fifoctx load
  drm/nouveau: create function for "dealing" with gpu lockup
  drm/nouveau: remove unused nouveau_channel_idle() function
  drm/nouveau: fix handling of fbcon colours in 8bpp
  drm/nv04: Context switching fixes.
  drm/nouveau: Use the software object for fencing.
  drm/nouveau: Allocate a per-channel instance of NV_SW.
  drm/nv50: make the blocksize depend on vram size
  drm/nouveau: better alignment of bo sizes and use roundup instead of ALIGN
  drm/nouveau: Don't skip card take down on nv0x.
  drm/nouveau: Implement nv42-nv43 TV load detection.
  drm/nouveau: Clean up the nv17-nv4x load detection code a bit.
  drm/nv50: fix fillrect color
  ...

14 years agoMerge remote branch 'korg/drm-radeon-next' into drm-linus
Dave Airlie [Mon, 11 Jan 2010 04:42:58 +0000 (14:42 +1000)]
Merge remote branch 'korg/drm-radeon-next' into drm-linus

* korg/drm-radeon-next:
  drm/radeon/kms: add additional safe regs for r4xx/rs6xx and r5xx
  drm/radeon/kms: Don't try to enable IRQ if we have no handler installed
  drm: Avoid calling vblank function is vblank wasn't initialized
  drm/radeon: mkregtable.c: close a file before exit
  drm/radeon/kms: Make sure we release AGP device if we acquired it
  drm/radeon/kms: Schedule host path read cache flush through the ring V2
  drm/radeon/kms: Workaround RV410/R420 CP errata (V3)
  drm/radeon/kms: detect sideport memory on IGP chips
  drm/radeon: fix a couple of array index errors
  drm/radeon/kms: add support for eDP (embedded DisplayPort)
  drm: Add eDP connector type
  drm/radeon/kms: pull in the latest upstream ObjectID.h changes
  drm/radeon/kms: whitespace changes to ObjectID.h
  drm/radeon/kms: fix typo in atom connector type handling

14 years agodrm/nv04: Fix set_operation software method.
Marcin Kościelnicki [Sun, 10 Jan 2010 17:09:14 +0000 (17:09 +0000)]
drm/nv04: Fix set_operation software method.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: initialise DMA tracking parameters earlier
Ben Skeggs [Fri, 8 Jan 2010 00:57:39 +0000 (10:57 +1000)]
drm/nouveau: initialise DMA tracking parameters earlier

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: use dma.max rather than pushbuf size for checking GET validity
Ben Skeggs [Fri, 8 Jan 2010 00:53:40 +0000 (10:53 +1000)]
drm/nouveau: use dma.max rather than pushbuf size for checking GET validity

Some upcoming G80 DMA changes will depend on this, but it's split out for
bisectibility just in case it causes some unexpected issues.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nv04: differentiate between nv04/nv05
Ben Skeggs [Thu, 7 Jan 2010 03:47:57 +0000 (13:47 +1000)]
drm/nv04: differentiate between nv04/nv05

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: Fix null deref in nouveau_fence_emit due to deleted fence
Luca Barbieri [Wed, 6 Jan 2010 03:02:45 +0000 (04:02 +0100)]
drm/nouveau: Fix null deref in nouveau_fence_emit due to deleted fence

Currently Nouveau will unvalidate all buffers if it is forced to wait on
one, and then start revalidating from the beginning.  While doing so, it
destroys the operation fence, causing nouveau_fence_emit to crash.

This patch fixes this bug by taking the fence object out of validate_op
and creating it just before emit.  The fence pointer is initialized to 0
and unref'ed unconditionally.

In addition to fixing the bug, this prevents its reintroduction and
simplifies the code.

Signed-off-by: Luca Barbieri <luca@luca-barbieri.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nv50: prevent a possible ctxprog hang
Ben Skeggs [Wed, 6 Jan 2010 02:00:02 +0000 (12:00 +1000)]
drm/nv50: prevent a possible ctxprog hang

The below is mainly an educated guess at what's going on, docs would
sure be handy...  NVIDIA? :P

It appears it's possible for a ctxprog to run even while a GPU exception
is pending.  The GF8 and up ctxprogs appear to have a small snippet of
code which detects this, and stalls the ctxprog until it's been handled,
which essentially looks like:

if (r2 & 0x00008000) {
r0 |= 0x80000000;
while (r0 & 0x80000000) {}
}

I don't know of any way that flag would get cleared unless the driver
intervenes (and indeed, in the cases I've seen the hang, nothing steps
in to automagically clear it for us).  This patch causes the driver to
clear the flag during the PGRAPH IRQ handler.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: have ttm's fault handler called directly
Ben Skeggs [Mon, 4 Jan 2010 05:52:20 +0000 (15:52 +1000)]
drm/nouveau: have ttm's fault handler called directly

There's no good reason for us to have our own anymore, this is left over
from an early port to these TTM interfaces.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nv50: restore correct cache1 get/put address on fifoctx load
Ben Skeggs [Mon, 4 Jan 2010 23:41:05 +0000 (09:41 +1000)]
drm/nv50: restore correct cache1 get/put address on fifoctx load

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: create function for "dealing" with gpu lockup
Marcin Slusarz [Mon, 4 Jan 2010 18:25:09 +0000 (19:25 +0100)]
drm/nouveau: create function for "dealing" with gpu lockup

It's mostly a cleanup, but in nv50_fbcon_accel_init gpu lockup
message was printed, but HWACCEL_DISBALED flag was not set.

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: remove unused nouveau_channel_idle() function
Ben Skeggs [Mon, 4 Jan 2010 02:53:01 +0000 (12:53 +1000)]
drm/nouveau: remove unused nouveau_channel_idle() function

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nouveau: fix handling of fbcon colours in 8bpp
Ben Skeggs [Sun, 3 Jan 2010 23:10:55 +0000 (09:10 +1000)]
drm/nouveau: fix handling of fbcon colours in 8bpp

Depending on the visual, the colours handed to us in fillrect() can either be
an actual colour, or an index into the pseudo-palette.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
14 years agodrm/nv04: Context switching fixes.
Francisco Jerez [Sat, 26 Dec 2009 13:39:46 +0000 (14:39 +0100)]
drm/nv04: Context switching fixes.

Signed-off-by: Francisco Jerez <currojerez@riseup.net>
14 years agodrm/nouveau: Use the software object for fencing.
Francisco Jerez [Sat, 26 Dec 2009 01:09:36 +0000 (02:09 +0100)]
drm/nouveau: Use the software object for fencing.

This should avoid a race condition on nv0x, if we're doing it with
actual PGRAPH objects and a there's a fence within the FIFO DMA fetch
area when a context switch kicks in.

In that case we get an ILLEGAL_MTHD interrupt as expected, but the
values in PGRAPH_TRAPPED_ADDR aren't calculated correctly and they're
almost useless (e.g. you can see ILLEGAL_MTHDs for the now inactive
channel, with a wrong offset/data pair).

Signed-off-by: Francisco Jerez <currojerez@riseup.net>
14 years agodrm/nouveau: Allocate a per-channel instance of NV_SW.
Francisco Jerez [Sat, 26 Dec 2009 01:42:45 +0000 (02:42 +0100)]
drm/nouveau: Allocate a per-channel instance of NV_SW.

It will be useful for various synchronization purposes, mostly stolen
from "[PATCH] drm/nv50: synchronize user channel after buffer object
move on kernel channel" by Maarten Maathuis.

Signed-off-by: Francisco Jerez <currojerez@riseup.net>
14 years agodrm/nv50: make the blocksize depend on vram size
Maarten Maathuis [Sat, 26 Dec 2009 20:46:36 +0000 (21:46 +0100)]
drm/nv50: make the blocksize depend on vram size

- This should be better than what we have now.
- I'm less sure about the non power of two path.

Signed-off-by: Maarten Maathuis <madman2003@gmail.com>
14 years agodrm/nouveau: better alignment of bo sizes and use roundup instead of ALIGN
Maarten Maathuis [Fri, 25 Dec 2009 17:51:17 +0000 (18:51 +0100)]
drm/nouveau: better alignment of bo sizes and use roundup instead of ALIGN

- Aligning to block size should ensure that the extra size is enough.
- Using roundup, because not all sizes are powers of two.

Signed-off-by: Maarten Maathuis <madman2003@gmail.com>
14 years agodrm/nouveau: Don't skip card take down on nv0x.
Francisco Jerez [Tue, 22 Dec 2009 17:24:09 +0000 (18:24 +0100)]
drm/nouveau: Don't skip card take down on nv0x.

Signed-off-by: Francisco Jerez <currojerez@riseup.net>
14 years agodrm/nouveau: Implement nv42-nv43 TV load detection.
Francisco Jerez [Thu, 17 Dec 2009 17:57:44 +0000 (18:57 +0100)]
drm/nouveau: Implement nv42-nv43 TV load detection.

Signed-off-by: Francisco Jerez <currojerez@riseup.net>
14 years agodrm/nouveau: Clean up the nv17-nv4x load detection code a bit.
Francisco Jerez [Thu, 17 Dec 2009 17:52:44 +0000 (18:52 +0100)]
drm/nouveau: Clean up the nv17-nv4x load detection code a bit.

Signed-off-by: Francisco Jerez <currojerez@riseup.net>