| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
|
|
| |
To avoid busting DRI2 while trying some different things for DRI3.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
|
|
| |
Older servers had no GPU support.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
| |
this avoids clashes with identifiers from other GPUs.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
| |
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
| |
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
| |
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
| |
add all the nvf0 stuff.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Initial reverse optimus didn't consider multiple screens, so
this overhauls the code to use the new X server interface,
and allows for multiple outputs on the nvidia to be used with
the intel doing the rendering.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
|
|
| |
Right now the log will output e.g. NVaa instead of NVAA, which is how we
usually talk about it.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
|
|
|
|
| |
memcpy() goes from taking 45% to 66% of total function time, which
translates to a 30% decrease in NVPutImage runtime.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
|
| |
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
|
|
|
|
|
| |
The DDX uses symbols from libdrm but currently does not link against
it, but rather relies on indirect linkage from libdrm_nouveau. This
could lead to insufficient dependencies in distribution packages, seen
in http://bugs.debian.org/633786.
Signed-off-by: Sven Joachim <svenjoac@gmx.de>
|
|
|
|
| |
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
|
| |
|
|
|
|
|
|
| |
Let the bikeshedding begin
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
|
|
|
|
|
|
| |
Building nouveau without dri2 is just silly
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
|
|
|
|
| |
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
|
|
|
|
|
|
|
|
|
| |
Nouveau does not use dri1, thus loading XFree86-DRI is not needed
On the other hand, we do use dri2
As a side effect, purge the 'set-but-unused' variable pLibDRMVersion
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
|
|
|
|
|
|
| |
Step 1 to completely rip out dri1 out of nouveau
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
|
| |
|
|
|
|
|
|
|
| |
Explicitly check for xorg-macros 1.8, as a subtle hint
to update your build deps if building your own ddx
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
|
|
|
|
| |
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
|
|
|
|
|
|
|
|
|
|
| |
nvd7 has no display engine, only enable sink output capability
if there is a crtc.
drmmode doesn't work without crtc's, so prevent crashes by skipping
the drmmode code in that case.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
|
|
| |
Far, far easier to maintain this way...
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
j_v on #nouveau bisected b1a630b48210d6a3c44994fce1b73273000ace5c has
breaking zaphod, on review it was trying to open the drm fd a second time
which was unnecessary.
Avoid the problem by storing the nv fd in an entity and have share it between
the two scrn info recs.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch let's cards with PCI class 0x30200 (3D controller) be
operated by the nouveau driver as well. The nv Quadro NVS 450 is
one such card, where the first GPU has PCI class 0x30000 and the
other GPU has PCI class 0x30200. By ignoring the 1 << 9 bit in
the PCI class mask, displays attached to the other GPU can also
be used now.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Dave Airlie <airlied@redhat.com>
Cc: Ben Skeggs <skeggsb@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The 'w' argument to NVCopyNV12ColorPlanes is used to index into the
source image. line_len is rounded up to 8 on NV50+, so if the source
image (+ left offset) is not rounded to 8, NVCopyNV12ColorPlanes could
read past the end of the array and crash X. This change can cause the
last few horizontal pixels of dst to not be initialized, but they
should be truncated by the renderer anyways.
See https://bugs.freedesktop.org/show_bug.cgi?id=63263
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
| |
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
|
|
|
|
|
| |
If forced close happens, all ioctl's will fail. Some of the handlers
also need to be unregistered before the module is unloaded entirely.
|
|
|
|
|
|
|
| |
Should fix next bit of
https://bugs.freedesktop.org/show_bug.cgi?id=60369
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Direct port of commit d1bc38b6673c54af61196056c489383fba8dced8
"nv40/exa: fix shaders on big-endian machines" to nv30.
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=60050
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
|
|
|
|
|
|
|
| |
should fix https://bugs.freedesktop.org/show_bug.cgi?id=60369
Reported-by: fabio.ped@libero.it
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
| |
This makes this fix more robust and less crashy.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
|
|
| |
We were getting 0xa5 error code on 2D transfers from tiled VRAM to untiled
GART when using USB offload devices, this fixes this by upping the alignment
on the 2D engine for shared buffers.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
| |
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
| |
this creates the shared pixmaps in GART, and makes sure they are untiled.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|