See also: http://bugzilla.redhat.com/514600Signed-off-by: NAdam Jackson <ajax@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Without an allocated colormap, FBIOGETCMAP fails.  This would make
programs restore an all-black colormap ("links -g") or fail to work
altogether ("mplayer -vo fbdev2").
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>
Signed-off-by: NDave Airlie <airlied@redhat.com>

When the framebuffer driver does not publish detailed timing information
for the current video mode, the correct value for the pixclock field is
zero, not -1.

Since pixclock is actually unsigned, the value -1 would be interpreted
as 4294967295 picoseconds (i.e., about 4 milliseconds) by
register_framebuffer() and userspace programs.

This patch allows X.org's fbdev driver to work.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>
Signed-off-by: NDave Airlie <airlied@redhat.com>

I examined several fbdev drivers and foudn the blanking code in
drm_fb_helper to be wrong. This patch fixes the fbdev blanking to behave
like other fbdev drivers.
Signed-off-by: NDave Airlie <airlied@redhat.com>

This adds a page flipping ioctl to the KMS API.  The ioctl takes an fb ID
and a ctrc ID and flips the crtc to the given fb at the next vblank.
The ioctl returns immediately but the flip doesn't happen until after
any rendering that's currently queued up against the new framebuffer
is done.  After submitting a page flip, any execbuffer involving the
old front buffer will block until the flip is completed.

Optionally, a vblank event can be generated when the swap eventually
happens.
Signed-off-by: NKristian Høgsberg <krh@bitplanet.net>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDave Airlie <airlied@redhat.com>

In drm_version, actually check the results from function calls so that
we're not potentially passing garbage back to userspace.
Signed-off-by: NAndres Salomon <dilinger@collabora.co.uk>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Don't inline it; the compiler can figure it out.  Comments added that are
based upon my interpretation of the code.  Hopefully they're correct. :)
Signed-off-by: NAndres Salomon <dilinger@collabora.co.uk>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NDave Airlie <airlied@redhat.com>

There are a few more macros in drmP.h that are unused; DRM_GET_PRIV_SAREA,
DRM_ARRAY_SIZE, and DRM_WAITCOUNT can go away completely.

Unfortunately, DRM_COPY is still used in one place, but we can at least
move it to where it's used.  It's an awful looking macro..

[akpm: fix overeagerness]
Signed-off-by: NAndres Salomon <dilinger@collabora.co.uk>
Cc: Dave Airlie <airlied@linux.ie>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NDave Airlie <airlied@redhat.com>

If we queue a vblank event but miss it, we should return the actual
sequence number we queued to userspace, so its event handling function
will know which event to look for.
Acked-by: NKristian Høgsberg <krh@bitplanet.net>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NDave Airlie <airlied@redhat.com>

This patch adds a new flag to the drmWaitVblank ioctl, which asks the drm
to return immediately and notify userspace when the specified vblank sequence
happens by sending an event back on the drm fd.

The event mechanism works with the other flags supported by the ioctls,
specifically, the vblank sequence can be specified relatively or absolutely,
and works for primary and seconday crtc.

The signal field of the vblank request is used to provide user data,
which will be sent back to user space in the vblank event.
Signed-off-by: NKristian Høgsberg <krh@redhat.com>
Reviewed-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NDave Airlie <airlied@redhat.com>

In commit c1c7af60 ("drm/i915: force
mode set at lid open time") the intel graphics driver was taught to
restore the LVDS mode on lid open.

That caused problems with interaction with the suspend/resume code,
which commonly runs at the same time (suspend is often caused by the lid
close event, while lid open is commonly a resume event), which was
worked around with in commit 06891e27
("drm/i915: fix suspend/resume breakage in lid notifier").

However, in the meantime the lid event code had also grown a user event
notifier (commit 06324194: "drm/i915:
generate a KMS uevent at lid open/close time"), and now _that_ causes
problems with suspend/resume and some versions of Xorg reacting to those
uevents by setting the mode.

So this effectively reverts that commit 06324194, and makes the lid
open protection logic against suspend/resume more explicit.  This fixes
at least one laptop. See

	http://bugzilla.kernel.org/show_bug.cgi?id=14484

for more details.
Acked-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Cc: Riccardo Magliocchetti <riccardo.magliocchetti@gmail.com>
Cc: Eric Anholt <eric@anholt.net>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This sets the fbcon to use TRUECOLOR by default, it then
only modifies the pseudo palette for fbcon, and only touches
the real palette when in 8-bit pseudo color mode.
Signed-off-by: NDave Airlie <airlied@redhat.com>

Sometimes we will get the incorrect display modeline when parsing the detailed
timing in EDID. For example:
   >hsync/vsync width is zero
   >sync is beyond the blank.

So add the basic check for the detailed timing in EDID to avoid the incorrect
display modeline.
Signed-off-by: NZhao Yakui <yakui.zhao@intel.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Since we register all radeon devices, and the arbiter only cares about
VGA class ones, we will fail to startup on display controller class devices.
We don't gain anything by using the return value here.

this helps kms on sparc64 get started.
Reported-by: NDavid S. Miller <davem@davemloft.net>
Signed-off-by: NDave Airlie <airlied@redhat.com>

This adds registers save/restore for Ironlake to make suspend work.
Signed-off-by: NGuo, Chaohong <chaohong.guo@intel.com>
[zhenyuw: some code re-orgnization, and add more save/restore for
FDI link and transcoder registers, also fix palette register for Ironlake]
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

Initialize clock.vco to silence gcc.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

When coming back from DPMS or turning on a display, make sure we have
the watermarks set up before turning on the display plane, otherwise we
may get underruns.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Tested-by: NDirk Hohndel <hohndel@infradead.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

Turns out G4x needs to have sensible watermarks set, especially for
self-refresh enabled modes.  Add support for it.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Tested-by: NDirk Hohndel <hohndel@infradead.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

Must set filter selection as hardcoded coefficients for medium 3x3
filtering, which matches vbios setting for Ironlake.

This fixes display corrupt issue on HP arrandale with new vbios.

Cc: Stable Team <stable@kernel.org>
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

For new stepping of PCH, the display reference clock
is fully under driver's control. This one trys to setup
all needed reference clock for different outputs. Older
stepping of PCH chipset should be ignoring this.

This fixes output failure issue on newer PCH which requires
driver to take control of reference clock enabling.

Cc: Stable Team <stable@kernel.org>
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

To enable framebuffer compression on a g4x, we not only need the buffer
to tiled (X only), we also need to hold a fence register for the buffer.
Currently we only install a fence register for pre-i965s when setting up
the scanout buffer. Rather than adding some convoluted logic to
g4x_enable_fbc() to acquire a fence register, and perhaps to
g4x_disable_fbc() to release it again, we can extend the acquisition
during setup to all chipsets.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

Turns out some machines, like the ThinkPad X40 don't come back if you
don't save/restore this register.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

FDI M/N calculation hasn't taken the current pipe color depth into account,
but always set as 24bpp. This one checks current pipe color depth setting,
and change FDI M/N calculation a little to use bits_per_pixel first, then
convert to bytes_per_pixel later.

This fixes display corrupt issue on Arrandle LVDS with 1600x900 panel
in 18bpp dual-channel mode.

Cc: Stable Team <stable@kernel.org>
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

Until we figure out the right setting for powersave features on
Ironlake, disable it for now. Also disable watermark update,
which has new registers for it on Ironlake too.
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Reviewed-by: NJesse Barnes <jbarnes@virtuousgeek.org>
[anholt: Resolved against the Pineview FBC changes]
Signed-off-by: NEric Anholt <eric@anholt.net>

If the device didn't support EDP, we would bail out too soon.
Signed-off-by: NAndy Lutomirski <luto@mit.edu>
[anholt: Pulled this patch out of the patch for adding quirks to
enable reclocking.]
Signed-off-by: NEric Anholt <eric@anholt.net>

This hasn't fixed the regressions we were testing against, but clearly
should be required.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

According to the spec the LVDS_BORDER_ENABLE bit decides whether the border
data should be included in the active display and data sent to the panel.
Border should be used when in VGA centered (un-scaled) mode or when scaling
a 4:3 source image to a wide screen panel (typical 16:9).

So when the LVDS scaling is used, decide whether the LVDS_BORDER should be
enabled or not according to the current scaling mode.

At the same time fix the typo error in LVDS center scaling mode.

https://bugs.freedesktop.org/show_bug.cgi?id=23789Signed-off-by: NZhao Yakui <yakui.zhao@intel.com>
tested-by: NZhao Jian <jian.zhao@intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

Pineview doesn't have this FBC mechanism, so this code doesn't apply.
Signed-off-by: NShaohua Li <shaohua.li@intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

Both r100/r600 had this wrong, use the macro to extract the register
to relocate.
Signed-off-by: NDave Airlie <airlied@redhat.com>

There is no need to assign vb before you know that space is available.

[agd5f: adapted for kernel tree.]
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

This is just a cleanup of the list macro usage.
Signed-off-by: NDave Airlie <airlied@redhat.com>

For AGP to work unmapped access must cover VRAM & AGP as
AGP is treated like VRAM by the GPU (ie physical address).
This patch properly setup the virtual memory system aperture
to cover AGP if AGP is enabled. It seems that there is memory
corruption after resume when using AGP (RV770 seems unaffected
thought). Version 2 just fix merge issue with updated AGP
fallback patch.
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

When GPU acceleration is not working with AGP try to fallback to non
AGP GART (either PCI or PCIE GART). This should make KMS failure on
AGP less painfull. We still need to find out what is wrong when AGP
fails but at least user have a lot of more chances to get a working
configuration with acceleration. This patch also cleanup R600/RV770
fallback path so they use same code as others asics. Version 2
factorize agp disabling logic to avoid code duplication and bugs.
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Bad generated header file leaded to use wrong register
to check IRQ status and acknowledge them. Fix the header
and use proper registers.
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Duplicate bits set
Signed-off-by: NRoel Kluin <roel.kluin@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

the global refcount wasn't being increased after the first reference.
this caused an oops on unload on a multi-gpu card.
Signed-off-by: NDave Airlie <airlied@redhat.com>

The previous patches had some unwanted side effects, I've fixed
the lack of 32bpp working, and fixed up 16bpp so it should also work.

this also adds the interface to allow the driver to set a preferred
console depth so for example low memory rn50 can set it to 8bpp.
It also catches 24bpp on cards that can't do it and forces 32bpp.

Tested on r100/r600/i945.
Signed-off-by: NDave Airlie <airlied@redhat.com>

This adds support for the setcmap api and fixes the 8bpp
support at least on radeon hardware. It adds a new load_lut
hook which can be called once the color map is setup.
Signed-off-by: NDave Airlie <airlied@redhat.com>

Also add single crtc for RN50 chips.

changes in v2:
fix vblank init to respect single crtc flag
fix r100 mode bandwidth to respect single crtc flag
Signed-off-by: NDave Airlie <airlied@redhat.com>