06 Dec, 2011
30 commits
-
As a mechanism to detect whether SWIOTLB is enabled or not.
We also fix the spelling - it was swioltb instead of
swiotlb.CC: FUJITA Tomonori
[v1: Ripped out swiotlb_enabled]
Signed-off-by: Konrad Rzeszutek Wilk -
Merge topic branch with some of Jesse's cleanups, the save/restore hooks
were being used by GMA500 so we can't just drop them.* drm-cleanups-jbarnes:
drm: remove some potentially dangerous DRM_ERRORs
drm: document the drm_mode_config structure
drm: document the drm_mode_group structure
drm: document and cleanup drm_mode_config_funcs
drm: document drm_mode_set structure
drm: remove unused fields in drm_connector and document the rest
drm: add drm_encoder comments
drm: add comments for drm_encoder_funcs
drm: fix comments for drm_crtc struct
drm: remove unused connector_count field from drm_display_mode -
Each of these error messages can be caused by a broken or malicious
userspace wanting to spam the dmesg with useless info. They're really
not worthy of DRM_DEBUG statements either; those are generally only
useful during bringup of new hardware or versions, and ought to be
removed before going upstream anyway.Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
Including a comment about what the locks are for.
Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
This is actually a core structure with a big future ahead of it. Make
it a little less mysterious.Signed-off-by: Jesse Barnes
Signed-off-by: Dave Airlie -
Just fix the wrapping mostly.
Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
This is a core mode setting structure that deserves a little verbiage.
Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
We never used initial_x/y or the force_encoder_id, so drop those fields
and proide a basic description of the others.Really, the ELD bits belong in drm_display_info rather than directly in
the connector, but that's a separate cleanup.Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
Just some basic comments about the place and function of the structure
and fields.Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
Just basic verbiage.
Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
Remove stale entries and update with the latest stuff.
Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
Doesn't really belong here anyway.
Signed-off-by: Jesse Barnes
Reviewed-by: Alex Deucher
Signed-off-by: Dave Airlie -
This merges a topic branch containing patches from Alan for the GMA500 driver.
* drm-gma500-alancox:
gma500: Oaktrail BIOS handling
gma500: Fix oaktrail probing part 1
gma500: Be smarter about layout
gma500: gtt based hardware scrolling console
gma500: frame buffer locking
gma500: Fix backlight crash
gma500: kill bogus code
gma500: Convert spaces to tabs in accel_2d.c.
gma500: do a pass over the FIXME tags
gma500: Add VBLANK support for Poulsbo hardware
gma500: Don't enable MSI on Poulsbo
gma500: Only register interrupt handler for poulsbo hardware
gma500: kill virtual mapping support
gma500: Move the API
gma500: kill off NUM_PIPE define
gma500: Rename the ioctls to avoid clashing with the legacy drivers
drm/gma500: begin pruning dead bits of API -
Now that we pull the right BIOS data out of the hat we need to use it when
doing our panel setup.Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
The Oaktrail platform does not use the GCT/VBT format that is used by the
Moorestowm (non PC legacy) equivalent device. It uses the BIOS tables which
means an opregion and the like.The current code uses the wrong table which breaks things like the Fujitsu
q550 tablets. Fix the table usage as a first step.The problem was found and diagnosed by Chia-I Wu
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
If we can't fit a page aligned display stride then it's not the end of the
world for a normal font, so try half a page and work down sizes.Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Add support for GTT based scrolling. Instead of pushing bits around we simply
use the GTT to change the mappings. This provides us with a very fast way to
scroll the display providing we have enough memory to allocate on 4K line
boundaries. In practice this seems to be the case except for very big displays
such as HDMI, and the usual configurations are netbooks/tablets.Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
If we are the console then a printk can hit us with a spin lock held (and
in fact the kernel will do its best to take the console printing lock).In that case we cannot politely sleep when synching after an accelerated op
but must behave obnoxiously to be sure of getting the bits out.Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Initial changes to get backlight behaviour we want and to fix backlight crashes
on suspend/resume paths.[Note: on some boxes this will now produce a warning about the backlight, this
isn't a regression it's an unfixed but non harmful case I still need to nail]Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
During the power split ups and work a chunk of code escaped into the
Poulsbo code path which it isn't for. On some devices such as the Dell
mini-10 this causes problems.Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Convert the spaces within the accel_2d.c file to tabs in order to comply
with the coding style of the kernel.Signed-off-by: Akshay Joshi
[Trimmed to subset relevant to current tree]
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Signed-off-by: Patrik Jakobsson
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Chipset reports MSI capabilities for Poulsbo even though it isn't really there.
Signed-off-by: Patrik Jakobsson
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
First step in adding proper irq handling. We'll start with poulsbo support so
make sure other chips don't touch drm_irq_install().Signed-off-by: Patrik Jakobsson
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
This isn't actually usable - we simply don't have the vmap space on a 32bit
system to do this stunt. Instead we will rely on the low level drivers
limiting the console resolution as before.The real fix is for someone to write a page table aware version of the
framebuffer console blit functions. Good university student project
perhaps..Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Finally move the API where it can be seen
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
We don't want this external in case someone adds more to the hardware. We
want it out of the ABI.Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
At this point we won't add an external set of definitions. We want to get
everything out before we admit to a public API beyond the standardised
ones.Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie
01 Dec, 2011
1 commit
-
Name the formats as DRM_FORMAT_X instead of DRM_FOURCC_X. Use consistent
names, especially for the RGB formats. Component order and byte order are
now strictly specified for each format.The RGB format naming follows a convention where the components names
and sizes are listed from left to right, matching the order within a
single pixel from most significant bit to least significant bit.The YUV format names vary more. For the 4:2:2 packed formats and 2
plane formats use the fourcc. For the three plane formats the
name includes the plane order and subsampling information using the
standard subsampling notation. Some of those also happen to match
the official fourcc definition.The fourccs for for all the RGB formats and some of the YUV formats
I invented myself. The idea was that looking at just the fourcc you
get some idea what the format is about without having to decode it
using some external reference.Signed-off-by: Ville Syrjälä
Signed-off-by: Dave Airlie
30 Nov, 2011
1 commit
-
This is used by nearly everyone including vmwgfx which doesn't generally
use the fb helper.Signed-off-by: Dave Airlie
29 Nov, 2011
1 commit
-
Commit 308e5bcbdb10 ("drm: add an fb creation ioctl that takes a pixel
format v5") missed one spot needing to be fixed up in the __BIG_ENDIAN
case.Fixes build error:
drivers/gpu/drm/radeon/radeon_fb.c: In function
'radeonfb_create_pinned_object':
drivers/gpu/drm/radeon/radeon_fb.c:144:18: error: 'struct drm_mode_fb_cmd2'
has no member named 'bpp'Signed-off-by: Stephen Rothwell
Signed-off-by: Dave Airlie
28 Nov, 2011
3 commits
-
fops field in drm_driver is a pointer to file_operations
struct, not embedded structureSigned-off-by: Ilija Hadzic
Signed-off-by: Dave Airlie -
psb_gfx.mod.c is a generated file and should not be
revision controlledSigned-off-by: Ilija Hadzic
Signed-off-by: Dave Airlie -
This takes over the staging change into the mainline driver.
Fixes -next part one.
Signed-off-by: Dave Airlie
25 Nov, 2011
1 commit
-
Here's a patch to move things over to the new addfb2 interfaces at
least.Signed-off-by: Dave Airlie
16 Nov, 2011
3 commits
-
* drm-gma500-alanc:
gma500: Now connect up to the DRM build to finish the job
gma500: fixup build versus latest header changes.
gma500: Add support for Cedarview
gma500: Add Oaktrail support
gma500: Add Poulsbo support
gma500: Add the core DRM files and headers
gma500: Add the i2c bus support
gma500: Add the glue to the various BIOS and firmware interfaces
gma500: Add device framework
gma500: introduce the framebuffer support code
gma500: introduce the GTT and MMU handling logic
gma500: GEM and GEM glue
gma500: Move the basic driver out of staging -
Signed-off-by: Alan Cox
Signed-off-by: Dave Airlie -
Signed-off-by: Dave Airlie