Dave Airlie [Mon, 9 Nov 2015 22:36:59 +0000 (08:36 +1000)]
drm/nouveau: fix build failures on all non ARM.
gk20a is an ARM only GPU, so we can just do the correct thing on
ARM but fail on other architectures. The other option was to use
SWIOTLB as the define, which means phys_to_page exists, but
this seems clearer.
Egbert Eich [Wed, 11 Jun 2014 12:59:55 +0000 (14:59 +0200)]
drm/ast: Initialized data needed to map fbdev memory
Due to a missing initialization there was no way to map fbdev memory.
Thus for example using the Xserver with the fbdev driver failed.
This fix adds initialization for fix.smem_start and fix.smem_len
in the fb_info structure, which fixes this problem.
Requested-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: stable@vger.kernel.org Signed-off-by: Egbert Eich <eich@suse.de>
[pulled from SuSE tree by me - airlied] Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Sat, 7 Nov 2015 07:16:59 +0000 (17:16 +1000)]
Merge tag 'drm-intel-next-fixes-2015-11-06' of git://anongit.freedesktop.org/drm-intel into drm-next
Here's a handful of i915 fixes for drm-next/v4.4. Imre's commit alone
should address the remaining warnings galore you experienced on
Skylake. Almost all of the rest are also fixes against user or QA
reported bugs, with references.
* tag 'drm-intel-next-fixes-2015-11-06' of git://anongit.freedesktop.org/drm-intel:
drm/i915/skl: disable display side power well support for now
drm/i915: Extend DSL readout fix to BDW and SKL.
drm/i915: Do graphics device reset under forcewake
drm/i915: Skip fence installation for objects with rotated views (v4)
drm/i915: add quirk to enable backlight on Dell Chromebook 11 (2015)
drm/i915/skl: Prevent unclaimed register writes on skylake.
drm/i915: disable CPU PWM also on LPT/SPT backlight disable
drm/i915: Fix maxfifo watermark calc on vlv cursor planes
drm/i915: add hotplug activation period to hotplug update mask
drm/vmwgfx: Properly flush cursor updates and page-flips
With the introduction of the new command buffer mechanism,
proper care wasn't taken to flush cursor image updates and
event-less screen-target page-flips.
Fix this by introducing explicit flush points.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Brian Paul <brianp@vmware.com>
Imre Deak [Thu, 5 Nov 2015 21:04:11 +0000 (23:04 +0200)]
drm/i915/skl: disable display side power well support for now
The display power well support on this platform is in a somewhat broken
state atm, so disable it by default.
This in effect will get rid of incorrect assert WARNs about the CSR/DMC
firmware not being loaded during power well toggling. It also removes a
problem during driver loading where a register is accessed while its
backing power well is down, resulting in another WARN. Until we come up
with the root cause of the second problem and the proper fix for both
issues, keep all display side power wells on.
Also clarify a bit the option description.
Reported-by: Dave Airlie <airlied@redhat.com>
Reference: http://mid.gmane.org/CAPM=9tyjBQjSBTKa49cRr6SYkpNW7Pq-fUFznZZ8Y1snvvk7mA@mail.gmail.com Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1446757451-2777-1-git-send-email-imre.deak@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Mika Kuoppala [Thu, 5 Nov 2015 11:11:38 +0000 (13:11 +0200)]
drm/i915: Do graphics device reset under forcewake
We have a timed release of a forcewake when using
I915_READ/WRITE macros. wait_for() macro will go to quite
long sleep if the first read doesn't satisfy the condition for
successful exit. With these two interacting, it is possible that
we lose the forcewake during the wait_for() and the subsequent read
will reaquire forcewake.
Further experiments with skl shows that when we lose forcewake,
we lose the reset request we submitted. So reset request register
is not power context saved.
Grab forcewakes for all engines before starting the reset/request
dance so that all requests stay valid for the duration of reset
requisition across all the engines.
v2: Add comment on power well sleeps. Wrap the reset handling
under forcewake instead of just reset requests (Chris)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92774 Cc: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Tested-by: Tomi Sarvela <tomix.p.sarvela@intel.com> (v1, v2) Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1446721898-1450-1-git-send-email-mika.kuoppala@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Vivek Kasireddy [Fri, 30 Oct 2015 01:54:38 +0000 (18:54 -0700)]
drm/i915: Skip fence installation for objects with rotated views (v4)
While pinning a fb object to the display plane, only install a fence
if the object is using a normal view. This corresponds with the
behavior found in i915_gem_object_do_pin() where the fencability
criteria is determined only for objects with normal views.
v2:
Look at the object's map_and_fenceable flag to determine whether to
install a fence or not (Chris).
v3:
Pin and unpin a fence only if the current view type is normal.
v4:
Extend the "view type is normal" check for pin_fence as well.
vga_switcheroo: Drop client power state VGA_SWITCHEROO_INIT
hda_intel.c:azx_probe() defers initialization of an audio controller
on the discrete GPU if the GPU is powered off. The power state of the
GPU is determined by calling vga_switcheroo_get_client_state().
vga_switcheroo_get_client_state() returns VGA_SWITCHEROO_INIT if
vga_switcheroo is not enabled, i.e. if no second GPU or no handler
has registered.
This can go wrong in the following scenario:
- Driver for the integrated GPU is not loaded.
- Driver for the discrete GPU registers with vga_switcheroo, uses driver
power control to power down the GPU, handler cuts power to the GPU.
- Driver for the audio controller gets loaded after the GPU was powered
down, calls vga_switcheroo_get_client_state() which returns
VGA_SWITCHEROO_INIT instead of VGA_SWITCHEROO_OFF.
- Consequence: azx_probe() tries to initialize the audio controller even
though the GPU is powered down.
The power state VGA_SWITCHEROO_INIT was introduced by c8e9cf7bb240
("vga_switcheroo: Add a helper function to get the client state").
It is not apparent what its benefit might be. The idea seems to
be to initialize the audio controller even if the power state is
VGA_SWITCHEROO_OFF (were vga_switcheroo enabled), but as shown
above this can fail.
Dave Airlie [Thu, 5 Nov 2015 01:07:03 +0000 (11:07 +1000)]
Merge tag 'drm/tegra/for-4.4-rc1' of git://anongit.freedesktop.org/tegra/linux into drm-next
drm/tegra: Changes for v4.4-rc1
Just a couple of minor fixes and cleanups for this cycle.
* tag 'drm/tegra/for-4.4-rc1' of git://anongit.freedesktop.org/tegra/linux:
drm/tegra: dc: Request/free syncpoint at init/exit
drm/tegra: fb: Remove gratuituous blank line
gpu: host1x: Fix MLOCK's debug info
Dave Airlie [Thu, 5 Nov 2015 01:06:56 +0000 (11:06 +1000)]
Merge tag 'imx-drm-next-2015-10-30' of git://git.pengutronix.de/git/pza/linux into drm-next
imx-drm fixes and color format updates
- Some correctness fixes found by coccinelle
- Add drivers/gpu/ipu-v3 directory to MAINTAINERS
- Add support for more color formats
- Fix a regression, making displays larger than FullHD work again
* tag 'imx-drm-next-2015-10-30' of git://git.pengutronix.de/git/pza/linux:
drm/imx: hdmi: fix HDMI setup to allow modes larger than FullHD
gpu: ipu-v3: fix div_ratio type
gpu: ipu-v3: csi: add support for 8 bpp grayscale sensors.
drm/imx: enable ARGB4444 16-bit color format
gpu: ipu-v3: add support for ARGB4444 16-bit color format
drm/imx: ipuv3-plane: enable support for RGBX8888 and RGBA8888 pixel formats
gpu: ipu-v3: add support for RGBX8888 and RGBA8888 pixel formats
drm/imx: enable 15-bit RGB with 1-bit alpha formats
gpu: ipu-v3: add support for 15-bit RGB with 1-bit alpha formats
MAINTAINERS: Add IPUv3 core driver to the i.MX DRM driver section
gpu: ipu-v3: ipu-csi: bool test doesn't need a comparison to false
Dave Airlie [Thu, 5 Nov 2015 00:57:25 +0000 (10:57 +1000)]
Merge branch 'drm-next-4.4' of git://people.freedesktop.org/~agd5f/linux into drm-next
- Updated register headers for GFX 8.1 for Stoney
- Add some new CZ revisions
- minor pageflip optimizations
- Fencing clean up
- Warning fix
- More fence cleanup
- oops fix
- Fiji fixes
* 'drm-next-4.4' of git://people.freedesktop.org/~agd5f/linux: (29 commits)
drm/amdgpu: group together common fence implementation
drm/amdgpu: remove AMDGPU_FENCE_OWNER_MOVE
drm/amdgpu: remove now unused fence functions
drm/amdgpu: fix fence fallback check
drm/amdgpu: fix stoping the scheduler timeout
drm/amdgpu: cleanup on error in amdgpu_cs_ioctl()
drm/amdgpu: update Fiji's Golden setting
drm/amdgpu: update Fiji's rev id
drm/amdgpu: extract common code in vi_common_early_init
drm/amd/scheduler: don't oops on failure to load
drm/amdgpu: don't oops on failure to load (v2)
drm/amdgpu: don't VT switch on suspend
drm/amdgpu: Make amdgpu_mn functions inline
drm/amdgpu: remove amdgpu_fence_ref/unref
drm/amdgpu: use common fence for sync
drm/amdgpu: use the new fence_is_later
drm/amdgpu: use common fences for VMID management v2
drm/amdgpu: move ring_from_fence to common code
drm/amdgpu: switch to common fence_wait_any_timeout v2
drm/amdgpu: remove unneeded fence functions
...
Christian König [Tue, 3 Nov 2015 11:21:57 +0000 (12:21 +0100)]
drm/amdgpu: group together common fence implementation
And also add some missing function documentation. No functional change.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 27 Oct 2015 16:28:24 +0000 (17:28 +0100)]
drm/amdgpu: remove AMDGPU_FENCE_OWNER_MOVE
Moves are exclusive operations anyway, just use the undefined owner for those.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 26 Oct 2015 12:22:33 +0000 (13:22 +0100)]
drm/amdgpu: remove now unused fence functions
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 3 Nov 2015 10:26:42 +0000 (11:26 +0100)]
drm/amdgpu: fix fence fallback check
Interrupts are notorious unreliable, enable the fallback at
a couple of more places.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 2 Nov 2015 15:25:10 +0000 (16:25 +0100)]
drm/amdgpu: fix stoping the scheduler timeout
cancel_delayed_work_sync is forbidden in interrupt context.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Dan Carpenter [Wed, 4 Nov 2015 13:25:09 +0000 (16:25 +0300)]
drm/amdgpu: cleanup on error in amdgpu_cs_ioctl()
We recently changed the locking in this function and now there is a
missing unlock on error. Also there are some other resources that we
should probably release as well...
Fixes: f48b2659f521 ('drm/amdgpu: fix the broken vm->mutex V2') Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Daniel Stone [Tue, 3 Nov 2015 21:42:31 +0000 (21:42 +0000)]
drm/i915: Fix locking around GuC firmware load
The GuC firmware load requires struct_mutex to create a GEM object,
but this collides badly with request_firmware. Move struct_mutex
locking down into the loader itself, so we don't hold it across the
entire load process, including request_firmware.
[ 20.451400] ======================================================
[ 20.451420] [ INFO: possible circular locking dependency detected ]
[ 20.451441] 4.3.0-rc5+ #1 Tainted: G W
[ 20.451457] -------------------------------------------------------
[ 20.451477] plymouthd/371 is trying to acquire lock:
[ 20.451494] (&dev->struct_mutex){+.+.+.}, at: [<ffffffffa0093c62>]
drm_gem_mmap+0x112/0x290 [drm]
[ 20.451538]
but task is already holding lock:
[ 20.451557] (&mm->mmap_sem){++++++}, at: [<ffffffff811fd9ac>]
vm_mmap_pgoff+0x8c/0xf0
[ 20.451591]
which lock already depends on the new lock.
Dave Airlie [Tue, 3 Nov 2015 16:10:03 +0000 (11:10 -0500)]
drm/amd/scheduler: don't oops on failure to load
In two places amdgpu tries to tear down something it hasn't
initalised when failing. This is what happens when you
enable experimental support on topaz which then fails in
ring init.
This patch allows it to fail cleanly.
agd: Split out from from the original patch since the
scheduler is a driver independent.
Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Cc: stable@vger.kernel.org
Dave Airlie [Tue, 3 Nov 2015 16:07:11 +0000 (11:07 -0500)]
drm/amdgpu: don't oops on failure to load (v2)
In two places amdgpu tries to tear down something it hasn't
initalised when failing. This is what happens when you
enable experimental support on topaz which then fails in
ring init.
This patch allows it to fail cleanly.
v2 (agd): split out scheduler change into a separate patch
Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Cc: stable@vger.kernel.org
Alex Deucher [Mon, 2 Nov 2015 15:52:32 +0000 (10:52 -0500)]
drm/amdgpu: don't VT switch on suspend
Reviewed-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Dave Airlie [Tue, 3 Nov 2015 05:42:59 +0000 (15:42 +1000)]
Merge branch 'linux-4.4' of git://anongit.freedesktop.org/git/nouveau/linux-2.6 into drm-next
- Vast improvements to gk20a instmem handling.
- Improved PGOB detection + GK107 support.
- Compatibility between old/new interfaces added, final missing piece to
finally enabling userspace to start using them.
- Kepler GDDR5 PLL stability improvements
- Support for non-GPIO (PWM) voltage controllers
- G8x/GT2xx memory clock improvements
- Misc other fixes
* 'linux-4.4' of git://anongit.freedesktop.org/git/nouveau/linux-2.6: (45 commits)
drm/nouveau: bump patchlevel to indicate availability of abi16/nvif interop
drm/nouveau/abi16: implement limited interoperability with usif/nvif
drm/nouveau/abi16: introduce locked variant of nouveau_abi16_get()
drm/nouveau/abi16: remove unused argument from nouveau_abi16_get()
drm/nouveau/pci: enable c800 magic for Medion Erazer X7827
drm/nouveau/pci: enable c800 magic for Lenovo Y510P
drm/nouveau/pll/gk104: fix PLL instability due to bad configuration with gddr5
drm/nouveau/clk/g84: Enable reclocking for GDDR3 G94-G200
drm/nouveau/bus/hwsq: Implement VBLANK waiting heuristic
drm/nouveau/fb/ramnv50: Script changes for G94 and up
drm/nouveau/fb/ramnv50: Deal with cards without timing entries
drm/nouveau/fb/ramnv50: Voltage GPIOs
drm/nouveau/fb/ramgt215: Restructure r111100 calculation for DDR2
drm/nouveau/fb/ramgt215: Change FBVDD/Q when BIOS asks for it
drm/nouveau/fb/ramgt215: Transform GPIO ramfuc method from FBVREF-specific to generic
drm/nouveau/bios/rammap: Identify DLLoff for >= GF100
drm/nouveau/pci: Handle 5-bit and 8-bit tag field
drm/nouveau/disp,pm: constify nvkm_object_func structures
drm/nouveau/gr: add FERMI_COMPUTE_B class to GF110+
drm/nouveau/gr: document mp error 0x10
...
Dave Airlie [Tue, 3 Nov 2015 05:42:10 +0000 (15:42 +1000)]
Merge branch 'exynos-drm-next' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos into drm-next
This pull request includes comprehensive cleanups to HDMI part and
several fixups. In addition, this pull request includes also a defconfig
patch which enables mixer driver as default. For this, I got already
Acked-by from Krzysztof Kozlowski who is a Exynos SoC maintainer.
* 'exynos-drm-next' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos: (34 commits)
drm/exynos/gem: remove DMA-mapping hacks used for constructing page array
ARM: exynos_defconfig: enable Exynos DRM Mixer driver
drm/exynos: simplify Kconfig component names
drm/exynos: re-arrange Kconfig entries
drm/exynos: abstract out common dependency
drm/exynos: separate Mixer and HDMI drivers
drm/exynos/mixer: replace direct cross-driver call with drm mode validation
drm/exynos: add atomic_check callback to exynos_crtc
drm/exynos/decon5433: add support for DECON-TV
drm/exynos/decon5433: remove duplicated initialization
drm/exynos/decon5433: merge different flag fields
drm/exynos/decon5433: add function to set particular register bits
drm/exynos/decon5433: fix timing registers writes
drm/exynos/decon5433: add PCLK clock
drm/exynos: cleanup name of gem object for exynos_drm
drm/exynos: fix to detach device of iommu
drm/exynos: add cursor plane support
drm/exynos: add global macro for the default primary plane
drm/exynos: fix spelling errors
drm: exynos: mixer: fix using usleep() in atomic context
...
Ilia Mirkin [Sat, 31 Oct 2015 19:06:11 +0000 (15:06 -0400)]
drm/nouveau/pci: enable c800 magic for Medion Erazer X7827
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91557 Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ilia Mirkin [Tue, 27 Oct 2015 21:39:49 +0000 (17:39 -0400)]
drm/nouveau/pci: enable c800 magic for Lenovo Y510P
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=70354#c75 Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Karol Herbst [Sun, 16 Aug 2015 08:19:25 +0000 (10:19 +0200)]
drm/nouveau/pll/gk104: fix PLL instability due to bad configuration with gddr5
This patch uses an approach closer to the nvidia driver to configure
both PLLs for high gddr5 memory clocks (usually above 2400MHz)
Previously nouveau used the one PLL as it was used for the lower clocks
and just adjusted the second PLL to get as close as possible to the
requested clock. This means for my card, that I got a 4050 MHz clock
although 4008 MHz was requested.
Now the driver iterates over a list of PLL configuration also used by
the nvidia driver and then adjust the second PLL to get near the
requested clock. Also it hold to some restriction I found while
analyzing the PLL configurations
This won't fix all gddr5 high clock issues itself, but it should be
fine on hybrid gpu systems as found on many laptops these days. Also
switching while normal desktop usage should be a lot more stable than
before.
v2: move the pll code into ramgk104
Signed-off-by: Karol Herbst <nouveau@karolherbst.de> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Roy Spliet [Tue, 29 Sep 2015 23:23:50 +0000 (00:23 +0100)]
drm/nouveau/fb/ramnv50: Script changes for G94 and up
10053c is not even read on some cards, and I have no idea exactly what the
criteria are. Likely NVIDIA pre-scans the VBIOS and in their driver disables
all features that are never used. The practical effect should be the same
as this implementation though.
Signed-off-by: Roy Spliet <rspliet@eclipso.eu> Tested-by: Pierre Moreau <pierre.morrow@free.fr> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Pierre Moreau [Sat, 3 Oct 2015 19:35:16 +0000 (21:35 +0200)]
drm/nouveau/pci: Handle 5-bit and 8-bit tag field
If the hardware supports extended tag field (8-bit ones), then enable it.
This is usually done by the VBIOS, but not on some MBPs (see fdo#86537).
In case extended tag field is not supported, 5-bit tag field is used which
limits the possible number of requests to 32. Apparently bits 7:0 of
0x08841c stores some number of outstanding requests, so cap it to 32 if
extended tag is unsupported.
Fixes: fdo#86537
v2: Restrict changes to chipsets >= 0x84
v3:
* Add nvkm_pci_mask to pci.h
* Mask bit 8 before setting it
v4:
* Rename `add` argument of nvkm_pci_mask to `value`
* Move code from nvkm_pci_init to g84_pci_init and remove PCIe and chipset
checks
v5:
* Rebase code on latest PCI structure
* Restore PCIe check
* Fix namings in nvkm_pci_mask
* Rephrase part of the commit message
Signed-off-by: Pierre Moreau <pierre.morrow@free.fr> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ilia Mirkin [Wed, 7 Oct 2015 22:39:32 +0000 (18:39 -0400)]
drm/nouveau/gr: document mp error 0x10
NVIDIA provided the documentation for mp error 0x10, INVALID_ADDR_SPACE,
which apparently happens when trying to use an atomic operation on
local or shared memory (instead of global memory).
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 1 Oct 2015 01:36:58 +0000 (11:36 +1000)]
drm/nouveau/pci/nv46: attempt to fix msi, and re-enable by default
Was not able to obtain a trace of NVRM due to kernel version annoyances,
however, experimentally confirmed that the WAR we use on NV50/G8x boards
works here too.
Samuel Pitoiset [Thu, 24 Sep 2015 18:26:15 +0000 (20:26 +0200)]
drm/nouveau/ibus/gf100: increase wait timeout to avoid read faults
Increase clock timeout of some unknown engines in order to avoid failure
at high gpcclk rate.
This fixes IBUS read faults on my GF119 when reclocking is manually
enabled. Note that memory reclocking is completely broken and NvMemExec
has to be disabled to allow core clock reclocking only.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Martin Peres [Wed, 16 Sep 2015 19:45:33 +0000 (22:45 +0300)]
drm/nouveau/gm204/6: add voltage control using the new gk104 volt class
I got confirmation that we can read and change the voltage with the same code.
The divider is also computed correctly on the gm204 we got our hands on.
Thanks to Yoshimo on IRC for executing the tests on his gm204!
Signed-off-by: Martin Peres <martin.peres@free.fr> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Martin Peres [Tue, 8 Sep 2015 22:34:33 +0000 (00:34 +0200)]
drm/nouveau/volt/gk104: add support for pwm and gpio modes
Most Keplers actually use the GPIO-based voltage management instead of the new
PWM-based one. Use the GPIO mode as a fallback as it already gracefully handles
the case where no GPIOs exist.
All the Maxwells seem to use the PWM method though.
v2:
- Do not forget to commit the PWM configuration change!
Signed-off-by: Martin Peres <martin.peres@free.fr>
drm/nouveau/ttm: set the DMA mask for platform devices
So far the DMA mask was not set for platform devices, which limited them
to a 32-bit physical space. Allow dma_set_mask() to be called for
non-PCI devices, and also take the IOMMU bit into account since it could
restrict the physically addressable space.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
The Great Nouveau Refactoring Take II brought us a lot of goodness,
including acquire/release methods that are called before and after an
instobj is modified. These functions can be used as synchronization
points to manage CPU/GPU coherency if we modify an instobj using the
CPU.
This patch replaces the legacy and slow PRAMIN access for gk20a instmem
with CPU mappings and writes. A LRU list is used to unmap unused
mappings after a certain threshold (currently 1MB) of mapped instobjs is
reached. This allows mappings to be reused most of the time.
Accessing instobjs using the CPU requires to maintain the GPU L2 cache,
which we do in the acquire/release functions. This triggers a lot of L2
flushes/invalidates, but most of them are performed on an empty cache
(and thus return immediately), and overall context setup performance
greatly benefits from this (from 250ms to 160ms on Jetson TK1 for a
simple libdrm program).
Making L2 management more explicit should allow us to grab some more
performance in the future.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Reintroduce macros allowing us to test a register against a certain
mask, since this is the most common usage pattern for the more generic
nvkm_xsec macros and makes the code more concise and readable.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ilia Mirkin [Tue, 20 Oct 2015 05:15:39 +0000 (01:15 -0400)]
drm/nouveau/gem: return only valid domain when there's only one
On nv50+, we restrict the valid domains to just the one where the buffer
was originally created. However after the buffer is evicted to system
memory, we might move it back to a different domain that was not
originally valid. When sharing the buffer and retrieving its GEM_INFO
data, we still want the domain that will be valid for this buffer in a
pushbuf, not the one where it currently happens to be.
This resolves fdo#92504 and several others. These are due to suspend
evicting all buffers, making it more likely that they temporarily end up
in the wrong place.
Cc: stable@vger.kernel.org
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92504 Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marek Szyprowski [Tue, 13 Oct 2015 11:47:20 +0000 (13:47 +0200)]
drm/exynos/gem: remove DMA-mapping hacks used for constructing page array
Exynos GEM objects contains an array of pointers to the pages, which the
allocated buffer consists of. Till now the code used some hacks (like
relying on DMA-mapping internal structures or using ARM-specific
dma_to_pfn helper) to build this array. This patch fixes this by adding
proper call to dma_get_sgtable_attrs() and using the acquired scatter-list
to construct needed array. This approach is more portable (work also for
ARM64) and finally fixes the layering violation that was present in this
code.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com>
Mixer driver is selected by CONFIG_DRM_EXYNOS_HDMI option. Since Exynos5433
HDMI does not require Mixer. There will be separate options to select Mixer
and HDMI. Adding new option to defconfig before Kconfig will allow to keep
bisectability.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com> Reviewed-by: Krzysztof Kozlowski <k.kozlowski@samsung.com> Acked-by: Krzysztof Kozlowski <k.kozlowski@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com>
Andrzej Hajda [Mon, 26 Oct 2015 12:03:45 +0000 (13:03 +0100)]
drm/exynos: simplify Kconfig component names
Many Exynos DRM sub-options mentions Exynos DRM in their titles.
It is redundant and can be safely shortened. The patch additionally
makes some entries more descriptive.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com>
Andrzej Hajda [Mon, 26 Oct 2015 12:03:44 +0000 (13:03 +0100)]
drm/exynos: re-arrange Kconfig entries
Exynos DRM driver have quite big number of components and options.
The patch re-arranges them into three logical groups:
- CRTCs,
- Encoders and Bridges,
- Sub-drivers.
It should make driver options more clear.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com>
Andrzej Hajda [Mon, 26 Oct 2015 12:03:40 +0000 (13:03 +0100)]
drm/exynos/mixer: replace direct cross-driver call with drm mode validation
HDMI driver called directly function from MIXER driver to invalidate modes
not supported by MIXER. The patch replaces the hack with proper .atomic_check
callback.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com>
Andrzej Hajda [Mon, 26 Oct 2015 12:03:39 +0000 (13:03 +0100)]
drm/exynos: add atomic_check callback to exynos_crtc
Some CRTCs needs mode validation, this patch adds neccessary
callback to Exynos DRM framework. It is called from DRM core
via atomic_check helper for drm_crtc.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com>
Dave Airlie [Mon, 2 Nov 2015 22:02:44 +0000 (08:02 +1000)]
Merge branch 'vmwgfx-next' of git://people.freedesktop.org/~thomash/linux into drm-next
Changes for vmwgfx for 4.4. If there is time, I'll follow up with a series
to move to threaded irqs.
* 'vmwgfx-next' of git://people.freedesktop.org/~thomash/linux:
drm/vmwgfx: Replace iowrite/ioread with volatile memory accesses
drm/vmwgfx: Turn off support for multisample count != 0 v2
drm/vmwgfx: switch from ioremap_cache to memremap
Dan Williams [Wed, 28 Oct 2015 07:19:45 +0000 (00:19 -0700)]
drm/vmwgfx: switch from ioremap_cache to memremap
Per commit 2e586a7e017a "drm/vmwgfx: Map the fifo as cached" the driver
expects the fifo registers to be cacheable. In preparation for
deprecating ioremap_cache() convert its usage in vmwgfx to memremap().
Cc: David Airlie <airlied@linux.ie> Cc: Thomas Hellstrom <thellstrom@vmware.com> Cc: Sinclair Yeh <syeh@vmware.com> Cc: dri-devel@lists.freedesktop.org Signed-off-by: Dan Williams <dan.j.williams@intel.com> Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
drm/i915/skl: Prevent unclaimed register writes on skylake.
I'm getting unclaimed register writes when checking the WM registers
after the crtc is disabled. So I would imagine those are guarded by
the crtc power well. Fix this by not reading out wm state when the
power well is off.
Cc: stable@vger.kernel.org # v4.3 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92181 Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Thierry Reding [Mon, 24 Aug 2015 12:47:10 +0000 (14:47 +0200)]
drm/tegra: dc: Request/free syncpoint at init/exit
syncpoints are resources provided by host1x and their lifetime is tied
to the host1x device. They are not properly reference counted either, so
removing the host1x device before any of its clients causes a use-after-
free error. Adding proper reference counting would be a major enterprise
so work around it for now by requesting and freeing the syncpoint at
init and exit time, respectively. The host1x device is guaranteed to be
around at this point.