wait_for_completion_timeout returns 0 in case of timeout and never
return < 0 so there is no additional information in printing the
value of time_left here as it will always be 0, thus it can be dropped.
Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
wait_for_completion_timeout returns >= 0 but never
negative - so the error check should be against equality
to 0 not <= 0.
Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
return type of wait_for_completion_timeout is unsigned long not int, this
patch assigns the return value of wait_for_completion_timeout to an
appropriately typed and named variable.
Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Some DSI peripherals rely on the HS clock on DSI clock lane as their clock
source. If the clock lane transitions between HS and LP states, it
can disrupt the functioning of such peripherals.
The mipi dsi mode flag MIPI_DSI_CLOCK_NON_CONTINUOUS already exists for
such peripheral drivers. Use it to configure the bit CLKLN_HS_FORCE_REQUEST
in DSI_LANE_CTRL.
Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
If the GEM object is imported, drm_prime_gem_destroy needs to be
called to clean up dma buffer related information.
Signed-off-by: Jilai Wang <jilaiw@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Fix build warning when building edp/edp_aux.o due to missing
prototype for edp_aux_transfer.
This function is only used in edp_aux.c so just make it static.
Reported-by: Nicholas Mc Guire <hofrat@osadl.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
wait_for_completion_timeout returns 0 in case of timeout so printing the
return value here will always yield 0 and is therefor redundant - dropped.
Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
The timeout is passed as a constant which makes it HZ dependent because
jiffies are expected so it should be converted to jiffies. The actual
value is not clear from the code - my best guess is that this should be
300 milliseconds given that other timeouts are in milliseconds based on
looking at other drm drivers (e.g. exynos_drm_dsi.c:356 300ms,
tegra/dpaux.c:188 250ms) - this needs to be confirmed by someone who
knows the details of the driver.
Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
wait_for_completion_timeout return >= 0 but never negative so the check
logic looks inconsistent. Further the return value of
wait_for_completion_timeout was being passed up the call chain but the
x call sites as drm_dp_i2c_do_msg()/drm_dp_dpcd_access() check for < 0
thus timeout was being treated as success case.
<snip> drivers/gpu/drm/drm_dp_helper.c:drm_dp_i2c_do_msg()
mutex_lock(&aux->hw_mutex);
ret = aux->transfer(aux, msg);
mutex_unlock(&aux->hw_mutex);
if (ret < 0) {
<snip>
logic in edp_aux_transfer() seems incorrect as it could return 0 (timeout)
but checks of <= 0 to indicate error so the return probably should be
-ETIMEDOUT in case wait_for_completion_timeout returns 0 (timeout
occurred).
Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Seems like disable can race with complete_flip() in process of disabling
a crtc, leading to:
[ 49.065364] Call trace:
[ 49.071441] [<ffffffc00041d5a0>] mdp5_ctl_blend+0x20/0x1c0
[ 49.073788] [<ffffffc00041ebcc>] mdp5_crtc_disable+0x3c/0xa8
[ 49.079348] [<ffffffc0003e7854>] disable_outputs.isra.4+0x11c/0x220
[ 49.085164] [<ffffffc0003e7afc>] drm_atomic_helper_commit_modeset_disables+0x14/0x38
[ 49.091155] [<ffffffc000425c80>] complete_commit+0x40/0xb8
[ 49.099136] [<ffffffc0004260ac>] msm_atomic_commit+0x364/0x398
[ 49.104430] [<ffffffc00040a614>] drm_atomic_commit+0x3c/0x70
[ 49.110249] [<ffffffc0003e67b8>] drm_atomic_helper_set_config+0x1b0/0x3e0
[ 49.116065] [<ffffffc0003f99bc>] drm_mode_set_config_internal+0x64/0xf8
[ 49.122746] [<ffffffc0003fa624>] drm_framebuffer_remove+0xe4/0x128
[ 49.129171] [<ffffffc0003feaf8>] drm_mode_rmfb+0xc0/0x100
[ 49.135420] [<ffffffc0003efba8>] drm_ioctl+0x258/0x4d0
[ 49.140889] [<ffffffc0001b0388>] do_vfs_ioctl+0x338/0x5d0
[ 49.145921] [<ffffffc0001b06a8>] SyS_ioctl+0x88/0xa0
It makes no sense to free the ctl without disabling all stages, so lets
just move them together to avoid the crash.
Signed-off-by: Rob Clark <robdclark@gmail.com>
As found in apq8016 (used in DragonBoard 410c) and msm8916.
Note that numerically a306 is actually 307 (since a305c already claimed
306). Nice and confusing.
Signed-off-by: Rob Clark <robdclark@gmail.com>
A few spots in the driver have support for downstream android
CONFIG_MSM_BUS_SCALING. This is mainly to simplify backporting the
driver for various devices which do not have sufficient upstream
kernel support. But the intentionally dead code seems to cause
some confusion. Rename the #define to make this more clear.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Using fb modifier flag, support NV12MT format in MDP4.
v2:
- rework the modifier's description [Daniel Vetter's comment]
- drop .set_mode_config() callback [Rob Clark's comment]
v3:
- change VENDOR's name and restrict usage to NV12 [pointed by Daniel]
Signed-off-by: Rob Clark <robdclark@gmail.com>
Dump a bit more info when the GPU hangs, without having hang_debug
enabled (which dumps a *lot* of registers). Also dump the scratch
registers, as they are useful for determining where in the cmdstream
the GPU hung (and they seem always safe to read when GPU has hung).
Note that the freedreno gallium driver emits increasing counter values
to SCRATCH6 (to identify tile #) and SCRATCH7 (to identify draw #), so
these two in particular can be used to "triangulate" where in the
cmdstream the GPU hung.
Signed-off-by: Rob Clark <robdclark@gmail.com>
The goto is correct, and we never reach the return statement so just
delete the dead code.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
virtio_gpu_alloc_object() returns an error pointer, it never returns
NULL.
Fixes: dc5698e80c ('Add virtio gpu driver.')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
More fixes for amdgpu for 4.2. We've integrated Jerome's comments
about the interface among other things. I'll be on vacation next week
so Christian will be handling any updates next week.
* 'drm-next-4.2-amdgpu' of git://people.freedesktop.org/~agd5f/linux: (23 commits)
drm/amdgpu: fix a amdgpu_dpm=0 bug
drm/amdgpu: don't enable/disable display twice on suspend/resume
drm/amdgpu: fix UVD/VCE VM emulation
drm/amdgpu: enable vce powergating
drm/amdgpu/iceland: don't call smu_init on resume
drm/amdgpu/tonga: don't call smu_init on resume
drm/amdgpu/cz: don't call smu_init on resume
drm/amdgpu: update to latest gfx8 golden register settings
drm/amdgpu: whitespace cleanup in gmc8 golden regs
drm/admgpu: move XDMA golden registers to dce code
drm/amdgpu: fix the build on big endian
drm/amdgpu: cleanup UAPI comments
drm/amdgpu: remove AMDGPU_CTX_OP_STATE_RUNNING
drm/amdgpu: remove the VI hardware semaphore in ring sync
drm/amdgpu: set the gfx config properly for all CZ variants (v2)
drm/amdgpu: also print the pci revision when printing the pci ids
drm/amdgpu: cleanup VA IOCTL
drm/amdgpu: fix saddr handling in amdgpu_vm_bo_unmap
drm/amdgpu: fix amdgpu_vm_bo_map
drm/amdgpu: remove unused AMDGPU_IB_FLAG_GDS
...
We were doing it in the common code and in the IP specific code.
Remove the IP specific code. The common code handles the
ordering properly.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Enable VCE dpm and powergating. VCE dpm dynamically scales the VCE clocks on
demand.
Signed-off-by: Sonny Jiang <sonny.jiang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
smu_init allocates buffers and initializes them. It does not
touch the hw. There is no need to do it again on resume. It
should really be part of sw_init (and smu_fini should be part
of sw_fini), but we need the firmware sizes from the other IPs
for firmware loading so we have to wait until sw init is done
for all other IPs.
Reviewed-by: Sonny Jiang <Sonny.Jiang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
smu_init allocates buffers and initializes them. It does not
touch the hw. There is no need to do it again on resume. It
should really be part of sw_init (and smu_fini should be part
of sw_fini), but we need the firmware sizes from the other IPs
for firmware loading so we have to wait until sw init is done
for all other IPs.
Reviewed-by: Sonny Jiang <Sonny.Jiang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
smu_init allocates buffers and initializes them. It does not
touch the hw. There is no need to do it again on resume. It
should really be part of sw_init (and smu_fini should be part
of sw_fini), but we need the firmware sizes from the other IPs
for firmware loading so we have to wait until sw init is done
for all other IPs.
Reviewed-by: Sonny Jiang <Sonny.Jiang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Some leftover copy and pastes from radeon that never
got updated.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Need to adjust the number of CUs and RBs.
v2: get proper values
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Samuel Li <samuel.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The driver makes use of this information so print if to aid in
debugging.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Samuel Li <samuel.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Remove the unnecessary returned status and make the IOCTL write only.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
We need to reset the bo_va address, otherwise new mappings
wouldn't be updated in the page table.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu submits both kernel and user fences, but just need one interrupt,
disable user fence interrupt and don't effect user fence.
v2: fix merge error
Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
This is the big pull request for amdgpu, the new driver for VI+ AMD
asics. I currently supports Tonga, Iceland, and Carrizo and also
contains a Kconfig option to build support for CI parts for testing.
All major functionality is supported (displays, gfx, compute, dma,
video decode/encode, etc.). Power management is working on Carrizo,
but is still being worked on for Tonga and Iceland.
* 'drm-next-4.2-amdgpu' of git://people.freedesktop.org/~agd5f/linux: (106 commits)
drm/amdgpu: only support IBs in the buffer list (v2)
drm/amdgpu: add vram_type and vram_bit_width for interface query (v2)
drm/amdgpu: add ib_size/start_alignment interface query
drm/amdgpu: add me/ce/pfp_feature_version interface query
drm/amdgpu add ce_ram_size for interface query
drm/amdgpu add max_memory_clock for interface query (v2)
drm/amdgpu: add hdp flush for gfx8 compute ring
drm/amdgpu: fix no hdp flush for compute ring
drm/amdgpu: add HEVC/H.265 UVD support
drm/amdgpu: stop loading firmware with pm.mutex locked
drm/amdgpu: remove mclk_lock
drm/amdgpu: fix description of vm_size module parameter (v2)
drm/amdgpu: remove all sh mem register modification in vm flush
drm/amdgpu: rename GEM_OP_SET_INITIAL_DOMAIN -> GEM_OP_SET_PLACEMENT
drm/amdgpu: fence should be added to shared slot
drm/amdgpu: sync fence of clear_invalids (v2)
drm/amdgpu: max_pde_used usage should be under protect
drm/amdgpu: fix bug of vm_bo_map (v2)
drm/amdgpu: implement the allocation range (v3)
drm/amdgpu: rename amdgpu_ip_funcs to amd_ip_funcs (v2)
...
This function could return a NULL pointer in case of handle not
present and in case of out of memory conditions however caller
function always returned EINVAL error hiding a possible ENOMEM.
This patch change the function to return the error instead to
be able to propagate the error instead of assuming EINVAL.
Signed-off-by: Frediano Ziglio <fziglio@redhat.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
In qxlhw_handle_to_bo we incremented counters twice, one time for release object
and one for reloc_info.
In the main function however reloc_info references was drop much earlier than
release so keeping the pointer only on release is safe and make cleaning
process easier.
Signed-off-by: Frediano Ziglio <fziglio@redhat.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
qxl_bo structure has two reference counters, one in the GEM object and
another in the TTM object. The GEM object keep a counter to the TTM object
so when GEM counter reached zero the TTM counter (using qxl_bo_unref) was
decremented. The qxl object is fully freed (both GEM and TTM part are cleaned)
when the TTM counter reach zero.
One issue was that surface idr structure has no owning on qxl_bo objects however
it contains a pointer to qxl_bo object. This caused some nasty race condition
for instance qxl_bo object was reaped even after counter was already zero.
This patch fix these races moving main counter (the one used by qxl_bo_(un)ref)
to GEM object which cleanup routine (qxl_gem_object_free) remove the idr pointer
(using qxl_surface_evict) when the counters are still valid.
Signed-off-by: Frediano Ziglio <fziglio@redhat.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Hai Li