platform/kernel/linux-rpi.git
8 years agodrm/amd/scheduler: Mark amdgpu_sched_ops const
Nils Wallménius [Sun, 10 Apr 2016 14:30:00 +0000 (16:30 +0200)]
drm/amd/scheduler: Mark amdgpu_sched_ops const

This marks the struct amdgpu_sched_ops const and
adjusts amd_sched_init to take a const pointer
for the ops param. The ops member of
struct amd_gpu_scheduler is also changed to const.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Nils Wallménius <nils.wallmenius@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amd: Mark some tables as const
Nils Wallménius [Sun, 10 Apr 2016 14:29:59 +0000 (16:29 +0200)]
drm/amd: Mark some tables as const

This patch marks some compile-time constant tables 'const'.
The tables marked in this patch are the low hanging fruit
where little other changes were necesary to avoid casting
away constness etc. Also mark some tables that are private
to a file as static.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Nils Wallménius <nils.wallmenius@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amd/powerplay: fix stutter setup in mclk level init
Alex Deucher [Fri, 8 Apr 2016 20:42:38 +0000 (16:42 -0400)]
drm/amd/powerplay: fix stutter setup in mclk level init

Stale ifdef.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: add support for SET_APPEND_CNT packet3 (v2)
Dave Airlie [Wed, 6 Apr 2016 20:50:25 +0000 (06:50 +1000)]
drm/radeon: add support for SET_APPEND_CNT packet3 (v2)

This adds support to the command parser for the set append counter
packet3, this is required to support atomic counters on
evergreen/cayman GPUs.

v2: fixup some of the hardcoded numbers with real register names
(Christian)

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>
8 years agodrm/radeon: handle more than 10 UVD sessions
Arindam Nath [Wed, 6 Apr 2016 19:33:52 +0000 (15:33 -0400)]
drm/radeon: handle more than 10 UVD sessions

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Arindam Nath <arindam.nath@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: add support for loading new UVD fw
Arindam Nath [Wed, 6 Apr 2016 19:33:51 +0000 (15:33 -0400)]
drm/radeon: add support for loading new UVD fw

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Arindam Nath <arindam.nath@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: reuse VMIDs already assigned to a process
Christian König [Wed, 9 Mar 2016 21:11:53 +0000 (22:11 +0100)]
drm/amdgpu: reuse VMIDs already assigned to a process

If we don't need to flush we can easily use another VMID
already assigned to the process.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: add a fence after the VM flush
Christian König [Tue, 1 Mar 2016 15:46:18 +0000 (16:46 +0100)]
drm/amdgpu: add a fence after the VM flush

This way we can track when the flush is done.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: use a sync object for VMID fences v2
Christian König [Mon, 15 Feb 2016 11:33:02 +0000 (12:33 +0100)]
drm/amdgpu: use a sync object for VMID fences v2

v2: rebase & cleanup

This way we can store more than one fence as user for each VMID.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com> (v1)
Reviewed-by: Chunming Zhou <david1.zhou@amd.com> (v1)
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: merge VM manager and VM context ID structure
Christian König [Tue, 8 Mar 2016 14:40:11 +0000 (15:40 +0100)]
drm/amdgpu: merge VM manager and VM context ID structure

No need to have two of them any more.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: Set PFLIP_SUBMITTED for crtc after address update
Andrey Grodzovsky [Wed, 30 Mar 2016 21:34:27 +0000 (17:34 -0400)]
drm/amdgpu: Set PFLIP_SUBMITTED for crtc after address update

Also add some pflip debug prints.

This change allows us to wait on pflip status until the new surface address
is actually submitted to the register.

This reverts ed3020e923240829dcdfd3343f6e91dc02c63775
drm/amdgpu: Move MMIO flip out of spinlocked region
The original change assumed DAL will aquire locks inside DAL
implemetion of page_flip callback which eventaully didn't happen.

This moves the flip before status update which makes sense for the
non-DAL code pathes as well.

Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: Support DRM_MODE_PAGE_FLIP_ASYNC
Michel Dänzer [Fri, 1 Apr 2016 09:51:34 +0000 (18:51 +0900)]
drm/radeon: Support DRM_MODE_PAGE_FLIP_ASYNC

When this flag is set, we program the hardware to execute the flip
during horizontal blank (i.e. for the next scanline) instead of during
vertical blank (i.e. for the next frame).

Currently this is only supported on ASICs which have a page flip
completion interrupt (>= R600), and only if the use_pflipirq parameter
has value 2 (the default).

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amd/powerplay: add deep sleep divider id into DPM table on Tonga
Eric Huang [Wed, 30 Mar 2016 20:30:12 +0000 (16:30 -0400)]
drm/amd/powerplay: add deep sleep divider id into DPM table on Tonga

Add a proper implementation for setting the deep sleep divider.

Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: optionally enable GART debugfs file
Christian König [Wed, 30 Mar 2016 12:42:57 +0000 (14:42 +0200)]
drm/amdgpu: optionally enable GART debugfs file

Keeping the pages array around can use a lot of system memory
when you want a large GART.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: remove GART page addr array
Christian König [Wed, 30 Mar 2016 08:54:16 +0000 (10:54 +0200)]
drm/amdgpu: remove GART page addr array

Not needed any more.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: use BO pages instead of GART array
Christian König [Wed, 30 Mar 2016 08:50:25 +0000 (10:50 +0200)]
drm/amdgpu: use BO pages instead of GART array

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: change parameter passing in the VM code
Christian König [Fri, 18 Mar 2016 20:00:35 +0000 (21:00 +0100)]
drm/amdgpu: change parameter passing in the VM code

Make it more flexible by passing src and page addresses
directly instead of the structures they contain.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: drop the GTT power of two limit
Christian König [Thu, 17 Mar 2016 15:25:15 +0000 (16:25 +0100)]
drm/amdgpu: drop the GTT power of two limit

As far as I can see that isn't neccessary any more.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amd/powerplay: use min_clock_in_sr for deep sleep feature.
Rex Zhu [Tue, 29 Mar 2016 11:32:37 +0000 (19:32 +0800)]
drm/amd/powerplay: use min_clock_in_sr for deep sleep feature.

This comes from the display handling code.

Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amd/amdgpu: Enable clockgating in UVD6 for Stoney
Tom St Denis [Wed, 23 Mar 2016 17:17:04 +0000 (13:17 -0400)]
drm/amd/amdgpu: Enable clockgating in UVD6 for Stoney

This patch enables clockgating for the UVD6 block in Stoney.

Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amd/amdgpu: Enable clockgating for UVD5 on Tonga
Tom St Denis [Wed, 23 Mar 2016 17:16:13 +0000 (13:16 -0400)]
drm/amd/amdgpu: Enable clockgating for UVD5 on Tonga

This patch enables clock gating for the UVD5 block with
Tonga.

Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amd/amdgpu: Add SW clock gating support to UVD 5 and 6
Tom St Denis [Wed, 23 Mar 2016 17:14:31 +0000 (13:14 -0400)]
drm/amd/amdgpu: Add SW clock gating support to UVD 5 and 6

This patch adds support for software clock gating to UVD 5
and UVD 6 blocks with a preliminary commented out hardware
gating routine.

Currently hardware gating does not work so it's not activated.

Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: don't include RADEON_HPD_NONE in HPD IRQ enable bitsets
Nicolai Stange [Tue, 22 Mar 2016 21:05:27 +0000 (22:05 +0100)]
drm/radeon: don't include RADEON_HPD_NONE in HPD IRQ enable bitsets

The values of all but the RADEON_HPD_NONE members of the radeon_hpd_id
enum transform 1:1 into bit positions within the 'enabled' bitset as
assembled by evergreen_hpd_init():

  enabled |= 1 << radeon_connector->hpd.hpd;

However, if ->hpd.hpd happens to equal RADEON_HPD_NONE == 0xff, UBSAN
reports

  UBSAN: Undefined behaviour in drivers/gpu/drm/radeon/evergreen.c:1867:16
  shift exponent 255 is too large for 32-bit type 'int'
  [...]
  Call Trace:
   [<ffffffff818c4d35>] dump_stack+0xbc/0x117
   [<ffffffff818c4c79>] ? _atomic_dec_and_lock+0x169/0x169
   [<ffffffff819411bb>] ubsan_epilogue+0xd/0x4e
   [<ffffffff81941cbc>] __ubsan_handle_shift_out_of_bounds+0x1fb/0x254
   [<ffffffffa0ba7f2e>] ? atom_execute_table+0x3e/0x50 [radeon]
   [<ffffffff81941ac1>] ? __ubsan_handle_load_invalid_value+0x158/0x158
   [<ffffffffa0b87700>] ? radeon_get_pll_use_mask+0x130/0x130 [radeon]
   [<ffffffff81219930>] ? wake_up_klogd_work_func+0x60/0x60
   [<ffffffff8121a35e>] ? vprintk_default+0x3e/0x60
   [<ffffffffa0c603c4>] evergreen_hpd_init+0x274/0x2d0 [radeon]
   [<ffffffffa0c603c4>] ? evergreen_hpd_init+0x274/0x2d0 [radeon]
   [<ffffffffa0bd196e>] radeon_modeset_init+0x8ce/0x18d0 [radeon]
   [<ffffffffa0b71d86>] radeon_driver_load_kms+0x186/0x350 [radeon]
   [<ffffffffa03b6b16>] drm_dev_register+0xc6/0x100 [drm]
   [<ffffffffa03bc8c4>] drm_get_pci_dev+0xe4/0x490 [drm]
   [<ffffffff814b83f0>] ? kfree+0x220/0x370
   [<ffffffffa0b687c2>] radeon_pci_probe+0x112/0x140 [radeon]
   [...]
  =====================================================================
  radeon 0000:01:00.0: No connectors reported connected with modes

At least on x86, there should be no user-visible impact as there

  1 << 0xff == 1 << (0xff & 31) == 1 << 31

holds and 31 > RADEON_MAX_HPD_PINS. Thus, this patch is a cosmetic one.

All of the above applies analogously to evergreen_hpd_fini(),
r100_hpd_init(), r100_hpd_fini(), r600_hpd_init(), r600_hpd_fini(),
rs600_hpd_init() and rs600_hpd_fini()

Silence UBSAN by checking ->hpd.hpd for RADEON_HPD_NONE before oring it
into the 'enabled' bitset in the *_init()- or the 'disabled' bitset in
the *_fini()-functions respectively.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: refine code for code style.
Rex Zhu [Wed, 16 Mar 2016 07:17:18 +0000 (15:17 +0800)]
drm/amdgpu: refine code for code style.

White space fix.

Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: No need to stop hw init although vce's state was not true.
Rex Zhu [Wed, 16 Mar 2016 06:48:18 +0000 (14:48 +0800)]
drm/amdgpu: No need to stop hw init although vce's state was not true.

This is not a fatal error.

v2: add comment why ignore the error here.

Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: fix issue that can't set vce clock gate.
Rex Zhu [Wed, 16 Mar 2016 06:45:40 +0000 (14:45 +0800)]
drm/amdgpu: fix issue that can't set vce clock gate.

Need to soft reset VCE as part of the clockgating
sequence.

Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: use ref to keep job alive
Monk Liu [Thu, 10 Mar 2016 04:14:44 +0000 (12:14 +0800)]
drm/amdgpu: use ref to keep job alive

this is to fix fatal page fault error that occured if:
job is signaled/released after its timeout work is already
put to the global queue (in this case the cancel_delayed_work
will return false), which will lead to NX-protection error
page fault during job_timeout_func.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: rework TDR in scheduler (v2)
Monk Liu [Fri, 4 Mar 2016 10:51:02 +0000 (18:51 +0800)]
drm/amdgpu: rework TDR in scheduler (v2)

Add two callbacks to scheduler to maintain jobs, and invoked for
job timeout calculations. Now TDR measures time gap from
job is processed by hw.

v2:
fix typo

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: get rid of incorrect TDR
Monk Liu [Fri, 4 Mar 2016 06:42:26 +0000 (14:42 +0800)]
drm/amdgpu: get rid of incorrect TDR

original time out detect routine is incorrect, cuz it measures
the gap from job scheduled, but we should only measure the
gap from processed by hw.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: put job to list before done
Monk Liu [Fri, 4 Mar 2016 06:33:44 +0000 (14:33 +0800)]
drm/amdgpu: put job to list before done

the mirror_list will be used for later time out detect
feature.  This is needed to properly detect a GPU
timeout with the scheduler.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: delay job free to when it's finished (v2)
Monk Liu [Thu, 3 Mar 2016 11:00:50 +0000 (19:00 +0800)]
drm/amdgpu: delay job free to when it's finished (v2)

for those jobs submitted through scheduler, do not
free it immediately after scheduled, instead free it
in global workqueue by its sched fence signaling
callback function.

v2:
call uf's bo_undef after job_run()
call job's sync free after job_run()
no static inline __amdgpu_job_free() anymore, just use
kfree(job) to replace it.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu/dce11: fix vertical bars appear on monitor
Vitaly Prosyak [Fri, 18 Mar 2016 19:49:41 +0000 (15:49 -0400)]
drm/amdgpu/dce11: fix vertical bars appear on monitor

Fixed mc stop and resume hardware programming sequence.

Signed-off-by: Vitaly Prosyak <vitaly.prosyak@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: use sched_job_init to initialize sched_job
Monk Liu [Mon, 7 Mar 2016 04:49:55 +0000 (12:49 +0800)]
drm/amdgpu: use sched_job_init to initialize sched_job

Consolidate job initialization in one place rather than
duplicating it in multiple places.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: patch cond exec for SDMA
Monk Liu [Thu, 14 Jan 2016 11:07:38 +0000 (19:07 +0800)]
drm/amdgpu: patch cond exec for SDMA

More ground work for conditional execution on SDMA
necessary for preemption.

Signed-off-by: Monk Liu <monk.liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: support cond exec
Monk Liu [Thu, 14 Jan 2016 10:08:16 +0000 (18:08 +0800)]
drm/amdgpu: support cond exec

This adds the groundwork for conditional execution on
SDMA which is necessary for preemption.

Signed-off-by: Monk Liu <monk.liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: improve vmid assigment V2
Chunming Zhou [Thu, 17 Mar 2016 03:41:37 +0000 (11:41 +0800)]
drm/amdgpu: improve vmid assigment V2

V2: the signaled items on the LRU maintain their order

Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: mark amdgpu_allowed_register_entry tables as 'const'
Nils Wallménius [Sat, 19 Mar 2016 15:12:17 +0000 (16:12 +0100)]
drm/amdgpu: mark amdgpu_allowed_register_entry tables as 'const'

Signed-off-by: Nils Wallménius <nils.wallmenius@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: do not store bios_header_start in amdgpu_device
Nils Wallménius [Sat, 19 Mar 2016 15:12:13 +0000 (16:12 +0100)]
drm/amdgpu: do not store bios_header_start in amdgpu_device

It is only used locally in amdgpu_get_bios

Signed-off-by: Nils Wallménius <nils.wallmenius@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: delete unused struct member suspend from radeon_device
Nils Wallménius [Sat, 19 Mar 2016 15:12:12 +0000 (16:12 +0100)]
drm/radeon: delete unused struct member suspend from radeon_device

Signed-off-by: Nils Wallménius <nils.wallmenius@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/amdgpu: delete unused struct member suspend from amdgpu_device
Nils Wallménius [Sat, 19 Mar 2016 15:12:11 +0000 (16:12 +0100)]
drm/amdgpu: delete unused struct member suspend from amdgpu_device

Signed-off-by: Nils Wallménius <nils.wallmenius@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: hard reset r600 and newer GPU when hibernating.
Jérome Glisse [Fri, 18 Mar 2016 15:58:39 +0000 (16:58 +0100)]
drm/radeon: hard reset r600 and newer GPU when hibernating.

Some GPU block like UVD and VCE require hard reset to be properly
resume if there is no real powerdown of the asic like during various
hibernation step. This patch perform such hard reset.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: allow to force hard GPU reset.
Jérome Glisse [Fri, 18 Mar 2016 15:58:38 +0000 (16:58 +0100)]
drm/radeon: allow to force hard GPU reset.

In some cases, like when freezing for hibernation, we need to be
able to force hard reset even if no engine are stuck. This patch
add a bool option to current asic reset callback to allow to force
hard reset on asic that supports it.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: add driver option to disable vce block.
Jérome Glisse [Fri, 18 Mar 2016 15:58:37 +0000 (16:58 +0100)]
drm/radeon: add driver option to disable vce block.

Quite few suspend/hibernation bugs are related to this block. Add
an option to disable those as a work around.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: add driver option to disable uvd block.
Jérome Glisse [Fri, 18 Mar 2016 15:58:36 +0000 (16:58 +0100)]
drm/radeon: add driver option to disable uvd block.

Quite few suspend/hibernation bugs are related to this block. Add
an option to disable those as a work around.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate cik vce initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:35 +0000 (16:58 +0100)]
drm/radeon: consolidate cik vce initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_vce doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate si vce initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:34 +0000 (16:58 +0100)]
drm/radeon: consolidate si vce initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_vce doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate ni vce initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:33 +0000 (16:58 +0100)]
drm/radeon: consolidate ni vce initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_vce doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: add a vce flag to know if need to initialize vce or not.
Jérome Glisse [Fri, 18 Mar 2016 15:58:32 +0000 (16:58 +0100)]
drm/radeon: add a vce flag to know if need to initialize vce or not.

This will later on serve for module option to disable vce.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate cik uvd initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:31 +0000 (16:58 +0100)]
drm/radeon: consolidate cik uvd initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_uvd doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate si uvd initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:30 +0000 (16:58 +0100)]
drm/radeon: consolidate si uvd initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_uvd doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate ni uvd initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:29 +0000 (16:58 +0100)]
drm/radeon: consolidate ni uvd initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_uvd doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate evergreen uvd initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:28 +0000 (16:58 +0100)]
drm/radeon: consolidate evergreen uvd initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_uvd doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate rv770 uvd initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:27 +0000 (16:58 +0100)]
drm/radeon: consolidate rv770 uvd initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_uvd doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agodrm/radeon: consolidate r600 uvd initialization and startup code.
Jérome Glisse [Fri, 18 Mar 2016 15:58:26 +0000 (16:58 +0100)]
drm/radeon: consolidate r600 uvd initialization and startup code.

This match the exact same control flow as existing code. It just
use goto instead of multiple levels of if/else. It also clarify
early initialization failures by clearing rdev->has_uvd doing so
does not change end result from hardware point of view, it only
avoids printing more error messages down the line and thus only
the original error is reported.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
8 years agoMerge branch 'for-next' of http://git.agner.ch/git/linux-drm-fsl-dcu into drm-next
Dave Airlie [Fri, 29 Apr 2016 04:57:51 +0000 (14:57 +1000)]
Merge branch 'for-next' of git.agner.ch/git/linux-drm-fsl-dcu into drm-next

This adds very rudimentary TCON (timing controller for raw LCD displays)
support to enable the bypass mode in order to use the DCU controller on
Freescale/NXP Vybrid SoC's.

Additionally the register clock and pixel clock has been separated, but
are currently still enabled and disabled pairwise.

Other than that, fixes and cleanups accross the driver.

* 'for-next' of http://git.agner.ch/git/linux-drm-fsl-dcu:
  drm/fsl-dcu: increment version and date
  drm/fsl-dcu: implement lastclose callback
  drm/fsl-dcu: disable output polling on driver unload
  drm/fsl-dcu: deallocate fbdev CMA on unload
  drm/fsl-dcu: use variable name dev for struct drm_device
  drm/fsl-dcu: handle missing panel gracefully
  drm/fsl-dcu: detach panel on destroy
  drm/layerscape: reduce excessive stack usage
  drm/fsl-dcu: add TCON driver
  drm/fsl-dcu: use common clock framework for pixel clock divider
  drm/fsl-dcu: add extra clock for pixel clock
  drm/fsl-dcu: disable clock on initialization failure and remove

8 years agoMerge tag 'sun4i-drm-for-4.7' of https://git.kernel.org/pub/scm/linux/kernel/git...
Dave Airlie [Fri, 29 Apr 2016 04:52:41 +0000 (14:52 +1000)]
Merge tag 'sun4i-drm-for-4.7' of https://git./linux/kernel/git/mripard/linux into drm-next

Allwinner DRM driver for 4.7

This pull request introduces the sun4i driver, meant to be used on the
older Allwinner SoCs (A10, A13, A20, A23, A31 and A33).

It currently supports only the A13, which has one of the simplest video
pipeline. Support for other video components and SoCs will be added
eventually.

It supports only a RGB or composite output. It doesn't do HDMI, VGA, LVDS
or power management yet, but that will come in time as well.

* tag 'sun4i-drm-for-4.7' of https://git.kernel.org/pub/scm/linux/kernel/git/mripard/linux:
  MAINTAINERS: Add a maintainer for the Allwinner DRM driver
  drm: sun4i: tv: Add NTSC output standard
  drm: sun4i: tv: Add PAL output standard
  drm: sun4i: Add composite output
  drm: sun4i: Add RGB output
  drm: Add Allwinner A10 Display Engine support
  drm: sun4i: Add DT bindings documentation
  drm: fb: Add seq_file definition

8 years agoMAINTAINERS: Add a maintainer for the Allwinner DRM driver
Maxime Ripard [Thu, 28 Apr 2016 08:12:33 +0000 (10:12 +0200)]
MAINTAINERS: Add a maintainer for the Allwinner DRM driver

Add myself as the maintainer of the new Allwinner DRM driver.

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
8 years agodrm: sun4i: tv: Add NTSC output standard
Maxime Ripard [Thu, 29 Oct 2015 08:39:56 +0000 (09:39 +0100)]
drm: sun4i: tv: Add NTSC output standard

Add the settings to support the NTSC standard.

Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
8 years agodrm: sun4i: tv: Add PAL output standard
Maxime Ripard [Thu, 29 Oct 2015 08:39:41 +0000 (09:39 +0100)]
drm: sun4i: tv: Add PAL output standard

Now that we have support for the composite output, we can start adding new
supported standards. Start with PAL, and we will add other eventually.

Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
8 years agodrm: sun4i: Add composite output
Maxime Ripard [Thu, 29 Oct 2015 08:39:01 +0000 (09:39 +0100)]
drm: sun4i: Add composite output

Some Allwinner SoCs have an IP called the TV encoder that is used to output
composite and VGA signals. In such a case, we need to use the second TCON
channel.

Add support for that TV encoder.

Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
8 years agodrm: sun4i: Add RGB output
Maxime Ripard [Thu, 29 Oct 2015 08:37:32 +0000 (09:37 +0100)]
drm: sun4i: Add RGB output

One of the A10 display pipeline possible output is an RGB interface to
drive LCD panels directly. This is done through the first channel of the
TCON that will output our video signals directly.

Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
8 years agodrm: Add Allwinner A10 Display Engine support
Maxime Ripard [Thu, 29 Oct 2015 08:36:23 +0000 (09:36 +0100)]
drm: Add Allwinner A10 Display Engine support

The Allwinner A10 and subsequent SoCs share the same display pipeline, with
variations in the number of controllers (1 or 2), or the presence or not of
some output (HDMI, TV, VGA) or not.

Add a driver with a limited set of features for now, and we will hopefully
support all of them eventually

Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
8 years agodrm: sun4i: Add DT bindings documentation
Maxime Ripard [Fri, 30 Oct 2015 13:25:05 +0000 (14:25 +0100)]
drm: sun4i: Add DT bindings documentation

The display pipeline of the Allwinner A10 is involving several loosely
coupled components.

Add a documentation for the bindings.

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
8 years agodrm: fb: Add seq_file definition
Maxime Ripard [Mon, 8 Feb 2016 12:58:56 +0000 (13:58 +0100)]
drm: fb: Add seq_file definition

Otherwise, building with DEBUG_FS enabled will trigger a build warning
because we're using a structure that has not been declared.

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
8 years agodrm: Switch blobs to the new generic modeset obj refcounting
Daniel Vetter [Fri, 22 Apr 2016 20:10:30 +0000 (22:10 +0200)]
drm: Switch blobs to the new generic modeset obj refcounting

Need to move the free function around a bit, but otherwise mostly
just removing code.

Specifically we can nuke all the _locked variants since the weak idr
reference is now protected by the idr_mutex, which we never hold
anywhere expect in the lookup/reg/unreg functions. And those never
call anything else.

Another benefit of this is that this patch switches the weak reference
logic from kref_put_mutex to kref_get_unless_zero. And the later is in
general more flexible wrt accomodating multiple weak references
protected by different locks, which might or might not come handy
eventually.

But one consequence of that switch is that we need to acquire the
blob_lock from the free function for the list_del calls. That's a bit
tricky to pull off, but works well if we pick the exact same scheme as
is already used for framebuffers. Most important changes:

- filp list is maintainer by create/destroy_blob ioctls directly
  (already the case, so we can just remove the redundant list_del from
  the free function).

- filp close handler walks the filp-private list lockless - works
  because we know no one else can access it. I copied the same comment
  from the fb code over to explain this.

- Otherwise we need to sufficiently restrict blob_lock critical
  sections to avoid all the unreference calls. Easy to do once the
  blob_lock only protects the list, and no longer the weak reference.

Cc: Dave Airlie <airlied@gmail.com>
Cc: Daniel Stone <daniels@collabora.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/atomic-helpers: Don't duplicate code in destroy helpers
Daniel Vetter [Fri, 22 Apr 2016 20:10:29 +0000 (22:10 +0200)]
drm/atomic-helpers: Don't duplicate code in destroy helpers

Random drive-by refactoring I spotted.

Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm: Fix fb leaks and WARN spew in get/set_prop ioctls
Daniel Vetter [Fri, 22 Apr 2016 20:10:28 +0000 (22:10 +0200)]
drm: Fix fb leaks and WARN spew in get/set_prop ioctls

Dave Airlie had at least the refcount leak fixed in a later patch (but
that patch does other things which need a bit more work). But we still
have the trouble that silly userspace could hit the WARN_ON in
drm_mode_object_find.

Fix this all up to make sure we don't leak objects, and don't spew
into demsg.

Fixes: d0f37cf62979 ("drm/mode: move framebuffer reference into object.")
Testcase: igt/kms_addfb_basic/invalid-*-prop*
Cc: Dave Airlie <airlied@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm: Improve kerneldoc for new mode object refcounting
Daniel Vetter [Fri, 22 Apr 2016 20:10:27 +0000 (22:10 +0200)]
drm: Improve kerneldoc for new mode object refcounting

Slipped through the cracks in my review. The one issue I spotted
is that drm_mode_object_find now acquires references and can be
used on FB objects, which caused follow-on bugs in get/set_prop ioctls.
Follow-up patches will fix that.

[airlied: fixup some incr fb/decr object mixups]

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agoMerge branch 'topic-arcpgu-v6' of https://github.com/foss-for-synopsys-dwc-arc-proces...
Dave Airlie [Tue, 26 Apr 2016 23:38:42 +0000 (09:38 +1000)]
Merge branch 'topic-arcpgu-v6' of https://github.com/foss-for-synopsys-dwc-arc-processors/linux into drm-next

This is DRM driver for ARC PGU - simple bitstreamer used on
Synopsys ARC SDP boards (both AXS101 and AXS103).

* 'topic-arcpgu-v6' of https://github.com/foss-for-synopsys-dwc-arc-processors/linux:
  arc: axs10x - add support of ARC PGU
  MAINTAINERS: Add maintainer for ARC PGU display controller
  drm: Add DT bindings documentation for ARC PGU display controller
  drm: Add support of ARC PGU display controller

8 years agodrm/sis: add missing include drm.h for the UAPI header
Emil Velikov [Thu, 21 Apr 2016 20:17:32 +0000 (21:17 +0100)]
drm/sis: add missing include drm.h for the UAPI header

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/qxl: remove XXX comment from the UAPI header
Emil Velikov [Thu, 21 Apr 2016 20:17:27 +0000 (21:17 +0100)]
drm/qxl: remove XXX comment from the UAPI header

One cannot rename the struct at this point, so might as well remove the
comment.

Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Dave Airlie <airlied@redhat.com>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agoMerge branch 'drm/next/du' of git://linuxtv.org/pinchartl/fbdev into drm-next
Dave Airlie [Tue, 26 Apr 2016 23:19:50 +0000 (09:19 +1000)]
Merge branch 'drm/next/du' of git://linuxtv.org/pinchartl/fbdev into drm-next

misc rcar changes.

* 'drm/next/du' of git://linuxtv.org/pinchartl/fbdev:
  drm: rcar-du: Fix compilation warning
  drm: rcar-du: Use ARCH_RENESAS
  drm: rcar-du: Clarify vsp dependency

8 years agoarc: axs10x - add support of ARC PGU
Alexey Brodkin [Fri, 19 Feb 2016 12:19:43 +0000 (15:19 +0300)]
arc: axs10x - add support of ARC PGU

Synopsys DesignWare ARC SDP boards sport ARC SDP display
controller attached to ADV7511 HDMI encoder.

That change adds desctiption of both ARC PGU and ADV7511 in
ARC SDP'd base-board Device Tree.

Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Pawel Moll <pawel.moll@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
Cc: Kumar Gala <galak@codeaurora.org>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: devicetree@vger.kernel.org
Cc: linux-snps-arc@lists.infradead.org
8 years agoMAINTAINERS: Add maintainer for ARC PGU display controller
Alexey Brodkin [Fri, 19 Feb 2016 12:34:30 +0000 (15:34 +0300)]
MAINTAINERS: Add maintainer for ARC PGU display controller

This updates MAINTEINERS file with information about maintainer of
ARC PGU display controller driver.

Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: linux-snps-arc@lists.infradead.org
8 years agodrm: Add DT bindings documentation for ARC PGU display controller
Alexey Brodkin [Fri, 19 Feb 2016 12:35:52 +0000 (15:35 +0300)]
drm: Add DT bindings documentation for ARC PGU display controller

This add DT bindings documentation for ARC PGU display controller.

Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Pawel Moll <pawel.moll@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
Cc: Kumar Gala <galak@codeaurora.org>
Cc: devicetree@vger.kernel.org
Cc: linux-snps-arc@lists.infradead.org
Acked-by: Rob Herring <robh@kernel.org>
8 years agodrm: Add support of ARC PGU display controller
Carlos Palminha [Fri, 19 Feb 2016 12:30:26 +0000 (15:30 +0300)]
drm: Add support of ARC PGU display controller

ARC PGU could be found on some development boards from Synopsys.
This is a simple byte streamer that reads data from a framebuffer
and sends data to the single encoder.

Signed-off-by: Carlos Palminha <palminha@synopsys.com>
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: David Airlie <airlied@linux.ie>
Cc: dri-devel@lists.freedesktop.org
Cc: linux-snps-arc@lists.infradead.org
8 years agodrm/fsl-dcu: increment version and date
Stefan Agner [Tue, 26 Apr 2016 03:47:57 +0000 (20:47 -0700)]
drm/fsl-dcu: increment version and date

The driver supports now a second platform and received several
fixes, hence a version increment is justified.

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: implement lastclose callback
Stefan Agner [Sun, 17 Apr 2016 04:18:54 +0000 (21:18 -0700)]
drm/fsl-dcu: implement lastclose callback

Use CMA helper drm_fbdev_cma_restore_mode to restore fbdev mode
in process which uses drm/kms dies.

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: disable output polling on driver unload
Stefan Agner [Sun, 17 Apr 2016 04:17:12 +0000 (21:17 -0700)]
drm/fsl-dcu: disable output polling on driver unload

Disabling output polling before unloading the driver.

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: deallocate fbdev CMA on unload
Stefan Agner [Sun, 17 Apr 2016 05:02:49 +0000 (22:02 -0700)]
drm/fsl-dcu: deallocate fbdev CMA on unload

Free fbdev CMA using drm_fbdev_cma_fini on unload. This fixes
a warning when unloading the driver:
WARNING: CPU: 0 PID: 164 at drivers/gpu/drm/drm_crtc.c:5930 drm_mode_config_cleanup+0x204/0x208

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: use variable name dev for struct drm_device
Stefan Agner [Wed, 13 Apr 2016 07:14:18 +0000 (00:14 -0700)]
drm/fsl-dcu: use variable name dev for struct drm_device

The driver uses different variable names for struct drm_device
across functions which is confusing. Stick to the more common
variable name dev. While at it, remove unnecessary if statement
in error handling.

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: handle missing panel gracefully
Stefan Agner [Thu, 3 Dec 2015 02:03:50 +0000 (18:03 -0800)]
drm/fsl-dcu: handle missing panel gracefully

If the device tree property fsl,panel is missing, drm_panel_attach
is called with a NULL pointer as first argument. Having a panel is
basically mandatory since RGB is the only supported connector.
Check if a panel node has been found, return -ENODEV and cleanup
otherwise.

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: detach panel on destroy
Stefan Agner [Thu, 24 Mar 2016 06:27:11 +0000 (23:27 -0700)]
drm/fsl-dcu: detach panel on destroy

Disable the earlier attached panel on connector destroy.

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/layerscape: reduce excessive stack usage
Arnd Bergmann [Mon, 22 Feb 2016 21:33:31 +0000 (22:33 +0100)]
drm/layerscape: reduce excessive stack usage

The fsl-dcu driver copies a drm_mode_config object to its
stack but then only accesses a single member (dpms_property)
once. The data structure is large enough to trigger a warning
about the amount of kernel stack being used:

drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c: In function 'fsl_dcu_drm_connector_create':
drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c:182:1: error: the frame size of 1040 bytes is larger than 1024 bytes [-Werror=frame-larger-than=]

This changes the fsl_dcu_drm_connector_create() function to
only access the drm_mode_config by reference, which is also
more efficient.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 109eee2f2a18 ("drm/layerscape: Add Freescale DCU DRM driver")
Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: add TCON driver
Stefan Agner [Wed, 2 Dec 2015 22:39:40 +0000 (14:39 -0800)]
drm/fsl-dcu: add TCON driver

Add driver for the TCON (timing controller) module. The TCON module
is a separate module attached after the DCU (display controller
unit). Each DCU instance has its own, directly connected TCON
instance. The DCU's RGB and timing signals are passing through
the TCON module. TCON can provide timing signals for raw TFT panels
or operate in a bypass mode which leaves all signals unaltered.

The driver currently only supports the bypass mode.

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: use common clock framework for pixel clock divider
Stefan Agner [Wed, 23 Mar 2016 01:06:08 +0000 (18:06 -0700)]
drm/fsl-dcu: use common clock framework for pixel clock divider

Use the common clock framework to calculate the pixel clock
dividier. The previous implementation rounded down the calculated
factor. Thanks to the CLK_DIVIDER_ROUND_CLOSEST flag using the
common clock framework divider implementation improves the pixel
clock accuracy in some cases. Ontop of that it also allows to see
the actual pixel clock in the sysfs clock summary.

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: add extra clock for pixel clock
Stefan Agner [Tue, 22 Mar 2016 22:45:29 +0000 (15:45 -0700)]
drm/fsl-dcu: add extra clock for pixel clock

The Vybrid DCU variant has two independent clock inputs, one
for the registers (IPG bus clock) and one for the pixel clock.
Support this distinction in the DCU DRM driver while staying
backward compatible for old device trees.

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm/fsl-dcu: disable clock on initialization failure and remove
Stefan Agner [Tue, 22 Mar 2016 22:13:05 +0000 (15:13 -0700)]
drm/fsl-dcu: disable clock on initialization failure and remove

Fix error handling during probe by reordering initialization and
adding a error path which disables clock again. Also disable the
clock on remove.

Signed-off-by: Stefan Agner <stefan@agner.ch>
8 years agodrm: rcar-du: Fix compilation warning
Laurent Pinchart [Fri, 22 Apr 2016 14:52:45 +0000 (17:52 +0300)]
drm: rcar-du: Fix compilation warning

Commit d63c25e4245a ("drm: rcar-du: Use generic
drm_connector_register_all() helper") left an unused local variable
behind. Remove it.

Fixes: d63c25e4245a ("drm: rcar-du: Use generic drm_connector_register_all() helper")
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
8 years agodrm: rcar-du: Use ARCH_RENESAS
Simon Horman [Fri, 25 Mar 2016 01:38:28 +0000 (10:38 +0900)]
drm: rcar-du: Use ARCH_RENESAS

Make use of ARCH_RENESAS in place of ARCH_SHMOBILE.

This is part of an ongoing process to migrate from ARCH_SHMOBILE to
ARCH_RENESAS the motivation for which being that RENESAS seems to be a more
appropriate name than SHMOBILE for the majority of Renesas ARM based SoCs.

Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
8 years agodrm: rcar-du: Clarify vsp dependency
Arnd Bergmann [Fri, 26 Feb 2016 11:52:20 +0000 (12:52 +0100)]
drm: rcar-du: Clarify vsp dependency

The VSP1 compositor code in DRM links against the respective V4L
driver, but the dependency is not expressed correctly in Kconfig,
which leads to a build error when the DRM driver is built-in
and the V4L driver is a module:

drivers/gpu/built-in.o: In function `rcar_du_vsp_plane_atomic_update':
rcar-du/rcar_du_vsp.c:183: undefined reference to `vsp1_du_atomic_update'

This patch avoids the problem by ensuring that the DRM VSP code can
only be enabled if the V4L driver is linked into the kernel, or
both are loadable modules.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 6d62ef3ac30b ("drm: rcar-du: Expose the VSP1 compositor through KMS planes")
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
8 years agodrm/modes: stop handling framebuffer special
Dave Airlie [Fri, 15 Apr 2016 05:10:42 +0000 (15:10 +1000)]
drm/modes: stop handling framebuffer special

Since ref counting is in the object now we can just call the
normal interfaces.

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/modes: reduce fb_lock to just protecting lists
Dave Airlie [Fri, 15 Apr 2016 05:10:41 +0000 (15:10 +1000)]
drm/modes: reduce fb_lock to just protecting lists

This reduces the fb_lock to just protecting the num_fb/fb_list.

"Previously fb refcounting, and especially the weak reference
(kref_get_unless_zero) used in fb lookups have been protected by fb_lock.
But with the refactoring to share refcounting in the drm_mode_object base
class that switched to being protected by idr_mutex, which means fb_lock
critical sections can be reduced."

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/modes: move reference taking into object lookup.
Dave Airlie [Fri, 15 Apr 2016 05:10:40 +0000 (15:10 +1000)]
drm/modes: move reference taking into object lookup.

When we lookup an ref counted object we now take a proper reference
using kref_get_unless_zero.

Framebuffer lookup no longer needs do this itself.

Convert rmfb to using framebuffer lookup and deal with the fact
it now gets an extra reference that we have to cleanup. This should
mean we can avoid holding fb_lock across rmfb. (if I'm wrong let me
know).

We also now only hold the fbs_lock around the list manipulation.

"Previously fb refcounting, and especially the weak reference
(kref_get_unless_zero) used in fb lookups have been protected by fb_lock.
But with the refactoring to share refcounting in the drm_mode_object base
class that switched to being protected by idr_mutex, which means fb_lock
critical sections can be reduced."

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/mode: reduce lock hold in addfb2
Dave Airlie [Fri, 15 Apr 2016 05:10:39 +0000 (15:10 +1000)]
drm/mode: reduce lock hold in addfb2

No need to hold the lock while assigning the variable.

Daniel wrote:
"Not sure why exactly I put that under the lock, but the only thing that
can race here is rmfb while addfb2 is still doing it's thing, with a
correctly guess (easy to do since they're fully deterministic) fb_id."

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/mode: reduce scope of fb_lock in framebuffer init
Dave Airlie [Fri, 15 Apr 2016 05:10:38 +0000 (15:10 +1000)]
drm/mode: reduce scope of fb_lock in framebuffer init

We don't need to hold the fb lock around the initialisation,
only around the list manipulaton.

So do the lock hold only around the register for now.

From Daniel:
Previously fb refcounting, and especially the weak reference
(kref_get_unless_zero) used in fb lookups have been protected by fb_lock.
But with the refactoring to share refcounting in the drm_mode_object base
class that switched to being protected by idr_mutex, which means fb_lock
critical sections can be reduced.

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/mode: use _object_find to find framebuffers.
Dave Airlie [Fri, 15 Apr 2016 05:10:37 +0000 (15:10 +1000)]
drm/mode: use _object_find to find framebuffers.

No point have this code dupliated at this point, use the
_object_find code instead now.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
8 years agodrm/mode: move framebuffer reference into object.
Dave Airlie [Fri, 15 Apr 2016 05:10:36 +0000 (15:10 +1000)]
drm/mode: move framebuffer reference into object.

This is the initial code to add references to some mode objects.
In the future we need to start reference counting connectors so
firstly I want to reorganise the code so the framebuffer ref counting
uses the same paths.

This patch shouldn't change any functionality, just moves the kref.

[airlied: move kerneldoc as well]
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/mode: introduce wrapper to read framebuffer refcount.
Dave Airlie [Fri, 15 Apr 2016 05:10:35 +0000 (15:10 +1000)]
drm/mode: introduce wrapper to read framebuffer refcount.

Avoids drivers knowing where the kref is stored.

[airlied: add kerneldoc]
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
8 years agodrm/modes: drop __drm_framebuffer_unregister.
Dave Airlie [Fri, 15 Apr 2016 05:10:34 +0000 (15:10 +1000)]
drm/modes: drop __drm_framebuffer_unregister.

Just use the generic function.

The main side effect of this is that the fb->base.id
is now protected by the idr mutex as well.

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>