DC Patches July 30, 2025
Subsystems (5)
cover DC Patches July 30, 2025
From: Roman Li <[email protected]>
This DC patchset brings improvements in multiple areas. In summary, we highlight:
* Add interface to log hw state when underflow happens
* Fix hubp programming of 3dlut fast load
* Avoid Read Remote DPCD Many Times
* More liberal vmin/vmax update for freesync
* Fix dmub access race condition
Aurabindo Pillai (2):
drm/amd/display: fix dmub access race condition drm/amd/display: more liberal vmin/vmax update for freesync
Duncan Ma (1):
drm/amd/display: Adjust AUX-less ALPM setting
Fangzhi Zuo (1):
drm/amd/display: Avoid Read Remote DPCD Many Times
Jingwen Zhu (1):
drm/amd/display: limited pll vco w/a v2
Mario Limonciello (1):
drm/amd/display: Revert "drm/amd/display: Fix AMDGPU_MAX_BL_LEVEL
value"
Muhammad Ahmed (1):
drm/amd/display: Adding interface to log hw state when underflow
happens
Paul Hsieh (1):
drm/amd/display: update dpp/disp clock from smu clock table
Reza Amini (1):
drm/amd/display: Fixing hubp programming of 3dlut fast load
Ryan Seto (1):
drm/amd/display: Toggle for Disable Force Pstate Allow on Disable
Siyang Liu (1):
drm/amd/display: fix a Null pointer dereference vulnerability
Taimur Hassan (1):
drm/amd/display: Promote DC to 3.2.344
.../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 81 ++++++++++++++++--- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 14 ++++ .../display/amdgpu_dm/amdgpu_dm_mst_types.c | 9 ++- .../display/dc/clk_mgr/dcn301/vg_clk_mgr.c | 16 ++++ drivers/gpu/drm/amd/display/dc/core/dc.c | 37 ++++++--- drivers/gpu/drm/amd/display/dc/dc.h | 59 +++++++++++++- .../amd/display/dc/dml/dcn301/dcn301_fpu.c | 20 ++++- .../display/dc/hubbub/dcn30/dcn30_hubbub.c | 31 +++++++ .../display/dc/hubbub/dcn30/dcn30_hubbub.h | 6 ++ .../display/dc/hubbub/dcn31/dcn31_hubbub.c | 2 + .../display/dc/hubbub/dcn32/dcn32_hubbub.c | 2 + .../display/dc/hubbub/dcn35/dcn35_hubbub.c | 2 + .../display/dc/hubbub/dcn401/dcn401_hubbub.c | 2 + .../amd/display/dc/hubp/dcn10/dcn10_hubp.h | 9 ++- .../amd/display/dc/hubp/dcn20/dcn20_hubp.h | 1 + .../amd/display/dc/hubp/dcn30/dcn30_hubp.c | 26 ++++++ .../amd/display/dc/hubp/dcn30/dcn30_hubp.h | 8 +- .../amd/display/dc/hubp/dcn31/dcn31_hubp.c | 15 ++++ .../amd/display/dc/hubp/dcn31/dcn31_hubp.h | 6 +- .../amd/display/dc/hubp/dcn32/dcn32_hubp.c | 3 + .../amd/display/dc/hubp/dcn35/dcn35_hubp.c | 3 + .../amd/display/dc/hubp/dcn401/dcn401_hubp.c | 41 ++++++++++ .../amd/display/dc/hubp/dcn401/dcn401_hubp.h | 8 +- .../amd/display/dc/hwss/dcn10/dcn10_hwseq.c | 63 +++++++++++---- .../amd/display/dc/hwss/dcn30/dcn30_hwseq.c | 48 +++++++++++ .../amd/display/dc/hwss/dcn30/dcn30_hwseq.h | 5 ++ .../amd/display/dc/hwss/dcn30/dcn30_init.c | 1 + .../amd/display/dc/hwss/dcn31/dcn31_init.c | 1 + .../amd/display/dc/hwss/dcn314/dcn314_init.c | 1 + .../amd/display/dc/hwss/dcn32/dcn32_init.c | 1 + .../amd/display/dc/hwss/dcn35/dcn35_init.c | 1 + .../amd/display/dc/hwss/dcn351/dcn351_init.c | 1 + .../amd/display/dc/hwss/dcn401/dcn401_hwseq.c | 36 +++++---- .../amd/display/dc/hwss/dcn401/dcn401_init.c | 1 + .../drm/amd/display/dc/hwss/hw_sequencer.h | 4 + .../gpu/drm/amd/display/dc/inc/hw/dchubbub.h | 2 + drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h | 21 ++++- drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h | 11 +++ .../link/protocols/link_edp_panel_control.c | 2 +- .../dc/resource/dcn32/dcn32_resource.h | 3 +- .../dc/resource/dcn401/dcn401_resource.c | 1 + .../dc/resource/dcn401/dcn401_resource.h | 3 +- drivers/gpu/drm/amd/display/dmub/dmub_srv.h | 1 + .../gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 3 +- .../gpu/drm/amd/display/dmub/src/dmub_dcn31.c | 1 + 45 files changed, 545 insertions(+), 67 deletions(-)
Signature · 1 line
2.34.1
PATCH 1/12 drm/amd/display: fix a Null pointer dereference vulnerability
From: Siyang Liu <[email protected]>
[Why]
A null pointer dereference vulnerability exists in the AMD display driver's (DC module) cleanup function dc_destruct().
When display control context (dc->ctx) construction fails
(due to memory allocation failure), this pointer remains NULL. During subsequent error handling when dc_destruct() is called, there's no NULL check before dereferencing the perf_trace member (dc->ctx->perf_trace), causing a kernel null pointer dereference crash.
[How]
Check if dc->ctx is non-NULL before dereferencing.
(Updated commit text and removed unnecessary error message)
--- drivers/gpu/drm/amd/display/dc/core/dc.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-)
drivers/gpu/drm/amd/display/dc/core/dc.c +10 -9
@@ -940,17 +940,18 @@ static void dc_destruct(struct dc *dc)if (dc->link_srv)link_destroy_link_service (&dc->link_srv); -if (dc->ctx->gpio_service) -dal_gpio_service_destroy (&dc->ctx->gpio_service); +if (dc->ctx) { +if (dc->ctx->gpio_service) +dal_gpio_service_destroy (&dc->ctx->gpio_service); -if (dc->ctx->created_bios) -dal_bios_parser_destroy (&dc->ctx->dc_bios); +if (dc->ctx->created_bios) +dal_bios_parser_destroy (&dc->ctx->dc_bios); +kfree (dc->ctx->logger); +dc_perf_trace_destroy (&dc->ctx->perf_trace); -kfree (dc->ctx->logger); -dc_perf_trace_destroy (&dc->ctx->perf_trace); - -kfree (dc->ctx); - dc->ctx =NULL ; +kfree (dc->ctx); + dc->ctx =NULL ; + }kfree (dc->bw_vbios); dc->bw_vbios =NULL ;
Signature · 1 line
2.34.1
PATCH 2/12 drm/amd/display: Adjust AUX-less ALPM setting
In reply to [email protected]From: Duncan Ma <[email protected]>
[Why & How]
Change ACDS period to support LTTPR.
--- .../drm/amd/display/dc/link/protocols/link_edp_panel_control.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c +1 -1
@@ -1046,7 +1046,7 @@ bool edp_setup_replay(struct dc_link *link, const struct dc_stream_state *streamif (link->replay_settings.config.alpm_mode ==DC_ALPM_AUXLESS ) { alpm_config.bits.ALPM_MODE_SEL =1 ; - alpm_config.bits.ACDS_PERIOD_DURATION =1 ; + alpm_config.bits.ACDS_PERIOD_DURATION =0 ; } dm_helpers_dp_write_dpcd(
Signature · 1 line
2.34.1
PATCH 3/12 drm/amd/display: fix dmub access race condition
In reply to [email protected]From: Aurabindo Pillai <[email protected]>
Accessing DC from amdgpu_dm is usually preceded by acquisition of dc_lock mutex. Most of the DC API that DM calls are under a DC lock. However, there are a few that are not. Some DC API called from interrupt context end up sending DMUB commands via a DC API, while other threads were using DMUB. This was apparent from a race between calls for setting idle optimization enable/disable and the DC API to set vmin/vmax.
Offload the call to dc_stream_adjust_vmin_vmax() to a thread instead of directly calling them from the interrupt handler such that it waits for dc_lock.
--- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 55 +++++++++++++++++-- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 14 +++++ 2 files changed, 63 insertions(+), 6 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +49 -6
@@ -530,6 +530,50 @@ static void dm_pflip_high_irq(void *interrupt_params) amdgpu_crtc->crtc_id, amdgpu_crtc, vrr_active, (int)!e); } +static void dm_handle_vmin_vmax_update (struct work_struct *offload_work) +{ + struct vupdate_offload_work *work = container_of(offload_work, struct vupdate_offload_work, work); +struct amdgpu_device *adev = work->adev; +struct dc_stream_state *stream = work->stream; +struct dc_crtc_timing_adjust *adjust = work->adjust; + +mutex_lock (&adev->dm.dc_lock); +dc_stream_adjust_vmin_vmax (adev->dm.dc, stream, adjust); +mutex_unlock (&adev->dm.dc_lock); + +dc_stream_release (stream); +kfree (work->adjust); +kfree (work); +} + +static void schedule_dc_vmin_vmax (struct amdgpu_device *adev, +struct dc_stream_state *stream, +struct dc_crtc_timing_adjust *adjust) +{ +struct vupdate_offload_work *offload_work =kzalloc (sizeof (*offload_work),GFP_KERNEL ); +if (!offload_work) { +drm_dbg_driver (adev_to_drm (adev),"Failed to allocate vupdate_offload_work\n" ); +return ; + } + +struct dc_crtc_timing_adjust *adjust_copy =kzalloc (sizeof (*adjust_copy),GFP_KERNEL ); +if (!adjust_copy) { +drm_dbg_driver (adev_to_drm (adev),"Failed to allocate adjust_copy\n" ); +kfree (offload_work); +return ; + } + +dc_stream_retain (stream); +memcpy (adjust_copy, adjust,sizeof (*adjust_copy)); + +INIT_WORK (&offload_work->work, dm_handle_vmin_vmax_update); + offload_work->adev = adev; + offload_work->stream = stream; + offload_work->adjust = adjust_copy; + +queue_work (system_wq, &offload_work->work); +} +static void dm_vupdate_high_irq (void *interrupt_params) {struct common_irq_params *irq_params = interrupt_params;
@@ -579,10 +623,9 @@ static void dm_vupdate_high_irq(void *interrupt_params) acrtc->dm_irq_params.stream, &acrtc->dm_irq_params.vrr_params); -dc_stream_adjust_vmin_vmax ( - adev->dm.dc, - acrtc->dm_irq_params.stream, - &acrtc->dm_irq_params.vrr_params.adjust); +schedule_dc_vmin_vmax (adev, + acrtc->dm_irq_params.stream, + &acrtc->dm_irq_params.vrr_params.adjust);spin_unlock_irqrestore (&adev_to_drm (adev)->event_lock, flags); } }
@@ -672,8 +715,8 @@ static void dm_crtc_high_irq(void *interrupt_params) acrtc->dm_irq_params.stream, &acrtc->dm_irq_params.vrr_params); -dc_stream_adjust_vmin_vmax (adev->dm.dc, acrtc->dm_irq_params.stream, - &acrtc->dm_irq_params.vrr_params.adjust); +schedule_dc_vmin_vmax (adev, acrtc->dm_irq_params.stream, + &acrtc->dm_irq_params.vrr_params.adjust); }/*
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +14 -0
@@ -153,6 +153,20 @@ struct idle_workqueue {bool running; }; +/** +* struct dm_vupdate_work - Work data for periodic action in idle +* @work: Kernel work data for the work event +* @adev: amdgpu_device back pointer +* @stream: DC stream associated with the crtc +* @adjust: DC CRTC timing adjust to be applied to the crtc +*/ +struct vupdate_offload_work { +struct work_struct work; +struct amdgpu_device *adev; +struct dc_stream_state *stream; +struct dc_crtc_timing_adjust *adjust; +}; +#define MAX_LUMINANCE_DATA_POINTS 99/**
Signature · 1 line
2.34.1
Re: [PATCH 03/12] drm/amd/display: fix dmub access race condition
Hi,
Quoted 9 lines · On 2025-07-30 14:58:54 -0400, [email protected] wrote:
From: Aurabindo Pillai <[email protected]>
Accessing DC from
amdgpu_dmis usually preceded by acquisition ofdc_lockmutex. Most of the DC API that DM calls are under a DC lock. However, there are a few that are not. Some DC API called from interrupt context end up sending DMUB commands via a DC API, while other threads were using DMUB. This was apparent from a race between calls for setting idle optimization enable/disable and the DC API to set vmin/vmax.
Quoted 7 lines
Offload the call to
dc_stream_adjust_vmin_vmax()to a thread instead of directly calling them from the interrupt handler such that it waits fordc_lock.Reviewed-by Nicholas Kazlauskas <[email protected]>Signed-off-by Aurabindo Pillai <[email protected]>Signed-off-by Roman Li <[email protected]>
With this patch I get sleeping function called from invalid context roughly every second. This occurs on at least PREEMPT_LAZY and PREEMPT_VOLUNTARY.
BUG: sleeping function called from invalid context at include/linux/sched/mm.h:321 in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 13841, name: cc1 preempt_count: 10002, expected: 0 CPU: 23 UID: 1000 PID: 13841 Comm: cc1 Kdump: loaded Tainted: G W 6.16.0-next-20250805 #631 PREEMPTLAZY Tainted: [W]=WARN Hardware name: Micro-Star International Co., Ltd. MS-7C91/MAG B550 TOMAHAWK (MS-7C91), BIOS A.I0 03/20/2025 Call Trace: <TASK> dump_stack_lvl+0x4b/0x70 __might_resched.cold+0xac/0xb9 __kmalloc_cache_noprof+0x2d4/0x430 ? schedule_dc_vmin_vmax+0x49/0x1c0 [amdgpu] ? __alloc_frozen_pages_noprof+0x18f/0x360 ? _raw_spin_unlock_irqrestore+0x12/0x40 ? schedule_dc_vmin_vmax+0x49/0x1c0 [amdgpu] schedule_dc_vmin_vmax+0x49/0x1c0 [amdgpu] dm_crtc_high_irq+0x2ab/0x310 [amdgpu] amdgpu_dm_irq_handler+0x8d/0x210 [amdgpu] amdgpu_irq_dispatch+0x166/0x1a0 [amdgpu] amdgpu_ih_process+0x60/0x160 [amdgpu] amdgpu_irq_handler+0x23/0x60 [amdgpu] __handle_irq_event_percpu+0x4a/0x1a0 handle_irq_event+0x38/0x90 handle_edge_irq+0xc4/0x190 __common_interrupt+0x44/0xe0 ? srso_return_thunk+0x5/0x5f common_interrupt+0x3a/0xa0 asm_common_interrupt+0x26/0x40 RIP: 0033:0x55d4848fab81 Code: 28 44 8b 5f 70 ff 87 80 00 00 00 49 89 d7 48 89 34 24 45 8d 73 ff 44 89 f3 21 cb 48 8b 4f 58 41 89 dd 49 c1 e5 03 4a 8b 2c 29 <48> 85 ed 0f 84 36 02 00 00 48 83 fd ff 0f 84 bc 00 00 00 44 39 4d RSP: 002b:00007fff71eeed50 EFLAGS: 00000212 RAX: 000000000000001f RBX: 0000000000002182 RCX: 00007f47b3f00010 RDX: 000000000000001f RSI: 000055d4bfce1e3c RDI: 000055d4bfbb2e80 RBP: 0000000000000000 R08: 0000000000000001 R09: 000000006d35e182 R10: 000055d4bfb94630 R11: 0000000000004000 R12: 000055d4bfbb2e80 R13: 0000000000010c10 R14: 0000000000003fff R15: 000000000000001f </TASK>
...
Quoted 18 lines
(quoted diff) +18
-0+static void schedule_dc_vmin_vmax(struct amdgpu_device *adev, + struct dc_stream_state *stream, + struct dc_crtc_timing_adjust *adjust) +{ + struct vupdate_offload_work *offload_work = kzalloc(sizeof(*offload_work), GFP_KERNEL); + if (!offload_work) { + drm_dbg_driver(adev_to_drm(adev), "Failed to allocate vupdate_offload_work\n"); + return; + } + + struct dc_crtc_timing_adjust *adjust_copy = kzalloc(sizeof(*adjust_copy), GFP_KERNEL); + if (!adjust_copy) { + drm_dbg_driver(adev_to_drm(adev), "Failed to allocate adjust_copy\n"); + kfree(offload_work); + return; + } + + dc_stream_retain(stream);
Review
Quoted 10 lines
(quoted diff) +10
-0+ memcpy(adjust_copy, adjust, sizeof(*adjust_copy)); + + INIT_WORK(&offload_work->work, dm_handle_vmin_vmax_update); + offload_work->adev = adev; + offload_work->stream = stream; + offload_work->adjust = adjust_copy; + + queue_work(system_wq, &offload_work->work); +} +
The allocations in this function seems to be the culprit. If I change them to GFP_ATOMIC the issue disappears:
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +2 -2
@@ -550,13 +550,13 @@ static void schedule_dc_vmin_vmax(struct amdgpu_device *adev,struct dc_stream_state *stream, struct dc_crtc_timing_adjust *adjust) { - struct vupdate_offload_work *offload_work = kzalloc(sizeof(*offload_work), GFP_KERNEL); + struct vupdate_offload_work *offload_work = kzalloc(sizeof(*offload_work), GFP_ATOMIC);if (!offload_work) {drm_dbg_driver (adev_to_drm (adev),"Failed to allocate vupdate_offload_work\n" );return ; } -struct dc_crtc_timing_adjust *adjust_copy =kzalloc (sizeof (*adjust_copy),GFP_KERNEL ); +struct dc_crtc_timing_adjust *adjust_copy =kzalloc (sizeof (*adjust_copy),GFP_ATOMIC );if (!adjust_copy) {drm_dbg_driver (adev_to_drm (adev),"Failed to allocate adjust_copy\n" );kfree (offload_work);
...
Regards,
Klara Modin
Re: [PATCH 03/12] drm/amd/display: fix dmub access race condition
Hi,
Quoted 24 lines · Klara Modin
Hi,
Quoted 9 lines · On 2025-07-30 14:58:54 -0400, [email protected] wrote:
From: Aurabindo Pillai <[email protected]>
Accessing DC from
amdgpu_dmis usually preceded by acquisition ofdc_lockmutex. Most of the DC API that DM calls are under a DC lock. However, there are a few that are not. Some DC API called from interrupt context end up sending DMUB commands via a DC API, while other threads were using DMUB. This was apparent from a race between calls for setting idle optimization enable/disable and the DC API to set vmin/vmax.Quoted 7 lines
Offload the call to
dc_stream_adjust_vmin_vmax()to a thread instead of directly calling them from the interrupt handler such that it waits fordc_lock.Reviewed-by Nicholas Kazlauskas <[email protected]>Signed-off-by Aurabindo Pillai <[email protected]>Signed-off-by Roman Li <[email protected]>With this patch I get sleeping function called from invalid context roughly every second. This occurs on at least
PREEMPT_LAZYandPREEMPT_VOLUNTARY.
...
Quoted 45 lines
...
Quoted 18 lines
(quoted diff) +18
-0+static void schedule_dc_vmin_vmax(struct amdgpu_device *adev, + struct dc_stream_state *stream, + struct dc_crtc_timing_adjust *adjust) +{ + struct vupdate_offload_work *offload_work = kzalloc(sizeof(*offload_work), GFP_KERNEL); + if (!offload_work) { + drm_dbg_driver(adev_to_drm(adev), "Failed to allocate vupdate_offload_work\n"); + return; + } + + struct dc_crtc_timing_adjust *adjust_copy = kzalloc(sizeof(*adjust_copy), GFP_KERNEL); + if (!adjust_copy) { + drm_dbg_driver(adev_to_drm(adev), "Failed to allocate adjust_copy\n"); + kfree(offload_work); + return; + } + + dc_stream_retain(stream);Review
Quoted 10 lines
(quoted diff) +10
-0+ memcpy(adjust_copy, adjust, sizeof(*adjust_copy)); + + INIT_WORK(&offload_work->work, dm_handle_vmin_vmax_update); + offload_work->adev = adev; + offload_work->stream = stream; + offload_work->adjust = adjust_copy; + + queue_work(system_wq, &offload_work->work); +} +The allocations in this function seems to be the culprit. If I change them to
GFP_ATOMICthe issue disappears:drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +1
-1@@ -550,13 +550,13 @@ static void schedule_dc_vmin_vmax(struct amdgpu_device *adev,struct dc_stream_state *stream, struct dc_crtc_timing_adjust *adjust) { - struct vupdate_offload_work *offload_work = kzalloc(sizeof(*offload_work), GFP_KERNEL); + struct vupdate_offload_work *offload_work = kzalloc(sizeof(*offload_work), GFP_ATOMIC);
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c static void schedule_dc_vmin_vmax(struct amdgpu_device *adev, · line 562
Quoted 10 lines
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +1
-1if (!offload_work) {drm_dbg_driver (adev_to_drm (adev),"Failed to allocate vupdate_offload_work\n" );return ; } -struct dc_crtc_timing_adjust *adjust_copy =kzalloc (sizeof (*adjust_copy),GFP_KERNEL ); +struct dc_crtc_timing_adjust *adjust_copy =kzalloc (sizeof (*adjust_copy),GFP_ATOMIC );if (!adjust_copy) {drm_dbg_driver (adev_to_drm (adev),"Failed to allocate adjust_copy\n" );kfree (offload_work);
Any thoughts? This is still an issue in the current next. I have also tried using GFP_NOWAIT instead and not seen any obvious issues under memory pressure.
Regards,
Klara Modin
PATCH 4/12 drm/amd/display: more liberal vmin/vmax update for freesync
In reply to [email protected]From: Aurabindo Pillai <[email protected]>
[Why]
FAMS2 expects vmin/vmax to be updated in the case when freesync is off, but supported. But we only update it when freesync is enabled.
[How]
Change the vsync handler such that dc_stream_adjust_vmin_vmax() its called irrespective of whether freesync is enabled. If freesync is supported, then there is no harm in updating vmin/vmax registers.
--- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 28 +++++++++++++------ 1 file changed, 20 insertions(+), 8 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +20 -8
@@ -612,6 +612,11 @@ static void dm_vupdate_high_irq(void *interrupt_params)* if a pageflip happened inside front-porch. */ if (vrr_active) { +bool replay_en = acrtc->dm_irq_params.stream->link->replay_settings.replay_feature_enabled; +bool psr_en = acrtc->dm_irq_params.stream->link->psr_settings.psr_feature_enabled; +bool fs_active_var_en = acrtc->dm_irq_params.freesync_config.state + ==VRR_STATE_ACTIVE_VARIABLE ; +amdgpu_dm_crtc_handle_vblank (acrtc);/* BTR processing for pre-DCE12 ASICs */
@@ -623,9 +628,11 @@ static void dm_vupdate_high_irq(void *interrupt_params) acrtc->dm_irq_params.stream, &acrtc->dm_irq_params.vrr_params); -schedule_dc_vmin_vmax (adev, - acrtc->dm_irq_params.stream, - &acrtc->dm_irq_params.vrr_params.adjust); +if (fs_active_var_en || (!fs_active_var_en && !replay_en && !psr_en)) { +schedule_dc_vmin_vmax (adev, + acrtc->dm_irq_params.stream, + &acrtc->dm_irq_params.vrr_params.adjust); + }spin_unlock_irqrestore (&adev_to_drm (adev)->event_lock, flags); } }
@@ -708,15 +715,20 @@ static void dm_crtc_high_irq(void *interrupt_params)spin_lock_irqsave (&adev_to_drm (adev)->event_lock, flags);if (acrtc->dm_irq_params.stream && - acrtc->dm_irq_params.vrr_params.supported && - acrtc->dm_irq_params.freesync_config.state == -VRR_STATE_ACTIVE_VARIABLE ) { + acrtc->dm_irq_params.vrr_params.supported) { +bool replay_en = acrtc->dm_irq_params.stream->link->replay_settings.replay_feature_enabled; +bool psr_en = acrtc->dm_irq_params.stream->link->psr_settings.psr_feature_enabled; +bool fs_active_var_en = acrtc->dm_irq_params.freesync_config.state ==VRR_STATE_ACTIVE_VARIABLE ; +mod_freesync_handle_v_update (adev->dm.freesync_module, acrtc->dm_irq_params.stream, &acrtc->dm_irq_params.vrr_params); -schedule_dc_vmin_vmax (adev, acrtc->dm_irq_params.stream, - &acrtc->dm_irq_params.vrr_params.adjust); +/* update vmin_vmax only if freesync is enabled, or only if PSR and REPLAY are disabled */ +if (fs_active_var_en || (!fs_active_var_en && !replay_en && !psr_en)) { +schedule_dc_vmin_vmax (adev, acrtc->dm_irq_params.stream, + &acrtc->dm_irq_params.vrr_params.adjust); + } }/*
Signature · 1 line
2.34.1
PATCH 5/12 drm/amd/display: update dpp/disp clock from smu clock table
In reply to [email protected]From: Paul Hsieh <[email protected]>
[Why]
The reason some high-resolution monitors fail to display properly is that this platform does not support sufficiently high DPP and DISP clock frequencies
[How]
Update DISP and DPP clocks from the smu clock table then DML can filter these mode if not support.
--- .../display/dc/clk_mgr/dcn301/vg_clk_mgr.c | 16 +++++++++++++++ .../amd/display/dc/dml/dcn301/dcn301_fpu.c | 20 ++++++++++++++++--- 2 files changed, 33 insertions(+), 3 deletions(-)
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c +16 -0
@@ -563,6 +563,7 @@ static void vg_clk_mgr_helper_populate_bw_params( {int i, j;struct clk_bw_params *bw_params = clk_mgr->base.bw_params; +uint32_t max_dispclk =0 , max_dppclk =0 ; j =-1 ;
@@ -584,6 +585,15 @@ static void vg_clk_mgr_helper_populate_bw_params(return ; } +/* dispclk and dppclk can be max at any voltage, same number of levels for both */ +if (clock_table->NumDispClkLevelsEnabled <=VG_NUM_DISPCLK_DPM_LEVELS && + clock_table->NumDispClkLevelsEnabled <=VG_NUM_DPPCLK_DPM_LEVELS ) { + max_dispclk =find_max_clk_value (clock_table->DispClocks, clock_table->NumDispClkLevelsEnabled); + max_dppclk =find_max_clk_value (clock_table->DppClocks, clock_table->NumDispClkLevelsEnabled); + }else { +ASSERT (0 ); + } + bw_params->clk_table.num_entries = j +1 ;for (i =0 ; i < bw_params->clk_table.num_entries -1 ; i++, j--) {
@@ -591,11 +601,17 @@ static void vg_clk_mgr_helper_populate_bw_params( bw_params->clk_table.entries[i].memclk_mhz = clock_table->DfPstateTable[j].memclk; bw_params->clk_table.entries[i].voltage = clock_table->DfPstateTable[j].voltage; bw_params->clk_table.entries[i].dcfclk_mhz =find_dcfclk_for_voltage (clock_table, clock_table->DfPstateTable[j].voltage); + +/* Now update clocks we do read */ + bw_params->clk_table.entries[i].dispclk_mhz = max_dispclk; + bw_params->clk_table.entries[i].dppclk_mhz = max_dppclk; } bw_params->clk_table.entries[i].fclk_mhz = clock_table->DfPstateTable[j].fclk; bw_params->clk_table.entries[i].memclk_mhz = clock_table->DfPstateTable[j].memclk; bw_params->clk_table.entries[i].voltage = clock_table->DfPstateTable[j].voltage; bw_params->clk_table.entries[i].dcfclk_mhz =find_max_clk_value (clock_table->DcfClocks,VG_NUM_DCFCLK_DPM_LEVELS ); + bw_params->clk_table.entries[i].dispclk_mhz =find_max_clk_value (clock_table->DispClocks,VG_NUM_DISPCLK_DPM_LEVELS ); + bw_params->clk_table.entries[i].dppclk_mhz =find_max_clk_value (clock_table->DppClocks,VG_NUM_DPPCLK_DPM_LEVELS ); bw_params->vram_type = bios_info->memory_type; bw_params->num_channels = bios_info->ma_channel_number;
drivers/gpu/drm/amd/display/dc/dml/dcn301/dcn301_fpu.c +17 -3
@@ -326,7 +326,7 @@ void dcn301_fpu_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_pstruct dcn301_resource_pool *pool =TO_DCN301_RES_POOL (dc->res_pool);struct clk_limit_table *clk_table = &bw_params->clk_table;unsigned int i, closest_clk_lvl; -int j; +int j =0 , max_dispclk_mhz =0 , max_dppclk_mhz =0 ;dc_assert_fp_enabled ();
@@ -338,6 +338,15 @@ void dcn301_fpu_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_p dcn3_01_soc.num_chans = bw_params->num_channels;ASSERT (clk_table->num_entries); + +/* Prepass to find max clocks independent of voltage level. */ +for (i =0 ; i < clk_table->num_entries; ++i) { +if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) + max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; +if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) + max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; + } +for (i =0 ; i < clk_table->num_entries; i++) {/* loop backwards*/ for (closest_clk_lvl =0 , j = dcn3_01_soc.num_states -1 ; j >=0 ; j--) {
@@ -353,8 +362,13 @@ void dcn301_fpu_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_p s[i].socclk_mhz = clk_table->entries[i].socclk_mhz; s[i].dram_speed_mts = clk_table->entries[i].memclk_mhz *2 ; - s[i].dispclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dispclk_mhz; - s[i].dppclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dppclk_mhz; +/* Clocks independent of voltage level. */ + s[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : + dcn3_01_soc.clock_limits[closest_clk_lvl].dispclk_mhz; + + s[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : + dcn3_01_soc.clock_limits[closest_clk_lvl].dppclk_mhz; + s[i].dram_bw_per_chan_gbps = dcn3_01_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; s[i].dscclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dscclk_mhz;
Signature · 1 line
2.34.1
PATCH 6/12 drm/amd/display: Revert "drm/amd/display: Fix AMDGPU_MAX_BL_LEVEL value"
In reply to [email protected]From: Mario Limonciello <[email protected]>
This reverts commit 66abb996999de0d440a02583a6e70c2c24deab45. This broke custom brightness curves but it wasn't obvious because of other related changes. Custom brightness curves are always from a 0-255 input signal. The correct fix was to fix the default value which was done by [1].
--- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +4 -4
@@ -4800,16 +4800,16 @@ static int get_brightness_range(const struct amdgpu_dm_backlight_caps *caps,return 1 ; } -/* Rescale from [min..max] to [0..MAX_BACKLIGHT_LEVEL] */ +/* Rescale from [min..max] to [0..AMDGPU_MAX_BL_LEVEL] */ static inline u32 scale_input_to_fw (int min,int max,u64 input) { -return DIV_ROUND_CLOSEST_ULL (input *MAX_BACKLIGHT_LEVEL , max - min); +return DIV_ROUND_CLOSEST_ULL (input *AMDGPU_MAX_BL_LEVEL , max - min); } -/* Rescale from [0..MAX_BACKLIGHT_LEVEL] to [min..max] */ +/* Rescale from [0..AMDGPU_MAX_BL_LEVEL] to [min..max] */ static inline u32 scale_fw_to_input (int min,int max,u64 input) { -return min +DIV_ROUND_CLOSEST_ULL (input * (max - min),MAX_BACKLIGHT_LEVEL ); +return min +DIV_ROUND_CLOSEST_ULL (input * (max - min),AMDGPU_MAX_BL_LEVEL ); } static void convert_custom_brightness(const struct amdgpu_dm_backlight_caps *caps,
Signature · 1 line
2.34.1
PATCH 7/12 drm/amd/display: Avoid Read Remote DPCD Many Times
In reply to [email protected]From: Fangzhi Zuo <[email protected]>
Reading remote dpcd is time consuming. Instead of reading each byte one by one, read 16 bytes together.
--- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +6 -3
@@ -1763,14 +1763,17 @@ static bool dp_get_link_current_set_bw(struct drm_dp_aux *aux, uint32_t *cur_linunion lane_count_set lane_count;u8 dp_link_encoding;u8 link_bw_set =0 ; +u8 data[16 ] = {0 }; *cur_link_bw =0 ; -if (drm_dp_dpcd_read (aux,DP_MAIN_LINK_CHANNEL_CODING_SET , &dp_link_encoding,1 ) !=1 || -drm_dp_dpcd_read (aux,DP_LANE_COUNT_SET , &lane_count.raw,1 ) !=1 || -drm_dp_dpcd_read (aux,DP_LINK_BW_SET , &link_bw_set,1 ) !=1 ) +if (drm_dp_dpcd_read (aux,DP_LINK_BW_SET , data,16 ) !=16 )return false; + dp_link_encoding = data[DP_MAIN_LINK_CHANNEL_CODING_SET -DP_LINK_BW_SET ]; + link_bw_set = data[DP_LINK_BW_SET -DP_LINK_BW_SET ]; + lane_count.raw = data[DP_LANE_COUNT_SET -DP_LINK_BW_SET ]; +switch (dp_link_encoding) {case DP_8b_10b_ENCODING: link_rate = link_bw_set;
Signature · 1 line
2.34.1
PATCH 8/12 drm/amd/display: limited pll vco w/a v2
In reply to [email protected]From: Jingwen Zhu <[email protected]>
[Why/How]
The w/a will cause reboot black screen issue.
--- drivers/gpu/drm/amd/display/dmub/dmub_srv.h | 1 + drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 3 ++- drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c | 1 + 3 files changed, 4 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dmub/dmub_srv.h +1 -0
@@ -316,6 +316,7 @@ struct dmub_srv_hw_params {bool disable_sldo_opt;bool enable_non_transparent_setconfig;bool lower_hbr3_phy_ssc; +bool override_hbr3_pll_vco; };/**
drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +2 -1
@@ -843,7 +843,8 @@ union dmub_fw_boot_options { uint32_t ips_sequential_ono: 1; /**< 1 to enable sequential ONO IPS sequence */ uint32_t disable_sldo_opt: 1; /**< 1 to disable SLDO optimizations */ uint32_t lower_hbr3_phy_ssc: 1; /**< 1 to lower hbr3 phy ssc to 0.125 percent */ - uint32_t reserved : 6; /**< reserved */ + uint32_t override_hbr3_pll_vco: 1; /**< 1 to override the hbr3 pll vco to 0 */ + uint32_t reserved : 5; /**< reserved */ } bits;/**< boot bits */ uint32_t all;/**< 32-bit access to bits */ };
drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c +1 -0
@@ -377,6 +377,7 @@ void dmub_dcn31_enable_dmub_boot_options(struct dmub_srv *dmub, const struct dmu boot_options.bits.dpia_hpd_int_enable_supported = params->dpia_hpd_int_enable_supported; boot_options.bits.power_optimization = params->power_optimization; boot_options.bits.lower_hbr3_phy_ssc = params->lower_hbr3_phy_ssc; + boot_options.bits.override_hbr3_pll_vco = params->override_hbr3_pll_vco; boot_options.bits.sel_mux_phy_c_d_phy_f_g = (dmub->asic ==DMUB_ASIC_DCN31B ) ?1 :0 ;
Signature · 1 line
2.34.1
PATCH 9/12 drm/amd/display: Fixing hubp programming of 3dlut fast load
In reply to [email protected]From: Reza Amini <[email protected]>
[why]
HUBP needs to know the size of the lut's destination in MPC. This is currently defaulted to 17, and needs to be set for specific lut size.
[how]
Define and apply the missing hubp field. Taking this opportunity to consolidate the programming of 3dlut into a hubp and mpc function.
--- drivers/gpu/drm/amd/display/dc/dc.h | 26 ++++++++ .../amd/display/dc/hubp/dcn10/dcn10_hubp.h | 1 + .../amd/display/dc/hubp/dcn20/dcn20_hubp.h | 1 + .../amd/display/dc/hubp/dcn401/dcn401_hubp.c | 38 +++++++++++ .../amd/display/dc/hubp/dcn401/dcn401_hubp.h | 4 ++ .../amd/display/dc/hwss/dcn10/dcn10_hwseq.c | 63 ++++++++++++++----- drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h | 19 +++++- drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h | 11 ++++ 8 files changed, 148 insertions(+), 15 deletions(-)
drivers/gpu/drm/amd/display/dc/dc.h +26 -0
@@ -1311,6 +1311,32 @@ union dc_3dlut_state { }; +#define MATRIX_9C__DIM_128_ALIGNED_LEN 16 // 9+8 : 9 * 8 + 7 * 8 = 72 + 56 = 128 % 128 = 0 +#define MATRIX_17C__DIM_128_ALIGNED_LEN 32 //17+15: 17 * 8 + 15 * 8 = 136 + 120 = 256 % 128 = 0 +#define MATRIX_33C__DIM_128_ALIGNED_LEN 64 //17+47: 17 * 8 + 47 * 8 = 136 + 376 = 512 % 128 = 0 + +struct lut_rgb { +uint16_t b; +uint16_t g; +uint16_t r; +uint16_t padding; +}; + +//this structure maps directly to how the lut will read it from memory +struct lut_mem_mapping { +union { +//NATIVE MODE 1, 2 +//RGB layout [b][g][r] //red is 128 byte aligned +//BGR layout [r][g][b] //blue is 128 byte aligned +struct lut_rgb rgb_17c[17 ][17 ][MATRIX_17C__DIM_128_ALIGNED_LEN ]; +struct lut_rgb rgb_33c[33 ][33 ][MATRIX_33C__DIM_128_ALIGNED_LEN ]; + +//TRANSFORMED +uint16_t linear_rgb[(33 *33 *33 *4 /128 +1 )*128 ]; + }; +uint16_t size; +}; +struct dc_rmcm_3dlut {bool isInUse;const struct dc_stream_state *stream;
drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.h +1 -0
@@ -671,6 +671,7 @@ struct dcn_fl_regs_st {uint32_t lut_done;uint32_t lut_addr_mode;uint32_t lut_width; +uint32_t lut_mpc_width;uint32_t lut_tmz;uint32_t lut_crossbar_sel_r;uint32_t lut_crossbar_sel_g;
drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.h +1 -0
@@ -264,6 +264,7 @@type HUBP_3DLUT_DONE ;\type HUBP_3DLUT_ADDRESSING_MODE ;\type HUBP_3DLUT_WIDTH ;\ +type HUBP_3DLUT_MPC_WIDTH ;\type HUBP_3DLUT_TMZ ;\type HUBP_3DLUT_CROSSBAR_SELECT_Y_G ;\type HUBP_3DLUT_CROSSBAR_SELECT_CB_B ;\
drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.c +38 -0
@@ -127,6 +127,43 @@ void hubp401_program_3dlut_fl_format(struct hubp *hubp, enum hubp_3dlut_fl_formaREG_UPDATE (_3DLUT_FL_CONFIG,HUBP0_3DLUT_FL_FORMAT , format); } +void hubp401_program_3dlut_fl_config ( +struct hubp *hubp, +struct hubp_fl_3dlut_config *cfg) +{ +struct dcn20_hubp *hubp2 =TO_DCN20_HUBP (hubp); + +uint32_t mpc_width = {(cfg->width ==17 ) ?0 :1 }; +uint32_t width = {cfg->width}; + +if (cfg->layout ==DC_CM2_GPU_MEM_LAYOUT_1D_PACKED_LINEAR ) + width = (cfg->width ==17 ) ?4916 :35940 ; + +REG_UPDATE_2 (_3DLUT_FL_CONFIG, +HUBP0_3DLUT_FL_MODE , cfg->mode, +HUBP0_3DLUT_FL_FORMAT , cfg->format); + +REG_UPDATE_2 (_3DLUT_FL_BIAS_SCALE, +HUBP0_3DLUT_FL_BIAS , cfg->bias, +HUBP0_3DLUT_FL_SCALE , cfg->scale); + +REG_UPDATE (HUBP_3DLUT_ADDRESS_HIGH , +HUBP_3DLUT_ADDRESS_HIGH , cfg->address.lut3d.addr.high_part); +REG_UPDATE (HUBP_3DLUT_ADDRESS_LOW , +HUBP_3DLUT_ADDRESS_LOW , cfg->address.lut3d.addr.low_part); + +//cross bar +REG_UPDATE_8 (HUBP_3DLUT_CONTROL , +HUBP_3DLUT_MPC_WIDTH , mpc_width, +HUBP_3DLUT_WIDTH , width, +HUBP_3DLUT_CROSSBAR_SELECT_CR_R , cfg->crossbar_bit_slice_cr_r, +HUBP_3DLUT_CROSSBAR_SELECT_Y_G , cfg->crossbar_bit_slice_y_g, +HUBP_3DLUT_CROSSBAR_SELECT_CB_B , cfg->crossbar_bit_slice_cb_b, +HUBP_3DLUT_ADDRESSING_MODE , cfg->addr_mode, +HUBP_3DLUT_TMZ , cfg->protection_bits, +HUBP_3DLUT_ENABLE , cfg->enabled ?1 :0 ); +} +void hubp401_update_mall_sel (struct hubp *hubp,uint32_t mall_sel,bool c_cursor) {struct dcn20_hubp *hubp2 =TO_DCN20_HUBP (hubp);
@@ -1033,6 +1070,7 @@ static struct hubp_funcs dcn401_hubp_funcs = { .hubp_program_3dlut_fl_crossbar = hubp401_program_3dlut_fl_crossbar, .hubp_get_3dlut_fl_done = hubp401_get_3dlut_fl_done, .hubp_clear_tiling = hubp401_clear_tiling, + .hubp_program_3dlut_fl_config = hubp401_program_3dlut_fl_config, }; bool hubp401_construct(
drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.h +4 -0
@@ -349,6 +349,10 @@ void hubp401_program_3dlut_fl_format(struct hubp *hubp, enum hubp_3dlut_fl_formavoid hubp401_program_3dlut_fl_mode (struct hubp *hubp,enum hubp_3dlut_fl_mode mode); +void hubp401_program_3dlut_fl_config ( +struct hubp *hubp, +struct hubp_fl_3dlut_config *cfg); +void hubp401_clear_tiling (struct hubp *hubp); void hubp401_vready_at_or_After_vsync(struct hubp *hubp,
drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c +49 -14
95 lines not on this page · raw message
drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h +18 -1
@@ -89,7 +89,7 @@ enum hubp_3dlut_fl_addressing_mode {enum hubp_3dlut_fl_width { hubp_3dlut_fl_width_17 =17 , hubp_3dlut_fl_width_33 =33 , - hubp_3dlut_fl_width_transformed =4916 + hubp_3dlut_fl_width_transformed =4916 ,//mpc default };enum hubp_3dlut_fl_crossbar_bit_slice {
@@ -99,6 +99,22 @@ enum hubp_3dlut_fl_crossbar_bit_slice { hubp_3dlut_fl_crossbar_bit_slice_48_63 =3 }; +struct hubp_fl_3dlut_config { +bool enabled; +enum hubp_3dlut_fl_width width; +enum hubp_3dlut_fl_mode mode; +enum hubp_3dlut_fl_format format; +uint16_t bias; +uint16_t scale; +struct dc_plane_address address; +enum hubp_3dlut_fl_addressing_mode addr_mode; +enum dc_cm2_gpu_mem_layout layout; +uint8_t protection_bits; +enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_y_g; +enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_cb_b; +enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_cr_r; +}; +struct hubp {const struct hubp_funcs *funcs;struct dc_context *ctx;
@@ -288,6 +304,7 @@ struct hubp_funcs {enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_cb_b, enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_cr_r);int (*hubp_get_3dlut_fl_done)(struct hubp *hubp); +void (*hubp_program_3dlut_fl_config)(struct hubp *hubp,struct hubp_fl_3dlut_config *cfg);void (*hubp_clear_tiling)(struct hubp *hubp); };
drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h +11 -0
@@ -115,6 +115,16 @@ enum MCM_LUT_ID {MCM_LUT_SHAPER }; +struct mpc_fl_3dlut_config { +bool enabled; +uint16_t width; +bool select_lut_bank_a; +uint16_t bit_depth; +int hubp_index; +uint16_t bias; +uint16_t scale; +}; +union mcm_lut_params {const struct pwl_params *pwl;const struct tetrahedral_params *lut3d;
@@ -1098,6 +1108,7 @@ struct mpc_funcs {* MPC RMCM new HW sequential programming functions */ struct { +void (*fl_3dlut_configure)(struct mpc *mpc,struct mpc_fl_3dlut_config *cfg,int mpcc_id);void (*enable_3dlut_fl)(struct mpc *mpc,bool enable,int mpcc_id);void (*update_3dlut_fast_load_select)(struct mpc *mpc,int mpcc_id,int hubp_idx); void (*program_lut_read_write_control)(struct mpc *mpc, const enum MCM_LUT_ID id,
Signature · 1 line
2.34.1
PATCH 10/12 drm/amd/display: Toggle for Disable Force Pstate Allow on Disable
In reply to [email protected]From: Ryan Seto <[email protected]>
[Why & How]
In theory, driver should be able to support disabling force pstate allow after hardware release however this behavior is not tested yet. Introducing a new toggle to disable the force on the fly.
--- drivers/gpu/drm/amd/display/dc/dc.h | 1 + .../amd/display/dc/hwss/dcn401/dcn401_hwseq.c | 36 +++++++++++-------- .../dc/resource/dcn401/dcn401_resource.c | 1 + 3 files changed, 24 insertions(+), 14 deletions(-)
drivers/gpu/drm/amd/display/dc/dc.h +1 -0
@@ -1073,6 +1073,7 @@ struct dc_debug_options {unsigned int force_mall_ss_num_ways;bool alloc_extra_way_for_cursor;uint32_t subvp_extra_lines; +bool disable_force_pstate_allow_on_hw_release;bool force_usr_allow;/* uses value at boot and disables switch */ bool disable_dtb_ref_clk_switch;
drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c +22 -14
@@ -1621,20 +1621,28 @@ void dcn401_unblank_stream(struct pipe_ctx *pipe_ctx,void dcn401_hardware_release (struct dc *dc) { -dc_dmub_srv_fams2_update_config (dc, dc->current_state, false); - -/* If pstate unsupported, or still supported -* by firmware, force it supported by dcn -*/ -if (dc->current_state) { -if ((!dc->clk_mgr->clks.p_state_change_support || - dc->current_state->bw_ctx.bw.dcn.fams2_global_config.features.bits.enable) && - dc->res_pool->hubbub->funcs->force_pstate_change_control) - dc->res_pool->hubbub->funcs->force_pstate_change_control ( - dc->res_pool->hubbub, true, true); - - dc->current_state->bw_ctx.bw.dcn.clk.p_state_change_support = true; - dc->clk_mgr->funcs->update_clocks (dc->clk_mgr, dc->current_state, true); +if (!dc->debug.disable_force_pstate_allow_on_hw_release) { +dc_dmub_srv_fams2_update_config (dc, dc->current_state, false); + +/* If pstate unsupported, or still supported +* by firmware, force it supported by dcn +*/ +if (dc->current_state) { +if ((!dc->clk_mgr->clks.p_state_change_support || + dc->current_state->bw_ctx.bw.dcn.fams2_global_config.features.bits.enable) && + dc->res_pool->hubbub->funcs->force_pstate_change_control) + dc->res_pool->hubbub->funcs->force_pstate_change_control ( + dc->res_pool->hubbub, true, true); + + dc->current_state->bw_ctx.bw.dcn.clk.p_state_change_support = true; + dc->clk_mgr->funcs->update_clocks (dc->clk_mgr, dc->current_state, true); + } + }else { +if (dc->current_state) { + dc->clk_mgr->clks.p_state_change_support = false; + dc->clk_mgr->funcs->update_clocks (dc->clk_mgr, dc->current_state, true); + } +dc_dmub_srv_fams2_update_config (dc, dc->current_state, false); } }
drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c +1 -0
@@ -708,6 +708,7 @@ static const struct dc_debug_options debug_defaults_drv = { }, .use_max_lb = true, .force_disable_subvp = false, + .disable_force_pstate_allow_on_hw_release = false, .exit_idle_opt_for_cursor_updates = true, .using_dml2 = true, .using_dml21 = true,
Signature · 1 line
2.34.1
PATCH 11/12 drm/amd/display: Adding interface to log hw state when underflow happens
In reply to [email protected]From: Muhammad Ahmed <[email protected]>
[why]
Will help us better debug underflow issues.
--- drivers/gpu/drm/amd/display/dc/core/dc.c | 18 +++++++ drivers/gpu/drm/amd/display/dc/dc.h | 30 ++++++++++++ .../display/dc/hubbub/dcn30/dcn30_hubbub.c | 31 ++++++++++++ .../display/dc/hubbub/dcn30/dcn30_hubbub.h | 6 +++ .../display/dc/hubbub/dcn31/dcn31_hubbub.c | 2 + .../display/dc/hubbub/dcn32/dcn32_hubbub.c | 2 + .../display/dc/hubbub/dcn35/dcn35_hubbub.c | 2 + .../display/dc/hubbub/dcn401/dcn401_hubbub.c | 2 + .../amd/display/dc/hubp/dcn10/dcn10_hubp.h | 8 +++- .../amd/display/dc/hubp/dcn30/dcn30_hubp.c | 26 ++++++++++ .../amd/display/dc/hubp/dcn30/dcn30_hubp.h | 8 +++- .../amd/display/dc/hubp/dcn31/dcn31_hubp.c | 15 ++++++ .../amd/display/dc/hubp/dcn31/dcn31_hubp.h | 6 ++- .../amd/display/dc/hubp/dcn32/dcn32_hubp.c | 3 ++ .../amd/display/dc/hubp/dcn35/dcn35_hubp.c | 3 ++ .../amd/display/dc/hubp/dcn401/dcn401_hubp.c | 3 ++ .../amd/display/dc/hubp/dcn401/dcn401_hubp.h | 4 +- .../amd/display/dc/hwss/dcn30/dcn30_hwseq.c | 48 +++++++++++++++++++ .../amd/display/dc/hwss/dcn30/dcn30_hwseq.h | 5 ++ .../amd/display/dc/hwss/dcn30/dcn30_init.c | 1 + .../amd/display/dc/hwss/dcn31/dcn31_init.c | 1 + .../amd/display/dc/hwss/dcn314/dcn314_init.c | 1 + .../amd/display/dc/hwss/dcn32/dcn32_init.c | 1 + .../amd/display/dc/hwss/dcn35/dcn35_init.c | 1 + .../amd/display/dc/hwss/dcn351/dcn351_init.c | 1 + .../amd/display/dc/hwss/dcn401/dcn401_init.c | 1 + .../drm/amd/display/dc/hwss/hw_sequencer.h | 4 ++ .../gpu/drm/amd/display/dc/inc/hw/dchubbub.h | 2 + drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h | 2 + .../dc/resource/dcn32/dcn32_resource.h | 3 +- .../dc/resource/dcn401/dcn401_resource.h | 3 +- 31 files changed, 236 insertions(+), 7 deletions(-)
drivers/gpu/drm/amd/display/dc/core/dc.c +18 -0
@@ -6339,3 +6339,21 @@ bool dc_can_clear_cursor_limit(struct dc *dc)return false; } + +void dc_get_underflow_debug_data_for_otg (struct dc *dc,int primary_otg_inst, +struct dc_underflow_debug_data *out_data) +{ +struct timing_generator *tg =NULL ; + +for (int i =0 ; i <MAX_PIPES ; i++) { +if (dc->res_pool->timing_generators[i] && + dc->res_pool->timing_generators[i]->inst == primary_otg_inst) { + tg = dc->res_pool->timing_generators[i]; +break ; + } + } + +dc_exit_ips_for_hw_access (dc); +if (dc->hwss.get_underflow_debug_data) + dc->hwss.get_underflow_debug_data (dc, tg, out_data); +}
drivers/gpu/drm/amd/display/dc/dc.h +30 -0
@@ -1819,6 +1819,23 @@ struct dc_surface_update {struct dc_bias_and_scale bias_and_scale; }; +struct dc_underflow_debug_data { +uint32_t otg_inst; +uint32_t otg_underflow; +uint32_t h_position; +uint32_t v_position; +uint32_t otg_frame_count; +struct dc_underflow_per_hubp_debug_data { +uint32_t hubp_underflow; +uint32_t hubp_in_blank; +uint32_t hubp_readline; +uint32_t det_config_error; + } hubps[MAX_PIPES ]; +uint32_t curr_det_sizes[MAX_PIPES ]; +uint32_t target_det_sizes[MAX_PIPES ]; +uint32_t compbuf_config_error; +}; +/* * Create a new surface with default parameters; */
@@ -2713,4 +2730,17 @@ bool dc_is_timing_changed(struct dc_stream_state *cur_stream,bool dc_is_cursor_limit_pending (struct dc *dc);bool dc_can_clear_cursor_limit (struct dc *dc); +/** +* dc_get_underflow_debug_data_for_otg() - Retrieve underflow debug data. +* +* @dc: Pointer to the display core context. +* @primary_otg_inst: Instance index of the primary OTG that underflowed. +* @out_data: Pointer to a dc_underflow_debug_data struct to be filled with debug information. +* +* This function collects and logs underflow-related HW states when underflow happens, +* including OTG underflow status, current read positions, frame count, and per-HUBP debug data. +* The results are stored in the provided out_data structure for further analysis or logging. +*/ +void dc_get_underflow_debug_data_for_otg (struct dc *dc,int primary_otg_inst,struct dc_underflow_debug_data *out_data); +#endif /* DC_INTERFACE_H_ */
drivers/gpu/drm/amd/display/dc/hubbub/dcn30/dcn30_hubbub.c +31 -0
@@ -440,6 +440,35 @@ void hubbub3_init_watermarks(struct hubbub *hubbub)REG_WRITE (DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D , reg); } +void hubbub3_get_det_sizes (struct hubbub *hubbub,uint32_t *curr_det_sizes,uint32_t *target_det_sizes) +{ +struct dcn20_hubbub *hubbub1 =TO_DCN20_HUBBUB (hubbub); + +REG_GET_2 (DCHUBBUB_DET0_CTRL ,DET0_SIZE_CURRENT , &curr_det_sizes[0 ], +DET0_SIZE , &target_det_sizes[0 ]); + +REG_GET_2 (DCHUBBUB_DET1_CTRL ,DET1_SIZE_CURRENT , &curr_det_sizes[1 ], +DET1_SIZE , &target_det_sizes[1 ]); + +REG_GET_2 (DCHUBBUB_DET2_CTRL ,DET2_SIZE_CURRENT , &curr_det_sizes[2 ], +DET2_SIZE , &target_det_sizes[2 ]); + +REG_GET_2 (DCHUBBUB_DET3_CTRL ,DET3_SIZE_CURRENT , &curr_det_sizes[3 ], +DET3_SIZE , &target_det_sizes[3 ]); + +} + +uint32_t hubbub3_compbuf_config_error (struct hubbub *hubbub) +{ +struct dcn20_hubbub *hubbub1 =TO_DCN20_HUBBUB (hubbub); +uint32_t compbuf_config_error =0 ; + +REG_GET (DCHUBBUB_COMPBUF_CTRL ,CONFIG_ERROR , + &compbuf_config_error); + +return compbuf_config_error; +} +static const struct hubbub_funcs hubbub30_funcs = { .update_dchub = hubbub2_update_dchub, .init_dchub_sys_ctx = hubbub3_init_dchub_sys_ctx,
@@ -457,6 +486,8 @@ static const struct hubbub_funcs hubbub30_funcs = { .force_pstate_change_control = hubbub3_force_pstate_change_control, .init_watermarks = hubbub3_init_watermarks, .hubbub_read_state = hubbub2_read_state, + .get_det_sizes = hubbub3_get_det_sizes, + .compbuf_config_error = hubbub3_compbuf_config_error, }; void hubbub3_construct(struct dcn20_hubbub *hubbub3,
drivers/gpu/drm/amd/display/dc/hubbub/dcn30/dcn30_hubbub.h +6 -0
@@ -133,4 +133,10 @@ void hubbub3_force_pstate_change_control(struct hubbub *hubbub,void hubbub3_init_watermarks (struct hubbub *hubbub); +void hubbub3_get_det_sizes (struct hubbub *hubbub, +uint32_t *curr_det_sizes, +uint32_t *target_det_sizes); + +uint32_t hubbub3_compbuf_config_error (struct hubbub *hubbub); +#endif
drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c +2 -0
@@ -1071,6 +1071,8 @@ static const struct hubbub_funcs hubbub31_funcs = { .program_compbuf_size = dcn31_program_compbuf_size, .init_crb = dcn31_init_crb, .hubbub_read_state = hubbub2_read_state, + .get_det_sizes = hubbub3_get_det_sizes, + .compbuf_config_error = hubbub3_compbuf_config_error, }; void hubbub31_construct(struct dcn20_hubbub *hubbub31,
drivers/gpu/drm/amd/display/dc/hubbub/dcn32/dcn32_hubbub.c +2 -0
@@ -1009,6 +1009,8 @@ static const struct hubbub_funcs hubbub32_funcs = { .force_usr_retraining_allow = hubbub32_force_usr_retraining_allow, .set_request_limit = hubbub32_set_request_limit, .get_mall_en = hubbub32_get_mall_en, + .get_det_sizes = hubbub3_get_det_sizes, + .compbuf_config_error = hubbub3_compbuf_config_error, }; void hubbub32_construct(struct dcn20_hubbub *hubbub2,
drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c +2 -0
@@ -589,6 +589,8 @@ static const struct hubbub_funcs hubbub35_funcs = { .hubbub_read_state = hubbub2_read_state, .force_usr_retraining_allow = hubbub32_force_usr_retraining_allow, .dchubbub_init = hubbub35_init, + .get_det_sizes = hubbub3_get_det_sizes, + .compbuf_config_error = hubbub3_compbuf_config_error, }; void hubbub35_construct(struct dcn20_hubbub *hubbub2,
drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.c +2 -0
@@ -1247,6 +1247,8 @@ static const struct hubbub_funcs hubbub4_01_funcs = { .program_compbuf_segments = dcn401_program_compbuf_segments, .wait_for_det_update = dcn401_wait_for_det_update, .program_arbiter = dcn401_program_arbiter, + .get_det_sizes = hubbub3_get_det_sizes, + .compbuf_config_error = hubbub3_compbuf_config_error, }; void hubbub401_construct(struct dcn20_hubbub *hubbub2,
drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.h +6 -2
@@ -104,7 +104,8 @@SRI (DCN_SURF1_TTU_CNTL1 ,HUBPREQ , id),\SRI (DCN_CUR0_TTU_CNTL0 ,HUBPREQ , id),\SRI (DCN_CUR0_TTU_CNTL1 ,HUBPREQ , id),\ -SRI (HUBP_CLK_CNTL ,HUBP , id) +SRI (HUBP_CLK_CNTL ,HUBP , id),\ +SRI (HUBPRET_READ_LINE_VALUE ,HUBPRET , id) /* Register address initialization macro for ASICs with VM */#define HUBP_REG_LIST_DCN_VM (id)\
@@ -249,7 +250,8 @@uint32_t CURSOR_POSITION ; \uint32_t CURSOR_HOT_SPOT ; \uint32_t CURSOR_DST_OFFSET ; \ - uint32_t HUBP_CLK_CNTL +uint32_t HUBP_CLK_CNTL ; \ + uint32_t HUBPRET_READ_LINE_VALUE#define HUBP_SF (reg_name, field_name, post_fix)\ .field_name = reg_name ## __ ## field_name ## post_fix
@@ -622,6 +624,8 @@type DCN_VM_SYSTEM_APERTURE_DEFAULT_SYSTEM ;\type DCN_VM_SYSTEM_APERTURE_DEFAULT_ADDR_MSB ;\type DCN_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB ;\ +type PIPE_READ_LINE ;\ +type HUBP_SEG_ALLOC_ERR_STATUS ;\/* todo: get these from GVM instead of reading registers ourselves */ \type PAGE_DIRECTORY_ENTRY_HI32 ;\type PAGE_DIRECTORY_ENTRY_LO32 ;\
drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.c +26 -0
@@ -505,6 +505,30 @@ void hubp3_init(struct hubp *hubp)hubp_reset (hubp); } +uint32_t hubp3_get_current_read_line (struct hubp *hubp) +{ +uint32_t read_line =0 ; +struct dcn20_hubp *hubp2 =TO_DCN20_HUBP (hubp); + +REG_GET (HUBPRET_READ_LINE_VALUE , +PIPE_READ_LINE , + &read_line); + +return read_line; +} + +unsigned int hubp3_get_underflow_status (struct hubp *hubp) +{ +uint32_t hubp_underflow =0 ; +struct dcn20_hubp *hubp2 =TO_DCN20_HUBP (hubp); + +REG_GET (DCHUBP_CNTL , +HUBP_UNDERFLOW_STATUS , + &hubp_underflow); + +return hubp_underflow; +} +static struct hubp_funcs dcn30_hubp_funcs = { .hubp_enable_tripleBuffer = hubp2_enable_triplebuffer, .hubp_is_triplebuffer_enabled = hubp2_is_triplebuffer_enabled,
@@ -534,6 +558,8 @@ static struct hubp_funcs dcn30_hubp_funcs = { .hubp_soft_reset = hubp1_soft_reset, .hubp_set_flip_int = hubp1_set_flip_int, .hubp_clear_tiling = hubp3_clear_tiling, + .hubp_get_underflow_status = hubp3_get_underflow_status, + .hubp_get_current_read_line = hubp3_get_current_read_line, }; bool hubp3_construct(
drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.h +7 -1
@@ -243,7 +243,8 @@HUBP_SF (HUBPREQ0_FLIP_PARAMETERS_6 ,REFCYC_PER_META_CHUNK_FLIP_C , mask_sh),\HUBP_SF (HUBPREQ0_VBLANK_PARAMETERS_5 ,REFCYC_PER_VM_GROUP_VBLANK , mask_sh),\HUBP_SF (HUBPREQ0_VBLANK_PARAMETERS_6 ,REFCYC_PER_VM_REQ_VBLANK , mask_sh),\ - HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, VM_GROUP_SIZE, mask_sh) +HUBP_SF (HUBP0_DCHUBP_REQ_SIZE_CONFIG ,VM_GROUP_SIZE , mask_sh),\ + HUBP_SF(HUBPRET0_HUBPRET_READ_LINE_VALUE, PIPE_READ_LINE, mask_sh) bool hubp3_construct( struct dcn20_hubp *hubp2,
@@ -299,6 +300,11 @@ void hubp3_init(struct hubp *hubp);void hubp3_clear_tiling (struct hubp *hubp); +uint32_t hubp3_get_current_read_line (struct hubp *hubp); + +uint32_t hubp3_get_underflow_status (struct hubp *hubp); + +#endif /* __DC_HUBP_DCN30_H__ */
drivers/gpu/drm/amd/display/dc/hubp/dcn31/dcn31_hubp.c +15 -0
@@ -68,6 +68,18 @@ void hubp31_program_extended_blank_value(hubp31_program_extended_blank (hubp, min_dst_y_next_start_optimized); } +uint32_t hubp31_get_det_config_error (struct hubp *hubp) +{ +uint32_t config_error =0 ; +struct dcn20_hubp *hubp2 =TO_DCN20_HUBP (hubp); + +REG_GET (DCHUBP_CNTL , +HUBP_SEG_ALLOC_ERR_STATUS , + &config_error); + +return config_error; +} +static struct hubp_funcs dcn31_hubp_funcs = { .hubp_enable_tripleBuffer = hubp2_enable_triplebuffer, .hubp_is_triplebuffer_enabled = hubp2_is_triplebuffer_enabled,
@@ -98,6 +110,9 @@ static struct hubp_funcs dcn31_hubp_funcs = { .hubp_in_blank = hubp1_in_blank, .program_extended_blank = hubp31_program_extended_blank, .hubp_clear_tiling = hubp3_clear_tiling, + .hubp_get_underflow_status = hubp3_get_underflow_status, + .hubp_get_current_read_line = hubp3_get_current_read_line, + .hubp_get_det_config_error = hubp31_get_det_config_error, }; bool hubp31_construct(
drivers/gpu/drm/amd/display/dc/hubp/dcn31/dcn31_hubp.h +5 -1
@@ -228,7 +228,9 @@HUBP_SF (HUBPREQ0_FLIP_PARAMETERS_6 ,REFCYC_PER_META_CHUNK_FLIP_C , mask_sh),\HUBP_SF (HUBPREQ0_VBLANK_PARAMETERS_5 ,REFCYC_PER_VM_GROUP_VBLANK , mask_sh),\HUBP_SF (HUBPREQ0_VBLANK_PARAMETERS_6 ,REFCYC_PER_VM_REQ_VBLANK , mask_sh),\ - HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, VM_GROUP_SIZE, mask_sh) +HUBP_SF (HUBP0_DCHUBP_REQ_SIZE_CONFIG ,VM_GROUP_SIZE , mask_sh),\ +HUBP_SF (HUBPRET0_HUBPRET_READ_LINE_VALUE ,PIPE_READ_LINE , mask_sh),\ + HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_SEG_ALLOC_ERR_STATUS, mask_sh) bool hubp31_construct(
@@ -246,4 +248,6 @@ void hubp31_set_unbounded_requesting(struct hubp *hubp, bool enable);void hubp31_program_extended_blank_value (struct hubp *hubp,unsigned int min_dst_y_next_start_optimized); +uint32_t hubp31_get_det_config_error (struct hubp *hubp); +#endif /* __DC_HUBP_DCN31_H__ */
drivers/gpu/drm/amd/display/dc/hubp/dcn32/dcn32_hubp.c +3 -0
@@ -206,6 +206,9 @@ static struct hubp_funcs dcn32_hubp_funcs = { .hubp_update_mall_sel = hubp32_update_mall_sel, .hubp_prepare_subvp_buffering = hubp32_prepare_subvp_buffering, .hubp_clear_tiling = hubp3_clear_tiling, + .hubp_get_underflow_status = hubp3_get_underflow_status, + .hubp_get_current_read_line = hubp3_get_current_read_line, + .hubp_get_det_config_error = hubp31_get_det_config_error, }; bool hubp32_construct(
drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.c +3 -0
@@ -218,6 +218,9 @@ static struct hubp_funcs dcn35_hubp_funcs = { .hubp_in_blank = hubp1_in_blank, .program_extended_blank = hubp31_program_extended_blank_value, .hubp_clear_tiling = hubp3_clear_tiling, + .hubp_get_underflow_status = hubp3_get_underflow_status, + .hubp_get_current_read_line = hubp3_get_current_read_line, + .hubp_get_det_config_error = hubp31_get_det_config_error, }; bool hubp35_construct(
drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.c +3 -0
@@ -1071,6 +1071,9 @@ static struct hubp_funcs dcn401_hubp_funcs = { .hubp_get_3dlut_fl_done = hubp401_get_3dlut_fl_done, .hubp_clear_tiling = hubp401_clear_tiling, .hubp_program_3dlut_fl_config = hubp401_program_3dlut_fl_config, + .hubp_get_underflow_status = hubp3_get_underflow_status, + .hubp_get_current_read_line = hubp3_get_current_read_line, + .hubp_get_det_config_error = hubp31_get_det_config_error, }; bool hubp401_construct(
drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.h +3 -1
@@ -252,7 +252,9 @@HUBP_SF (HUBP0_DCHUBP_MCACHEID_CONFIG ,MCACHEID_MALL_PREF_1H_P0 , mask_sh),\HUBP_SF (HUBP0_DCHUBP_MCACHEID_CONFIG ,MCACHEID_MALL_PREF_2H_P0 , mask_sh),\HUBP_SF (HUBP0_DCHUBP_MCACHEID_CONFIG ,MCACHEID_MALL_PREF_1H_P1 , mask_sh),\ - HUBP_SF(HUBP0_DCHUBP_MCACHEID_CONFIG, MCACHEID_MALL_PREF_2H_P1, mask_sh) +HUBP_SF (HUBP0_DCHUBP_MCACHEID_CONFIG ,MCACHEID_MALL_PREF_2H_P1 , mask_sh),\ +HUBP_SF (HUBPRET0_HUBPRET_READ_LINE_VALUE ,PIPE_READ_LINE , mask_sh),\ + HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_SEG_ALLOC_ERR_STATUS, mask_sh)void hubp401_update_mall_sel (struct hubp *hubp,uint32_t mall_sel,bool c_cursor);
drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c +48 -0
51 lines not on this page · raw message
drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h +5 -0
@@ -29,6 +29,7 @@#include "hw_sequencer_private.h" struct dc ; +struct dc_underflow_debug_data ;void dcn30_init_hw (struct dc *dc); void dcn30_program_all_writeback_pipes_in_tree(
@@ -98,4 +99,8 @@ void dcn30_prepare_bandwidth(struct dc *dc,void dcn30_wait_for_all_pending_updates (const struct pipe_ctx *pipe_ctx); +void dcn30_get_underflow_debug_data (const struct dc *dc, +struct timing_generator *tg, +struct dc_underflow_debug_data *out_data); +#endif /* __DC_HWSS_DCN30_H__ */
drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c +1 -0
@@ -110,6 +110,7 @@ static const struct hw_sequencer_funcs dcn30_funcs = { .update_visual_confirm_color = dcn10_update_visual_confirm_color, .is_abm_supported = dcn21_is_abm_supported, .wait_for_all_pending_updates = dcn30_wait_for_all_pending_updates, + .get_underflow_debug_data = dcn30_get_underflow_debug_data, };static const struct hwseq_private_funcs dcn30_private_funcs = {
drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c +1 -0
@@ -112,6 +112,7 @@ static const struct hw_sequencer_funcs dcn31_funcs = { .exit_optimized_pwr_state = dcn21_exit_optimized_pwr_state, .update_visual_confirm_color = dcn10_update_visual_confirm_color, .setup_hpo_hw_control = dcn31_setup_hpo_hw_control, + .get_underflow_debug_data = dcn30_get_underflow_debug_data, };static const struct hwseq_private_funcs dcn31_private_funcs = {
drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c +1 -0
@@ -115,6 +115,7 @@ static const struct hw_sequencer_funcs dcn314_funcs = { .update_visual_confirm_color = dcn10_update_visual_confirm_color, .calculate_pix_rate_divider = dcn314_calculate_pix_rate_divider, .setup_hpo_hw_control = dcn31_setup_hpo_hw_control, + .get_underflow_debug_data = dcn30_get_underflow_debug_data, };static const struct hwseq_private_funcs dcn314_private_funcs = {
drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c +1 -0
@@ -121,6 +121,7 @@ static const struct hw_sequencer_funcs dcn32_funcs = { .calculate_pix_rate_divider = dcn32_calculate_pix_rate_divider, .program_outstanding_updates = dcn32_program_outstanding_updates, .wait_for_all_pending_updates = dcn30_wait_for_all_pending_updates, + .get_underflow_debug_data = dcn30_get_underflow_debug_data, };static const struct hwseq_private_funcs dcn32_private_funcs = {
drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c +1 -0
@@ -128,6 +128,7 @@ static const struct hw_sequencer_funcs dcn35_funcs = { .enable_plane = dcn20_enable_plane, .update_dchubp_dpp = dcn20_update_dchubp_dpp, .post_unlock_reset_opp = dcn20_post_unlock_reset_opp, + .get_underflow_debug_data = dcn30_get_underflow_debug_data, };static const struct hwseq_private_funcs dcn35_private_funcs = {
drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c +1 -0
@@ -123,6 +123,7 @@ static const struct hw_sequencer_funcs dcn351_funcs = { .set_long_vtotal = dcn35_set_long_vblank, .calculate_pix_rate_divider = dcn32_calculate_pix_rate_divider, .setup_hpo_hw_control = dcn35_setup_hpo_hw_control, + .get_underflow_debug_data = dcn30_get_underflow_debug_data, };static const struct hwseq_private_funcs dcn351_private_funcs = {
drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c +1 -0
@@ -104,6 +104,7 @@ static const struct hw_sequencer_funcs dcn401_funcs = { .enable_plane = dcn20_enable_plane, .update_dchubp_dpp = dcn20_update_dchubp_dpp, .post_unlock_reset_opp = dcn20_post_unlock_reset_opp, + .get_underflow_debug_data = dcn30_get_underflow_debug_data, };static const struct hwseq_private_funcs dcn401_private_funcs = {
drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h +4 -0
@@ -47,6 +47,7 @@ struct link_resource;struct dc_dmub_cmd ;struct pg_block_update ;struct drr_params ; +struct dc_underflow_debug_data ;struct subvp_pipe_control_lock_fast_params {struct dc *dc;
@@ -475,6 +476,9 @@ struct hw_sequencer_funcs { struct dc_state *context);void (*post_unlock_reset_opp)(struct dc *dc,struct pipe_ctx *opp_head); +void (*get_underflow_debug_data)(const struct dc *dc, +struct timing_generator *tg, +struct dc_underflow_debug_data *out_data); }; void color_space_to_black_color(
drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h +2 -0
@@ -229,6 +229,8 @@ struct hubbub_funcs {void (*program_compbuf_segments)(struct hubbub *hubbub,unsigned compbuf_size_seg,bool safe_to_increase);void (*wait_for_det_update)(struct hubbub *hubbub,int hubp_inst);bool (*program_arbiter)(struct hubbub *hubbub,struct dml2_display_arb_regs *arb_regs,bool safe_to_lower); +void (*get_det_sizes)(struct hubbub *hubbub,uint32_t *curr_det_sizes,uint32_t *target_det_sizes); +uint32_t (*compbuf_config_error)(struct hubbub *hubbub); };struct hubbub {
drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h +2 -0
@@ -306,6 +306,8 @@ struct hubp_funcs {int (*hubp_get_3dlut_fl_done)(struct hubp *hubp);void (*hubp_program_3dlut_fl_config)(struct hubp *hubp,struct hubp_fl_3dlut_config *cfg);void (*hubp_clear_tiling)(struct hubp *hubp); +uint32_t (*hubp_get_current_read_line)(struct hubp *hubp); + uint32_t (*hubp_get_det_config_error)(struct hubp *hubp); };#endif
drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h +2 -1
@@ -1141,7 +1141,8 @@ unsigned int dcn32_get_max_hw_cursor_size(const struct dc *dc,SRI_ARR (DCN_SURF1_TTU_CNTL1 ,HUBPREQ , id), \SRI_ARR (DCN_CUR0_TTU_CNTL0 ,HUBPREQ , id), \SRI_ARR (DCN_CUR0_TTU_CNTL1 ,HUBPREQ , id), \ - SRI_ARR(HUBP_CLK_CNTL, HUBP, id) +SRI_ARR (HUBP_CLK_CNTL ,HUBP , id), \ + SRI_ARR(HUBPRET_READ_LINE_VALUE, HUBPRET, id)#define HUBP_REG_LIST_DCN2_COMMON_RI (id) \ HUBP_REG_LIST_DCN_RI(id), HUBP_REG_LIST_DCN_VM_RI(id), \ SRI_ARR(PREFETCH_SETTINGS, HUBPREQ, id), \
drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h +2 -1
@@ -140,7 +140,8 @@ void dcn401_prepare_mcache_programming(struct dc *dc, struct dc_state *context);SRI_ARR (UCLK_PSTATE_FORCE ,HUBPREQ , id), \HUBP_3DLUT_FL_REG_LIST_DCN401 (id), \SRI_ARR (DCSURF_VIEWPORT_MCACHE_SPLIT_COORDINATE ,HUBP , id), \ -SRI_ARR (DCHUBP_MCACHEID_CONFIG ,HUBP , id) +SRI_ARR (DCHUBP_MCACHEID_CONFIG ,HUBP , id), \ +SRI_ARR (HUBPRET_READ_LINE_VALUE ,HUBPRET , id) /* ABM */#define ABM_DCN401_REG_LIST_RI (id) \
Signature · 1 line
2.34.1
PATCH 12/12 drm/amd/display: Promote DC to 3.2.344
In reply to [email protected]From: Taimur Hassan <[email protected]>
Summary:
* Add interface to log hw state when underflow happens
* Fix hubp programming of 3dlut fast load
* Avoid Read Remote DPCD Many Times
* More liberal vmin/vmax update for freesync
* Fix dmub access race condition
--- drivers/gpu/drm/amd/display/dc/dc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dc/dc.h +1 -1
@@ -55,7 +55,7 @@ struct aux_payload;struct set_config_cmd_payload ;struct dmub_notification ; -#define DC_VER "3.2.343" +#define DC_VER "3.2.344"/** * MAX_SURFACES - representative of the upper bound of surfaces that can be piped to a single CRTC
Signature · 1 line
2.34.1
[Public]
Hi all,
This week this patchset was tested on 4 systems, two dGPU and two APU based, and tested across multiple display and connection types.
APU
* Single Display eDP -> 1080p 60hz, 1920x1200 165hz
* Single Display DP (SST DSC) -> 4k144hz, 4k240hz
* Multi display -> eDP + DP/HDMI/USB-C -> 1080p 60hz eDP + 4k 144hz, 4k 240hz (Includes USB-C to DP/HDMI adapters)
* Thunderbolt -> LG Ultrafine 5k
* MST DSC -> Cable Matters 101075 (DP to 3x DP) with 3x 4k60hz displays, HP Hook G2 with 2x 4k60hz displays
* USB 4 -> HP Hook G4, Lenovo Thunderbolt Dock, both with 2x 4k60hz DP and 1x 4k60hz HDMI displays
* SST PCON -> Club3D CAC-1085 + 1x 4k 144hz, FRL3, at a max resolution supported by the dongle of 4k 120hz YUV420 12bpc.
* MST PCON -> 1x 4k 144hz, FRL3, at a max resolution supported by the adapter of 4k 120hz RGB 8bpc.
DGPU
* Single Display DP (SST DSC) -> 4k144hz, 4k240hz
* Multiple Display DP -> 4k240hz + 4k144hz
* MST (Startech MST14DP123DP [DP to 3x DP] and 2x 4k 60hz displays)
* MST DSC (with Cable Matters 101075 [DP to 3x DP] with 3x 4k60hz displays)
The testing is a mix of automated and manual tests. Manual testing includes (but is not limited to)
* Changing display configurations and settings
* Video/Audio playback
* Benchmark testing
* Suspend/Resume testing
* Feature testing (Freesync, HDCP, etc.)
Automated testing includes (but is not limited to)
* Script testing (scripts to automate some of the manual checks)
* IGT testing
The testing is mainly tested on the following displays, but occasionally there are tests with other displays
* Samsung G8 Neo 4k240hz
* Samsung QN55QN95B 4k 120hz
* Acer XV322QKKV 4k144hz
* HP U27 4k Wireless 4k60hz
* LG 27UD58B 4k60hz
* LG 32UN650WA 4k60hz
* LG Ultrafine 5k 5k60hz
* AU Optronics B140HAN01.1 1080p 60hz eDP
* AU Optronics B160UAN01.J 1920x1200 165hz eDP
The patchset consists of the amd-staging-drm-next branch (Head commit - 6795226cf20783be8a3ede94a26c1bc8976ad697 -> drm/amdgpu: Add chain runlists support to GC9.4.2) with new patches added on top of it.
Tested on Ubuntu 24.04.3, on Wayland and X11, using KDE Plasma and Gnome.
Thank you,
Dan Wheeler
Sr. Technologist | AMD
SW Display
------------------------------------------------------------------------------------------------------------------
1 Commerce Valley Dr E, Thornhill, ON L3T 7X6
amd.com
-----Original Message-----
From: [email protected] <[email protected]>
Sent: Wednesday, July 30, 2025 2:59 PM
To: [email protected]
Cc: Wentland, Harry <[email protected]>; Li, Sun peng (Leo) <[email protected]>; Pillai, Aurabindo <[email protected]>; Li, Roman <[email protected]>; Lin, Wayne <[email protected]>; Chung, ChiaHsuan (Tom) <[email protected]>; Zuo, Jerry <[email protected]>; Wheeler, Daniel <[email protected]>; Wu, Ray <[email protected]>; LIPSKI, IVAN <[email protected]>; Hung, Alex <[email protected]>; Li, Roman <[email protected]>
Subject: [PATCH 00/12] DC Patches July 30, 2025
From: Roman Li <[email protected]>
This DC patchset brings improvements in multiple areas. In summary, we highlight:
* Add interface to log hw state when underflow happens
* Fix hubp programming of 3dlut fast load
* Avoid Read Remote DPCD Many Times
* More liberal vmin/vmax update for freesync
* Fix dmub access race condition
Aurabindo Pillai (2):
drm/amd/display: fix dmub access race condition drm/amd/display: more liberal vmin/vmax update for freesync
Duncan Ma (1):
drm/amd/display: Adjust AUX-less ALPM setting
Fangzhi Zuo (1):
drm/amd/display: Avoid Read Remote DPCD Many Times
Jingwen Zhu (1):
drm/amd/display: limited pll vco w/a v2
Mario Limonciello (1):
drm/amd/display: Revert "drm/amd/display: Fix AMDGPU_MAX_BL_LEVEL
value"
Muhammad Ahmed (1):
drm/amd/display: Adding interface to log hw state when underflow
happens
Paul Hsieh (1):
drm/amd/display: update dpp/disp clock from smu clock table
Reza Amini (1):
drm/amd/display: Fixing hubp programming of 3dlut fast load
Ryan Seto (1):
drm/amd/display: Toggle for Disable Force Pstate Allow on Disable
Siyang Liu (1):
drm/amd/display: fix a Null pointer dereference vulnerability
Taimur Hassan (1):
drm/amd/display: Promote DC to 3.2.344
.../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 81 ++++++++++++++++--- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 14 ++++ .../display/amdgpu_dm/amdgpu_dm_mst_types.c | 9 ++- .../display/dc/clk_mgr/dcn301/vg_clk_mgr.c | 16 ++++ drivers/gpu/drm/amd/display/dc/core/dc.c | 37 ++++++--- drivers/gpu/drm/amd/display/dc/dc.h | 59 +++++++++++++- .../amd/display/dc/dml/dcn301/dcn301_fpu.c | 20 ++++- .../display/dc/hubbub/dcn30/dcn30_hubbub.c | 31 +++++++ .../display/dc/hubbub/dcn30/dcn30_hubbub.h | 6 ++ .../display/dc/hubbub/dcn31/dcn31_hubbub.c | 2 + .../display/dc/hubbub/dcn32/dcn32_hubbub.c | 2 + .../display/dc/hubbub/dcn35/dcn35_hubbub.c | 2 + .../display/dc/hubbub/dcn401/dcn401_hubbub.c | 2 + .../amd/display/dc/hubp/dcn10/dcn10_hubp.h | 9 ++- .../amd/display/dc/hubp/dcn20/dcn20_hubp.h | 1 + .../amd/display/dc/hubp/dcn30/dcn30_hubp.c | 26 ++++++ .../amd/display/dc/hubp/dcn30/dcn30_hubp.h | 8 +- .../amd/display/dc/hubp/dcn31/dcn31_hubp.c | 15 ++++ .../amd/display/dc/hubp/dcn31/dcn31_hubp.h | 6 +- .../amd/display/dc/hubp/dcn32/dcn32_hubp.c | 3 + .../amd/display/dc/hubp/dcn35/dcn35_hubp.c | 3 + .../amd/display/dc/hubp/dcn401/dcn401_hubp.c | 41 ++++++++++ .../amd/display/dc/hubp/dcn401/dcn401_hubp.h | 8 +- .../amd/display/dc/hwss/dcn10/dcn10_hwseq.c | 63 +++++++++++---- .../amd/display/dc/hwss/dcn30/dcn30_hwseq.c | 48 +++++++++++ .../amd/display/dc/hwss/dcn30/dcn30_hwseq.h | 5 ++ .../amd/display/dc/hwss/dcn30/dcn30_init.c | 1 + .../amd/display/dc/hwss/dcn31/dcn31_init.c | 1 + .../amd/display/dc/hwss/dcn314/dcn314_init.c | 1 + .../amd/display/dc/hwss/dcn32/dcn32_init.c | 1 + .../amd/display/dc/hwss/dcn35/dcn35_init.c | 1 + .../amd/display/dc/hwss/dcn351/dcn351_init.c | 1 + .../amd/display/dc/hwss/dcn401/dcn401_hwseq.c | 36 +++++---- .../amd/display/dc/hwss/dcn401/dcn401_init.c | 1 + .../drm/amd/display/dc/hwss/hw_sequencer.h | 4 + .../gpu/drm/amd/display/dc/inc/hw/dchubbub.h | 2 + drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h | 21 ++++- drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h | 11 +++ .../link/protocols/link_edp_panel_control.c | 2 +- .../dc/resource/dcn32/dcn32_resource.h | 3 +- .../dc/resource/dcn401/dcn401_resource.c | 1 + .../dc/resource/dcn401/dcn401_resource.h | 3 +- drivers/gpu/drm/amd/display/dmub/dmub_srv.h | 1 + .../gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 3 +- .../gpu/drm/amd/display/dmub/src/dmub_dcn31.c | 1 + 45 files changed, 545 insertions(+), 67 deletions(-)
--
2.34.1