* [REGRESSION 6.14][BISECTED] AMD Carrizo shutdown hang since amdgpu commit ff69bba05f08
@ 2025-02-10 18:15 Michał Pecio
2025-02-10 23:28 ` Alex Deucher
0 siblings, 1 reply; 3+ messages in thread
From: Michał Pecio @ 2025-02-10 18:15 UTC (permalink / raw)
To: amd-gfx; +Cc: regressions, Kenneth Feng, Alex Deucher, Boyuan Zhang
Hi,
I have a machine with this APU:
AMD PRO A10-8770 R7, 10 COMPUTE CORES 4C+6G
6.13 was OK, with 6.14-rc2 it hangs on shutdown after the display
turns off.
git-bisect says:
# first bad commit: [ff69bba05f085cd6d4277c27ac7600160167b384] drm/amd/pm: add inst to dpm_set_powergating_by_smu
The preceding commit does indeed shutdown normally.
AMDGPU is built into the kernel image, not sure if it matters.
Final lines of dynamic debug over serial pasted below. Any ideas?
Thanks,
Michal
[ 79.232113] [SURFACE]:Pipe 0: width, height, x, y
[ 79.232113] viewport:1600, 1200, 0, 0
[ 79.232113] recout: 1600, 1200, 0, 0
[ 79.251418] amdgpu: [IPS]: dc_allow_idle_optimizations_internal: allow_idle old=0 new=1 (caller=amdgpu_dm_crtc_vblank_control_worker)
[ 79.253727] [SCALER]:resource_build_scaling_params pipe 0:
[ 79.253727] Viewport: height:1200 width:1600 x:0 y:0 Recout: height:1200 width:1600 x:0 y:0 HACTIVE:1600 VACTIVE:1200
[ 79.253727] src_rect: height:1200 width:1600 x:0 y:0 dst_rect: height:1200 width:1600 x:0 y:0 clip_rect: height:1200 width:1600 x:0 y:0
[ 79.303043] [SCALER]:resource_build_scaling_params pipe 0:
[ 79.303043] Viewport: height:1200 width:1600 x:0 y:0 Recout: height:1200 width:1600 x:0 y:0 HACTIVE:1600 VACTIVE:1200
[ 79.303043] src_rect: height:1200 width:1600 x:0 y:0 dst_rect: height:1200 width:1600 x:0 y:0 clip_rect: height:1200 width:1600 x:0 y:0
[ 79.339233] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: start
[ 79.339691] [SCALER]:resource_build_scaling_params pipe 0:
[ 79.339691] Viewport: height:1200 width:1600 x:0 y:0 Recout: height:1200 width:1600 x:0 y:0 HACTIVE:1600 VACTIVE:1200
[ 79.339691] src_rect: height:1200 width:1600 x:0 y:0 dst_rect: height:1200 width:1600 x:0 y:0 clip_rect: height:1200 width:1600 x:0 y:0
[ 79.384264] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: start
[ 79.384601] [SURFACE]:Pipe:0 ffff888104900000: addr hi:0xff, addr low:0xed3000, src: 0, 0, 1600, 1200; dst: 0, 0, 1600, 1200;clip: 0, 0, 1600, 1200
[ 79.408652] [SURFACE]:Pipe 0: width, height, x, y
[ 79.408652] viewport:1600, 1200, 0, 0
[ 79.408652] recout: 1600, 1200, 0, 0
[ 79.428580] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: start
[ 79.428597] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: finish,
[ 79.428597] nbpMark_b: 0 nbpMark_a: 31870 urgentMark_b: 0 urgentMark_a: 16615
[ 79.428597] stutMark_b: 0 stutMark_a: 27915
[ 79.428597] nbpMark_b: 0 nbpMark_a: 0 urgentMark_b: 0 urgentMark_a: 0
[ 79.428597] stutMark_b: 0 stutMark_a: 0
[ 79.428597] nbpMark_b: 0 nbpMark_a: 0 urgentMark_b: 0 urgentMark_a: 0
[ 79.428597] stutMark_b: 0 stutMark_a: 0 stutter_mode_enable: 1
[ 79.428597] cstate: 1 pstate: 1 nbpstate: 1 sync: 0 dispclk: 0
[ 79.428597] sclk: 0 sclk_sleep: 12225 yclk: 2668000 blackout_recovery_time_us: 0
[ 79.429071] amdgpu: [IPS]: dc_allow_idle_optimizations_internal: allow_idle old=0 new=1 (caller=amdgpu_dm_crtc_vblank_control_worker)
[ 79.527445] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: start
[ 79.527463] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: finish,
[ 79.527463] nbpMark_b: 0 nbpMark_a: 31870 urgentMark_b: 0 urgentMark_a: 16615
[ 79.527463] stutMark_b: 0 stutMark_a: 27915
[ 79.527463] nbpMark_b: 0 nbpMark_a: 0 urgentMark_b: 0 urgentMark_a: 0
[ 79.527463] stutMark_b: 0 stutMark_a: 0
[ 79.527463] nbpMark_b: 0 nbpMark_a: 0 urgentMark_b: 0 urgentMark_a: 0
[ 79.527463] stutMark_b: 0 stutMark_a: 0 stutter_mode_enable: 1
[ 79.527463] cstate: 1 pstate: 1 nbpstate: 1 sync: 0 dispclk: 0
[ 79.527463] sclk: 0 sclk_sleep: 12225 yclk: 2668000 blackout_recovery_time_us: 0
[ 79.609435] amdgpu: Setting Deep Sleep Clock: 1222
[ 79.616505] amdgpu: enable Low Memory PState.
[ 79.708792] amdgpu: Setting Deep Sleep Clock: 1222
[ 79.715842] amdgpu: enable Low Memory PState.
[ 79.722763] amdgpu 0000:00:01.0: amdgpu: IP block9 already in the target gate state!
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [REGRESSION 6.14][BISECTED] AMD Carrizo shutdown hang since amdgpu commit ff69bba05f08
2025-02-10 18:15 [REGRESSION 6.14][BISECTED] AMD Carrizo shutdown hang since amdgpu commit ff69bba05f08 Michał Pecio
@ 2025-02-10 23:28 ` Alex Deucher
2025-02-11 0:23 ` Michał Pecio
0 siblings, 1 reply; 3+ messages in thread
From: Alex Deucher @ 2025-02-10 23:28 UTC (permalink / raw)
To: Michał Pecio
Cc: amd-gfx, regressions, Kenneth Feng, Alex Deucher, Boyuan Zhang
[-- Attachment #1: Type: text/plain, Size: 4240 bytes --]
On Mon, Feb 10, 2025 at 6:03 PM Michał Pecio <michal.pecio@gmail.com> wrote:
>
> Hi,
>
> I have a machine with this APU:
> AMD PRO A10-8770 R7, 10 COMPUTE CORES 4C+6G
>
> 6.13 was OK, with 6.14-rc2 it hangs on shutdown after the display
> turns off.
>
> git-bisect says:
> # first bad commit: [ff69bba05f085cd6d4277c27ac7600160167b384] drm/amd/pm: add inst to dpm_set_powergating_by_smu
>
> The preceding commit does indeed shutdown normally.
> AMDGPU is built into the kernel image, not sure if it matters.
>
> Final lines of dynamic debug over serial pasted below. Any ideas?
Does the attached patch fix it?
Alex
>
> Thanks,
> Michal
>
> [ 79.232113] [SURFACE]:Pipe 0: width, height, x, y
> [ 79.232113] viewport:1600, 1200, 0, 0
> [ 79.232113] recout: 1600, 1200, 0, 0
> [ 79.251418] amdgpu: [IPS]: dc_allow_idle_optimizations_internal: allow_idle old=0 new=1 (caller=amdgpu_dm_crtc_vblank_control_worker)
> [ 79.253727] [SCALER]:resource_build_scaling_params pipe 0:
> [ 79.253727] Viewport: height:1200 width:1600 x:0 y:0 Recout: height:1200 width:1600 x:0 y:0 HACTIVE:1600 VACTIVE:1200
> [ 79.253727] src_rect: height:1200 width:1600 x:0 y:0 dst_rect: height:1200 width:1600 x:0 y:0 clip_rect: height:1200 width:1600 x:0 y:0
> [ 79.303043] [SCALER]:resource_build_scaling_params pipe 0:
> [ 79.303043] Viewport: height:1200 width:1600 x:0 y:0 Recout: height:1200 width:1600 x:0 y:0 HACTIVE:1600 VACTIVE:1200
> [ 79.303043] src_rect: height:1200 width:1600 x:0 y:0 dst_rect: height:1200 width:1600 x:0 y:0 clip_rect: height:1200 width:1600 x:0 y:0
> [ 79.339233] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: start
> [ 79.339691] [SCALER]:resource_build_scaling_params pipe 0:
> [ 79.339691] Viewport: height:1200 width:1600 x:0 y:0 Recout: height:1200 width:1600 x:0 y:0 HACTIVE:1600 VACTIVE:1200
> [ 79.339691] src_rect: height:1200 width:1600 x:0 y:0 dst_rect: height:1200 width:1600 x:0 y:0 clip_rect: height:1200 width:1600 x:0 y:0
> [ 79.384264] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: start
> [ 79.384601] [SURFACE]:Pipe:0 ffff888104900000: addr hi:0xff, addr low:0xed3000, src: 0, 0, 1600, 1200; dst: 0, 0, 1600, 1200;clip: 0, 0, 1600, 1200
> [ 79.408652] [SURFACE]:Pipe 0: width, height, x, y
> [ 79.408652] viewport:1600, 1200, 0, 0
> [ 79.408652] recout: 1600, 1200, 0, 0
> [ 79.428580] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: start
> [ 79.428597] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: finish,
> [ 79.428597] nbpMark_b: 0 nbpMark_a: 31870 urgentMark_b: 0 urgentMark_a: 16615
> [ 79.428597] stutMark_b: 0 stutMark_a: 27915
> [ 79.428597] nbpMark_b: 0 nbpMark_a: 0 urgentMark_b: 0 urgentMark_a: 0
> [ 79.428597] stutMark_b: 0 stutMark_a: 0
> [ 79.428597] nbpMark_b: 0 nbpMark_a: 0 urgentMark_b: 0 urgentMark_a: 0
> [ 79.428597] stutMark_b: 0 stutMark_a: 0 stutter_mode_enable: 1
> [ 79.428597] cstate: 1 pstate: 1 nbpstate: 1 sync: 0 dispclk: 0
> [ 79.428597] sclk: 0 sclk_sleep: 12225 yclk: 2668000 blackout_recovery_time_us: 0
> [ 79.429071] amdgpu: [IPS]: dc_allow_idle_optimizations_internal: allow_idle old=0 new=1 (caller=amdgpu_dm_crtc_vblank_control_worker)
> [ 79.527445] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: start
> [ 79.527463] [BANDWIDTH_CALCS]:dce110_validate_bandwidth: finish,
> [ 79.527463] nbpMark_b: 0 nbpMark_a: 31870 urgentMark_b: 0 urgentMark_a: 16615
> [ 79.527463] stutMark_b: 0 stutMark_a: 27915
> [ 79.527463] nbpMark_b: 0 nbpMark_a: 0 urgentMark_b: 0 urgentMark_a: 0
> [ 79.527463] stutMark_b: 0 stutMark_a: 0
> [ 79.527463] nbpMark_b: 0 nbpMark_a: 0 urgentMark_b: 0 urgentMark_a: 0
> [ 79.527463] stutMark_b: 0 stutMark_a: 0 stutter_mode_enable: 1
> [ 79.527463] cstate: 1 pstate: 1 nbpstate: 1 sync: 0 dispclk: 0
> [ 79.527463] sclk: 0 sclk_sleep: 12225 yclk: 2668000 blackout_recovery_time_us: 0
> [ 79.609435] amdgpu: Setting Deep Sleep Clock: 1222
> [ 79.616505] amdgpu: enable Low Memory PState.
> [ 79.708792] amdgpu: Setting Deep Sleep Clock: 1222
> [ 79.715842] amdgpu: enable Low Memory PState.
> [ 79.722763] amdgpu 0000:00:01.0: amdgpu: IP block9 already in the target gate state!
[-- Attachment #2: 0001-drm-amdgpu-pm-fix-UVD-handing-in-amdgpu_dpm_set_powe.patch --]
[-- Type: text/x-patch, Size: 1611 bytes --]
From 70210c5a79b4840967b26cb97abbc09b87d2ae75 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Mon, 10 Feb 2025 18:18:29 -0500
Subject: [PATCH] drm/amdgpu/pm: fix UVD handing in
amdgpu_dpm_set_powergating_by_smu()
UVD and VCN were split into separate dpm helpers in commit
ff69bba05f08 ("drm/amd/pm: add inst to dpm_set_powergating_by_smu")
but the check at the top was still included UVD from an earlier
version of the patch. Fix the check.
Fixes: ff69bba05f08 ("drm/amd/pm: add inst to dpm_set_powergating_by_smu")
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3959
Link: https://lists.freedesktop.org/archives/amd-gfx/2025-February/119827.html
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Boyuan Zhang <boyuan.zhang@amd.com>
---
drivers/gpu/drm/amd/pm/amdgpu_dpm.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c
index 6a9e26905edfc..7a22aef6e59c3 100644
--- a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c
+++ b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c
@@ -78,7 +78,7 @@ int amdgpu_dpm_set_powergating_by_smu(struct amdgpu_device *adev,
int ret = 0;
const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
enum ip_power_state pwr_state = gate ? POWER_STATE_OFF : POWER_STATE_ON;
- bool is_vcn = (block_type == AMD_IP_BLOCK_TYPE_UVD || block_type == AMD_IP_BLOCK_TYPE_VCN);
+ bool is_vcn = block_type == AMD_IP_BLOCK_TYPE_VCN;
if (atomic_read(&adev->pm.pwr_state[block_type]) == pwr_state &&
(!is_vcn || adev->vcn.num_vcn_inst == 1)) {
--
2.48.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [REGRESSION 6.14][BISECTED] AMD Carrizo shutdown hang since amdgpu commit ff69bba05f08
2025-02-10 23:28 ` Alex Deucher
@ 2025-02-11 0:23 ` Michał Pecio
0 siblings, 0 replies; 3+ messages in thread
From: Michał Pecio @ 2025-02-11 0:23 UTC (permalink / raw)
To: Alex Deucher
Cc: amd-gfx, regressions, Kenneth Feng, Alex Deucher, Boyuan Zhang
On Mon, 10 Feb 2025 18:28:16 -0500, Alex Deucher wrote:
> Does the attached patch fix it?
Yep, it prints a few more lines and shuts down now.
Thanks!
[ 53.525125] cstate: 1 pstate: 1 nbpstate: 1 sync: 0 dispclk: 0
[ 53.525125] sclk: 0 sclk_sleep: 12225 yclk: 2668000 blackout_recovery_time_us: 0
[ 53.605942] amdgpu: Setting Deep Sleep Clock: 1222
[ 53.613220] amdgpu: enable Low Memory PState.
[ 53.707787] amdgpu: Setting Deep Sleep Clock: 1222
[ 53.714787] amdgpu: enable Low Memory PState.
[ 53.721659] amdgpu 0000:00:01.0: amdgpu: IP block9 already in the target gate state!
[ 53.732460] amdgpu 0000:00:01.0: amdgpu: IP block8 already in the target gate state!
[ 53.742564] amdgpu: Setting Deep Sleep Clock: 1222
[ 53.750680] amdgpu: enable Low Memory PState.
[ 53.759500] amdgpu: enable Low Memory PState.
[ 53.767937] ACPI: PM: Preparing to enter system sleep state S5
[ 53.777664] ACPI Warning: Time parameter 255 us > 100 us violating ACPI spec, please fix the firmware. (20240827/exsystem-142)
[ 53.792988] kvm: exiting hardware virtualization
[ 53.801592] reboot: Power down
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2025-02-11 7:57 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-02-10 18:15 [REGRESSION 6.14][BISECTED] AMD Carrizo shutdown hang since amdgpu commit ff69bba05f08 Michał Pecio
2025-02-10 23:28 ` Alex Deucher
2025-02-11 0:23 ` Michał Pecio
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.