[openchrome-devel] drm-openchrome: Branch 'drm-next-6.4' - 53 commits - drivers/gpu/drm
Kevin Brace
kevinbrace at kemper.freedesktop.org
Fri May 19 02:00:06 UTC 2023
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 13
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 39 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c | 6
drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c | 1
drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c | 1
drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 1
drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c | 2
drivers/gpu/drm/amd/amdgpu/nv.c | 23 -
drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c | 5
drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c | 2
drivers/gpu/drm/amd/amdgpu/soc15.c | 25 -
drivers/gpu/drm/amd/amdgpu/soc21.c | 23 -
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 34 +
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c | 1
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c | 17
drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c | 2
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c | 5
drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 3
drivers/gpu/drm/amd/display/dc/dc.h | 1
drivers/gpu/drm/amd/display/dc/dc_stream.h | 2
drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c | 19 -
drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c | 9
drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c | 25 +
drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c | 4
drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c | 4
drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c | 2
drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c | 2
drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c | 1
drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c | 56 +--
drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c | 10
drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c | 178 +++++-----
drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c | 18 -
drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c | 4
drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c | 17
drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h | 2
drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c | 24 -
drivers/gpu/drm/amd/display/dc/link/link_dpms.c | 5
drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c | 5
drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c | 1
drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c | 3
drivers/gpu/drm/amd/display/include/signal_types.h | 1
drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h | 4
drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c | 3
drivers/gpu/drm/i915/display/icl_dsi.c | 2
drivers/gpu/drm/i915/display/intel_dsi_vbt.c | 11
drivers/gpu/drm/i915/display/intel_dsi_vbt.h | 1
drivers/gpu/drm/i915/display/skl_scaler.c | 17
drivers/gpu/drm/i915/display/vlv_dsi.c | 22 -
drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c | 20 -
drivers/gpu/drm/i915/i915_pci.c | 2
50 files changed, 415 insertions(+), 263 deletions(-)
New commits:
commit dee1b5f74550a1bc6bbc99c6aaf1e363fdab4b6b
Merge: efa0ff28cf70 1bef84af084e
Author: Kevin Brace <kevinbrace at bracecomputerlab.com>
Date: Thu May 18 18:55:26 2023 -0700
Merge tag 'drm-next-2023-05-05' of git://anongit.freedesktop.org/drm/drm into drm-next-6.4
drm fixes part 2 for 6.4-rc1
amdgpu:
- SR-IOV fixes
- DCN 3.2 fixes
- DC mclk handling fixes
- eDP fixes
- SubVP fixes
- HDCP regression fix
- DSC fixes
- DC FP fixes
- DCN 3.x fixes
- Display flickering fix when switching between vram and gtt
- Z8 power saving fix
- Fix hang when skipping modeset
- GPU reset fixes
- Doorbell fix when resizing BARs
- Fix spurious warnings in gmc
- Locking fix for AMDGPU_SCHED IOCTL
- SR-IOV fix
- DCN 3.1.4 fix
- DCN 3.2 fix
- Fix job cleanup when CS is aborted
i915:
- skl pipe source size check
- mtl transcoder mask fix
- DSI power on sequence fix
- GuC versioning corner case fix
commit 1bef84af084e981550d9ecc3359baa22533d7b99
Merge: f4c41a7fd7f9 c8c2969bfcba
Author: Dave Airlie <airlied at redhat.com>
Date: Fri May 5 11:27:25 2023 +1000
Merge tag 'drm-intel-next-fixes-2023-05-04-1' of git://anongit.freedesktop.org/drm/drm-intel into drm-next
Add missing GPU transcoder masks for MTL and fix DSI power on sequence
for Nextbook Ares 8A. Fix GuC version corner case.
Signed-off-by: Dave Airlie <airlied at redhat.com>
From: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/ZFOskabVuN45dNaA@jlahtine-mobl.ger.corp.intel.com
commit f4c41a7fd7f99329e5af0ac0a236504a60bfb17c
Merge: fa0d9c066dee 1253685f0d3e
Author: Dave Airlie <airlied at redhat.com>
Date: Fri May 5 11:27:08 2023 +1000
Merge tag 'amd-drm-fixes-6.4-2023-05-03' of https://gitlab.freedesktop.org/agd5f/linux into drm-next
amd-drm-fixes-6.4-2023-05-03:
amdgpu:
- GPU reset fixes
- Doorbell fix when resizing BARs
- Fix spurious warnings in gmc
- Locking fix for AMDGPU_SCHED IOCTL
- SR-IOV fix
- DCN 3.1.4 fix
- DCN 3.2 fix
- Fix job cleanup when CS is aborted
Signed-off-by: Dave Airlie <airlied at redhat.com>
From: Alex Deucher <alexander.deucher at amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230504034018.7950-1-alexander.deucher@amd.com
commit fa0d9c066dee8f52eabcb8416459aa0568b832f9
Merge: 6f5a5e867058 d944eafed618
Author: Dave Airlie <airlied at redhat.com>
Date: Fri May 5 11:23:11 2023 +1000
Merge tag 'drm-intel-next-fixes-2023-04-27' of git://anongit.freedesktop.org/drm/drm-intel into drm-next
One cc stable for pipe source size check on SKL+
Signed-off-by: Dave Airlie <airlied at redhat.com>
From: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/ZEpbSG1ZOSVqzGLx@jlahtine-mobl.ger.corp.intel.com
commit 6f5a5e8670587d5066aacd0235071a166ee458fc
Merge: cf03e2956af3 d893f39320e1
Author: Dave Airlie <airlied at redhat.com>
Date: Fri May 5 09:13:21 2023 +1000
Merge tag 'amd-drm-fixes-6.4-2023-04-26' of https://gitlab.freedesktop.org/agd5f/linux into drm-next
amd-drm-fixes-6.4-2023-04-26:
amdgpu:
- SR-IOV fixes
- DCN 3.2 fixes
- DC mclk handling fixes
- eDP fixes
- SubVP fixes
- HDCP regression fix
- DSC fixes
- DC FP fixes
- DCN 3.x fixes
- Display flickering fix when switching between vram and gtt
- Z8 power saving fix
- Fix hang when skipping modeset
Signed-off-by: Dave Airlie <airlied at redhat.com>
From: Alex Deucher <alexander.deucher at amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230427033012.7668-1-alexander.deucher@amd.com
commit 1253685f0d3eb3eab0bfc4bf15ab341a5f3da0c8
Author: Guchun Chen <guchun.chen at amd.com>
Date: Wed Apr 26 09:46:54 2023 +0800
drm/amdgpu: drop redundant sched job cleanup when cs is aborted
Once command submission failed due to userptr invalidation in
amdgpu_cs_submit, legacy code will perform cleanup of scheduler
job. However, it's not needed at all, as former commit has integrated
job cleanup stuff into amdgpu_job_free. Otherwise, because of double
free, a NULL pointer dereference will occur in such scenario.
Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/2457
Fixes: f7d66fb2ea43 ("drm/amdgpu: cleanup scheduler job initialization v2")
Signed-off-by: Guchun Chen <guchun.chen at amd.com>
Reviewed-by: Christian König <christian.koenig at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
Cc: stable at vger.kernel.org
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 08eced097bd8..2eb2c66843a8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -1276,7 +1276,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
r = drm_sched_job_add_dependency(&leader->base, fence);
if (r) {
dma_fence_put(fence);
- goto error_cleanup;
+ return r;
}
}
@@ -1303,7 +1303,8 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
}
if (r) {
r = -EAGAIN;
- goto error_unlock;
+ mutex_unlock(&p->adev->notifier_lock);
+ return r;
}
p->fence = dma_fence_get(&leader->base.s_fence->finished);
@@ -1350,14 +1351,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
mutex_unlock(&p->adev->notifier_lock);
mutex_unlock(&p->bo_list->bo_list_mutex);
return 0;
-
-error_unlock:
- mutex_unlock(&p->adev->notifier_lock);
-
-error_cleanup:
- for (i = 0; i < p->gang_size; ++i)
- drm_sched_job_cleanup(&p->jobs[i]->base);
- return r;
}
/* Cleanup the parser structure */
commit 682439fffad9fa9a38d37dd1b1318e9374232213
Author: Samson Tam <Samson.Tam at amd.com>
Date: Wed Apr 19 18:17:14 2023 -0400
drm/amd/display: filter out invalid bits in pipe_fuses
[Why]
Reading pipe_fuses from register may have invalid bits set, which may
affect the num_pipes erroneously.
[How]
Add read_pipes_fuses() call and filter bits based on expected number
of pipes.
Reviewed-by: Alvin Lee <Alvin.Lee2 at amd.com>
Acked-by: Alan Liu <HaoPing.Liu at amd.com>
Signed-off-by: Samson Tam <Samson.Tam at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
Cc: stable at vger.kernel.org # 6.1.x
diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
index a876e6eb6cd8..22dd1ebea618 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
@@ -2079,6 +2079,14 @@ static struct resource_funcs dcn32_res_pool_funcs = {
.restore_mall_state = dcn32_restore_mall_state,
};
+static uint32_t read_pipe_fuses(struct dc_context *ctx)
+{
+ uint32_t value = REG_READ(CC_DC_PIPE_DIS);
+ /* DCN32 support max 4 pipes */
+ value = value & 0xf;
+ return value;
+}
+
static bool dcn32_resource_construct(
uint8_t num_virtual_links,
@@ -2122,7 +2130,7 @@ static bool dcn32_resource_construct(
pool->base.res_cap = &res_cap_dcn32;
/* max number of pipes for ASIC before checking for pipe fuses */
num_pipes = pool->base.res_cap->num_timing_generator;
- pipe_fuses = REG_READ(CC_DC_PIPE_DIS);
+ pipe_fuses = read_pipe_fuses(ctx);
for (i = 0; i < pool->base.res_cap->num_timing_generator; i++)
if (pipe_fuses & 1 << i)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c
index e5ab7f3077c4..a60ddb343d13 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c
@@ -1632,6 +1632,14 @@ static struct resource_funcs dcn321_res_pool_funcs = {
.restore_mall_state = dcn32_restore_mall_state,
};
+static uint32_t read_pipe_fuses(struct dc_context *ctx)
+{
+ uint32_t value = REG_READ(CC_DC_PIPE_DIS);
+ /* DCN321 support max 4 pipes */
+ value = value & 0xf;
+ return value;
+}
+
static bool dcn321_resource_construct(
uint8_t num_virtual_links,
@@ -1674,7 +1682,7 @@ static bool dcn321_resource_construct(
pool->base.res_cap = &res_cap_dcn321;
/* max number of pipes for ASIC before checking for pipe fuses */
num_pipes = pool->base.res_cap->num_timing_generator;
- pipe_fuses = REG_READ(CC_DC_PIPE_DIS);
+ pipe_fuses = read_pipe_fuses(ctx);
for (i = 0; i < pool->base.res_cap->num_timing_generator; i++)
if (pipe_fuses & 1 << i)
commit 8f586cc16c1fc3c2202c9d54563db8c7ed365f82
Author: Leo Chen <sancchen at amd.com>
Date: Thu Apr 13 17:34:24 2023 -0400
drm/amd/display: Change default Z8 watermark values
[Why & How]
Previous Z8 watermark values were causing flickering and OTC underflow.
Updating Z8 watermark values based on the measurement.
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas at amd.com>
Cc: Mario Limonciello <mario.limonciello at amd.com>
Cc: Alex Deucher <alexander.deucher at amd.com>
Cc: stable at vger.kernel.org
Acked-by: Alan Liu <HaoPing.Liu at amd.com>
Signed-off-by: Leo Chen <sancchen at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c
index 44082f65de1f..9e54e3d0eb78 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c
@@ -149,8 +149,8 @@ static struct _vcs_dpi_soc_bounding_box_st dcn3_14_soc = {
.num_states = 5,
.sr_exit_time_us = 16.5,
.sr_enter_plus_exit_time_us = 18.5,
- .sr_exit_z8_time_us = 210.0,
- .sr_enter_plus_exit_z8_time_us = 310.0,
+ .sr_exit_z8_time_us = 268.0,
+ .sr_enter_plus_exit_z8_time_us = 393.0,
.writeback_latency_us = 12.0,
.dram_channel_width_bytes = 4,
.round_trip_ping_latency_dcfclk_cycles = 106,
commit 100bd00881f8553d0ccfc99a575966d990c455eb
Author: Horace Chen <horace.chen at amd.com>
Date: Tue Apr 25 13:15:32 2023 +0800
drm/amdgpu: disable SDMA WPTR_POLL_ENABLE for SR-IOV
[Why]
This WPTR_POLL_ENABLE is a hardware contigious polling which will cause
FCLK and UCLK to keep on a high level.
Mostly its case can be covered by F32_WPTR_POLL_ENABLE which polls by
firmware.
So to save power, SR-IOV also needs to disable this bit
Signed-off-by: Horace Chen <horace.chen at amd.com>
Acked-by: Alex Deucher <alexander.deucher at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
index eb722830531f..3d9a80511a45 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
@@ -510,10 +510,7 @@ static int sdma_v6_0_gfx_resume(struct amdgpu_device *adev)
lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC);
rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1);
- if (amdgpu_sriov_vf(adev))
- rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, WPTR_POLL_ENABLE, 1);
- else
- rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, WPTR_POLL_ENABLE, 0);
+ rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, WPTR_POLL_ENABLE, 0);
rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, F32_WPTR_POLL_ENABLE, 1);
WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_BASE), ring->gpu_addr >> 8);
commit 2397e3d8d2e120355201a8310b61929f5a8bd2c0
Author: Chia-I Wu <olvaffe at gmail.com>
Date: Wed Apr 26 15:54:55 2023 -0700
drm/amdgpu: add a missing lock for AMDGPU_SCHED
mgr->ctx_handles should be protected by mgr->lock.
v2: improve commit message
v3: add a Fixes tag
Signed-off-by: Chia-I Wu <olvaffe at gmail.com>
Reviewed-by: Christian König <christian.koenig at amd.com>
Fixes: 52c6a62c64fa ("drm/amdgpu: add interface for editing a foreign process's priority v3")
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c
index e9b45089a28a..863b2a34b2d6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c
@@ -38,6 +38,7 @@ static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev,
{
struct fd f = fdget(fd);
struct amdgpu_fpriv *fpriv;
+ struct amdgpu_ctx_mgr *mgr;
struct amdgpu_ctx *ctx;
uint32_t id;
int r;
@@ -51,8 +52,11 @@ static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev,
return r;
}
- idr_for_each_entry(&fpriv->ctx_mgr.ctx_handles, ctx, id)
+ mgr = &fpriv->ctx_mgr;
+ mutex_lock(&mgr->lock);
+ idr_for_each_entry(&mgr->ctx_handles, ctx, id)
amdgpu_ctx_priority_override(ctx, priority);
+ mutex_unlock(&mgr->lock);
fdput(f);
return 0;
commit 922a76ba31adf84e72bc947267385be420c689ee
Author: Hamza Mahfooz <hamza.mahfooz at amd.com>
Date: Tue May 2 11:59:08 2023 -0400
drm/amdgpu: fix an amdgpu_irq_put() issue in gmc_v9_0_hw_fini()
As made mention of in commit 08c677cb0b43 ("drm/amdgpu: fix
amdgpu_irq_put call trace in gmc_v10_0_hw_fini") and commit 13af556104fa
("drm/amdgpu: fix amdgpu_irq_put call trace in gmc_v11_0_hw_fini"). It
is meaningless to call amdgpu_irq_put() for gmc.ecc_irq. So, remove it
from gmc_v9_0_hw_fini().
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2522
Fixes: 3029c855d79f ("drm/amdgpu: Fix desktop freezed after gpu-reset")
Reviewed-by: Mario Limonciello <mario.limonciello at amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
Cc: stable at vger.kernel.org
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index 64ab1a306dfe..2fe21cefd772 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -1999,7 +1999,6 @@ static int gmc_v9_0_hw_fini(void *handle)
if (adev->mmhub.funcs->update_power_gating)
adev->mmhub.funcs->update_power_gating(adev, false);
- amdgpu_irq_put(adev, &adev->gmc.ecc_irq, 0);
amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0);
return 0;
commit 08c677cb0b436a96a836792bb35a8ec5de4999c2
Author: Horatio Zhang <Hongkun.Zhang at amd.com>
Date: Tue Apr 25 10:52:28 2023 +0800
drm/amdgpu: fix amdgpu_irq_put call trace in gmc_v10_0_hw_fini
The gmc.ecc_irq is enabled by firmware per IFWI setting,
and the host driver is not privileged to enable/disable
the interrupt. So, it is meaningless to use the amdgpu_irq_put
function in gmc_v10_0_hw_fini, which also leads to the call
trace.
[ 82.340264] Call Trace:
[ 82.340265] <TASK>
[ 82.340269] gmc_v10_0_hw_fini+0x83/0xa0 [amdgpu]
[ 82.340447] gmc_v10_0_suspend+0xe/0x20 [amdgpu]
[ 82.340623] amdgpu_device_ip_suspend_phase2+0x127/0x1c0 [amdgpu]
[ 82.340789] amdgpu_device_ip_suspend+0x3d/0x80 [amdgpu]
[ 82.340955] amdgpu_device_pre_asic_reset+0xdd/0x2b0 [amdgpu]
[ 82.341122] amdgpu_device_gpu_recover.cold+0x4dd/0xbb2 [amdgpu]
[ 82.341359] amdgpu_debugfs_reset_work+0x4c/0x70 [amdgpu]
[ 82.341529] process_one_work+0x21d/0x3f0
[ 82.341535] worker_thread+0x1fa/0x3c0
[ 82.341538] ? process_one_work+0x3f0/0x3f0
[ 82.341540] kthread+0xff/0x130
[ 82.341544] ? kthread_complete_and_exit+0x20/0x20
[ 82.341547] ret_from_fork+0x22/0x30
Signed-off-by: Horatio Zhang <Hongkun.Zhang at amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang at amd.com>
Reviewed-by: Guchun Chen <guchun.chen at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2522
Fixes: c8b5a95b5709 ("drm/amdgpu: Fix desktop freezed after gpu-reset")
Cc: stable at vger.kernel.org
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
index 7d6f4a68f416..b213dcf8ca06 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
@@ -1143,7 +1143,6 @@ static int gmc_v10_0_hw_fini(void *handle)
return 0;
}
- amdgpu_irq_put(adev, &adev->gmc.ecc_irq, 0);
amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0);
return 0;
commit 13af556104fa93b1945c70bbf8a0a62cd2c92879
Author: Horatio Zhang <Hongkun.Zhang at amd.com>
Date: Tue Apr 25 13:16:32 2023 +0800
drm/amdgpu: fix amdgpu_irq_put call trace in gmc_v11_0_hw_fini
The gmc.ecc_irq is enabled by firmware per IFWI setting,
and the host driver is not privileged to enable/disable
the interrupt. So, it is meaningless to use the amdgpu_irq_put
function in gmc_v11_0_hw_fini, which also leads to the call
trace.
[ 102.980303] Call Trace:
[ 102.980303] <TASK>
[ 102.980304] gmc_v11_0_hw_fini+0x54/0x90 [amdgpu]
[ 102.980357] gmc_v11_0_suspend+0xe/0x20 [amdgpu]
[ 102.980409] amdgpu_device_ip_suspend_phase2+0x240/0x460 [amdgpu]
[ 102.980459] amdgpu_device_ip_suspend+0x3d/0x80 [amdgpu]
[ 102.980520] amdgpu_device_pre_asic_reset+0xd9/0x490 [amdgpu]
[ 102.980573] amdgpu_device_gpu_recover.cold+0x548/0xce6 [amdgpu]
[ 102.980687] amdgpu_debugfs_reset_work+0x4c/0x70 [amdgpu]
[ 102.980740] process_one_work+0x21f/0x3f0
[ 102.980741] worker_thread+0x200/0x3e0
[ 102.980742] ? process_one_work+0x3f0/0x3f0
[ 102.980743] kthread+0xfd/0x130
[ 102.980743] ? kthread_complete_and_exit+0x20/0x20
[ 102.980744] ret_from_fork+0x22/0x30
Signed-off-by: Horatio Zhang <Hongkun.Zhang at amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang at amd.com>
Reviewed-by: Guchun Chen <guchun.chen at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2522
Fixes: c8b5a95b5709 ("drm/amdgpu: Fix desktop freezed after gpu-reset")
Cc: stable at vger.kernel.org
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
index d809f2ed5600..d95f9fe8f1c5 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
@@ -951,7 +951,6 @@ static int gmc_v11_0_hw_fini(void *handle)
return 0;
}
- amdgpu_irq_put(adev, &adev->gmc.ecc_irq, 0);
amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0);
gmc_v11_0_gart_disable(adev);
commit b03f38b9bd90d9eb29951e56f5a4375984c8dffb
Author: Shane Xiao <shane.xiao at amd.com>
Date: Tue Apr 25 22:39:08 2023 +0800
drm/amdgpu: Enable doorbell selfring after resize FB BAR
[Why]
The selfring doorbell aperture will change when resize FB
BAR successfully during gmc sw init, we should reorder
the sequence of enabling doorbell selfring aperture.
[How]
Move enable_doorbell_selfring_aperture from *_common_hw_init
to *_common_late_init.
This fixes the potential issue that GPU ring its own
doorbell when this device is in translated mode when
iommu is on.
v2: Remove *_enable_doorbell_aperture functions (Christian)
v3: Add comments to note that why we need enable doorbell
selfring late (Christian)
Signed-off-by: Shane Xiao <shane.xiao at amd.com>
Signed-off-by: Aaron Liu <aaron.liu at amd.com>
Tested-by: Xiaomeng Hou <Xiaomeng.Hou at amd.com>
Reviewed-by: Christian K�nig <christian.koenig at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c
index 47420b403871..98c826f1f89b 100644
--- a/drivers/gpu/drm/amd/amdgpu/nv.c
+++ b/drivers/gpu/drm/amd/amdgpu/nv.c
@@ -531,13 +531,6 @@ static void nv_program_aspm(struct amdgpu_device *adev)
}
-static void nv_enable_doorbell_aperture(struct amdgpu_device *adev,
- bool enable)
-{
- adev->nbio.funcs->enable_doorbell_aperture(adev, enable);
- adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, enable);
-}
-
const struct amdgpu_ip_block_version nv_common_ip_block =
{
.type = AMD_IP_BLOCK_TYPE_COMMON,
@@ -999,6 +992,11 @@ static int nv_common_late_init(void *handle)
}
}
+ /* Enable selfring doorbell aperture late because doorbell BAR
+ * aperture will change if resize BAR successfully in gmc sw_init.
+ */
+ adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, true);
+
return 0;
}
@@ -1038,7 +1036,7 @@ static int nv_common_hw_init(void *handle)
if (adev->nbio.funcs->remap_hdp_registers && !amdgpu_sriov_vf(adev))
adev->nbio.funcs->remap_hdp_registers(adev);
/* enable the doorbell aperture */
- nv_enable_doorbell_aperture(adev, true);
+ adev->nbio.funcs->enable_doorbell_aperture(adev, true);
return 0;
}
@@ -1047,8 +1045,13 @@ static int nv_common_hw_fini(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
- /* disable the doorbell aperture */
- nv_enable_doorbell_aperture(adev, false);
+ /* Disable the doorbell aperture and selfring doorbell aperture
+ * separately in hw_fini because nv_enable_doorbell_aperture
+ * has been removed and there is no need to delay disabling
+ * selfring doorbell.
+ */
+ adev->nbio.funcs->enable_doorbell_aperture(adev, false);
+ adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, false);
return 0;
}
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index bc5dd80f10c1..6d15d5cd9e07 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -619,13 +619,6 @@ static void soc15_program_aspm(struct amdgpu_device *adev)
adev->nbio.funcs->program_aspm(adev);
}
-static void soc15_enable_doorbell_aperture(struct amdgpu_device *adev,
- bool enable)
-{
- adev->nbio.funcs->enable_doorbell_aperture(adev, enable);
- adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, enable);
-}
-
const struct amdgpu_ip_block_version vega10_common_ip_block =
{
.type = AMD_IP_BLOCK_TYPE_COMMON,
@@ -1125,6 +1118,11 @@ static int soc15_common_late_init(void *handle)
if (amdgpu_sriov_vf(adev))
xgpu_ai_mailbox_get_irq(adev);
+ /* Enable selfring doorbell aperture late because doorbell BAR
+ * aperture will change if resize BAR successfully in gmc sw_init.
+ */
+ adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, true);
+
return 0;
}
@@ -1182,7 +1180,8 @@ static int soc15_common_hw_init(void *handle)
adev->nbio.funcs->remap_hdp_registers(adev);
/* enable the doorbell aperture */
- soc15_enable_doorbell_aperture(adev, true);
+ adev->nbio.funcs->enable_doorbell_aperture(adev, true);
+
/* HW doorbell routing policy: doorbell writing not
* in SDMA/IH/MM/ACV range will be routed to CP. So
* we need to init SDMA doorbell range prior
@@ -1198,8 +1197,14 @@ static int soc15_common_hw_fini(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
- /* disable the doorbell aperture */
- soc15_enable_doorbell_aperture(adev, false);
+ /* Disable the doorbell aperture and selfring doorbell aperture
+ * separately in hw_fini because soc15_enable_doorbell_aperture
+ * has been removed and there is no need to delay disabling
+ * selfring doorbell.
+ */
+ adev->nbio.funcs->enable_doorbell_aperture(adev, false);
+ adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, false);
+
if (amdgpu_sriov_vf(adev))
xgpu_ai_mailbox_put_irq(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/soc21.c b/drivers/gpu/drm/amd/amdgpu/soc21.c
index 514bfc705d5a..744be2a05623 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc21.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc21.c
@@ -450,13 +450,6 @@ static void soc21_program_aspm(struct amdgpu_device *adev)
adev->nbio.funcs->program_aspm(adev);
}
-static void soc21_enable_doorbell_aperture(struct amdgpu_device *adev,
- bool enable)
-{
- adev->nbio.funcs->enable_doorbell_aperture(adev, enable);
- adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, enable);
-}
-
const struct amdgpu_ip_block_version soc21_common_ip_block =
{
.type = AMD_IP_BLOCK_TYPE_COMMON,
@@ -764,6 +757,11 @@ static int soc21_common_late_init(void *handle)
amdgpu_irq_get(adev, &adev->nbio.ras_err_event_athub_irq, 0);
}
+ /* Enable selfring doorbell aperture late because doorbell BAR
+ * aperture will change if resize BAR successfully in gmc sw_init.
+ */
+ adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, true);
+
return 0;
}
@@ -797,7 +795,7 @@ static int soc21_common_hw_init(void *handle)
if (adev->nbio.funcs->remap_hdp_registers)
adev->nbio.funcs->remap_hdp_registers(adev);
/* enable the doorbell aperture */
- soc21_enable_doorbell_aperture(adev, true);
+ adev->nbio.funcs->enable_doorbell_aperture(adev, true);
return 0;
}
@@ -806,8 +804,13 @@ static int soc21_common_hw_fini(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
- /* disable the doorbell aperture */
- soc21_enable_doorbell_aperture(adev, false);
+ /* Disable the doorbell aperture and selfring doorbell aperture
+ * separately in hw_fini because soc21_enable_doorbell_aperture
+ * has been removed and there is no need to delay disabling
+ * selfring doorbell.
+ */
+ adev->nbio.funcs->enable_doorbell_aperture(adev, false);
+ adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, false);
if (amdgpu_sriov_vf(adev)) {
xgpu_nv_mailbox_put_irq(adev);
commit 4eea7fb980dc44545a32eec92e2662053b34cd9d
Author: lyndonli <Lyndon.Li at amd.com>
Date: Sun Apr 23 17:05:15 2023 +0800
drm/amdgpu: Use the default reset when loading or reloading the driver
Below call trace and errors are observed when reloading
amdgpu driver with the module parameter reset_method=3.
It should do a default reset when loading or reloading the
driver, regardless of the module parameter reset_method.
v2: add comments inside and modify commit messages.
[ +2.180243] [drm] psp gfx command ID_LOAD_TOC(0x20) failed
and response status is (0x0)
[ +0.000011] [drm:psp_hw_start [amdgpu]] *ERROR* Failed to load toc
[ +0.000890] [drm:psp_hw_start [amdgpu]] *ERROR* PSP tmr init failed!
[ +0.020683] [drm:amdgpu_fill_buffer [amdgpu]] *ERROR* Trying to
clear memory with ring turned off.
[ +0.000003] RIP: 0010:amdgpu_bo_release_notify+0x1ef/0x210 [amdgpu]
[ +0.000004] Call Trace:
[ +0.000003] <TASK>
[ +0.000008] ttm_bo_release+0x2c4/0x330 [amdttm]
[ +0.000026] amdttm_bo_put+0x3c/0x70 [amdttm]
[ +0.000020] amdgpu_bo_free_kernel+0xe6/0x140 [amdgpu]
[ +0.000728] psp_v11_0_ring_destroy+0x34/0x60 [amdgpu]
[ +0.000826] psp_hw_init+0xe7/0x2f0 [amdgpu]
[ +0.000813] amdgpu_device_fw_loading+0x1ad/0x2d0 [amdgpu]
[ +0.000731] amdgpu_device_init.cold+0x108e/0x2002 [amdgpu]
[ +0.001071] ? do_pci_enable_device+0xe1/0x110
[ +0.000011] amdgpu_driver_load_kms+0x1a/0x160 [amdgpu]
[ +0.000729] amdgpu_pci_probe+0x179/0x3a0 [amdgpu]
Signed-off-by: lyndonli <Lyndon.Li at amd.com>
Signed-off-by: Yunxiang Li <Yunxiang.Li at amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu at amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 9b1eaba85bbd..981a9cfb63b5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -3578,6 +3578,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
int r, i;
bool px = false;
u32 max_MBps;
+ int tmp;
adev->shutdown = false;
adev->flags = flags;
@@ -3799,7 +3800,13 @@ int amdgpu_device_init(struct amdgpu_device *adev,
}
}
} else {
+ tmp = amdgpu_reset_method;
+ /* It should do a default reset when loading or reloading the driver,
+ * regardless of the module parameter reset_method.
+ */
+ amdgpu_reset_method = AMD_RESET_METHOD_NONE;
r = amdgpu_asic_reset(adev);
+ amdgpu_reset_method = tmp;
if (r) {
dev_err(adev->dev, "asic reset on init failed\n");
goto failed;
commit 74a49415144035f171751d55b11ba04c9f348f9f
Author: lyndonli <Lyndon.Li at amd.com>
Date: Sun Apr 23 16:46:30 2023 +0800
drm/amdgpu: Fix mode2 reset for sienna cichlid
Before this change, sienna_cichlid_get_reset_handler will always
return NULL, although the module parameter reset_method is 3
when loading amdgpu driver.
Signed-off-by: lyndonli <Lyndon.Li at amd.com>
Signed-off-by: Yunxiang Li <Yunxiang.Li at amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu at amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c b/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c
index 81a6d5b94987..8b8086d5c864 100644
--- a/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c
+++ b/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c
@@ -40,7 +40,7 @@ static bool sienna_cichlid_is_mode2_default(struct amdgpu_reset_control *reset_c
adev->pm.fw_version >= 0x3a5500 && !amdgpu_sriov_vf(adev))
return true;
#endif
- return false;
+ return amdgpu_reset_method == AMD_RESET_METHOD_MODE2;
}
static struct amdgpu_reset_handler *
commit c8c2969bfcba5fcba3a5b078315c1b586d927d9f
Author: Hans de Goede <hdegoede at redhat.com>
Date: Tue Apr 25 21:44:41 2023 +0200
drm/i915/dsi: Use unconditional msleep() instead of intel_dsi_msleep()
The intel_dsi_msleep() helper skips sleeping if the MIPI-sequences have
a version of 3 or newer and the panel is in vid-mode.
This is based on the big comment around line 730 which starts with
"Panel enable/disable sequences from the VBT spec.", where
the "v3 video mode seq" column does not have any wait t# entries.
Checking the Windows driver shows that it does always honor
the VBT delays independent of the version of the VBT sequences.
Commit 6fdb335f1c9c ("drm/i915/dsi: Use unconditional msleep for
the panel_on_delay when there is no reset-deassert MIPI-sequence")
switched to a direct msleep() instead of intel_dsi_msleep()
when there is no MIPI_SEQ_DEASSERT_RESET sequence, to fix
the panel on an Acer Aspire Switch 10 E SW3-016 not turning on.
And now testing on a Nextbook Ares 8A shows that panel_on_delay
must always be honored otherwise the panel will not turn on.
Instead of only always using regular msleep() for panel_on_delay
do as Windows does and always use regular msleep() everywhere
were intel_dsi_msleep() is used and drop the intel_dsi_msleep()
helper.
Changes in v2:
- Replace all intel_dsi_msleep() calls instead of just
the intel_dsi_msleep(panel_on_delay) call
Cc: stable at vger.kernel.org
Fixes: 6fdb335f1c9c ("drm/i915/dsi: Use unconditional msleep for the panel_on_delay when there is no reset-deassert MIPI-sequence")
Signed-off-by: Hans de Goede <hdegoede at redhat.com>
Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230425194441.68086-1-hdegoede@redhat.com
(cherry picked from commit fa83c12132f71302f7d4b02758dc0d46048d3f5f)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
diff --git a/drivers/gpu/drm/i915/display/icl_dsi.c b/drivers/gpu/drm/i915/display/icl_dsi.c
index ad78148e0788..c9aeba0ecf91 100644
--- a/drivers/gpu/drm/i915/display/icl_dsi.c
+++ b/drivers/gpu/drm/i915/display/icl_dsi.c
@@ -1140,7 +1140,7 @@ static void gen11_dsi_powerup_panel(struct intel_encoder *encoder)
/* panel power on related mipi dsi vbt sequences */
intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_POWER_ON);
- intel_dsi_msleep(intel_dsi, intel_dsi->panel_on_delay);
+ msleep(intel_dsi->panel_on_delay);
intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_DEASSERT_RESET);
intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_INIT_OTP);
intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_DISPLAY_ON);
diff --git a/drivers/gpu/drm/i915/display/intel_dsi_vbt.c b/drivers/gpu/drm/i915/display/intel_dsi_vbt.c
index 695b0d69a4cb..c7935ea498c4 100644
--- a/drivers/gpu/drm/i915/display/intel_dsi_vbt.c
+++ b/drivers/gpu/drm/i915/display/intel_dsi_vbt.c
@@ -763,17 +763,6 @@ void intel_dsi_vbt_exec_sequence(struct intel_dsi *intel_dsi,
gpiod_set_value_cansleep(intel_dsi->gpio_backlight, 0);
}
-void intel_dsi_msleep(struct intel_dsi *intel_dsi, int msec)
-{
- struct intel_connector *connector = intel_dsi->attached_connector;
-
- /* For v3 VBTs in vid-mode the delays are part of the VBT sequences */
- if (is_vid_mode(intel_dsi) && connector->panel.vbt.dsi.seq_version >= 3)
- return;
-
- msleep(msec);
-}
-
void intel_dsi_log_params(struct intel_dsi *intel_dsi)
{
struct drm_i915_private *i915 = to_i915(intel_dsi->base.base.dev);
diff --git a/drivers/gpu/drm/i915/display/intel_dsi_vbt.h b/drivers/gpu/drm/i915/display/intel_dsi_vbt.h
index dc642c1fe7ef..468d873fab1a 100644
--- a/drivers/gpu/drm/i915/display/intel_dsi_vbt.h
+++ b/drivers/gpu/drm/i915/display/intel_dsi_vbt.h
@@ -16,7 +16,6 @@ void intel_dsi_vbt_gpio_init(struct intel_dsi *intel_dsi, bool panel_is_on);
void intel_dsi_vbt_gpio_cleanup(struct intel_dsi *intel_dsi);
void intel_dsi_vbt_exec_sequence(struct intel_dsi *intel_dsi,
enum mipi_seq seq_id);
-void intel_dsi_msleep(struct intel_dsi *intel_dsi, int msec);
void intel_dsi_log_params(struct intel_dsi *intel_dsi);
#endif /* __INTEL_DSI_VBT_H__ */
diff --git a/drivers/gpu/drm/i915/display/vlv_dsi.c b/drivers/gpu/drm/i915/display/vlv_dsi.c
index 028965ab442d..61d008d4e5f1 100644
--- a/drivers/gpu/drm/i915/display/vlv_dsi.c
+++ b/drivers/gpu/drm/i915/display/vlv_dsi.c
@@ -737,7 +737,6 @@ static void intel_dsi_pre_enable(struct intel_atomic_state *state,
{
struct intel_dsi *intel_dsi = enc_to_intel_dsi(encoder);
struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
- struct intel_connector *connector = to_intel_connector(conn_state->connector);
struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
enum pipe pipe = crtc->pipe;
enum port port;
@@ -779,21 +778,10 @@ static void intel_dsi_pre_enable(struct intel_atomic_state *state,
if (!IS_GEMINILAKE(dev_priv))
intel_dsi_prepare(encoder, pipe_config);
+ /* Give the panel time to power-on and then deassert its reset */
intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_POWER_ON);
-
- /*
- * Give the panel time to power-on and then deassert its reset.
- * Depending on the VBT MIPI sequences version the deassert-seq
- * may contain the necessary delay, intel_dsi_msleep() will skip
- * the delay in that case. If there is no deassert-seq, then an
- * unconditional msleep is used to give the panel time to power-on.
- */
- if (connector->panel.vbt.dsi.sequence[MIPI_SEQ_DEASSERT_RESET]) {
- intel_dsi_msleep(intel_dsi, intel_dsi->panel_on_delay);
- intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_DEASSERT_RESET);
- } else {
- msleep(intel_dsi->panel_on_delay);
- }
+ msleep(intel_dsi->panel_on_delay);
+ intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_DEASSERT_RESET);
if (IS_GEMINILAKE(dev_priv)) {
glk_cold_boot = glk_dsi_enable_io(encoder);
@@ -827,7 +815,7 @@ static void intel_dsi_pre_enable(struct intel_atomic_state *state,
msleep(20); /* XXX */
for_each_dsi_port(port, intel_dsi->ports)
dpi_send_cmd(intel_dsi, TURN_ON, false, port);
- intel_dsi_msleep(intel_dsi, 100);
+ msleep(100);
intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_DISPLAY_ON);
@@ -949,7 +937,7 @@ static void intel_dsi_post_disable(struct intel_atomic_state *state,
/* Assert reset */
intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_ASSERT_RESET);
- intel_dsi_msleep(intel_dsi, intel_dsi->panel_off_delay);
+ msleep(intel_dsi->panel_off_delay);
intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_POWER_OFF);
intel_dsi->panel_power_off_time = ktime_get_boottime();
commit 6ece90e3665a9b7fb2637fcca26cebd42991580b
Author: Radhakrishna Sripada <radhakrishna.sripada at intel.com>
Date: Thu Apr 20 15:12:47 2023 -0700
drm/i915/mtl: Add the missing CPU transcoder mask in intel_device_info
CPU transcoder mask is used to iterate over the available
CPU transcoders in the macro for_each_cpu_transcoder().
The macro is broken on MTL and got highlighted when audio
state was being tracked for each transcoder added in [1].
Add the missing CPU transcoder mask which is similar to ADL-P
mask but without DSI transcoders.
[1]: https://patchwork.freedesktop.org/patch/523723/
Fixes: 7835303982d1 ("drm/i915/mtl: Add MeteorLake PCI IDs")
Cc: Ville Syrjälä <ville.syrjala at linux.intel.com>
Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada at intel.com>
Acked-by: Haridhar Kalvala <haridhar.kalvala at intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa at intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230420221248.2511314-1-radhakrishna.sripada@intel.com
(cherry picked from commit bddc18913bd44adae5c828fd514d570f43ba1576)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
index cddb6e197972..2a012da8ccfa 100644
--- a/drivers/gpu/drm/i915/i915_pci.c
+++ b/drivers/gpu/drm/i915/i915_pci.c
@@ -1134,6 +1134,8 @@ static const struct intel_gt_definition xelpmp_extra_gt[] = {
static const struct intel_device_info mtl_info = {
XE_HP_FEATURES,
XE_LPDP_FEATURES,
+ .__runtime.cpu_transcoder_mask = BIT(TRANSCODER_A) | BIT(TRANSCODER_B) |
+ BIT(TRANSCODER_C) | BIT(TRANSCODER_D),
/*
* Real graphics IP version will be obtained from hardware GMD_ID
* register. Value provided here is just for sanity checking.
commit 1816f4a17f54a01afa2f06d6571c39890b97d282
Author: John Harrison <John.C.Harrison at Intel.com>
Date: Fri Apr 21 15:47:42 2023 -0700
drm/i915/guc: Actually return an error if GuC version range check fails
Dan Carpenter pointed out that 'err' was not being set in the case
where the GuC firmware version range check fails. Fix that.
Note that while this is a bug fix for a previous patch (see Fixes tag
below). It is an exceedingly low risk bug. The range check is
asserting that the GuC firmware version is within spec. So it should
not be possible to ever have a firmware file that fails this check. If
larger version numbers are required in the future, that would be a
backwards breaking spec change and thus require a major version bump,
in which case an old i915 driver would not load that new version anyway.
Fixes: 9bbba0667f37 ("drm/i915/guc: Use GuC submission API version number")
Reported-by: Dan Carpenter <error27 at gmail.com>
Signed-off-by: John Harrison <John.C.Harrison at Intel.com>
Cc: John Harrison <John.C.Harrison at Intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio at intel.com>
Cc: Alan Previn <alan.previn.teres.alexis at intel.com>
Cc: Umesh Nerlige Ramappa <umesh.nerlige.ramappa at intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi at intel.com>
Cc: Matthew Brost <matthew.brost at intel.com>
Cc: Andi Shyti <andi.shyti at linux.intel.com>
Cc: Matthew Auld <matthew.auld at intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
Cc: Lucas De Marchi <lucas.demarchi at intel.com>
Cc: Jani Nikula <jani.nikula at intel.com>
Reviewed-by: Andi Shyti <andi.shyti at linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230421224742.2357198-1-John.C.Harrison@Intel.com
(cherry picked from commit 80ab31799002166ac7c660bacfbff4f85bc29107)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
index 24765c30a0e1..c36e68e23a14 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
@@ -635,9 +635,10 @@ static bool is_ver_8bit(struct intel_uc_fw_ver *ver)
return ver->major < 0xFF && ver->minor < 0xFF && ver->patch < 0xFF;
}
-static bool guc_check_version_range(struct intel_uc_fw *uc_fw)
+static int guc_check_version_range(struct intel_uc_fw *uc_fw)
{
struct intel_guc *guc = container_of(uc_fw, struct intel_guc, fw);
+ struct intel_gt *gt = __uc_fw_to_gt(uc_fw);
/*
* GuC version number components are defined as being 8-bits.
@@ -646,24 +647,24 @@ static bool guc_check_version_range(struct intel_uc_fw *uc_fw)
*/
if (!is_ver_8bit(&uc_fw->file_selected.ver)) {
- gt_warn(__uc_fw_to_gt(uc_fw), "%s firmware: invalid file version: 0x%02X:%02X:%02X\n",
+ gt_warn(gt, "%s firmware: invalid file version: 0x%02X:%02X:%02X\n",
intel_uc_fw_type_repr(uc_fw->type),
uc_fw->file_selected.ver.major,
uc_fw->file_selected.ver.minor,
uc_fw->file_selected.ver.patch);
- return false;
+ return -EINVAL;
}
if (!is_ver_8bit(&guc->submission_version)) {
- gt_warn(__uc_fw_to_gt(uc_fw), "%s firmware: invalid submit version: 0x%02X:%02X:%02X\n",
+ gt_warn(gt, "%s firmware: invalid submit version: 0x%02X:%02X:%02X\n",
intel_uc_fw_type_repr(uc_fw->type),
guc->submission_version.major,
guc->submission_version.minor,
guc->submission_version.patch);
- return false;
+ return -EINVAL;
}
- return true;
+ return i915_inject_probe_error(gt->i915, -EINVAL);
}
static int check_fw_header(struct intel_gt *gt,
@@ -772,8 +773,11 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw)
if (err)
goto fail;
- if (uc_fw->type == INTEL_UC_FW_TYPE_GUC && !guc_check_version_range(uc_fw))
- goto fail;
+ if (uc_fw->type == INTEL_UC_FW_TYPE_GUC) {
+ err = guc_check_version_range(uc_fw);
+ if (err)
+ goto fail;
+ }
if (uc_fw->file_wanted.ver.major && uc_fw->file_selected.ver.major) {
/* Check the file's major version was as it claimed */
commit d893f39320e1248d1c97fde0d6e51e5ea008a76b
Author: Leo Chen <sancchen at amd.com>
Date: Tue Apr 11 10:49:38 2023 -0400
drm/amd/display: Lowering min Z8 residency time
[Why & How]
Per HW team request, we're lowering the minimum Z8
residency time to 2000us. This enables Z8 support for additional
modes we were previously blocking like 2k>60hz
Cc: stable at vger.kernel.org
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas at amd.com>
Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Signed-off-by: Leo Chen <sancchen at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c
index 50ed7e09d5ba..2f7df8d34a91 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c
@@ -885,7 +885,7 @@ static const struct dc_plane_cap plane_cap = {
static const struct dc_debug_options debug_defaults_drv = {
.disable_z10 = false,
.enable_z9_disable_interface = true,
- .minimum_z8_residency_time = 3080,
+ .minimum_z8_residency_time = 2000,
.psr_skip_crtc_disable = true,
.disable_dmcu = true,
.force_abm_enable = false,
commit 08da182175db4c7f80850354849d95f2670e8cd9
Author: Hamza Mahfooz <hamza.mahfooz at amd.com>
Date: Fri Apr 14 14:26:27 2023 -0400
drm/amd/display: fix flickering caused by S/G mode
Currently, on a handful of ASICs. We allow the framebuffer for a given
plane to exist in either VRAM or GTT. However, if the plane's new
framebuffer is in a different memory domain than it's previous
framebuffer, flipping between them can cause the screen to flicker. So,
to fix this, don't perform an immediate flip in the aforementioned case.
Cc: stable at vger.kernel.org
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2354
Reviewed-by: Roman Li <Roman.Li at amd.com>
Fixes: 81d0bcf99009 ("drm/amdgpu: make display pinning more flexible (v2)")
Signed-off-by: Hamza Mahfooz <hamza.mahfooz at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index b619d7cdb525..8d17fd5a817e 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -7901,6 +7901,13 @@ static void amdgpu_dm_commit_cursors(struct drm_atomic_state *state)
amdgpu_dm_plane_handle_cursor_update(plane, old_plane_state);
}
+static inline uint32_t get_mem_type(struct drm_framebuffer *fb)
+{
+ struct amdgpu_bo *abo = gem_to_amdgpu_bo(fb->obj[0]);
+
+ return abo->tbo.resource ? abo->tbo.resource->mem_type : 0;
+}
+
static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
struct dc_state *dc_state,
struct drm_device *dev,
@@ -8043,11 +8050,13 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
/*
* Only allow immediate flips for fast updates that don't
- * change FB pitch, DCC state, rotation or mirroing.
+ * change memory domain, FB pitch, DCC state, rotation or
+ * mirroring.
*/
bundle->flip_addrs[planes_count].flip_immediate =
crtc->state->async_flip &&
- acrtc_state->update_type == UPDATE_TYPE_FAST;
+ acrtc_state->update_type == UPDATE_TYPE_FAST &&
+ get_mem_type(old_plane_state->fb) == get_mem_type(fb);
timestamp_ns = ktime_get_ns();
bundle->flip_addrs[planes_count].flip_timestamp_in_us = div_u64(timestamp_ns, 1000);
commit 9675b3ba99ec79273d94afa09e9b69e2b8c0d238
Author: Igor Kravchenko <Igor.Kravchenko at amd.com>
Date: Fri Jul 10 16:24:30 2020 -0400
drm/amd/display: Set min_width and min_height capability for DCN30
Add min_width, min_height fields to dc_plane_cap structure. Set values
to 16x16 for discrete ASICs, and 64x64 for others.
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Signed-off-by: Igor Kravchenko <Igor.Kravchenko at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c
index 965f5ceb33f7..67a34cda3774 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c
@@ -701,7 +701,9 @@ static const struct dc_plane_cap plane_cap = {
.argb8888 = 167,
.nv12 = 167,
.fp16 = 167
- }
+ },
+ 16,
+ 16
};
static const struct dc_debug_options debug_defaults_drv = {
commit 3caab67db1f69e077fb12ac194d3cd2a4de06d8d
Author: Jasdeep Dhillon <jasdeep.dhillon at amd.com>
Date: Tue Feb 28 11:46:31 2023 -0500
drm/amd/display: Isolate remaining FPU code in DCN32
[Why]
DCN32 resource contains code that uses FPU.
[How]
Moved code into DCN32 FPU
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Signed-off-by: Jasdeep Dhillon <jasdeep.dhillon at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
index 23a972f2885f..47beb4ea779d 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
@@ -2876,3 +2876,9 @@ bool dcn32_find_vactive_pipe(struct dc *dc, const struct dc_state *context, uint
}
return vactive_found;
}
+
+void dcn32_set_clock_limits(const struct _vcs_dpi_soc_bounding_box_st *soc_bb)
+{
+ dc_assert_fp_enabled();
+ dcn3_2_soc.clock_limits[0].dcfclk_mhz = 1200.0;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h
index 9a0806a0e2ef..dcf512cd3072 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h
@@ -80,4 +80,6 @@ void dcn32_assign_fpo_vactive_candidate(struct dc *dc, const struct dc_state *co
bool dcn32_find_vactive_pipe(struct dc *dc, const struct dc_state *context, uint32_t vactive_margin_req);
+void dcn32_set_clock_limits(const struct _vcs_dpi_soc_bounding_box_st *soc_bb);
+
#endif
commit 989cd3e76a4aab76fe7dd50090ac3fa501c537f6
Author: Aurabindo Pillai <aurabindo.pillai at amd.com>
Date: Thu Apr 6 15:59:45 2023 -0400
drm/amd/display: Update bounding box values for DCN321
[Why&how]
Update bounding box values as per hardware spec
Fixes: 197485c69543 ("drm/amd/display: Create dcn321_fpu file")
Acked-by: Leo Li <sunpeng.li at amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c
index 57b9bd896678..342a1bcb4927 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c
@@ -106,16 +106,16 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_21_soc = {
.clock_limits = {
{
.state = 0,
- .dcfclk_mhz = 1564.0,
- .fabricclk_mhz = 400.0,
- .dispclk_mhz = 2150.0,
- .dppclk_mhz = 2150.0,
+ .dcfclk_mhz = 1434.0,
+ .fabricclk_mhz = 2250.0,
+ .dispclk_mhz = 1720.0,
+ .dppclk_mhz = 1720.0,
.phyclk_mhz = 810.0,
.phyclk_d18_mhz = 667.0,
- .phyclk_d32_mhz = 625.0,
+ .phyclk_d32_mhz = 313.0,
.socclk_mhz = 1200.0,
- .dscclk_mhz = 716.667,
- .dram_speed_mts = 1600.0,
+ .dscclk_mhz = 573.333,
+ .dram_speed_mts = 16000.0,
.dtbclk_mhz = 1564.0,
},
},
@@ -125,14 +125,14 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_21_soc = {
.sr_exit_z8_time_us = 285.0,
.sr_enter_plus_exit_z8_time_us = 320,
.writeback_latency_us = 12.0,
- .round_trip_ping_latency_dcfclk_cycles = 263,
+ .round_trip_ping_latency_dcfclk_cycles = 207,
.urgent_latency_pixel_data_only_us = 4,
.urgent_latency_pixel_mixed_with_vm_data_us = 4,
.urgent_latency_vm_data_only_us = 4,
- .fclk_change_latency_us = 20,
- .usr_retraining_latency_us = 2,
- .smn_latency_us = 2,
- .mall_allocated_for_dcn_mbytes = 64,
+ .fclk_change_latency_us = 7,
+ .usr_retraining_latency_us = 0,
+ .smn_latency_us = 0,
+ .mall_allocated_for_dcn_mbytes = 32,
.urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096,
.urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096,
.urgent_out_of_order_return_per_channel_vm_only_bytes = 4096,
commit 99d92eaca5d915763b240aae24669f5bf3227ecf
Author: Aurabindo Pillai <aurabindo.pillai at amd.com>
Date: Thu Apr 6 15:48:48 2023 -0400
drm/amd/display: Do not clear GPINT register when releasing DMUB from reset
[Why & How]
There's no need to clear GPINT register for DMUB
when releasing it from reset. Fix that.
Fixes: ac2e555e0a7f ("drm/amd/display: Add DMCUB source files and changes for DCN32/321")
Reviewed-by: Leo Li <sunpeng.li at amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
index b0adbf783aae..9c20516be066 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
@@ -137,7 +137,6 @@ void dmub_dcn32_reset(struct dmub_srv *dmub)
void dmub_dcn32_reset_release(struct dmub_srv *dmub)
{
- REG_WRITE(DMCUB_GPINT_DATAIN1, 0);
REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 0);
REG_WRITE(DMCUB_SCRATCH15, dmub->psp_version & 0x001100FF);
REG_UPDATE_2(DMCUB_CNTL, DMCUB_ENABLE, 1, DMCUB_TRACEPORT_EN, 1);
commit 425afa0ac99a05b39e6cd00704fa0e3e925cee2b
Author: Cruise Hung <Cruise.Hung at amd.com>
Date: Fri May 13 09:16:42 2022 +0800
drm/amd/display: Reset OUTBOX0 r/w pointer on DMUB reset
[Why & How]
We missed resetting OUTBOX0 mailbox r/w pointer on DMUB reset.
Fix it.
Fixes: 6ecf9773a503 ("drm/amd/display: Fix DMUB outbox trace in S4 (#4465)")
Signed-off-by: Cruise Hung <Cruise.Hung at amd.com>
Acked-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
index a76da0131add..b0adbf783aae 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
@@ -130,6 +130,8 @@ void dmub_dcn32_reset(struct dmub_srv *dmub)
REG_WRITE(DMCUB_INBOX1_WPTR, 0);
REG_WRITE(DMCUB_OUTBOX1_RPTR, 0);
REG_WRITE(DMCUB_OUTBOX1_WPTR, 0);
+ REG_WRITE(DMCUB_OUTBOX0_RPTR, 0);
+ REG_WRITE(DMCUB_OUTBOX0_WPTR, 0);
REG_WRITE(DMCUB_SCRATCH0, 0);
}
commit d1c5c3e252b8a911a524e6ee33b82aca81397745
Author: Aurabindo Pillai <aurabindo.pillai at amd.com>
Date: Thu Apr 6 12:28:59 2023 -0400
drm/amd/display: Fixes for dcn32_clk_mgr implementation
[Why&How]
Fix CLK MGR early initialization and add logging.
Fixes: 265280b99822 ("drm/amd/display: add CLKMGR changes for DCN32/321")
Reviewed-by: Leo Li <sunpeng.li at amd.com>
Reviewed-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c
index ea753f8fa175..8d9444db092a 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c
@@ -878,6 +878,8 @@ void dcn32_clk_mgr_construct(
struct pp_smu_funcs *pp_smu,
struct dccg *dccg)
{
+ struct clk_log_info log_info = {0};
+
clk_mgr->base.ctx = ctx;
clk_mgr->base.funcs = &dcn32_funcs;
if (ASICREV_IS_GC_11_0_2(clk_mgr->base.ctx->asic_id.hw_internal_rev)) {
@@ -911,6 +913,7 @@ void dcn32_clk_mgr_construct(
clk_mgr->base.clks.ref_dtbclk_khz = 268750;
}
+
/* integer part is now VCO frequency in kHz */
clk_mgr->base.dentist_vco_freq_khz = dcn32_get_vco_frequency_from_reg(clk_mgr);
@@ -918,6 +921,8 @@ void dcn32_clk_mgr_construct(
if (clk_mgr->base.dentist_vco_freq_khz == 0)
clk_mgr->base.dentist_vco_freq_khz = 4300000; /* Updated as per HW docs */
+ dcn32_dump_clk_registers(&clk_mgr->base.boot_snapshot, &clk_mgr->base, &log_info);
+
if (ctx->dc->debug.disable_dtb_ref_clk_switch &&
clk_mgr->base.clks.ref_dtbclk_khz != clk_mgr->base.boot_snapshot.dtbclk) {
clk_mgr->base.clks.ref_dtbclk_khz = clk_mgr->base.boot_snapshot.dtbclk;
commit b1bcdd409d2d158867ce0b71cfa9bcefe83ce07f
Author: Tianci Yin <tianci.yin at amd.com>
Date: Mon Feb 6 15:58:46 2023 +0800
drm/amd/display: Disable migration to ensure consistency of per-CPU variable
[why]
Since the variable fpu_recursion_depth is per-CPU type, it has one copy
on each CPU, thread migration causes data consistency issue, then the
call trace shows up. And preemption disabling can't prevent migration.
[how]
Disable migration to ensure consistency of fpu_recursion_depth.
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Tianci Yin <tianci.yin at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c
index 1743ca0a3641..c42aa947c969 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c
@@ -89,6 +89,7 @@ void dc_fpu_begin(const char *function_name, const int line)
if (*pcpu == 1) {
#if defined(CONFIG_X86)
+ migrate_disable();
kernel_fpu_begin();
#elif defined(CONFIG_PPC64)
if (cpu_has_feature(CPU_FTR_VSX_COMP)) {
@@ -129,6 +130,7 @@ void dc_fpu_end(const char *function_name, const int line)
if (*pcpu <= 0) {
#if defined(CONFIG_X86)
kernel_fpu_end();
+ migrate_enable();
#elif defined(CONFIG_PPC64)
if (cpu_has_feature(CPU_FTR_VSX_COMP)) {
disable_kernel_vsx();
commit fc3888fe2c63b35a22db8234d142823a5ffda9d8
Author: Aurabindo Pillai <aurabindo.pillai at amd.com>
Date: Wed Apr 5 16:17:42 2023 -0400
drm/amd/display: remove incorrect early return
[Why&How]
Remove incorrect early return in a device specific fifo reset workaround
Reviewed-by: Leo Li <sunpeng.li at amd.com>
Reviewed-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
index 330ab036c830..c6ce2b7123b7 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -687,7 +687,6 @@ static void apply_synaptics_fifo_reset_wa(struct drm_dp_aux *aux)
return;
data[0] |= (1 << 1); // set bit 1 to 1
- return;
if (!execute_synaptics_rc_command(aux, false, 0x31, 4, 0x221198, data))
return;
commit dd24662d9dfbad281bbf030f06d68c7938fa0c66
Author: Hersen Wu <hersenxs.wu at amd.com>
Date: Sun May 29 10:54:30 2022 -0400
drm/amd/display: Return error code on DSC atomic check failure
[Why&How]
We were not returning -EINVAL on DSC atomic check fail. Add it.
Fixes: 71be4b16d39a ("drm/amd/display: dsc validate fail not pass to atomic check")
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Hersen Wu <hersenxs.wu at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index fc08b4191e24..b619d7cdb525 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -10170,6 +10170,7 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
ret = compute_mst_dsc_configs_for_state(state, dm_state->context, vars);
if (ret) {
DRM_DEBUG_DRIVER("compute_mst_dsc_configs_for_state() failed\n");
+ ret = -EINVAL;
goto fail;
}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 5dc79b753d5f..810ab682f424 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -1410,6 +1410,7 @@ int pre_validate_dsc(struct drm_atomic_state *state,
ret = pre_compute_mst_dsc_configs_for_state(state, local_dc_state, vars);
if (ret != 0) {
DRM_INFO_ONCE("pre_compute_mst_dsc_configs_for_state() failed\n");
+ ret = -EINVAL;
goto clean_exit;
}
commit e0cce122514ff76c3c986103c94de68fbb401949
Author: Jingwen Zhu <Jingwen.Zhu at amd.com>
Date: Thu Mar 30 16:38:59 2023 +0800
drm/amd/display: Improvement for handling edp link training fails
[Why]
The eDP retrain will cause the DPCD 300 to be reset to default.
And cause the brightness can't be set correctly.
[How]
delete the call to edp panel power control in both
enable_link_output/disable_link_output entirely and
only call edp panel control in enable_link_dp and
disable_link_dp once.
Reviewed-by: Charlene Liu <Charlene.Liu at amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Jingwen Zhu <Jingwen.Zhu at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 9fe0ce91db00..8d2460d06bce 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -3031,10 +3031,12 @@ void dce110_enable_dp_link_output(
const struct link_hwss *link_hwss = get_link_hwss(link, link_res);
unsigned int i;
-
+ /*
+ * Add the logic to extract BOTH power up and power down sequences
+ * from enable/disable link output and only call edp panel control
+ * in enable_link_dp and disable_link_dp once.
+ */
if (link->connector_signal == SIGNAL_TYPE_EDP) {
- if (!link->dc->config.edp_no_power_sequencing)
- link->dc->hwss.edp_power_control(link, true);
link->dc->hwss.edp_wait_for_hpd_ready(link, true);
}
@@ -3096,11 +3098,12 @@ void dce110_disable_link_output(struct dc_link *link,
link_hwss->disable_link_output(link, link_res, signal);
link->phy_state.symclk_state = SYMCLK_OFF_TX_OFF;
-
- if (signal == SIGNAL_TYPE_EDP &&
- link->dc->hwss.edp_backlight_control)
- link->dc->hwss.edp_power_control(link, false);
- else if (dmcu != NULL && dmcu->funcs->lock_phy)
+ /*
+ * Add the logic to extract BOTH power up and power down sequences
+ * from enable/disable link output and only call edp panel control
+ * in enable_link_dp and disable_link_dp once.
+ */
+ if (dmcu != NULL && dmcu->funcs->lock_phy)
dmcu->funcs->unlock_phy(dmcu);
dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_DISABLE_LINK_PHY);
}
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
index 027ad1f0144d..2267fb097830 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
@@ -1927,6 +1927,11 @@ static void disable_link_dp(struct dc_link *link,
dp_disable_link_phy(link, link_res, signal);
+ if (link->connector_signal == SIGNAL_TYPE_EDP) {
+ if (!link->dc->config.edp_no_power_sequencing)
+ link->dc->hwss.edp_power_control(link, false);
+ }
+
if (signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
/* set the sink to SST mode after disabling the link */
enable_mst_on_sink(link, false);
commit 56d8ce9d8c17bea955b0c2551ee86149486890ae
Author: Michael Mityushkin <michael.mityushkin at amd.com>
Date: Thu Mar 30 11:35:08 2023 -0400
drm/amd/display: Apply correct panel mode when reinitializing hardware
[Why]
When link training during engine recovery, ASSR might fail causing panel
mode to be reset to default. This should not happen for eDP as it
will prevent the panel from turning back on.
[How]
Added dp_panel_mode to struct dc_link to remember previously applied
panel mode. Do not reset panel mode to default while performing link
training if previously used panel mode = eDP.
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas at amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Michael Mityushkin <michael.mityushkin at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 23ee63b98dcd..30f0ba05a6e6 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -1454,6 +1454,7 @@ struct dc_link {
struct ddc_service *ddc;
+ enum dp_panel_mode panel_mode;
bool aux_mode;
/* Private to DC core */
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c
index 170f33835930..579fa222810d 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c
@@ -1596,7 +1596,10 @@ bool perform_link_training_with_retries(
* Report and continue with eDP panel mode to
* perform eDP link training with right settings
*/
- cp_psp->funcs.enable_assr(cp_psp->handle, link);
+ bool result;
+ result = cp_psp->funcs.enable_assr(cp_psp->handle, link);
+ if (!result && link->panel_mode != DP_PANEL_MODE_EDP)
+ panel_mode = DP_PANEL_MODE_DEFAULT;
}
}
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
index d895046787bc..8d1df863659c 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
@@ -83,6 +83,7 @@ void dp_set_panel_mode(struct dc_link *link, enum dp_panel_mode panel_mode)
ASSERT(result == DC_OK);
}
}
+ link->panel_mode = panel_mode;
DC_LOG_DETECTION_DP_CAPS("Link: %d eDP panel mode supported: %d "
"eDP panel mode enabled: %d \n",
link->link_index,
commit 3cf7cd3f770a0b89dc5f06e19edb52e65b93b214
Author: Hersen Wu <hersenxs.wu at amd.com>
Date: Tue Mar 28 10:45:24 2023 -0400
drm/amd/display: fix access hdcp_workqueue assert
[Why] hdcp are enabled for asics from raven. for old asics
which hdcp are not enabled, hdcp_workqueue are null. some
access to hdcp work queue are not guarded with pointer check.
[How] add hdcp_workqueue pointer check before access workqueue.
Fixes: 82986fd631fa ("drm/amd/display: save restore hdcp state when display is unplugged from mst hub")
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2444
Reported-by: Niklāvs Koļesņikovs <89q1r14hd at relay.firefox.com>
Reviewed-by: Bhawanpreet Lakha <Bhawanpreet.Lakha at amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Hersen Wu <hersenxs.wu at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
Cc: stable at vger.kernel.org
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 8b03c8d8f0b8..fc08b4191e24 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -8559,6 +8559,9 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(dm_new_con_state->base.crtc);
struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+ if (!adev->dm.hdcp_workqueue)
+ continue;
+
pr_debug("[HDCP_DM] -------------- i : %x ----------\n", i);
if (!connector)
@@ -8607,6 +8610,9 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(dm_new_con_state->base.crtc);
struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+ if (!adev->dm.hdcp_workqueue)
+ continue;
+
new_crtc_state = NULL;
old_crtc_state = NULL;
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 994ba426ca66..5dc79b753d5f 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -379,13 +379,17 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector)
if (aconnector->dc_sink && connector->state) {
struct drm_device *dev = connector->dev;
struct amdgpu_device *adev = drm_to_adev(dev);
- struct hdcp_workqueue *hdcp_work = adev->dm.hdcp_workqueue;
- struct hdcp_workqueue *hdcp_w = &hdcp_work[aconnector->dc_link->link_index];
- connector->state->hdcp_content_type =
- hdcp_w->hdcp_content_type[connector->index];
- connector->state->content_protection =
- hdcp_w->content_protection[connector->index];
+ if (adev->dm.hdcp_workqueue) {
+ struct hdcp_workqueue *hdcp_work = adev->dm.hdcp_workqueue;
+ struct hdcp_workqueue *hdcp_w =
+ &hdcp_work[aconnector->dc_link->link_index];
+
+ connector->state->hdcp_content_type =
+ hdcp_w->hdcp_content_type[connector->index];
+ connector->state->content_protection =
+ hdcp_w->content_protection[connector->index];
+ }
}
if (aconnector->dc_sink) {
commit 0c0463ff010b80a0c03937ca8cf85587ded2f20e
Author: Alvin Lee <Alvin.Lee2 at amd.com>
Date: Tue Mar 28 09:53:33 2023 -0400
drm/amd/display: Reduce SubVP + DRR stretch margin
[Description]
- Having excessively large margin causes failure in the static
schedulability check in some cases for SubVP + DRR
- 100us of DRR margin is sufficient based on a weeks worth of
stress testing on different display configs
Reviewed-by: Michael Strauss <Michael.Strauss at amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Alvin Lee <Alvin.Lee2 at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 181a3408cc61..25284006019c 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -144,7 +144,7 @@ struct test_pattern {
unsigned int cust_pattern_size;
};
-#define SUBVP_DRR_MARGIN_US 600 // 600us for DRR margin (SubVP + DRR)
+#define SUBVP_DRR_MARGIN_US 100 // 100us for DRR margin (SubVP + DRR)
enum mall_stream_type {
SUBVP_NONE, // subvp not in use
commit 025ce392b5f213696ca0af3e07735d0fae020694
Author: Hersen Wu <hersenxs.wu at amd.com>
Date: Mon Mar 27 09:10:48 2023 -0400
drm/amd/display: fix memleak in aconnector->timing_requested
[Why]
when amdgpu_dm_update_connector_after_detect is called
two times successively with valid sink, memory allocated of
aconnector->timing_requested for the first call is not free.
this causes memeleak.
[How]
allocate memory only when aconnector->timing_requested
is null.
Reviewed-by: Qingqing Zhuo <Qingqing.Zhuo at amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Hersen Wu <hersenxs.wu at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index c432436cd66d..8b03c8d8f0b8 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -3128,9 +3128,12 @@ void amdgpu_dm_update_connector_after_detect(
aconnector->edid);
}
- aconnector->timing_requested = kzalloc(sizeof(struct dc_crtc_timing), GFP_KERNEL);
- if (!aconnector->timing_requested)
- dm_error("%s: failed to create aconnector->requested_timing\n", __func__);
+ if (!aconnector->timing_requested) {
+ aconnector->timing_requested =
+ kzalloc(sizeof(struct dc_crtc_timing), GFP_KERNEL);
+ if (!aconnector->timing_requested)
+ dm_error("failed to create aconnector->requested_timing\n");
+ }
drm_connector_update_edid_property(connector, aconnector->edid);
amdgpu_dm_update_freesync_caps(connector, aconnector->edid);
commit da5e14909776edea4462672fb4a3007802d262e7
Author: Aurabindo Pillai <aurabindo.pillai at amd.com>
Date: Fri Mar 24 10:42:37 2023 -0400
drm/amd/display: Fix hang when skipping modeset
[Why&How]
When skipping full modeset since the only state change was a front porch
change, the DC commit sequence requires extra checks to handle non
existant plane states being asked to be removed from context.
Reviewed-by: Alvin Lee <Alvin.Lee2 at amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 6cacb76f389e..c432436cd66d 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -7972,6 +7972,8 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
continue;
dc_plane = dm_new_plane_state->dc_state;
+ if (!dc_plane)
+ continue;
bundle->surface_updates[planes_count].surface = dc_plane;
if (new_pcrtc_state->color_mgmt_changed) {
@@ -9620,8 +9622,9 @@ static int dm_update_plane_state(struct dc *dc,
return -EINVAL;
}
+ if (dm_old_plane_state->dc_state)
+ dc_plane_state_release(dm_old_plane_state->dc_state);
- dc_plane_state_release(dm_old_plane_state->dc_state);
dm_new_plane_state->dc_state = NULL;
*lock_and_validation_needed = true;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 85d54bfb595c..117d80cb36fb 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1707,6 +1707,9 @@ bool dc_remove_plane_from_context(
struct dc_stream_status *stream_status = NULL;
struct resource_pool *pool = dc->res_pool;
+ if (!plane_state)
+ return true;
+
for (i = 0; i < context->stream_count; i++)
if (context->streams[i] == stream) {
stream_status = &context->stream_status[i];
commit 3fb7efd6866e5d43770e999b33d619a3b345dc2f
Author: Alex Hung <alex.hung at amd.com>
Date: Wed Mar 15 19:09:15 2023 +0800
drm/amd/display: allow edp updates for virtual signal
[Why]
When IGT's kms_hdmi_inject forces EDID for HDMI audio, dc rejects the
request because virtual signal is not in dc_is_audio_capable_signal
function.
[How]
Includes SIGNAL_TYPE_VIRTUAL as audio capable.
Reviewed-by: Chao-kai Wang <Stylon.Wang at amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Alex Hung <alex.hung at amd.com>
Signed-off-by: Wenchieh Chien <wenchieh.chien at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/include/signal_types.h b/drivers/gpu/drm/amd/display/include/signal_types.h
index beed70179bb5..23a308c3eccb 100644
--- a/drivers/gpu/drm/amd/display/include/signal_types.h
+++ b/drivers/gpu/drm/amd/display/include/signal_types.h
@@ -104,6 +104,7 @@ static inline bool dc_is_audio_capable_signal(enum signal_type signal)
{
return (signal == SIGNAL_TYPE_DISPLAY_PORT ||
signal == SIGNAL_TYPE_DISPLAY_PORT_MST ||
+ signal == SIGNAL_TYPE_VIRTUAL ||
dc_is_hdmi_signal(signal));
}
commit f11aee97b13ea6817287cd8dbed9b09a260ff0e7
Author: Josip Pavic <Josip.Pavic at amd.com>
Date: Fri May 14 14:04:02 2021 -0400
drm/amd/display: copy dmub caps to dc on dcn31
[Why & How]
Add code path to copy dmub caps to dc, which is missing on dcn31
Acked-by: Qingqing Zhuo <qingqing.zhuo at amd.com>
Signed-off-by: Josip Pavic <Josip.Pavic at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
index 7ac6e69cff37..62ce36c75c4d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
@@ -295,6 +295,10 @@ void dcn31_init_hw(struct dc *dc)
if (dc->res_pool->hubbub->funcs->init_crb)
dc->res_pool->hubbub->funcs->init_crb(dc->res_pool->hubbub);
#endif
+
+ // Get DMCUB capabilities
+ dc_dmub_srv_query_caps_cmd(dc->ctx->dmub_srv->dmub);
+ dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr;
}
void dcn31_dsc_pg_control(
commit 822b84ecfc646da0f87fd947fa00dc3be5e45ecc
Author: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Date: Tue Apr 4 14:54:05 2023 -0600
drm/amd/display: Add missing WA and MCLK validation
When the commit fff7eb56b376 ("drm/amd/display: Don't set dram clock
change requirement for SubVP") was merged, we missed some parts
associated with the MCLK switch. This commit adds all the missing parts.
Fixes: fff7eb56b376 ("drm/amd/display: Don't set dram clock change requirement for SubVP")
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c
index db0974fe58ab..1f5ee5cde6e1 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c
@@ -948,6 +948,7 @@ void dcn32_init_hw(struct dc *dc)
if (dc->ctx->dmub_srv) {
dc_dmub_srv_query_caps_cmd(dc->ctx->dmub_srv->dmub);
dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr;
+ dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch;
}
}
diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
index 0beb11d95eb7..a876e6eb6cd8 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
@@ -2023,7 +2023,7 @@ int dcn32_populate_dml_pipes_from_context(
// In general cases we want to keep the dram clock change requirement
// (prefer configs that support MCLK switch). Only override to false
// for SubVP
- if (subvp_in_use)
+ if (context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching || subvp_in_use)
context->bw_ctx.dml.soc.dram_clock_change_requirement_final = false;
else
context->bw_ctx.dml.soc.dram_clock_change_requirement_final = true;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c
index 80972ee5e55b..a352c703e258 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c
@@ -368,7 +368,9 @@ void dcn30_fpu_update_soc_for_wm_a(struct dc *dc, struct dc_state *context)
dc_assert_fp_enabled();
if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].valid) {
- context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.pstate_latency_us;
+ if (!context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching ||
+ context->bw_ctx.dml.soc.dram_clock_change_latency_us == 0)
+ context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.pstate_latency_us;
context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.sr_enter_plus_exit_time_us;
context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.sr_exit_time_us;
}
@@ -563,6 +565,20 @@ void dcn30_fpu_calculate_wm_and_dlg(
pipe_idx++;
}
+ // WA: restrict FPO to use first non-strobe mode (NV24 BW issue)
+ if (context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching &&
+ dc->dml.soc.num_chans <= 4 &&
+ context->bw_ctx.dml.vba.DRAMSpeed <= 1700 &&
+ context->bw_ctx.dml.vba.DRAMSpeed >= 1500) {
+
+ for (i = 0; i < dc->dml.soc.num_states; i++) {
+ if (dc->dml.soc.clock_limits[i].dram_speed_mts > 1700) {
+ context->bw_ctx.dml.vba.DRAMSpeed = dc->dml.soc.clock_limits[i].dram_speed_mts;
+ break;
+ }
+ }
+ }
+
dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel);
if (!pstate_en)
commit ce560ac40272a5c8b5b68a9d63a75edd9e66aed2
Author: Wesley Chalmers <Wesley.Chalmers at amd.com>
Date: Tue Feb 28 13:48:00 2023 -0500
drm/amd/display: Block optimize on consecutive FAMS enables
[WHY]
It is possible to commit state multiple times in rapid succession with
FAMS enabled; if each of these commits were to set optimized_required,
then the user may see latency.
[HOW]
fw_based_mclk_switching is currently not used in dc->clk_mgr; use it
to track whether the current state has FAMS enabled;
if it has, then do not disable FAMS in prepare_bandwidth, and do not set
optimized_required.
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Signed-off-by: Wesley Chalmers <Wesley.Chalmers at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
index 6ce10fd4bb1a..422fbf79da64 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
@@ -2117,6 +2117,9 @@ void dcn20_optimize_bandwidth(
dc_dmub_srv_p_state_delegate(dc,
true, context);
context->bw_ctx.bw.dcn.clk.p_state_change_support = true;
+ dc->clk_mgr->clks.fw_based_mclk_switching = true;
+ } else {
+ dc->clk_mgr->clks.fw_based_mclk_switching = false;
}
dc->clk_mgr->funcs->update_clocks(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
index 0411867654dd..8263a07f265f 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
@@ -983,9 +983,13 @@ void dcn30_set_disp_pattern_generator(const struct dc *dc,
}
void dcn30_prepare_bandwidth(struct dc *dc,
- struct dc_state *context)
+ struct dc_state *context)
{
- if (context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching) {
+ bool p_state_change_support = context->bw_ctx.bw.dcn.clk.p_state_change_support;
+ /* Any transition into an FPO config should disable MCLK switching first to avoid
+ * driver and FW P-State synchronization issues.
+ */
+ if (context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching || dc->clk_mgr->clks.fw_based_mclk_switching) {
dc->optimized_required = true;
context->bw_ctx.bw.dcn.clk.p_state_change_support = false;
}
@@ -996,7 +1000,19 @@ void dcn30_prepare_bandwidth(struct dc *dc,
dc->clk_mgr->funcs->set_max_memclk(dc->clk_mgr, dc->clk_mgr->bw_params->clk_table.entries[dc->clk_mgr->bw_params->clk_table.num_entries - 1].memclk_mhz);
dcn20_prepare_bandwidth(dc, context);
+ /*
+ * enabled -> enabled: do not disable
+ * enabled -> disabled: disable
+ * disabled -> enabled: don't care
+ * disabled -> disabled: don't care
+ */
+ if (!context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching)
+ dc_dmub_srv_p_state_delegate(dc, false, context);
- dc_dmub_srv_p_state_delegate(dc, false, context);
+ if (context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching || dc->clk_mgr->clks.fw_based_mclk_switching) {
+ /* After disabling P-State, restore the original value to ensure we get the correct P-State
+ * on the next optimize. */
+ context->bw_ctx.bw.dcn.clk.p_state_change_support = p_state_change_support;
+ }
}
commit 474f01015ffdb74e01c2eb3584a2822c64e7b2be
Author: Wesley Chalmers <Wesley.Chalmers at amd.com>
Date: Thu Nov 3 22:29:31 2022 -0400
drm/amd/display: Do not set drr on pipe commit
[WHY]
Writing to DRR registers such as OTG_V_TOTAL_MIN on the same frame as a
pipe commit can cause underflow.
[HOW]
Move DMUB p-state delegate into optimze_bandwidth; enabling FAMS sets
optimized_required.
This change expects that Freesync requests are blocked when
optimized_required is true.
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Signed-off-by: Wesley Chalmers <Wesley.Chalmers at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
index 5403e9399a46..6ce10fd4bb1a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
@@ -2113,6 +2113,12 @@ void dcn20_optimize_bandwidth(
if (hubbub->funcs->program_compbuf_size)
hubbub->funcs->program_compbuf_size(hubbub, context->bw_ctx.bw.dcn.compbuf_size_kb, true);
+ if (context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching) {
+ dc_dmub_srv_p_state_delegate(dc,
+ true, context);
+ context->bw_ctx.bw.dcn.clk.p_state_change_support = true;
+ }
+
dc->clk_mgr->funcs->update_clocks(
dc->clk_mgr,
context,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
index 0e071fbc9154..0411867654dd 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
@@ -985,11 +985,18 @@ void dcn30_set_disp_pattern_generator(const struct dc *dc,
void dcn30_prepare_bandwidth(struct dc *dc,
struct dc_state *context)
{
+ if (context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching) {
+ dc->optimized_required = true;
+ context->bw_ctx.bw.dcn.clk.p_state_change_support = false;
+ }
+
if (dc->clk_mgr->dc_mode_softmax_enabled)
if (dc->clk_mgr->clks.dramclk_khz <= dc->clk_mgr->bw_params->dc_mode_softmax_memclk * 1000 &&
context->bw_ctx.bw.dcn.clk.dramclk_khz > dc->clk_mgr->bw_params->dc_mode_softmax_memclk * 1000)
dc->clk_mgr->funcs->set_max_memclk(dc->clk_mgr, dc->clk_mgr->bw_params->clk_table.entries[dc->clk_mgr->bw_params->clk_table.num_entries - 1].memclk_mhz);
dcn20_prepare_bandwidth(dc, context);
+
+ dc_dmub_srv_p_state_delegate(dc, false, context);
}
commit d944eafed618a8507270b324ad9d5405bb7f0b3e
Author: Ville Syrjälä <ville.syrjala at linux.intel.com>
Date: Tue Apr 18 20:55:14 2023 +0300
drm/i915: Check pipe source size when using skl+ scalers
The skl+ scalers only sample 12 bits of PIPESRC so we can't
do any plane scaling at all when the pipe source size is >4k.
Make sure the pipe source size is also below the scaler's src
size limits. Might not be 100% accurate, but should at least be
safe. We can refine the limits later if we discover that recent
hw is less restricted.
Cc: stable at vger.kernel.org
Tested-by: Ross Zwisler <zwisler at google.com>
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/8357
Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230418175528.13117-2-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula at intel.com>
(cherry picked from commit 691248d4135fe3fae64b4ee0676bc96a7fd6950c)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
diff --git a/drivers/gpu/drm/i915/display/skl_scaler.c b/drivers/gpu/drm/i915/display/skl_scaler.c
index 473d53610b92..0e7e014fcc71 100644
--- a/drivers/gpu/drm/i915/display/skl_scaler.c
+++ b/drivers/gpu/drm/i915/display/skl_scaler.c
@@ -111,6 +111,8 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
const struct drm_display_mode *adjusted_mode =
&crtc_state->hw.adjusted_mode;
+ int pipe_src_w = drm_rect_width(&crtc_state->pipe_src);
+ int pipe_src_h = drm_rect_height(&crtc_state->pipe_src);
int min_src_w, min_src_h, min_dst_w, min_dst_h;
int max_src_w, max_src_h, max_dst_w, max_dst_h;
@@ -207,6 +209,21 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
return -EINVAL;
}
+ /*
+ * The pipe scaler does not use all the bits of PIPESRC, at least
+ * on the earlier platforms. So even when we're scaling a plane
+ * the *pipe* source size must not be too large. For simplicity
+ * we assume the limits match the scaler source size limits. Might
+ * not be 100% accurate on all platforms, but good enough for now.
+ */
+ if (pipe_src_w > max_src_w || pipe_src_h > max_src_h) {
+ drm_dbg_kms(&dev_priv->drm,
+ "scaler_user index %u.%u: pipe src size %ux%u "
+ "is out of scaler range\n",
+ crtc->pipe, scaler_user, pipe_src_w, pipe_src_h);
+ return -EINVAL;
+ }
+
/* mark this plane as a scaler user in crtc_state */
scaler_state->scaler_users |= (1 << scaler_user);
drm_dbg_kms(&dev_priv->drm, "scaler_user index %u.%u: "
commit 0c1f033159712b3d071cfe4a3ec0f36f1914453b
Author: Tom Rix <trix at redhat.com>
Date: Sat Apr 15 11:17:22 2023 -0400
drm/amd/display: set variable dccg314_init storage-class-specifier to static
smatch reports
drivers/gpu/drm/amd/amdgpu/../display/dc/dcn314/dcn314_dccg.c:277:6: warning: symbol
'dccg314_init' was not declared. Should it be static?
This variable is only used in one file so should be static.
Signed-off-by: Tom Rix <trix at redhat.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c
index 6f879265ad9c..de7bfba2c179 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c
@@ -274,7 +274,7 @@ static void dccg314_set_dpstreamclk(
}
}
-void dccg314_init(struct dccg *dccg)
+static void dccg314_init(struct dccg *dccg)
{
int otg_inst;
commit 64626c0ee13257e330bc09fa6a169385c0eaf9ca
Author: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Date: Mon Apr 3 19:23:29 2023 -0600
drm/amd/display: Use pointer in the memcpy
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
index 03718cc148e0..f1c1a4b5fcac 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
@@ -2333,7 +2333,7 @@ void dcn21_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params
k++;
}
- memcpy(dcn2_1_soc.clock_limits, s, sizeof(dcn2_1_soc.clock_limits));
+ memcpy(&dcn2_1_soc.clock_limits, s, sizeof(dcn2_1_soc.clock_limits));
if (clk_table->num_entries) {
dcn2_1_soc.num_states = clk_table->num_entries + 1;
commit b62f91569f9aa54b0a60d46a022482415cb968a9
Author: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Date: Mon Apr 3 15:40:21 2023 -0600
drm/amd/display: Remove wrong assignment of DP link rate
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
index 9fadac1b4c64..03718cc148e0 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
@@ -1360,7 +1360,6 @@ int dcn20_populate_dml_pipes_from_context(struct dc *dc,
pipes[pipe_cnt].dout.is_virtual = 1;
pipes[pipe_cnt].dout.output_type = dm_dp;
pipes[pipe_cnt].dout.dp_lanes = 4;
- pipes[pipe_cnt].dout.dp_rate = dm_dp_rate_hbr2;
}
switch (res_ctx->pipe_ctx[i].stream->timing.display_color_depth) {
commit ac7485cc363f2c603a3e1a7a609ef065ad56b19b
Author: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Date: Mon Apr 3 15:38:09 2023 -0600
drm/amd/display: Set dp_rate to dm_dp_rate_na by default
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
index b79014f04cef..9fadac1b4c64 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
@@ -1300,8 +1300,7 @@ int dcn20_populate_dml_pipes_from_context(struct dc *dc,
pipes[pipe_cnt].pipe.dest.pixel_rate_mhz *= 2;
pipes[pipe_cnt].pipe.dest.otg_inst = res_ctx->pipe_ctx[i].stream_res.tg->inst;
pipes[pipe_cnt].dout.dp_lanes = 4;
- if (res_ctx->pipe_ctx[i].stream->link)
- pipes[pipe_cnt].dout.dp_rate = dm_dp_rate_na;
+ pipes[pipe_cnt].dout.dp_rate = dm_dp_rate_na;
pipes[pipe_cnt].dout.is_virtual = 0;
pipes[pipe_cnt].pipe.dest.vtotal_min = res_ctx->pipe_ctx[i].stream->adjust.v_total_min;
pipes[pipe_cnt].pipe.dest.vtotal_max = res_ctx->pipe_ctx[i].stream->adjust.v_total_max;
commit 0fdf06e449b6d6d970c0709c71a8738cfe551ecc
Author: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Date: Mon Apr 3 15:37:46 2023 -0600
drm/amd/display: Set maximum VStartup if is DCN201
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
index 6e32dc68f7bc..b79014f04cef 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
@@ -1259,6 +1259,8 @@ int dcn20_populate_dml_pipes_from_context(struct dc *dc,
pipes[pipe_cnt].clks_cfg.refclk_mhz = dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000.0;
+ pipes[pipe_cnt].pipe.dest.use_maximum_vstartup = dc->ctx->dce_version == DCN_VERSION_2_01;
+
pipes[pipe_cnt].dout.dsc_enable = res_ctx->pipe_ctx[i].stream->timing.flags.DSC;
/* todo: rotation?*/
pipes[pipe_cnt].dout.dsc_slices = res_ctx->pipe_ctx[i].stream->timing.dsc_cfg.num_slices_h;
commit 83aeb49c8c467e9fe77c4f01c80472a4329db49c
Author: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Date: Mon Apr 3 15:29:36 2023 -0600
drm/amd/display: Adjust code identation and other minor details
This commit replaces spaces with tabs in multiple functions and adjusts
the indentation in some other parts of the code to improve readability.
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
index e30d1f60695d..0beb11d95eb7 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c
@@ -324,7 +324,6 @@ static const struct dcn10_link_enc_shift le_shift = {
static const struct dcn10_link_enc_mask le_mask = {
LINK_ENCODER_MASK_SH_LIST_DCN31(_MASK), \
-
//DPCS_DCN31_MASK_SH_LIST(_MASK)
};
@@ -2093,27 +2092,28 @@ static bool dcn32_resource_construct(
uint32_t pipe_fuses = 0;
uint32_t num_pipes = 4;
- #undef REG_STRUCT
- #define REG_STRUCT bios_regs
- bios_regs_init();
-
- #undef REG_STRUCT
- #define REG_STRUCT clk_src_regs
- clk_src_regs_init(0, A),
- clk_src_regs_init(1, B),
- clk_src_regs_init(2, C),
- clk_src_regs_init(3, D),
- clk_src_regs_init(4, E);
- #undef REG_STRUCT
- #define REG_STRUCT abm_regs
- abm_regs_init(0),
- abm_regs_init(1),
- abm_regs_init(2),
- abm_regs_init(3);
-
- #undef REG_STRUCT
- #define REG_STRUCT dccg_regs
- dccg_regs_init();
+#undef REG_STRUCT
+#define REG_STRUCT bios_regs
+ bios_regs_init();
+
+#undef REG_STRUCT
+#define REG_STRUCT clk_src_regs
+ clk_src_regs_init(0, A),
+ clk_src_regs_init(1, B),
+ clk_src_regs_init(2, C),
+ clk_src_regs_init(3, D),
+ clk_src_regs_init(4, E);
+
+#undef REG_STRUCT
+#define REG_STRUCT abm_regs
+ abm_regs_init(0),
+ abm_regs_init(1),
+ abm_regs_init(2),
+ abm_regs_init(3);
+
+#undef REG_STRUCT
+#define REG_STRUCT dccg_regs
+ dccg_regs_init();
DC_FP_START();
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
index 38d1f2be8cf3..6e32dc68f7bc 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
@@ -917,19 +917,19 @@ void dcn20_populate_dml_writeback_from_context(struct dc *dc,
}
void dcn20_fpu_set_wb_arb_params(struct mcif_arb_params *wb_arb_params,
- struct dc_state *context,
- display_e2e_pipe_params_st *pipes,
- int pipe_cnt, int i)
+ struct dc_state *context,
+ display_e2e_pipe_params_st *pipes,
+ int pipe_cnt, int i)
{
- int k;
+ int k;
- dc_assert_fp_enabled();
+ dc_assert_fp_enabled();
- for (k = 0; k < sizeof(wb_arb_params->cli_watermark)/sizeof(wb_arb_params->cli_watermark[0]); k++) {
- wb_arb_params->cli_watermark[k] = get_wm_writeback_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
- wb_arb_params->pstate_watermark[k] = get_wm_writeback_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
- }
- wb_arb_params->time_per_pixel = 16.0 * 1000 / (context->res_ctx.pipe_ctx[i].stream->phy_pix_clk / 1000); /* 4 bit fraction, ms */
+ for (k = 0; k < sizeof(wb_arb_params->cli_watermark)/sizeof(wb_arb_params->cli_watermark[0]); k++) {
+ wb_arb_params->cli_watermark[k] = get_wm_writeback_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+ wb_arb_params->pstate_watermark[k] = get_wm_writeback_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+ }
+ wb_arb_params->time_per_pixel = 16.0 * 1000 / (context->res_ctx.pipe_ctx[i].stream->phy_pix_clk / 1000); /* 4 bit fraction, ms */
}
static bool is_dtbclk_required(struct dc *dc, struct dc_state *context)
@@ -1037,11 +1037,11 @@ static void dcn20_adjust_freesync_v_startup(
*vstartup_start = ((newVstartup > *vstartup_start) ? newVstartup : *vstartup_start);
}
-void dcn20_calculate_dlg_params(
- struct dc *dc, struct dc_state *context,
- display_e2e_pipe_params_st *pipes,
- int pipe_cnt,
- int vlevel)
+void dcn20_calculate_dlg_params(struct dc *dc,
+ struct dc_state *context,
+ display_e2e_pipe_params_st *pipes,
+ int pipe_cnt,
+ int vlevel)
{
int i, pipe_idx;
@@ -1083,6 +1083,7 @@ void dcn20_calculate_dlg_params(
pipes[pipe_idx].pipe.dest.vupdate_offset = get_vupdate_offset(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
pipes[pipe_idx].pipe.dest.vupdate_width = get_vupdate_width(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
pipes[pipe_idx].pipe.dest.vready_offset = get_vready_offset(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
+
if (context->res_ctx.pipe_ctx[i].stream->mall_stream_config.type == SUBVP_PHANTOM) {
// Phantom pipe requires that DET_SIZE = 0 and no unbounded requests
context->res_ctx.pipe_ctx[i].det_buffer_size_kb = 0;
@@ -1091,6 +1092,7 @@ void dcn20_calculate_dlg_params(
context->res_ctx.pipe_ctx[i].det_buffer_size_kb = context->bw_ctx.dml.ip.det_buffer_size_kbytes;
context->res_ctx.pipe_ctx[i].unbounded_req = pipes[pipe_idx].pipe.src.unbounded_req_mode;
}
+
if (context->bw_ctx.bw.dcn.clk.dppclk_khz < pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000)
context->bw_ctx.bw.dcn.clk.dppclk_khz = pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000;
context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz =
@@ -1118,6 +1120,7 @@ void dcn20_calculate_dlg_params(
if (!context->res_ctx.pipe_ctx[i].stream)
continue;
+ /* cstate disabled on 201 */
if (dc->ctx->dce_version == DCN_VERSION_2_01)
cstate_en = false;
@@ -1201,11 +1204,10 @@ static void swizzle_to_dml_params(
}
}
-int dcn20_populate_dml_pipes_from_context(
- struct dc *dc,
- struct dc_state *context,
- display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+int dcn20_populate_dml_pipes_from_context(struct dc *dc,
+ struct dc_state *context,
+ display_e2e_pipe_params_st *pipes,
+ bool fast_validate)
{
int pipe_cnt, i;
bool synchronized_vblank = true;
@@ -1507,6 +1509,7 @@ int dcn20_populate_dml_pipes_from_context(
default:
break;
}
+
pipes[pipe_cnt].pipe.src.viewport_y_y = scl->viewport.y;
pipes[pipe_cnt].pipe.src.viewport_y_c = scl->viewport_c.y;
pipes[pipe_cnt].pipe.src.viewport_x_y = scl->viewport.x;
@@ -1615,13 +1618,12 @@ int dcn20_populate_dml_pipes_from_context(
return pipe_cnt;
}
-void dcn20_calculate_wm(
- struct dc *dc, struct dc_state *context,
- display_e2e_pipe_params_st *pipes,
- int *out_pipe_cnt,
- int *pipe_split_from,
- int vlevel,
- bool fast_validate)
+void dcn20_calculate_wm(struct dc *dc, struct dc_state *context,
+ display_e2e_pipe_params_st *pipes,
+ int *out_pipe_cnt,
+ int *pipe_split_from,
+ int vlevel,
+ bool fast_validate)
{
int pipe_cnt, i, pipe_idx;
@@ -1733,8 +1735,11 @@ void dcn20_calculate_wm(
context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
}
-void dcn20_update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb,
- struct pp_smu_nv_clock_table *max_clocks, unsigned int *uclk_states, unsigned int num_states)
+void dcn20_update_bounding_box(struct dc *dc,
+ struct _vcs_dpi_soc_bounding_box_st *bb,
+ struct pp_smu_nv_clock_table *max_clocks,
+ unsigned int *uclk_states,
+ unsigned int num_states)
{
int num_calculated_states = 0;
int min_dcfclk = 0;
@@ -1796,9 +1801,8 @@ void dcn20_update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_s
bb->clock_limits[num_calculated_states].state = bb->num_states;
}
-void dcn20_cap_soc_clocks(
- struct _vcs_dpi_soc_bounding_box_st *bb,
- struct pp_smu_nv_clock_table max_clocks)
+void dcn20_cap_soc_clocks(struct _vcs_dpi_soc_bounding_box_st *bb,
+ struct pp_smu_nv_clock_table max_clocks)
{
int i;
@@ -1954,80 +1958,80 @@ validate_out:
}
bool dcn20_validate_bandwidth_fp(struct dc *dc,
- struct dc_state *context,
- bool fast_validate)
+ struct dc_state *context,
+ bool fast_validate)
{
- bool voltage_supported = false;
- bool full_pstate_supported = false;
- bool dummy_pstate_supported = false;
- double p_state_latency_us;
+ bool voltage_supported = false;
+ bool full_pstate_supported = false;
+ bool dummy_pstate_supported = false;
+ double p_state_latency_us;
- dc_assert_fp_enabled();
+ dc_assert_fp_enabled();
- p_state_latency_us = context->bw_ctx.dml.soc.dram_clock_change_latency_us;
- context->bw_ctx.dml.soc.disable_dram_clock_change_vactive_support =
- dc->debug.disable_dram_clock_change_vactive_support;
- context->bw_ctx.dml.soc.allow_dram_clock_one_display_vactive =
- dc->debug.enable_dram_clock_change_one_display_vactive;
+ p_state_latency_us = context->bw_ctx.dml.soc.dram_clock_change_latency_us;
+ context->bw_ctx.dml.soc.disable_dram_clock_change_vactive_support =
+ dc->debug.disable_dram_clock_change_vactive_support;
+ context->bw_ctx.dml.soc.allow_dram_clock_one_display_vactive =
+ dc->debug.enable_dram_clock_change_one_display_vactive;
- /*Unsafe due to current pipe merge and split logic*/
- ASSERT(context != dc->current_state);
+ /*Unsafe due to current pipe merge and split logic*/
+ ASSERT(context != dc->current_state);
- if (fast_validate) {
- return dcn20_validate_bandwidth_internal(dc, context, true);
- }
+ if (fast_validate) {
+ return dcn20_validate_bandwidth_internal(dc, context, true);
+ }
- // Best case, we support full UCLK switch latency
- voltage_supported = dcn20_validate_bandwidth_internal(dc, context, false);
- full_pstate_supported = context->bw_ctx.bw.dcn.clk.p_state_change_support;
+ // Best case, we support full UCLK switch latency
+ voltage_supported = dcn20_validate_bandwidth_internal(dc, context, false);
+ full_pstate_supported = context->bw_ctx.bw.dcn.clk.p_state_change_support;
- if (context->bw_ctx.dml.soc.dummy_pstate_latency_us == 0 ||
- (voltage_supported && full_pstate_supported)) {
- context->bw_ctx.bw.dcn.clk.p_state_change_support = full_pstate_supported;
- goto restore_dml_state;
- }
+ if (context->bw_ctx.dml.soc.dummy_pstate_latency_us == 0 ||
+ (voltage_supported && full_pstate_supported)) {
+ context->bw_ctx.bw.dcn.clk.p_state_change_support = full_pstate_supported;
+ goto restore_dml_state;
+ }
- // Fallback: Try to only support G6 temperature read latency
- context->bw_ctx.dml.soc.dram_clock_change_latency_us = context->bw_ctx.dml.soc.dummy_pstate_latency_us;
+ // Fallback: Try to only support G6 temperature read latency
+ context->bw_ctx.dml.soc.dram_clock_change_latency_us = context->bw_ctx.dml.soc.dummy_pstate_latency_us;
- voltage_supported = dcn20_validate_bandwidth_internal(dc, context, false);
- dummy_pstate_supported = context->bw_ctx.bw.dcn.clk.p_state_change_support;
+ voltage_supported = dcn20_validate_bandwidth_internal(dc, context, false);
+ dummy_pstate_supported = context->bw_ctx.bw.dcn.clk.p_state_change_support;
- if (voltage_supported && (dummy_pstate_supported || !(context->stream_count))) {
- context->bw_ctx.bw.dcn.clk.p_state_change_support = false;
- goto restore_dml_state;
- }
+ if (voltage_supported && (dummy_pstate_supported || !(context->stream_count))) {
+ context->bw_ctx.bw.dcn.clk.p_state_change_support = false;
+ goto restore_dml_state;
+ }
- // ERROR: fallback is supposed to always work.
- ASSERT(false);
+ // ERROR: fallback is supposed to always work.
+ ASSERT(false);
restore_dml_state:
- context->bw_ctx.dml.soc.dram_clock_change_latency_us = p_state_latency_us;
- return voltage_supported;
+ context->bw_ctx.dml.soc.dram_clock_change_latency_us = p_state_latency_us;
+ return voltage_supported;
}
void dcn20_fpu_set_wm_ranges(int i,
- struct pp_smu_wm_range_sets *ranges,
- struct _vcs_dpi_soc_bounding_box_st *loaded_bb)
+ struct pp_smu_wm_range_sets *ranges,
+ struct _vcs_dpi_soc_bounding_box_st *loaded_bb)
{
- dc_assert_fp_enabled();
+ dc_assert_fp_enabled();
- ranges->reader_wm_sets[i].min_fill_clk_mhz = (i > 0) ? (loaded_bb->clock_limits[i - 1].dram_speed_mts / 16) + 1 : 0;
- ranges->reader_wm_sets[i].max_fill_clk_mhz = loaded_bb->clock_limits[i].dram_speed_mts / 16;
+ ranges->reader_wm_sets[i].min_fill_clk_mhz = (i > 0) ? (loaded_bb->clock_limits[i - 1].dram_speed_mts / 16) + 1 : 0;
+ ranges->reader_wm_sets[i].max_fill_clk_mhz = loaded_bb->clock_limits[i].dram_speed_mts / 16;
}
void dcn20_fpu_adjust_dppclk(struct vba_vars_st *v,
- int vlevel,
- int max_mpc_comb,
- int pipe_idx,
- bool is_validating_bw)
+ int vlevel,
+ int max_mpc_comb,
+ int pipe_idx,
+ bool is_validating_bw)
{
- dc_assert_fp_enabled();
+ dc_assert_fp_enabled();
- if (is_validating_bw)
- v->RequiredDPPCLK[vlevel][max_mpc_comb][pipe_idx] *= 2;
- else
- v->RequiredDPPCLK[vlevel][max_mpc_comb][pipe_idx] /= 2;
+ if (is_validating_bw)
+ v->RequiredDPPCLK[vlevel][max_mpc_comb][pipe_idx] *= 2;
+ else
+ v->RequiredDPPCLK[vlevel][max_mpc_comb][pipe_idx] /= 2;
}
int dcn21_populate_dml_pipes_from_context(struct dc *dc,
commit ef3d74aa7e5d0ba4e9fc00f1409652e29f46fc59
Author: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Date: Mon Apr 3 14:10:27 2023 -0600
drm/amd/display: Add missing mclk update
When using FPO, there is some misconfiguration that happens for the lack
of configuration of the MCLK switch in some circumstances. This commit
adds the required field update when using the MCLK switch.
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
index f0037cb43dca..23a972f2885f 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
@@ -1331,6 +1331,11 @@ static void dcn32_calculate_dlg_params(struct dc *dc, struct dc_state *context,
context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb]
!= dm_dram_clock_change_unsupported;
+ /* Pstate change might not be supported by hardware, but it might be
+ * possible with firmware driven vertical blank stretching.
+ */
+ context->bw_ctx.bw.dcn.clk.p_state_change_support |= context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching;
+
context->bw_ctx.bw.dcn.clk.dppclk_khz = 0;
context->bw_ctx.bw.dcn.clk.dtbclk_en = is_dtbclk_required(dc, context);
context->bw_ctx.bw.dcn.clk.ref_dtbclk_khz = context->bw_ctx.dml.vba.DTBCLKPerState[vlevel] * 1000;
commit 764ba43d34ac5fd16e0e377643f89a7208f1f67b
Author: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Date: Mon Apr 3 14:06:16 2023 -0600
drm/amd/display: Update bouding box values for DCN32
All clock values came from firmware, but bounding box values can be
helpful in some debug situations. This commit updates some of the values
associated with clock speed and memory channels.
Reviewed-by: Aurabindo Pillai <aurabindo.pillai at amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
index 4548320217fc..f0037cb43dca 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
@@ -109,7 +109,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_2_soc = {
{
.state = 0,
.dcfclk_mhz = 1564.0,
- .fabricclk_mhz = 400.0,
+ .fabricclk_mhz = 2500.0,
.dispclk_mhz = 2150.0,
.dppclk_mhz = 2150.0,
.phyclk_mhz = 810.0,
@@ -117,7 +117,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_2_soc = {
.phyclk_d32_mhz = 625.0,
.socclk_mhz = 1200.0,
.dscclk_mhz = 716.667,
- .dram_speed_mts = 16000.0,
+ .dram_speed_mts = 18000.0,
.dtbclk_mhz = 1564.0,
},
},
@@ -148,7 +148,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_2_soc = {
.max_avg_fabric_bw_use_normal_percent = 60.0,
.max_avg_dram_bw_use_normal_strobe_percent = 50.0,
.max_avg_dram_bw_use_normal_percent = 15.0,
- .num_chans = 8,
+ .num_chans = 24,
.dram_channel_width_bytes = 2,
.fabric_datapath_to_dcn_data_return_bytes = 64,
.return_bus_width_bytes = 64,
commit 38eecbe086a4e52f54b2bbda8feba65d44addbef
Author: Chong Li <chongli2 at amd.com>
Date: Fri Apr 14 13:51:19 2023 +0800
drm/amdgpu: release gpu full access after "amdgpu_device_ip_late_init"
[WHY]
Function "amdgpu_irq_update()" called by "amdgpu_device_ip_late_init()" is an atomic context.
We shouldn't access registers through KIQ since "msleep()" may be called in "amdgpu_kiq_rreg()".
[HOW]
Move function "amdgpu_virt_release_full_gpu()" after function "amdgpu_device_ip_late_init()",
to ensure that registers be accessed through RLCG instead of KIQ.
Call Trace:
<TASK>
show_stack+0x52/0x69
dump_stack_lvl+0x49/0x6d
dump_stack+0x10/0x18
__schedule_bug.cold+0x4f/0x6b
__schedule+0x473/0x5d0
? __wake_up_klogd.part.0+0x40/0x70
? vprintk_emit+0xbe/0x1f0
schedule+0x68/0x110
schedule_timeout+0x87/0x160
? timer_migration_handler+0xa0/0xa0
msleep+0x2d/0x50
amdgpu_kiq_rreg+0x18d/0x1f0 [amdgpu]
amdgpu_device_rreg.part.0+0x59/0xd0 [amdgpu]
amdgpu_device_rreg+0x3a/0x50 [amdgpu]
amdgpu_sriov_rreg+0x3c/0xb0 [amdgpu]
gfx_v10_0_set_gfx_eop_interrupt_state.constprop.0+0x16c/0x190 [amdgpu]
gfx_v10_0_set_eop_interrupt_state+0xa5/0xb0 [amdgpu]
amdgpu_irq_update+0x53/0x80 [amdgpu]
amdgpu_irq_get+0x7c/0xb0 [amdgpu]
amdgpu_fence_driver_hw_init+0x58/0x90 [amdgpu]
amdgpu_device_init.cold+0x16b7/0x2022 [amdgpu]
Signed-off-by: Chong Li <chongli2 at amd.com>
Reviewed-by: JingWen.Chen2 at amd.com
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index a2292acf06d0..9b1eaba85bbd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2539,8 +2539,6 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
amdgpu_fru_get_product_info(adev);
init_failed:
- if (amdgpu_sriov_vf(adev))
- amdgpu_virt_release_full_gpu(adev, true);
return r;
}
@@ -3859,18 +3857,6 @@ fence_driver_init:
r = amdgpu_device_ip_init(adev);
if (r) {
- /* failed in exclusive mode due to timeout */
- if (amdgpu_sriov_vf(adev) &&
- !amdgpu_sriov_runtime(adev) &&
- amdgpu_virt_mmio_blocked(adev) &&
- !amdgpu_virt_wait_reset(adev)) {
- dev_err(adev->dev, "VF exclusive mode timeout\n");
- /* Don't send request since VF is inactive. */
- adev->virt.caps &= ~AMDGPU_SRIOV_CAPS_RUNTIME;
- adev->virt.ops = NULL;
- r = -EAGAIN;
- goto release_ras_con;
- }
dev_err(adev->dev, "amdgpu_device_ip_init failed\n");
amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, 0, 0);
goto release_ras_con;
@@ -3939,8 +3925,10 @@ fence_driver_init:
msecs_to_jiffies(AMDGPU_RESUME_MS));
}
- if (amdgpu_sriov_vf(adev))
+ if (amdgpu_sriov_vf(adev)) {
+ amdgpu_virt_release_full_gpu(adev, true);
flush_delayed_work(&adev->delayed_init_work);
+ }
r = sysfs_create_files(&adev->dev->kobj, amdgpu_dev_attributes);
if (r)
@@ -3980,6 +3968,20 @@ fence_driver_init:
return 0;
release_ras_con:
+ if (amdgpu_sriov_vf(adev))
+ amdgpu_virt_release_full_gpu(adev, true);
+
+ /* failed in exclusive mode due to timeout */
+ if (amdgpu_sriov_vf(adev) &&
+ !amdgpu_sriov_runtime(adev) &&
+ amdgpu_virt_mmio_blocked(adev) &&
+ !amdgpu_virt_wait_reset(adev)) {
+ dev_err(adev->dev, "VF exclusive mode timeout\n");
+ /* Don't send request since VF is inactive. */
+ adev->virt.caps &= ~AMDGPU_SRIOV_CAPS_RUNTIME;
+ adev->virt.ops = NULL;
+ r = -EAGAIN;
+ }
amdgpu_release_ras_context(adev);
failed:
commit 8d9cdb4674f6e4e7fc789f8184a58c73eeadc16c
Author: Tom Rix <trix at redhat.com>
Date: Fri Apr 14 08:03:44 2023 -0400
drm/amd/pm: change pmfw_decoded_link_width, speed variables to globals
gcc with W=1 reports
In file included from drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu13/smu_v13_0.c:36:
./drivers/gpu/drm/amd/amdgpu/../pm/swsmu/inc/smu_v13_0.h:66:18: error:
‘pmfw_decoded_link_width’ defined but not used [-Werror=unused-const-variable=]
66 | static const int pmfw_decoded_link_width[7] = {0, 1, 2, 4, 8, 12, 16};
| ^~~~~~~~~~~~~~~~~~~~~~~
./drivers/gpu/drm/amd/amdgpu/../pm/swsmu/inc/smu_v13_0.h:65:18: error:
‘pmfw_decoded_link_speed’ defined but not used [-Werror=unused-const-variable=]
65 | static const int pmfw_decoded_link_speed[5] = {1, 2, 3, 4, 5};
| ^~~~~~~~~~~~~~~~~~~~~~~
These variables are defined and used in smu_v13_0_7_ppt.c and smu_v13_0_0_ppt.c.
There should be only one definition. So define the variables as globals
in smu_v13_0.c
Signed-off-by: Tom Rix <trix at redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h
index 7944ce80e5c3..df3baaab0037 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h
@@ -62,8 +62,8 @@
#define CTF_OFFSET_HOTSPOT 5
#define CTF_OFFSET_MEM 5
-static const int pmfw_decoded_link_speed[5] = {1, 2, 3, 4, 5};
-static const int pmfw_decoded_link_width[7] = {0, 1, 2, 4, 8, 12, 16};
+extern const int pmfw_decoded_link_speed[5];
+extern const int pmfw_decoded_link_width[7];
#define DECODE_GEN_SPEED(gen_speed_idx) (pmfw_decoded_link_speed[gen_speed_idx])
#define DECODE_LANE_WIDTH(lane_width_idx) (pmfw_decoded_link_width[lane_width_idx])
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c
index 73175c993da9..393c6a7b9609 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c
@@ -85,6 +85,9 @@ MODULE_FIRMWARE("amdgpu/smu_13_0_10.bin");
static const int link_width[] = {0, 1, 2, 4, 8, 12, 16};
static const int link_speed[] = {25, 50, 80, 160};
+const int pmfw_decoded_link_speed[5] = {1, 2, 3, 4, 5};
+const int pmfw_decoded_link_width[7] = {0, 1, 2, 4, 8, 12, 16};
+
int smu_v13_0_init_microcode(struct smu_context *smu)
{
struct amdgpu_device *adev = smu->adev;
commit 4de867fc237487ce2951a8231d7390237d3f3be8
Author: Jane Jian <Jane.Jian at amd.com>
Date: Thu Apr 13 10:49:06 2023 +0800
drm/amdgpu/vcn: fix mmsch ctx table size
add jpeg table size to ctx table size rather than override it
Signed-off-by: Jane Jian <Jane.Jian at amd.com>
Reviewed-by: JingWen Chen <JingWen.Chen2 at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c
index a6ad678fd507..77e1e64aa1d1 100644
--- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c
@@ -430,7 +430,7 @@ static int jpeg_v4_0_start_sriov(struct amdgpu_device *adev)
MMSCH_COMMAND__END;
header.version = MMSCH_VERSION;
- header.total_size = sizeof(struct mmsch_v4_0_init_header) >> 2;
+ header.total_size = RREG32_SOC15(VCN, 0, regMMSCH_VF_CTX_SIZE);
header.jpegdec.init_status = 0;
header.jpegdec.table_offset = 0;
More information about the openchrome-devel
mailing list