Merge tag 'drm-fixes-2025-09-05' of https://gitlab.freedesktop.org/drm/kernel
Pull drm fixes from Dave Airlie:
"Weekly drm fixes roundup, nouveau has two fixes for fence/irq racing
problems that should fix a bunch of instability in userspace.
Otherwise amdgpu along with some single fixes to bridge, xe, ivpu.
Looks about usual for this time in the release.
scheduler:
- fix race in unschedulable tracepoint
bridge:
- ti-sn65dsi86: fix REFCLK setting
xe:
- Fix incorrect migration of backed-up object to VRAM
amdgpu:
- UserQ fixes
- MES 11 fix
- eDP/LVDS fix
- Fix non-DC audio clean up
- Fix duplicate cursor issue
- Fix error path in PSP init
nouveau:
- fix nonstall interrupt handling
- fix race on fence vs irq emission
- update MAINTAINERS entry
ivpu:
- prevent recovery work during device remove"
* tag 'drm-fixes-2025-09-05' of https://gitlab.freedesktop.org/drm/kernel:
drm/amd/amdgpu: Fix missing error return on kzalloc failure
drm/bridge: ti-sn65dsi86: fix REFCLK setting
MAINTAINERS: Update git entry for nouveau
drm/xe: Fix incorrect migration of backed-up object to VRAM
drm/sched: Fix racy access to drm_sched_entity.dependency
accel/ivpu: Prevent recovery work from being queued during device removal
nouveau: Membar before between semaphore writes and the interrupt
nouveau: fix disabling the nonstall irq due to storm code
drm/amd/display: Clear the CUR_ENABLE register on DCN314 w/out DPP PG
drm/amdgpu: drop hw access in non-DC audio fini
drm/amd: Re-enable common modes for eDP and LVDS
drm/amdgpu/mes11: make MES_MISC_OP_CHANGE_CONFIG failure non-fatal
drm/amdgpu/sdma: bump firmware version checks for user queue support
diff --git a/MAINTAINERS b/MAINTAINERS
index 2df02e4..cd7ff55 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -7821,7 +7821,7 @@
Q: https://gitlab.freedesktop.org/drm/nouveau/-/merge_requests
B: https://gitlab.freedesktop.org/drm/nouveau/-/issues
C: irc://irc.oftc.net/nouveau
-T: git https://gitlab.freedesktop.org/drm/nouveau.git
+T: git https://gitlab.freedesktop.org/drm/misc/kernel.git
F: drivers/gpu/drm/nouveau/
F: include/uapi/drm/nouveau_drm.h
diff --git a/drivers/accel/ivpu/ivpu_drv.c b/drivers/accel/ivpu/ivpu_drv.c
index 3d6d524..3289751 100644
--- a/drivers/accel/ivpu/ivpu_drv.c
+++ b/drivers/accel/ivpu/ivpu_drv.c
@@ -677,7 +677,7 @@ static void ivpu_bo_unbind_all_user_contexts(struct ivpu_device *vdev)
static void ivpu_dev_fini(struct ivpu_device *vdev)
{
ivpu_jobs_abort_all(vdev);
- ivpu_pm_cancel_recovery(vdev);
+ ivpu_pm_disable_recovery(vdev);
ivpu_pm_disable(vdev);
ivpu_prepare_for_reset(vdev);
ivpu_shutdown(vdev);
diff --git a/drivers/accel/ivpu/ivpu_pm.c b/drivers/accel/ivpu/ivpu_pm.c
index eacda1d..475ddc9 100644
--- a/drivers/accel/ivpu/ivpu_pm.c
+++ b/drivers/accel/ivpu/ivpu_pm.c
@@ -417,10 +417,10 @@ void ivpu_pm_init(struct ivpu_device *vdev)
ivpu_dbg(vdev, PM, "Autosuspend delay = %d\n", delay);
}
-void ivpu_pm_cancel_recovery(struct ivpu_device *vdev)
+void ivpu_pm_disable_recovery(struct ivpu_device *vdev)
{
drm_WARN_ON(&vdev->drm, delayed_work_pending(&vdev->pm->job_timeout_work));
- cancel_work_sync(&vdev->pm->recovery_work);
+ disable_work_sync(&vdev->pm->recovery_work);
}
void ivpu_pm_enable(struct ivpu_device *vdev)
diff --git a/drivers/accel/ivpu/ivpu_pm.h b/drivers/accel/ivpu/ivpu_pm.h
index 89b264c..a2aa7a2 100644
--- a/drivers/accel/ivpu/ivpu_pm.h
+++ b/drivers/accel/ivpu/ivpu_pm.h
@@ -25,7 +25,7 @@ struct ivpu_pm_info {
void ivpu_pm_init(struct ivpu_device *vdev);
void ivpu_pm_enable(struct ivpu_device *vdev);
void ivpu_pm_disable(struct ivpu_device *vdev);
-void ivpu_pm_cancel_recovery(struct ivpu_device *vdev);
+void ivpu_pm_disable_recovery(struct ivpu_device *vdev);
int ivpu_pm_suspend_cb(struct device *dev);
int ivpu_pm_resume_cb(struct device *dev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index 2348431..693357c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -448,7 +448,7 @@ static int psp_sw_init(struct amdgpu_ip_block *ip_block)
psp->cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
if (!psp->cmd) {
dev_err(adev->dev, "Failed to allocate memory to command buffer!\n");
- ret = -ENOMEM;
+ return -ENOMEM;
}
adev->psp.xgmi_context.supports_extended_data =
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
index bf7c22f..ba73518 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -1462,17 +1462,12 @@ static int dce_v10_0_audio_init(struct amdgpu_device *adev)
static void dce_v10_0_audio_fini(struct amdgpu_device *adev)
{
- int i;
-
if (!amdgpu_audio)
return;
if (!adev->mode_info.audio.enabled)
return;
- for (i = 0; i < adev->mode_info.audio.num_pins; i++)
- dce_v10_0_audio_enable(adev, &adev->mode_info.audio.pin[i], false);
-
adev->mode_info.audio.enabled = false;
}
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
index 47e0578..b01d88d 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -1511,17 +1511,12 @@ static int dce_v11_0_audio_init(struct amdgpu_device *adev)
static void dce_v11_0_audio_fini(struct amdgpu_device *adev)
{
- int i;
-
if (!amdgpu_audio)
return;
if (!adev->mode_info.audio.enabled)
return;
- for (i = 0; i < adev->mode_info.audio.num_pins; i++)
- dce_v11_0_audio_enable(adev, &adev->mode_info.audio.pin[i], false);
-
adev->mode_info.audio.enabled = false;
}
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
index 276c025..81760a2 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
@@ -1451,17 +1451,12 @@ static int dce_v6_0_audio_init(struct amdgpu_device *adev)
static void dce_v6_0_audio_fini(struct amdgpu_device *adev)
{
- int i;
-
if (!amdgpu_audio)
return;
if (!adev->mode_info.audio.enabled)
return;
- for (i = 0; i < adev->mode_info.audio.num_pins; i++)
- dce_v6_0_audio_enable(adev, &adev->mode_info.audio.pin[i], false);
-
adev->mode_info.audio.enabled = false;
}
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
index e62ccf9..19a265b 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -1443,17 +1443,12 @@ static int dce_v8_0_audio_init(struct amdgpu_device *adev)
static void dce_v8_0_audio_fini(struct amdgpu_device *adev)
{
- int i;
-
if (!amdgpu_audio)
return;
if (!adev->mode_info.audio.enabled)
return;
- for (i = 0; i < adev->mode_info.audio.num_pins; i++)
- dce_v8_0_audio_enable(adev, &adev->mode_info.audio.pin[i], false);
-
adev->mode_info.audio.enabled = false;
}
diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
index 28eb846..3f6a828c 100644
--- a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
@@ -641,8 +641,9 @@ static int mes_v11_0_misc_op(struct amdgpu_mes *mes,
break;
case MES_MISC_OP_CHANGE_CONFIG:
if ((mes->adev->mes.sched_version & AMDGPU_MES_VERSION_MASK) < 0x63) {
- dev_err(mes->adev->dev, "MES FW version must be larger than 0x63 to support limit single process feature.\n");
- return -EINVAL;
+ dev_warn_once(mes->adev->dev,
+ "MES FW version must be larger than 0x63 to support limit single process feature.\n");
+ return 0;
}
misc_pkt.opcode = MESAPI_MISC__CHANGE_CONFIG;
misc_pkt.change_config.opcode =
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
index e6d8edd..db6e419 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
@@ -1377,7 +1377,7 @@ static int sdma_v6_0_sw_init(struct amdgpu_ip_block *ip_block)
switch (amdgpu_ip_version(adev, SDMA0_HWIP, 0)) {
case IP_VERSION(6, 0, 0):
- if ((adev->sdma.instance[0].fw_version >= 24) && !adev->sdma.disable_uq)
+ if ((adev->sdma.instance[0].fw_version >= 27) && !adev->sdma.disable_uq)
adev->userq_funcs[AMDGPU_HW_IP_DMA] = &userq_mes_funcs;
break;
case IP_VERSION(6, 0, 1):
@@ -1385,11 +1385,11 @@ static int sdma_v6_0_sw_init(struct amdgpu_ip_block *ip_block)
adev->userq_funcs[AMDGPU_HW_IP_DMA] = &userq_mes_funcs;
break;
case IP_VERSION(6, 0, 2):
- if ((adev->sdma.instance[0].fw_version >= 21) && !adev->sdma.disable_uq)
+ if ((adev->sdma.instance[0].fw_version >= 23) && !adev->sdma.disable_uq)
adev->userq_funcs[AMDGPU_HW_IP_DMA] = &userq_mes_funcs;
break;
case IP_VERSION(6, 0, 3):
- if ((adev->sdma.instance[0].fw_version >= 25) && !adev->sdma.disable_uq)
+ if ((adev->sdma.instance[0].fw_version >= 27) && !adev->sdma.disable_uq)
adev->userq_funcs[AMDGPU_HW_IP_DMA] = &userq_mes_funcs;
break;
case IP_VERSION(6, 1, 0):
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 a0ca3b2..7808a64 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -8381,8 +8381,7 @@ static int amdgpu_dm_connector_get_modes(struct drm_connector *connector)
drm_add_modes_noedid(connector, 1920, 1080);
} else {
amdgpu_dm_connector_ddc_get_modes(connector, drm_edid);
- if (encoder && (connector->connector_type != DRM_MODE_CONNECTOR_eDP) &&
- (connector->connector_type != DRM_MODE_CONNECTOR_LVDS))
+ if (encoder)
amdgpu_dm_connector_add_common_modes(encoder, connector);
amdgpu_dm_connector_add_freesync_modes(connector, drm_edid);
}
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c
index 75fb77bc..01480a0 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c
@@ -520,6 +520,15 @@ void dpp1_dppclk_control(
REG_UPDATE(DPP_CONTROL, DPP_CLOCK_ENABLE, 0);
}
+void dpp_force_disable_cursor(struct dpp *dpp_base)
+{
+ struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
+
+ /* Force disable cursor */
+ REG_UPDATE(CURSOR0_CONTROL, CUR0_ENABLE, 0);
+ dpp_base->pos.cur0_ctl.bits.cur0_enable = 0;
+}
+
static const struct dpp_funcs dcn10_dpp_funcs = {
.dpp_read_state = dpp_read_state,
.dpp_reset = dpp_reset,
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.h
index c48139b..f466182 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.h
@@ -1525,4 +1525,6 @@ void dpp1_construct(struct dcn10_dpp *dpp1,
void dpp1_cm_get_gamut_remap(struct dpp *dpp_base,
struct dpp_grph_csc_adjustment *adjust);
+void dpp_force_disable_cursor(struct dpp *dpp_base);
+
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c
index 2d70586..09be2a9 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c
@@ -1494,6 +1494,7 @@ static struct dpp_funcs dcn30_dpp_funcs = {
.dpp_dppclk_control = dpp1_dppclk_control,
.dpp_set_hdr_multiplier = dpp3_set_hdr_multiplier,
.dpp_get_gamut_remap = dpp3_cm_get_gamut_remap,
+ .dpp_force_disable_cursor = dpp_force_disable_cursor,
};
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c
index e68f21f..5609845 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c
@@ -528,3 +528,75 @@ void dcn314_disable_link_output(struct dc_link *link,
apply_symclk_on_tx_off_wa(link);
}
+
+/**
+ * dcn314_dpp_pg_control - DPP power gate control.
+ *
+ * @hws: dce_hwseq reference.
+ * @dpp_inst: DPP instance reference.
+ * @power_on: true if we want to enable power gate, false otherwise.
+ *
+ * Enable or disable power gate in the specific DPP instance.
+ * If power gating is disabled, will force disable cursor in the DPP instance.
+ */
+void dcn314_dpp_pg_control(
+ struct dce_hwseq *hws,
+ unsigned int dpp_inst,
+ bool power_on)
+{
+ uint32_t power_gate = power_on ? 0 : 1;
+ uint32_t pwr_status = power_on ? 0 : 2;
+
+
+ if (hws->ctx->dc->debug.disable_dpp_power_gate) {
+ /* Workaround for DCN314 with disabled power gating */
+ if (!power_on) {
+
+ /* Force disable cursor if power gating is disabled */
+ struct dpp *dpp = hws->ctx->dc->res_pool->dpps[dpp_inst];
+ if (dpp && dpp->funcs->dpp_force_disable_cursor)
+ dpp->funcs->dpp_force_disable_cursor(dpp);
+ }
+ return;
+ }
+ if (REG(DOMAIN1_PG_CONFIG) == 0)
+ return;
+
+ switch (dpp_inst) {
+ case 0: /* DPP0 */
+ REG_UPDATE(DOMAIN1_PG_CONFIG,
+ DOMAIN1_POWER_GATE, power_gate);
+
+ REG_WAIT(DOMAIN1_PG_STATUS,
+ DOMAIN1_PGFSM_PWR_STATUS, pwr_status,
+ 1, 1000);
+ break;
+ case 1: /* DPP1 */
+ REG_UPDATE(DOMAIN3_PG_CONFIG,
+ DOMAIN3_POWER_GATE, power_gate);
+
+ REG_WAIT(DOMAIN3_PG_STATUS,
+ DOMAIN3_PGFSM_PWR_STATUS, pwr_status,
+ 1, 1000);
+ break;
+ case 2: /* DPP2 */
+ REG_UPDATE(DOMAIN5_PG_CONFIG,
+ DOMAIN5_POWER_GATE, power_gate);
+
+ REG_WAIT(DOMAIN5_PG_STATUS,
+ DOMAIN5_PGFSM_PWR_STATUS, pwr_status,
+ 1, 1000);
+ break;
+ case 3: /* DPP3 */
+ REG_UPDATE(DOMAIN7_PG_CONFIG,
+ DOMAIN7_POWER_GATE, power_gate);
+
+ REG_WAIT(DOMAIN7_PG_STATUS,
+ DOMAIN7_PGFSM_PWR_STATUS, pwr_status,
+ 1, 1000);
+ break;
+ default:
+ BREAK_TO_DEBUGGER();
+ break;
+ }
+}
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.h
index 2305ad2..6c072d0 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.h
@@ -47,4 +47,6 @@ void dcn314_dpp_root_clock_control(struct dce_hwseq *hws, unsigned int dpp_inst,
void dcn314_disable_link_output(struct dc_link *link, const struct link_resource *link_res, enum signal_type signal);
+void dcn314_dpp_pg_control(struct dce_hwseq *hws, unsigned int dpp_inst, bool power_on);
+
#endif /* __DC_HWSS_DCN314_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c
index f511274..9f454fa 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c
@@ -141,6 +141,7 @@ static const struct hwseq_private_funcs dcn314_private_funcs = {
.enable_power_gating_plane = dcn314_enable_power_gating_plane,
.dpp_root_clock_control = dcn314_dpp_root_clock_control,
.hubp_pg_control = dcn31_hubp_pg_control,
+ .dpp_pg_control = dcn314_dpp_pg_control,
.program_all_writeback_pipes_in_tree = dcn30_program_all_writeback_pipes_in_tree,
.update_odm = dcn314_update_odm,
.dsc_pg_control = dcn314_dsc_pg_control,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
index 0c5675d1..1b7c085 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
@@ -349,6 +349,9 @@ struct dpp_funcs {
struct dpp *dpp_base,
enum dc_color_space color_space,
struct dc_csc_transform cursor_csc_color_matrix);
+
+ void (*dpp_force_disable_cursor)(struct dpp *dpp_base);
+
};
diff --git a/drivers/gpu/drm/bridge/ti-sn65dsi86.c b/drivers/gpu/drm/bridge/ti-sn65dsi86.c
index 4643903..ae0d08e 100644
--- a/drivers/gpu/drm/bridge/ti-sn65dsi86.c
+++ b/drivers/gpu/drm/bridge/ti-sn65dsi86.c
@@ -393,6 +393,17 @@ static int __maybe_unused ti_sn65dsi86_resume(struct device *dev)
gpiod_set_value_cansleep(pdata->enable_gpio, 1);
/*
+ * After EN is deasserted and an external clock is detected, the bridge
+ * will sample GPIO3:1 to determine its frequency. The driver will
+ * overwrite this setting in ti_sn_bridge_set_refclk_freq(). But this is
+ * racy. Thus we have to wait a couple of us. According to the datasheet
+ * the GPIO lines has to be stable at least 5 us (td5) but it seems that
+ * is not enough and the refclk frequency value is still lost or
+ * overwritten by the bridge itself. Waiting for 20us seems to work.
+ */
+ usleep_range(20, 30);
+
+ /*
* If we have a reference clock we can enable communication w/ the
* panel (including the aux channel) w/out any need for an input clock
* so we can do it in resume which lets us read the EDID before
diff --git a/drivers/gpu/drm/nouveau/gv100_fence.c b/drivers/gpu/drm/nouveau/gv100_fence.c
index cccdeca..317e516 100644
--- a/drivers/gpu/drm/nouveau/gv100_fence.c
+++ b/drivers/gpu/drm/nouveau/gv100_fence.c
@@ -18,7 +18,7 @@ gv100_fence_emit32(struct nouveau_channel *chan, u64 virtual, u32 sequence)
struct nvif_push *push = &chan->chan.push;
int ret;
- ret = PUSH_WAIT(push, 8);
+ ret = PUSH_WAIT(push, 13);
if (ret)
return ret;
@@ -32,6 +32,11 @@ gv100_fence_emit32(struct nouveau_channel *chan, u64 virtual, u32 sequence)
NVDEF(NVC36F, SEM_EXECUTE, PAYLOAD_SIZE, 32BIT) |
NVDEF(NVC36F, SEM_EXECUTE, RELEASE_TIMESTAMP, DIS));
+ PUSH_MTHD(push, NVC36F, MEM_OP_A, 0,
+ MEM_OP_B, 0,
+ MEM_OP_C, NVDEF(NVC36F, MEM_OP_C, MEMBAR_TYPE, SYS_MEMBAR),
+ MEM_OP_D, NVDEF(NVC36F, MEM_OP_D, OPERATION, MEMBAR));
+
PUSH_MTHD(push, NVC36F, NON_STALL_INTERRUPT, 0);
PUSH_KICK(push);
diff --git a/drivers/gpu/drm/nouveau/include/nvhw/class/clc36f.h b/drivers/gpu/drm/nouveau/include/nvhw/class/clc36f.h
index 8735dda..338f74b 100644
--- a/drivers/gpu/drm/nouveau/include/nvhw/class/clc36f.h
+++ b/drivers/gpu/drm/nouveau/include/nvhw/class/clc36f.h
@@ -7,6 +7,91 @@
#define NVC36F_NON_STALL_INTERRUPT (0x00000020)
#define NVC36F_NON_STALL_INTERRUPT_HANDLE 31:0
+// NOTE - MEM_OP_A and MEM_OP_B have been replaced in gp100 with methods for
+// specifying the page address for a targeted TLB invalidate and the uTLB for
+// a targeted REPLAY_CANCEL for UVM.
+// The previous MEM_OP_A/B functionality is in MEM_OP_C/D, with slightly
+// rearranged fields.
+#define NVC36F_MEM_OP_A (0x00000028)
+#define NVC36F_MEM_OP_A_TLB_INVALIDATE_CANCEL_TARGET_CLIENT_UNIT_ID 5:0 // only relevant for REPLAY_CANCEL_TARGETED
+#define NVC36F_MEM_OP_A_TLB_INVALIDATE_INVALIDATION_SIZE 5:0 // Used to specify size of invalidate, used for invalidates which are not of the REPLAY_CANCEL_TARGETED type
+#define NVC36F_MEM_OP_A_TLB_INVALIDATE_CANCEL_TARGET_GPC_ID 10:6 // only relevant for REPLAY_CANCEL_TARGETED
+#define NVC36F_MEM_OP_A_TLB_INVALIDATE_CANCEL_MMU_ENGINE_ID 6:0 // only relevant for REPLAY_CANCEL_VA_GLOBAL
+#define NVC36F_MEM_OP_A_TLB_INVALIDATE_SYSMEMBAR 11:11
+#define NVC36F_MEM_OP_A_TLB_INVALIDATE_SYSMEMBAR_EN 0x00000001
+#define NVC36F_MEM_OP_A_TLB_INVALIDATE_SYSMEMBAR_DIS 0x00000000
+#define NVC36F_MEM_OP_A_TLB_INVALIDATE_TARGET_ADDR_LO 31:12
+#define NVC36F_MEM_OP_B (0x0000002c)
+#define NVC36F_MEM_OP_B_TLB_INVALIDATE_TARGET_ADDR_HI 31:0
+#define NVC36F_MEM_OP_C (0x00000030)
+#define NVC36F_MEM_OP_C_MEMBAR_TYPE 2:0
+#define NVC36F_MEM_OP_C_MEMBAR_TYPE_SYS_MEMBAR 0x00000000
+#define NVC36F_MEM_OP_C_MEMBAR_TYPE_MEMBAR 0x00000001
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PDB 0:0
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PDB_ONE 0x00000000
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PDB_ALL 0x00000001 // Probably nonsensical for MMU_TLB_INVALIDATE_TARGETED
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_GPC 1:1
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_GPC_ENABLE 0x00000000
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_GPC_DISABLE 0x00000001
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_REPLAY 4:2 // only relevant if GPC ENABLE
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_REPLAY_NONE 0x00000000
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_REPLAY_START 0x00000001
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_REPLAY_START_ACK_ALL 0x00000002
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_REPLAY_CANCEL_TARGETED 0x00000003
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_REPLAY_CANCEL_GLOBAL 0x00000004
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_REPLAY_CANCEL_VA_GLOBAL 0x00000005
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACK_TYPE 6:5 // only relevant if GPC ENABLE
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACK_TYPE_NONE 0x00000000
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACK_TYPE_GLOBALLY 0x00000001
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACK_TYPE_INTRANODE 0x00000002
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE 9:7 //only relevant for REPLAY_CANCEL_VA_GLOBAL
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE_VIRT_READ 0
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE_VIRT_WRITE 1
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE_VIRT_ATOMIC_STRONG 2
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE_VIRT_RSVRVD 3
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE_VIRT_ATOMIC_WEAK 4
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE_VIRT_ATOMIC_ALL 5
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE_VIRT_WRITE_AND_ATOMIC 6
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_ACCESS_TYPE_VIRT_ALL 7
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL 9:7 // Invalidate affects this level and all below
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL_ALL 0x00000000 // Invalidate tlb caches at all levels of the page table
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL_PTE_ONLY 0x00000001
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL_UP_TO_PDE0 0x00000002
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL_UP_TO_PDE1 0x00000003
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL_UP_TO_PDE2 0x00000004
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL_UP_TO_PDE3 0x00000005
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL_UP_TO_PDE4 0x00000006
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PAGE_TABLE_LEVEL_UP_TO_PDE5 0x00000007
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PDB_APERTURE 11:10 // only relevant if PDB_ONE
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PDB_APERTURE_VID_MEM 0x00000000
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PDB_APERTURE_SYS_MEM_COHERENT 0x00000002
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PDB_APERTURE_SYS_MEM_NONCOHERENT 0x00000003
+#define NVC36F_MEM_OP_C_TLB_INVALIDATE_PDB_ADDR_LO 31:12 // only relevant if PDB_ONE
+#define NVC36F_MEM_OP_C_ACCESS_COUNTER_CLR_TARGETED_NOTIFY_TAG 19:0
+// MEM_OP_D MUST be preceded by MEM_OPs A-C.
+#define NVC36F_MEM_OP_D (0x00000034)
+#define NVC36F_MEM_OP_D_TLB_INVALIDATE_PDB_ADDR_HI 26:0 // only relevant if PDB_ONE
+#define NVC36F_MEM_OP_D_OPERATION 31:27
+#define NVC36F_MEM_OP_D_OPERATION_MEMBAR 0x00000005
+#define NVC36F_MEM_OP_D_OPERATION_MMU_TLB_INVALIDATE 0x00000009
+#define NVC36F_MEM_OP_D_OPERATION_MMU_TLB_INVALIDATE_TARGETED 0x0000000a
+#define NVC36F_MEM_OP_D_OPERATION_L2_PEERMEM_INVALIDATE 0x0000000d
+#define NVC36F_MEM_OP_D_OPERATION_L2_SYSMEM_INVALIDATE 0x0000000e
+// CLEAN_LINES is an alias for Tegra/GPU IP usage
+#define NVC36F_MEM_OP_B_OPERATION_L2_INVALIDATE_CLEAN_LINES 0x0000000e
+#define NVC36F_MEM_OP_D_OPERATION_L2_CLEAN_COMPTAGS 0x0000000f
+#define NVC36F_MEM_OP_D_OPERATION_L2_FLUSH_DIRTY 0x00000010
+#define NVC36F_MEM_OP_D_OPERATION_L2_WAIT_FOR_SYS_PENDING_READS 0x00000015
+#define NVC36F_MEM_OP_D_OPERATION_ACCESS_COUNTER_CLR 0x00000016
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TYPE 1:0
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TYPE_MIMC 0x00000000
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TYPE_MOMC 0x00000001
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TYPE_ALL 0x00000002
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TYPE_TARGETED 0x00000003
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TARGETED_TYPE 2:2
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TARGETED_TYPE_MIMC 0x00000000
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TARGETED_TYPE_MOMC 0x00000001
+#define NVC36F_MEM_OP_D_ACCESS_COUNTER_CLR_TARGETED_BANK 6:3
#define NVC36F_SEM_ADDR_LO (0x0000005c)
#define NVC36F_SEM_ADDR_LO_OFFSET 31:2
#define NVC36F_SEM_ADDR_HI (0x00000060)
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
index fdffa03..6fd4e60 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
@@ -350,6 +350,8 @@ nvkm_fifo_dtor(struct nvkm_engine *engine)
nvkm_chid_unref(&fifo->chid);
nvkm_event_fini(&fifo->nonstall.event);
+ if (fifo->func->nonstall_dtor)
+ fifo->func->nonstall_dtor(fifo);
mutex_destroy(&fifo->mutex);
if (fifo->func->dtor)
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/ga100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/ga100.c
index e74493a..6848a56f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/ga100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/ga100.c
@@ -517,19 +517,11 @@ ga100_fifo_nonstall_intr(struct nvkm_inth *inth)
static void
ga100_fifo_nonstall_block(struct nvkm_event *event, int type, int index)
{
- struct nvkm_fifo *fifo = container_of(event, typeof(*fifo), nonstall.event);
- struct nvkm_runl *runl = nvkm_runl_get(fifo, index, 0);
-
- nvkm_inth_block(&runl->nonstall.inth);
}
static void
ga100_fifo_nonstall_allow(struct nvkm_event *event, int type, int index)
{
- struct nvkm_fifo *fifo = container_of(event, typeof(*fifo), nonstall.event);
- struct nvkm_runl *runl = nvkm_runl_get(fifo, index, 0);
-
- nvkm_inth_allow(&runl->nonstall.inth);
}
const struct nvkm_event_func
@@ -564,12 +556,26 @@ ga100_fifo_nonstall_ctor(struct nvkm_fifo *fifo)
if (ret)
return ret;
+ nvkm_inth_allow(&runl->nonstall.inth);
+
nr = max(nr, runl->id + 1);
}
return nr;
}
+void
+ga100_fifo_nonstall_dtor(struct nvkm_fifo *fifo)
+{
+ struct nvkm_runl *runl;
+
+ nvkm_runl_foreach(runl, fifo) {
+ if (runl->nonstall.vector < 0)
+ continue;
+ nvkm_inth_block(&runl->nonstall.inth);
+ }
+}
+
int
ga100_fifo_runl_ctor(struct nvkm_fifo *fifo)
{
@@ -599,6 +605,7 @@ ga100_fifo = {
.runl_ctor = ga100_fifo_runl_ctor,
.mmu_fault = &tu102_fifo_mmu_fault,
.nonstall_ctor = ga100_fifo_nonstall_ctor,
+ .nonstall_dtor = ga100_fifo_nonstall_dtor,
.nonstall = &ga100_fifo_nonstall,
.runl = &ga100_runl,
.runq = &ga100_runq,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/ga102.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/ga102.c
index 755235f..18a0b1f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/ga102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/ga102.c
@@ -30,6 +30,7 @@ ga102_fifo = {
.runl_ctor = ga100_fifo_runl_ctor,
.mmu_fault = &tu102_fifo_mmu_fault,
.nonstall_ctor = ga100_fifo_nonstall_ctor,
+ .nonstall_dtor = ga100_fifo_nonstall_dtor,
.nonstall = &ga100_fifo_nonstall,
.runl = &ga100_runl,
.runq = &ga100_runq,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
index 5e81ae1..fff1428 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
@@ -41,6 +41,7 @@ struct nvkm_fifo_func {
void (*start)(struct nvkm_fifo *, unsigned long *);
int (*nonstall_ctor)(struct nvkm_fifo *);
+ void (*nonstall_dtor)(struct nvkm_fifo *);
const struct nvkm_event_func *nonstall;
const struct nvkm_runl_func *runl;
@@ -200,6 +201,7 @@ u32 tu102_chan_doorbell_handle(struct nvkm_chan *);
int ga100_fifo_runl_ctor(struct nvkm_fifo *);
int ga100_fifo_nonstall_ctor(struct nvkm_fifo *);
+void ga100_fifo_nonstall_dtor(struct nvkm_fifo *);
extern const struct nvkm_event_func ga100_fifo_nonstall;
extern const struct nvkm_runl_func ga100_runl;
extern const struct nvkm_runq_func ga100_runq;
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fifo.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fifo.c
index 1ac5628..4ed54b3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fifo.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fifo.c
@@ -601,6 +601,7 @@ r535_fifo_new(const struct nvkm_fifo_func *hw, struct nvkm_device *device,
rm->chan.func = &r535_chan;
rm->nonstall = &ga100_fifo_nonstall;
rm->nonstall_ctor = ga100_fifo_nonstall_ctor;
+ rm->nonstall_dtor = ga100_fifo_nonstall_dtor;
return nvkm_fifo_new_(rm, device, type, inst, pfifo);
}
diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
index 8867b95..3d06f72 100644
--- a/drivers/gpu/drm/scheduler/sched_entity.c
+++ b/drivers/gpu/drm/scheduler/sched_entity.c
@@ -391,7 +391,8 @@ EXPORT_SYMBOL(drm_sched_entity_set_priority);
* Add a callback to the current dependency of the entity to wake up the
* scheduler when the entity becomes available.
*/
-static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
+static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity,
+ struct drm_sched_job *sched_job)
{
struct drm_gpu_scheduler *sched = entity->rq->sched;
struct dma_fence *fence = entity->dependency;
@@ -421,6 +422,10 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
entity->dependency = fence;
}
+ if (trace_drm_sched_job_unschedulable_enabled() &&
+ !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &entity->dependency->flags))
+ trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
+
if (!dma_fence_add_callback(entity->dependency, &entity->cb,
drm_sched_entity_wakeup))
return true;
@@ -461,10 +466,8 @@ struct drm_sched_job *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
while ((entity->dependency =
drm_sched_job_dependency(sched_job, entity))) {
- if (drm_sched_entity_add_dependency_cb(entity)) {
- trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
+ if (drm_sched_entity_add_dependency_cb(entity, sched_job))
return NULL;
- }
}
/* skip jobs from entity that marked guilty */
diff --git a/drivers/gpu/drm/xe/xe_bo.c b/drivers/gpu/drm/xe/xe_bo.c
index 1be2415..9954bb4 100644
--- a/drivers/gpu/drm/xe/xe_bo.c
+++ b/drivers/gpu/drm/xe/xe_bo.c
@@ -819,8 +819,7 @@ static int xe_bo_move(struct ttm_buffer_object *ttm_bo, bool evict,
return ret;
}
- tt_has_data = ttm && (ttm_tt_is_populated(ttm) ||
- (ttm->page_flags & TTM_TT_FLAG_SWAPPED));
+ tt_has_data = ttm && (ttm_tt_is_populated(ttm) || ttm_tt_is_swapped(ttm));
move_lacks_source = !old_mem || (handle_system_ccs ? (!bo->ccs_cleared) :
(!mem_type_is_vram(old_mem_type) && !tt_has_data));
