Commit 5a773869 authored by Linus Torvalds's avatar Linus Torvalds

Merge tag 'drm-fixes-2022-09-30-1' of git://anongit.freedesktop.org/drm/drm

Pull drm fixes from Dave Airlie:
 "Last set of fixes for 6.0 hopefully - minor bridge fixes, i915 fixes,
  and a bunch of amdgpu fixes for new IP blocks, along with a couple of
  regression fixes. Should be all set for merge window next week.

  amdgpu:
   - GC 11.x fixes
   - SMU 13.x fixes
   - DCN 3.1.4 fixes
   - DCN 3.2.x fixes
   - GC 9.x fix
   - Fence fix
   - SR-IOV supend/resume fix
   - PSR regression fix

  i915:
   - Restrict forced preemption to the active context
   - Restrict perf_limit_reasons to the supported platforms - gen11+

  bridge:
   - analogix: Revert earlier suspend fix
   - lt8912b: Fix corrupt display output"

* tag 'drm-fixes-2022-09-30-1' of git://anongit.freedesktop.org/drm/drm: (26 commits)
  drm/amd/display: Prevent OTG shutdown during PSR SU
  drm/i915/gt: Perf_limit_reasons are only available for Gen11+
  drm/amdgpu: Add amdgpu suspend-resume code path under SRIOV
  drm/amdgpu: Remove fence_process in count_emitted
  drm/amdgpu: Correct the position in patch_cond_exec
  drm/amd/display: fill in clock values when DPM is not enabled
  drm/amd/display: Avoid unnecessary pixel rate divider programming
  drm/amd/display: Remove assert for odm transition case
  drm/amd/display: Fix typo in get_pixel_rate_div
  drm/amd/display: Fix audio on display after unplugging another
  drm/amd/display: Add explicit FIFO disable for DP blank
  drm/amd/display: Wrap OTG disable workaround with FIFO control
  drm/amd/display: Do DIO FIFO enable after DP video stream enable
  drm/amd/display: Update DCN32 to use new SR latencies
  drm/amd/display: Avoid avoid unnecessary pixel rate divider programming
  drm/amdkfd: fix dropped interrupt in kfd_int_process_v11
  drm/amdgpu: pass queue size and is_aql_queue to MES
  drm/amdkfd: fix MQD init for GFX11 in init_mqd
  drm/amd/pm: use adverse selection for dpm features unsupported by driver
  drm/amd/pm: enable gfxoff feature for SMU 13.0.0
  ...
parents 987a926c 6643b383
......@@ -1050,6 +1050,10 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev)
{
if (adev->flags & AMD_IS_APU)
return false;
if (amdgpu_sriov_vf(adev))
return false;
return pm_suspend_target_state != PM_SUSPEND_TO_IDLE;
}
......
......@@ -3152,7 +3152,8 @@ static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev)
continue;
if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON ||
adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC ||
adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH) {
adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH ||
(adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP && amdgpu_sriov_vf(adev))) {
r = adev->ip_blocks[i].version->funcs->resume(adev);
if (r) {
......@@ -4064,12 +4065,20 @@ static void amdgpu_device_evict_resources(struct amdgpu_device *adev)
int amdgpu_device_suspend(struct drm_device *dev, bool fbcon)
{
struct amdgpu_device *adev = drm_to_adev(dev);
int r = 0;
if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
adev->in_suspend = true;
if (amdgpu_sriov_vf(adev)) {
amdgpu_virt_fini_data_exchange(adev);
r = amdgpu_virt_request_full_gpu(adev, false);
if (r)
return r;
}
if (amdgpu_acpi_smart_shift_update(dev, AMDGPU_SS_DEV_D3))
DRM_WARN("smart shift update failed\n");
......@@ -4093,6 +4102,9 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon)
amdgpu_device_ip_suspend_phase2(adev);
if (amdgpu_sriov_vf(adev))
amdgpu_virt_release_full_gpu(adev, false);
return 0;
}
......@@ -4111,6 +4123,12 @@ int amdgpu_device_resume(struct drm_device *dev, bool fbcon)
struct amdgpu_device *adev = drm_to_adev(dev);
int r = 0;
if (amdgpu_sriov_vf(adev)) {
r = amdgpu_virt_request_full_gpu(adev, true);
if (r)
return r;
}
if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
......@@ -4125,6 +4143,13 @@ int amdgpu_device_resume(struct drm_device *dev, bool fbcon)
}
r = amdgpu_device_ip_resume(adev);
/* no matter what r is, always need to properly release full GPU */
if (amdgpu_sriov_vf(adev)) {
amdgpu_virt_init_data_exchange(adev);
amdgpu_virt_release_full_gpu(adev, true);
}
if (r) {
dev_err(adev->dev, "amdgpu_device_ip_resume failed (%d).\n", r);
return r;
......
......@@ -400,7 +400,6 @@ unsigned amdgpu_fence_count_emitted(struct amdgpu_ring *ring)
/* We are not protected by ring lock when reading the last sequence
* but it's ok to report slightly wrong fence count here.
*/
amdgpu_fence_process(ring);
emitted = 0x100000000ull;
emitted -= atomic_read(&ring->fence_drv.last_seq);
emitted += READ_ONCE(ring->fence_drv.sync_seq);
......
......@@ -222,6 +222,8 @@ struct mes_add_queue_input {
uint64_t tba_addr;
uint64_t tma_addr;
uint32_t is_kfd_process;
uint32_t is_aql_queue;
uint32_t queue_size;
};
struct mes_remove_queue_input {
......
......@@ -5260,6 +5260,8 @@ static void gfx_v11_0_update_spm_vmid(struct amdgpu_device *adev, unsigned vmid)
{
u32 reg, data;
amdgpu_gfx_off_ctrl(adev, false);
reg = SOC15_REG_OFFSET(GC, 0, regRLC_SPM_MC_CNTL);
if (amdgpu_sriov_is_pp_one_vf(adev))
data = RREG32_NO_KIQ(reg);
......@@ -5273,6 +5275,8 @@ static void gfx_v11_0_update_spm_vmid(struct amdgpu_device *adev, unsigned vmid)
WREG32_SOC15_NO_KIQ(GC, 0, regRLC_SPM_MC_CNTL, data);
else
WREG32_SOC15(GC, 0, regRLC_SPM_MC_CNTL, data);
amdgpu_gfx_off_ctrl(adev, true);
}
static const struct amdgpu_rlc_funcs gfx_v11_0_rlc_funcs = {
......
......@@ -5597,7 +5597,7 @@ static void gfx_v9_0_ring_emit_patch_cond_exec(struct amdgpu_ring *ring, unsigne
BUG_ON(offset > ring->buf_mask);
BUG_ON(ring->ring[offset] != 0x55aa55aa);
cur = (ring->wptr & ring->buf_mask) - 1;
cur = (ring->wptr - 1) & ring->buf_mask;
if (likely(cur > offset))
ring->ring[offset] = cur - offset;
else
......
......@@ -185,6 +185,10 @@ static int mes_v11_0_add_hw_queue(struct amdgpu_mes *mes,
mes_add_queue_pkt.is_kfd_process = input->is_kfd_process;
mes_add_queue_pkt.trap_en = 1;
/* For KFD, gds_size is re-used for queue size (needed in MES for AQL queues) */
mes_add_queue_pkt.is_aql_queue = input->is_aql_queue;
mes_add_queue_pkt.gds_size = input->queue_size;
return mes_v11_0_submit_pkt_and_poll_completion(mes,
&mes_add_queue_pkt, sizeof(mes_add_queue_pkt),
offsetof(union MESAPI__ADD_QUEUE, api_status));
......
......@@ -205,6 +205,8 @@ static int add_queue_mes(struct device_queue_manager *dqm, struct queue *q,
}
queue_input.is_kfd_process = 1;
queue_input.is_aql_queue = (q->properties.format == KFD_QUEUE_FORMAT_AQL);
queue_input.queue_size = q->properties.queue_size >> 2;
queue_input.paging = false;
queue_input.tba_addr = qpd->tba_addr;
......
......@@ -350,11 +350,11 @@ static void event_interrupt_wq_v11(struct kfd_dev *dev,
print_sq_intr_info_inst(context_id0, context_id1);
sq_int_priv = REG_GET_FIELD(context_id0,
SQ_INTERRUPT_WORD_WAVE_CTXID0, PRIV);
if (sq_int_priv /*&& (kfd_set_dbg_ev_from_interrupt(dev, pasid,
/*if (sq_int_priv && (kfd_set_dbg_ev_from_interrupt(dev, pasid,
KFD_CTXID0_DOORBELL_ID(context_id0),
KFD_CTXID0_TRAP_CODE(context_id0),
NULL, 0))*/)
return;
NULL, 0)))
return;*/
break;
case SQ_INTERRUPT_WORD_ENCODING_ERROR:
print_sq_intr_info_error(context_id0, context_id1);
......
......@@ -126,6 +126,10 @@ static void init_mqd(struct mqd_manager *mm, void **mqd,
m->compute_static_thread_mgmt_se1 = 0xFFFFFFFF;
m->compute_static_thread_mgmt_se2 = 0xFFFFFFFF;
m->compute_static_thread_mgmt_se3 = 0xFFFFFFFF;
m->compute_static_thread_mgmt_se4 = 0xFFFFFFFF;
m->compute_static_thread_mgmt_se5 = 0xFFFFFFFF;
m->compute_static_thread_mgmt_se6 = 0xFFFFFFFF;
m->compute_static_thread_mgmt_se7 = 0xFFFFFFFF;
m->cp_hqd_persistent_state = CP_HQD_PERSISTENT_STATE__PRELOAD_REQ_MASK |
0x55 << CP_HQD_PERSISTENT_STATE__PRELOAD_SIZE__SHIFT;
......
......@@ -170,7 +170,13 @@ bool amdgpu_dm_psr_enable(struct dc_stream_state *stream)
&stream, 1,
&params);
power_opt |= psr_power_opt_z10_static_screen;
/*
* Only enable static-screen optimizations for PSR1. For PSR SU, this
* causes vstartup interrupt issues, used by amdgpu_dm to send vblank
* events.
*/
if (link->psr_settings.psr_version < DC_PSR_VERSION_SU_1)
power_opt |= psr_power_opt_z10_static_screen;
return dc_link_set_psr_allow_active(link, &psr_enable, false, false, &power_opt);
}
......
......@@ -130,11 +130,20 @@ static void dcn314_disable_otg_wa(struct clk_mgr *clk_mgr_base, struct dc_state
if (pipe->top_pipe || pipe->prev_odm_pipe)
continue;
if (pipe->stream && (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal))) {
struct stream_encoder *stream_enc = pipe->stream_res.stream_enc;
if (disable) {
if (stream_enc && stream_enc->funcs->disable_fifo)
pipe->stream_res.stream_enc->funcs->disable_fifo(stream_enc);
pipe->stream_res.tg->funcs->immediate_disable_crtc(pipe->stream_res.tg);
reset_sync_context_for_pipe(dc, context, i);
} else
} else {
pipe->stream_res.tg->funcs->enable_crtc(pipe->stream_res.tg);
if (stream_enc && stream_enc->funcs->enable_fifo)
pipe->stream_res.stream_enc->funcs->enable_fifo(stream_enc);
}
}
}
}
......
......@@ -156,12 +156,14 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base)
{
struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
unsigned int num_levels;
unsigned int num_dcfclk_levels, num_dtbclk_levels, num_dispclk_levels;
memset(&(clk_mgr_base->clks), 0, sizeof(struct dc_clocks));
clk_mgr_base->clks.p_state_change_support = true;
clk_mgr_base->clks.prev_p_state_change_support = true;
clk_mgr_base->clks.fclk_prev_p_state_change_support = true;
clk_mgr->smu_present = false;
clk_mgr->dpm_present = false;
if (!clk_mgr_base->bw_params)
return;
......@@ -179,6 +181,7 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base)
dcn32_init_single_clock(clk_mgr, PPCLK_DCFCLK,
&clk_mgr_base->bw_params->clk_table.entries[0].dcfclk_mhz,
&num_levels);
num_dcfclk_levels = num_levels;
/* SOCCLK */
dcn32_init_single_clock(clk_mgr, PPCLK_SOCCLK,
......@@ -189,11 +192,16 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base)
dcn32_init_single_clock(clk_mgr, PPCLK_DTBCLK,
&clk_mgr_base->bw_params->clk_table.entries[0].dtbclk_mhz,
&num_levels);
num_dtbclk_levels = num_levels;
/* DISPCLK */
dcn32_init_single_clock(clk_mgr, PPCLK_DISPCLK,
&clk_mgr_base->bw_params->clk_table.entries[0].dispclk_mhz,
&num_levels);
num_dispclk_levels = num_levels;
if (num_dcfclk_levels && num_dtbclk_levels && num_dispclk_levels)
clk_mgr->dpm_present = true;
if (clk_mgr_base->ctx->dc->debug.min_disp_clk_khz) {
unsigned int i;
......@@ -658,6 +666,12 @@ static void dcn32_get_memclk_states_from_smu(struct clk_mgr *clk_mgr_base)
&num_levels);
clk_mgr_base->bw_params->clk_table.num_entries = num_levels ? num_levels : 1;
if (clk_mgr->dpm_present && !num_levels)
clk_mgr->dpm_present = false;
if (!clk_mgr->dpm_present)
dcn32_patch_dpm_table(clk_mgr_base->bw_params);
DC_FP_START();
/* Refresh bounding box */
clk_mgr_base->ctx->dc->res_pool->funcs->update_bw_bounding_box(
......
......@@ -2164,8 +2164,7 @@ static void dce110_setup_audio_dto(
continue;
if (pipe_ctx->stream->signal != SIGNAL_TYPE_HDMI_TYPE_A)
continue;
if (pipe_ctx->stream_res.audio != NULL &&
pipe_ctx->stream_res.audio->enabled == false) {
if (pipe_ctx->stream_res.audio != NULL) {
struct audio_output audio_output;
build_audio_output(context, pipe_ctx, &audio_output);
......@@ -2205,8 +2204,7 @@ static void dce110_setup_audio_dto(
if (!dc_is_dp_signal(pipe_ctx->stream->signal))
continue;
if (pipe_ctx->stream_res.audio != NULL &&
pipe_ctx->stream_res.audio->enabled == false) {
if (pipe_ctx->stream_res.audio != NULL) {
struct audio_output audio_output;
build_audio_output(context, pipe_ctx, &audio_output);
......
......@@ -45,6 +45,48 @@
#define DC_LOGGER \
dccg->ctx->logger
static void dccg314_get_pixel_rate_div(
struct dccg *dccg,
uint32_t otg_inst,
enum pixel_rate_div *k1,
enum pixel_rate_div *k2)
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
uint32_t val_k1 = PIXEL_RATE_DIV_NA, val_k2 = PIXEL_RATE_DIV_NA;
*k1 = PIXEL_RATE_DIV_NA;
*k2 = PIXEL_RATE_DIV_NA;
switch (otg_inst) {
case 0:
REG_GET_2(OTG_PIXEL_RATE_DIV,
OTG0_PIXEL_RATE_DIVK1, &val_k1,
OTG0_PIXEL_RATE_DIVK2, &val_k2);
break;
case 1:
REG_GET_2(OTG_PIXEL_RATE_DIV,
OTG1_PIXEL_RATE_DIVK1, &val_k1,
OTG1_PIXEL_RATE_DIVK2, &val_k2);
break;
case 2:
REG_GET_2(OTG_PIXEL_RATE_DIV,
OTG2_PIXEL_RATE_DIVK1, &val_k1,
OTG2_PIXEL_RATE_DIVK2, &val_k2);
break;
case 3:
REG_GET_2(OTG_PIXEL_RATE_DIV,
OTG3_PIXEL_RATE_DIVK1, &val_k1,
OTG3_PIXEL_RATE_DIVK2, &val_k2);
break;
default:
BREAK_TO_DEBUGGER();
return;
}
*k1 = (enum pixel_rate_div)val_k1;
*k2 = (enum pixel_rate_div)val_k2;
}
static void dccg314_set_pixel_rate_div(
struct dccg *dccg,
uint32_t otg_inst,
......@@ -52,6 +94,11 @@ static void dccg314_set_pixel_rate_div(
enum pixel_rate_div k2)
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
enum pixel_rate_div cur_k1 = PIXEL_RATE_DIV_NA, cur_k2 = PIXEL_RATE_DIV_NA;
dccg314_get_pixel_rate_div(dccg, otg_inst, &cur_k1, &cur_k2);
if (k1 == PIXEL_RATE_DIV_NA || k2 == PIXEL_RATE_DIV_NA || (k1 == cur_k1 && k2 == cur_k2))
return;
switch (otg_inst) {
case 0:
......
......@@ -56,7 +56,8 @@ static void enc314_enable_fifo(struct stream_encoder *enc)
/* TODO: Confirm if we need to wait for DIG_SYMCLK_FE_ON */
REG_WAIT(DIG_FE_CNTL, DIG_SYMCLK_FE_ON, 1, 10, 5000);
REG_UPDATE_2(DIG_FIFO_CTRL0, DIG_FIFO_RESET, 1, DIG_FIFO_READ_START_LEVEL, 0x7);
REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_READ_START_LEVEL, 0x7);
REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_RESET, 1);
REG_WAIT(DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, 1, 10, 5000);
REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_RESET, 0);
REG_WAIT(DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, 0, 10, 5000);
......@@ -261,6 +262,16 @@ static bool is_two_pixels_per_containter(const struct dc_crtc_timing *timing)
return two_pix;
}
void enc314_stream_encoder_dp_blank(
struct dc_link *link,
struct stream_encoder *enc)
{
/* New to DCN314 - disable the FIFO before VID stream disable. */
enc314_disable_fifo(enc);
enc1_stream_encoder_dp_blank(link, enc);
}
static void enc314_stream_encoder_dp_unblank(
struct dc_link *link,
struct stream_encoder *enc,
......@@ -316,15 +327,11 @@ static void enc314_stream_encoder_dp_unblank(
/* switch DP encoder to CRTC data, but reset it the fifo first. It may happen
* that it overflows during mode transition, and sometimes doesn't recover.
*/
REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_READ_START_LEVEL, 0x7);
REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 1);
udelay(10);
REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 0);
/* DIG Resync FIFO now needs to be explicitly enabled. */
enc314_enable_fifo(enc);
/* wait 100us for DIG/DP logic to prime
* (i.e. a few video lines)
*/
......@@ -340,6 +347,12 @@ static void enc314_stream_encoder_dp_unblank(
REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, true);
/*
* DIG Resync FIFO now needs to be explicitly enabled.
* This should come after DP_VID_STREAM_ENABLE per HW docs.
*/
enc314_enable_fifo(enc);
dp_source_sequence_trace(link, DPCD_SOURCE_SEQ_AFTER_ENABLE_DP_VID_STREAM);
}
......@@ -408,7 +421,7 @@ static const struct stream_encoder_funcs dcn314_str_enc_funcs = {
.stop_dp_info_packets =
enc1_stream_encoder_stop_dp_info_packets,
.dp_blank =
enc1_stream_encoder_dp_blank,
enc314_stream_encoder_dp_blank,
.dp_unblank =
enc314_stream_encoder_dp_unblank,
.audio_mute_control = enc3_audio_mute_control,
......
......@@ -42,6 +42,48 @@
#define DC_LOGGER \
dccg->ctx->logger
static void dccg32_get_pixel_rate_div(
struct dccg *dccg,
uint32_t otg_inst,
enum pixel_rate_div *k1,
enum pixel_rate_div *k2)
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
uint32_t val_k1 = PIXEL_RATE_DIV_NA, val_k2 = PIXEL_RATE_DIV_NA;
*k1 = PIXEL_RATE_DIV_NA;
*k2 = PIXEL_RATE_DIV_NA;
switch (otg_inst) {
case 0:
REG_GET_2(OTG_PIXEL_RATE_DIV,
OTG0_PIXEL_RATE_DIVK1, &val_k1,
OTG0_PIXEL_RATE_DIVK2, &val_k2);
break;
case 1:
REG_GET_2(OTG_PIXEL_RATE_DIV,
OTG1_PIXEL_RATE_DIVK1, &val_k1,
OTG1_PIXEL_RATE_DIVK2, &val_k2);
break;
case 2:
REG_GET_2(OTG_PIXEL_RATE_DIV,
OTG2_PIXEL_RATE_DIVK1, &val_k1,
OTG2_PIXEL_RATE_DIVK2, &val_k2);
break;
case 3:
REG_GET_2(OTG_PIXEL_RATE_DIV,
OTG3_PIXEL_RATE_DIVK1, &val_k1,
OTG3_PIXEL_RATE_DIVK2, &val_k2);
break;
default:
BREAK_TO_DEBUGGER();
return;
}
*k1 = (enum pixel_rate_div)val_k1;
*k2 = (enum pixel_rate_div)val_k2;
}
static void dccg32_set_pixel_rate_div(
struct dccg *dccg,
uint32_t otg_inst,
......@@ -50,6 +92,17 @@ static void dccg32_set_pixel_rate_div(
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
enum pixel_rate_div cur_k1 = PIXEL_RATE_DIV_NA, cur_k2 = PIXEL_RATE_DIV_NA;
// Don't program 0xF into the register field. Not valid since
// K1 / K2 field is only 1 / 2 bits wide
if (k1 == PIXEL_RATE_DIV_NA || k2 == PIXEL_RATE_DIV_NA)
return;
dccg32_get_pixel_rate_div(dccg, otg_inst, &cur_k1, &cur_k2);
if (k1 == cur_k1 && k2 == cur_k2)
return;
switch (otg_inst) {
case 0:
REG_UPDATE_2(OTG_PIXEL_RATE_DIV,
......
......@@ -98,9 +98,13 @@ static void dcn32_program_det_size(struct hubbub *hubbub, int hubp_inst, unsigne
default:
break;
}
/* Should never be hit, if it is we have an erroneous hw config*/
ASSERT(hubbub2->det0_size + hubbub2->det1_size + hubbub2->det2_size
+ hubbub2->det3_size + hubbub2->compbuf_size_segments <= hubbub2->crb_size_segs);
if (hubbub2->det0_size + hubbub2->det1_size + hubbub2->det2_size
+ hubbub2->det3_size + hubbub2->compbuf_size_segments > hubbub2->crb_size_segs) {
/* This may happen during seamless transition from ODM 2:1 to ODM4:1 */
DC_LOG_WARNING("CRB Config Warning: DET size (%d,%d,%d,%d) + Compbuf size (%d) > CRB segments (%d)\n",
hubbub2->det0_size, hubbub2->det1_size, hubbub2->det2_size, hubbub2->det3_size,
hubbub2->compbuf_size_segments, hubbub2->crb_size_segs);
}
}
static void dcn32_program_compbuf_size(struct hubbub *hubbub, unsigned int compbuf_size_kb, bool safe_to_increase)
......
......@@ -121,8 +121,8 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_2_soc = {
},
},
.num_states = 1,
.sr_exit_time_us = 20.16,
.sr_enter_plus_exit_time_us = 27.13,
.sr_exit_time_us = 42.97,
.sr_enter_plus_exit_time_us = 49.94,
.sr_exit_z8_time_us = 285.0,
.sr_enter_plus_exit_z8_time_us = 320,
.writeback_latency_us = 12.0,
......@@ -1926,6 +1926,45 @@ static void remove_entry_from_table_at_index(struct _vcs_dpi_voltage_scaling_st
memset(&table[--(*num_entries)], 0, sizeof(struct _vcs_dpi_voltage_scaling_st));
}
void dcn32_patch_dpm_table(struct clk_bw_params *bw_params)
{
int i;
unsigned int max_dcfclk_mhz = 0, max_dispclk_mhz = 0, max_dppclk_mhz = 0,
max_phyclk_mhz = 0, max_dtbclk_mhz = 0, max_fclk_mhz = 0, max_uclk_mhz = 0;
for (i = 0; i < MAX_NUM_DPM_LVL; i++) {
if (bw_params->clk_table.entries[i].dcfclk_mhz > max_dcfclk_mhz)
max_dcfclk_mhz = bw_params->clk_table.entries[i].dcfclk_mhz;
if (bw_params->clk_table.entries[i].fclk_mhz > max_fclk_mhz)
max_fclk_mhz = bw_params->clk_table.entries[i].fclk_mhz;
if (bw_params->clk_table.entries[i].memclk_mhz > max_uclk_mhz)
max_uclk_mhz = bw_params->clk_table.entries[i].memclk_mhz;
if (bw_params->clk_table.entries[i].dispclk_mhz > max_dispclk_mhz)
max_dispclk_mhz = bw_params->clk_table.entries[i].dispclk_mhz;
if (bw_params->clk_table.entries[i].dppclk_mhz > max_dppclk_mhz)
max_dppclk_mhz = bw_params->clk_table.entries[i].dppclk_mhz;
if (bw_params->clk_table.entries[i].phyclk_mhz > max_phyclk_mhz)
max_phyclk_mhz = bw_params->clk_table.entries[i].phyclk_mhz;
if (bw_params->clk_table.entries[i].dtbclk_mhz > max_dtbclk_mhz)
max_dtbclk_mhz = bw_params->clk_table.entries[i].dtbclk_mhz;
}
/* Scan through clock values we currently have and if they are 0,
* then populate it with dcn3_2_soc.clock_limits[] value.
*
* Do it for DCFCLK, DISPCLK, DTBCLK and UCLK as any of those being
* 0, will cause it to skip building the clock table.
*/
if (max_dcfclk_mhz == 0)
bw_params->clk_table.entries[0].dcfclk_mhz = dcn3_2_soc.clock_limits[0].dcfclk_mhz;
if (max_dispclk_mhz == 0)
bw_params->clk_table.entries[0].dispclk_mhz = dcn3_2_soc.clock_limits[0].dispclk_mhz;
if (max_dtbclk_mhz == 0)
bw_params->clk_table.entries[0].dtbclk_mhz = dcn3_2_soc.clock_limits[0].dtbclk_mhz;
if (max_uclk_mhz == 0)
bw_params->clk_table.entries[0].memclk_mhz = dcn3_2_soc.clock_limits[0].dram_speed_mts / 16;
}
static int build_synthetic_soc_states(struct clk_bw_params *bw_params,
struct _vcs_dpi_voltage_scaling_st *table, unsigned int *num_entries)
{
......
......@@ -77,4 +77,6 @@ int dcn32_find_dummy_latency_index_for_fw_based_mclk_switch(struct dc *dc,
int pipe_cnt,
int vlevel);
void dcn32_patch_dpm_table(struct clk_bw_params *bw_params);
#endif
......@@ -340,6 +340,8 @@ struct clk_mgr_internal {
bool smu_present;
void *wm_range_table;
long long wm_range_table_addr;
bool dpm_present;
};
struct clk_mgr_internal_funcs {
......
......@@ -269,7 +269,8 @@ union MESAPI__ADD_QUEUE {
uint32_t map_kiq_utility_queue : 1;
uint32_t is_kfd_process : 1;
uint32_t trap_en : 1;
uint32_t reserved : 21;
uint32_t is_aql_queue : 1;
uint32_t reserved : 20;
};
struct MES_API_STATUS api_status;
uint64_t tma_addr;
......
......@@ -239,82 +239,47 @@ smu_v13_0_0_get_allowed_feature_mask(struct smu_context *smu,
uint32_t *feature_mask, uint32_t num)
{
struct amdgpu_device *adev = smu->adev;
u32 smu_version;
if (num > 2)
return -EINVAL;
memset(feature_mask, 0, sizeof(uint32_t) * num);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FW_DATA_READ_BIT);
memset(feature_mask, 0xff, sizeof(uint32_t) * num);
if (adev->pm.pp_feature & PP_SCLK_DPM_MASK) {
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_GFXCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFX_IMU_BIT);
if (!(adev->pm.pp_feature & PP_SCLK_DPM_MASK)) {
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_GFXCLK_BIT);
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_GFX_IMU_BIT);
}
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_MM_DPM_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_VCN_BIT);
if ((adev->pg_flags & AMD_PG_SUPPORT_ATHUB) &&
(adev->pg_flags & AMD_PG_SUPPORT_MMHUB))
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_ATHUB_MMHUB_PG_BIT);
if (!(adev->pg_flags & AMD_PG_SUPPORT_ATHUB) ||
!(adev->pg_flags & AMD_PG_SUPPORT_MMHUB))
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_ATHUB_MMHUB_PG_BIT);
if (adev->pm.pp_feature & PP_SOCCLK_DPM_MASK)
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT);
#if 0
if (adev->pm.pp_feature & PP_GFXOFF_MASK)
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFXOFF_BIT);
#endif
if (!(adev->pm.pp_feature & PP_SOCCLK_DPM_MASK))
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_THROTTLERS_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FAN_CONTROL_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DF_CSTATE_BIT);
if (adev->pm.pp_feature & PP_MCLK_DPM_MASK) {
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_UCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_VMEMP_SCALING_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_VDDIO_MEM_SCALING_BIT);
/* PMFW 78.58 contains a critical fix for gfxoff feature */
smu_cmn_get_smc_version(smu, NULL, &smu_version);
if ((smu_version < 0x004e3a00) ||
!(adev->pm.pp_feature & PP_GFXOFF_MASK))
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_GFXOFF_BIT);
if (!(adev->pm.pp_feature & PP_MCLK_DPM_MASK)) {
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_UCLK_BIT);
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_VMEMP_SCALING_BIT);
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_VDDIO_MEM_SCALING_BIT);
}
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_MEM_TEMP_READ_BIT);
if (adev->pm.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_GFXCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_SOCCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_SOC_MPCLK_DS_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_BACO_MPCLK_DS_BIT);
if (!(adev->pm.pp_feature & PP_SCLK_DEEP_SLEEP_MASK))
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DS_GFXCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_FCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_DCN_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_DCFCLK_BIT);
if (adev->pm.pp_feature & PP_PCIE_DPM_MASK) {
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_LINK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_LCLK_BIT);
if (!(adev->pm.pp_feature & PP_PCIE_DPM_MASK)) {
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_LINK_BIT);
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DS_LCLK_BIT);
}
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_BACO_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_MP0CLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FW_DSTATE_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_OUT_OF_BAND_MONITOR_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_SOC_CG_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_FCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FW_CTF_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_UCLK_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_VR0HOT_BIT);
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_GFX_POWER_OPTIMIZER_BIT);
if (adev->pm.pp_feature & PP_ULV_MASK)
*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFX_ULV_BIT);
if (!(adev->pm.pp_feature & PP_ULV_MASK))
*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_GFX_ULV_BIT);
return 0;
}
......
......@@ -1864,12 +1864,6 @@ EXPORT_SYMBOL_GPL(analogix_dp_remove);
int analogix_dp_suspend(struct analogix_dp_device *dp)
{
clk_disable_unprepare(dp->clock);
if (dp->plat_data->panel) {
if (drm_panel_unprepare(dp->plat_data->panel))
DRM_ERROR("failed to turnoff the panel\n");
}
return 0;
}
EXPORT_SYMBOL_GPL(analogix_dp_suspend);
......@@ -1884,13 +1878,6 @@ int analogix_dp_resume(struct analogix_dp_device *dp)
return ret;
}
if (dp->plat_data->panel) {
if (drm_panel_prepare(dp->plat_data->panel)) {
DRM_ERROR("failed to setup the panel\n");
return -EBUSY;
}
}
return 0;
}
EXPORT_SYMBOL_GPL(analogix_dp_resume);
......
......@@ -188,7 +188,7 @@ static int lt8912_write_lvds_config(struct lt8912 *lt)
{0x03, 0xff},
};
return regmap_multi_reg_write(lt->regmap[I2C_CEC_DSI], seq, ARRAY_SIZE(seq));
return regmap_multi_reg_write(lt->regmap[I2C_MAIN], seq, ARRAY_SIZE(seq));
};
static inline struct lt8912 *bridge_to_lt8912(struct drm_bridge *b)
......@@ -268,7 +268,7 @@ static int lt8912_video_setup(struct lt8912 *lt)
u32 hactive, h_total, hpw, hfp, hbp;
u32 vactive, v_total, vpw, vfp, vbp;
u8 settle = 0x08;
int ret;
int ret, hsync_activehigh, vsync_activehigh;
if (!lt)
return -EINVAL;
......@@ -278,12 +278,14 @@ static int lt8912_video_setup(struct lt8912 *lt)
hpw = lt->mode.hsync_len;
hbp = lt->mode.hback_porch;
h_total = hactive + hfp + hpw + hbp;
hsync_activehigh = lt->mode.flags & DISPLAY_FLAGS_HSYNC_HIGH;
vactive = lt->mode.vactive;
vfp = lt->mode.vfront_porch;
vpw = lt->mode.vsync_len;
vbp = lt->mode.vback_porch;
v_total = vactive + vfp + vpw + vbp;
vsync_activehigh = lt->mode.flags & DISPLAY_FLAGS_VSYNC_HIGH;
if (vactive <= 600)
settle = 0x04;
......@@ -317,6 +319,13 @@ static int lt8912_video_setup(struct lt8912 *lt)
ret |= regmap_write(lt->regmap[I2C_CEC_DSI], 0x3e, hfp & 0xff);
ret |= regmap_write(lt->regmap[I2C_CEC_DSI], 0x3f, hfp >> 8);
ret |= regmap_update_bits(lt->regmap[I2C_MAIN], 0xab, BIT(0),
vsync_activehigh ? BIT(0) : 0);
ret |= regmap_update_bits(lt->regmap[I2C_MAIN], 0xab, BIT(1),
hsync_activehigh ? BIT(1) : 0);
ret |= regmap_update_bits(lt->regmap[I2C_MAIN], 0xb2, BIT(0),
lt->connector.display_info.is_hdmi ? BIT(0) : 0);
return ret;
}
......
......@@ -165,6 +165,21 @@ struct intel_engine_execlists {
*/
struct timer_list preempt;
/**
* @preempt_target: active request at the time of the preemption request
*
* We force a preemption to occur if the pending contexts have not
* been promoted to active upon receipt of the CS ack event within
* the timeout. This timeout maybe chosen based on the target,
* using a very short timeout if the context is no longer schedulable.
* That short timeout may not be applicable to other contexts, so
* if a context switch should happen within before the preemption
* timeout, we may shoot early at an innocent context. To prevent this,
* we record which context was active at the time of the preemption
* request and only reset that context upon the timeout.
*/
const struct i915_request *preempt_target;
/**
* @ccid: identifier for contexts submitted to this engine
*/
......
......@@ -1241,6 +1241,9 @@ static unsigned long active_preempt_timeout(struct intel_engine_cs *engine,
if (!rq)
return 0;
/* Only allow ourselves to force reset the currently active context */
engine->execlists.preempt_target = rq;
/* Force a fast reset for terminated contexts (ignoring sysfs!) */
if (unlikely(intel_context_is_banned(rq->context) || bad_request(rq)))
return INTEL_CONTEXT_BANNED_PREEMPT_TIMEOUT_MS;
......@@ -2427,8 +2430,24 @@ static void execlists_submission_tasklet(struct tasklet_struct *t)
GEM_BUG_ON(inactive - post > ARRAY_SIZE(post));
if (unlikely(preempt_timeout(engine))) {
const struct i915_request *rq = *engine->execlists.active;
/*
* If after the preempt-timeout expired, we are still on the
* same active request/context as before we initiated the
* preemption, reset the engine.
*
* However, if we have processed a CS event to switch contexts,
* but not yet processed the CS event for the pending
* preemption, reset the timer allowing the new context to
* gracefully exit.
*/
cancel_timer(&engine->execlists.preempt);
engine->execlists.error_interrupt |= ERROR_PREEMPT;
if (rq == engine->execlists.preempt_target)
engine->execlists.error_interrupt |= ERROR_PREEMPT;
else
set_timer_ms(&engine->execlists.preempt,
active_preempt_timeout(engine, rq));
}
if (unlikely(READ_ONCE(engine->execlists.error_interrupt))) {
......
......@@ -545,8 +545,7 @@ static INTEL_GT_RPS_BOOL_ATTR_RO(throttle_reason_ratl, RATL_MASK);
static INTEL_GT_RPS_BOOL_ATTR_RO(throttle_reason_vr_thermalert, VR_THERMALERT_MASK);
static INTEL_GT_RPS_BOOL_ATTR_RO(throttle_reason_vr_tdc, VR_TDC_MASK);
static const struct attribute *freq_attrs[] = {
&dev_attr_punit_req_freq_mhz.attr,
static const struct attribute *throttle_reason_attrs[] = {
&attr_throttle_reason_status.attr,
&attr_throttle_reason_pl1.attr,
&attr_throttle_reason_pl2.attr,
......@@ -763,12 +762,20 @@ void intel_gt_sysfs_pm_init(struct intel_gt *gt, struct kobject *kobj)
if (!is_object_gt(kobj))
return;
ret = sysfs_create_files(kobj, freq_attrs);
ret = sysfs_create_file(kobj, &dev_attr_punit_req_freq_mhz.attr);
if (ret)
drm_warn(&gt->i915->drm,
"failed to create gt%u throttle sysfs files (%pe)",
"failed to create gt%u punit_req_freq_mhz sysfs (%pe)",
gt->info.id, ERR_PTR(ret));
if (GRAPHICS_VER(gt->i915) >= 11) {
ret = sysfs_create_files(kobj, throttle_reason_attrs);
if (ret)
drm_warn(&gt->i915->drm,
"failed to create gt%u throttle sysfs files (%pe)",
gt->info.id, ERR_PTR(ret));
}
if (HAS_MEDIA_RATIO_MODE(gt->i915) && intel_uc_uses_guc_slpc(&gt->uc)) {
ret = sysfs_create_files(kobj, media_perf_power_attrs);
if (ret)
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment