summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915/gt/intel_rps.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/i915/gt/intel_rps.c')
-rw-r--r--drivers/gpu/drm/i915/gt/intel_rps.c241
1 files changed, 161 insertions, 80 deletions
diff --git a/drivers/gpu/drm/i915/gt/intel_rps.c b/drivers/gpu/drm/i915/gt/intel_rps.c
index b2671ac59dc0..b01c837ab646 100644
--- a/drivers/gpu/drm/i915/gt/intel_rps.c
+++ b/drivers/gpu/drm/i915/gt/intel_rps.c
@@ -5,21 +5,28 @@
#include <linux/string_helpers.h>
-#include <drm/i915_drm.h>
+#include <drm/intel/i915_drm.h>
+
+#include "display/intel_display_rps.h"
+#include "display/vlv_clock.h"
+#include "soc/intel_dram.h"
-#include "display/intel_display.h"
#include "i915_drv.h"
#include "i915_irq.h"
+#include "i915_reg.h"
+#include "i915_wait_util.h"
#include "intel_breadcrumbs.h"
#include "intel_gt.h"
#include "intel_gt_clock_utils.h"
#include "intel_gt_irq.h"
+#include "intel_gt_pm.h"
#include "intel_gt_pm_irq.h"
+#include "intel_gt_print.h"
#include "intel_gt_regs.h"
#include "intel_mchbar_regs.h"
#include "intel_pcode.h"
#include "intel_rps.h"
-#include "vlv_sideband.h"
+#include "vlv_iosf_sb.h"
#include "../../../platform/x86/intel_ips.h"
#define BUSY_MAX_EI 20u /* ms */
@@ -48,7 +55,7 @@ static struct intel_guc_slpc *rps_to_slpc(struct intel_rps *rps)
{
struct intel_gt *gt = rps_to_gt(rps);
- return &gt->uc.guc.slpc;
+ return &gt_to_guc(gt)->slpc;
}
static bool rps_uses_slpc(struct intel_rps *rps)
@@ -70,14 +77,15 @@ static void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
static void rps_timer(struct timer_list *t)
{
- struct intel_rps *rps = from_timer(rps, t, timer);
+ struct intel_rps *rps = timer_container_of(rps, t, timer);
+ struct intel_gt *gt = rps_to_gt(rps);
struct intel_engine_cs *engine;
ktime_t dt, last, timestamp;
enum intel_engine_id id;
s64 max_busy[3] = {};
timestamp = 0;
- for_each_engine(engine, rps_to_gt(rps), id) {
+ for_each_engine(engine, gt, id) {
s64 busy;
int i;
@@ -121,7 +129,7 @@ static void rps_timer(struct timer_list *t)
busy += div_u64(max_busy[i], 1 << i);
}
- GT_TRACE(rps_to_gt(rps),
+ GT_TRACE(gt,
"busy:%lld [%d%%], max:[%lld, %lld, %lld], interval:%d\n",
busy, (int)div64_u64(100 * busy, dt),
max_busy[0], max_busy[1], max_busy[2],
@@ -131,12 +139,12 @@ static void rps_timer(struct timer_list *t)
rps->cur_freq < rps->max_freq_softlimit) {
rps->pm_iir |= GEN6_PM_RP_UP_THRESHOLD;
rps->pm_interval = 1;
- schedule_work(&rps->work);
+ queue_work(gt->i915->unordered_wq, &rps->work);
} else if (100 * busy < rps->power.down_threshold * dt &&
rps->cur_freq > rps->min_freq_softlimit) {
rps->pm_iir |= GEN6_PM_RP_DOWN_THRESHOLD;
rps->pm_interval = 1;
- schedule_work(&rps->work);
+ queue_work(gt->i915->unordered_wq, &rps->work);
} else {
rps->last_adj = 0;
}
@@ -156,7 +164,7 @@ static void rps_start_timer(struct intel_rps *rps)
static void rps_stop_timer(struct intel_rps *rps)
{
- del_timer_sync(&rps->timer);
+ timer_delete_sync(&rps->timer);
rps->pm_timestamp = ktime_sub(ktime_get(), rps->pm_timestamp);
cancel_work_sync(&rps->work);
}
@@ -260,10 +268,10 @@ static const struct cparams {
u16 c;
} cparams[] = {
{ 1, 1333, 301, 28664 },
- { 1, 1066, 294, 24460 },
+ { 1, 1067, 294, 24460 },
{ 1, 800, 294, 25192 },
{ 0, 1333, 276, 27605 },
- { 0, 1066, 276, 27605 },
+ { 0, 1067, 276, 27605 },
{ 0, 800, 231, 23784 },
};
@@ -271,19 +279,24 @@ static void gen5_rps_init(struct intel_rps *rps)
{
struct drm_i915_private *i915 = rps_to_i915(rps);
struct intel_uncore *uncore = rps_to_uncore(rps);
+ unsigned int fsb_freq, mem_freq;
u8 fmax, fmin, fstart;
u32 rgvmodectl;
int c_m, i;
- if (i915->fsb_freq <= 3200)
+ fsb_freq = intel_fsb_freq(i915);
+ mem_freq = intel_mem_freq(i915);
+
+ if (fsb_freq <= 3200000)
c_m = 0;
- else if (i915->fsb_freq <= 4800)
+ else if (fsb_freq <= 4800000)
c_m = 1;
else
c_m = 2;
for (i = 0; i < ARRAY_SIZE(cparams); i++) {
- if (cparams[i].i == c_m && cparams[i].t == i915->mem_freq) {
+ if (cparams[i].i == c_m &&
+ cparams[i].t == DIV_ROUND_CLOSEST(mem_freq, 1000)) {
rps->ips.m = cparams[i].m;
rps->ips.c = cparams[i].c;
break;
@@ -544,6 +557,7 @@ static unsigned int init_emon(struct intel_uncore *uncore)
static bool gen5_rps_enable(struct intel_rps *rps)
{
struct drm_i915_private *i915 = rps_to_i915(rps);
+ struct intel_display *display = i915->display;
struct intel_uncore *uncore = rps_to_uncore(rps);
u8 fstart, vstart;
u32 rgvmodectl;
@@ -601,9 +615,7 @@ static bool gen5_rps_enable(struct intel_rps *rps)
rps->ips.last_count2 = intel_uncore_read(uncore, GFXEC);
rps->ips.last_time2 = ktime_get_raw_ns();
- spin_lock(&i915->irq_lock);
- ilk_enable_display_irq(i915, DE_PCU_EVENT);
- spin_unlock(&i915->irq_lock);
+ ilk_display_rps_enable(display);
spin_unlock_irq(&mchdev_lock);
@@ -615,14 +627,13 @@ static bool gen5_rps_enable(struct intel_rps *rps)
static void gen5_rps_disable(struct intel_rps *rps)
{
struct drm_i915_private *i915 = rps_to_i915(rps);
+ struct intel_display *display = i915->display;
struct intel_uncore *uncore = rps_to_uncore(rps);
u16 rgvswctl;
spin_lock_irq(&mchdev_lock);
- spin_lock(&i915->irq_lock);
- ilk_disable_display_irq(i915, DE_PCU_EVENT);
- spin_unlock(&i915->irq_lock);
+ ilk_display_rps_disable(display);
rgvswctl = intel_uncore_read16(uncore, MEMSWCTL);
@@ -669,7 +680,6 @@ static void rps_set_power(struct intel_rps *rps, int new_power)
{
struct intel_gt *gt = rps_to_gt(rps);
struct intel_uncore *uncore = gt->uncore;
- u32 threshold_up = 0, threshold_down = 0; /* in % */
u32 ei_up = 0, ei_down = 0;
lockdep_assert_held(&rps->power.mutex);
@@ -677,9 +687,6 @@ static void rps_set_power(struct intel_rps *rps, int new_power)
if (new_power == rps->power.mode)
return;
- threshold_up = 95;
- threshold_down = 85;
-
/* Note the units here are not exactly 1us, but 1280ns. */
switch (new_power) {
case LOW_POWER:
@@ -706,17 +713,22 @@ static void rps_set_power(struct intel_rps *rps, int new_power)
GT_TRACE(gt,
"changing power mode [%d], up %d%% @ %dus, down %d%% @ %dus\n",
- new_power, threshold_up, ei_up, threshold_down, ei_down);
+ new_power,
+ rps->power.up_threshold, ei_up,
+ rps->power.down_threshold, ei_down);
set(uncore, GEN6_RP_UP_EI,
intel_gt_ns_to_pm_interval(gt, ei_up * 1000));
set(uncore, GEN6_RP_UP_THRESHOLD,
- intel_gt_ns_to_pm_interval(gt, ei_up * threshold_up * 10));
+ intel_gt_ns_to_pm_interval(gt,
+ ei_up * rps->power.up_threshold * 10));
set(uncore, GEN6_RP_DOWN_EI,
intel_gt_ns_to_pm_interval(gt, ei_down * 1000));
set(uncore, GEN6_RP_DOWN_THRESHOLD,
- intel_gt_ns_to_pm_interval(gt, ei_down * threshold_down * 10));
+ intel_gt_ns_to_pm_interval(gt,
+ ei_down *
+ rps->power.down_threshold * 10));
set(uncore, GEN6_RP_CONTROL,
(GRAPHICS_VER(gt->i915) > 9 ? 0 : GEN6_RP_MEDIA_TURBO) |
@@ -728,8 +740,6 @@ static void rps_set_power(struct intel_rps *rps, int new_power)
skip_hw_write:
rps->power.mode = new_power;
- rps->power.up_threshold = threshold_up;
- rps->power.down_threshold = threshold_down;
}
static void gen6_rps_set_thresholds(struct intel_rps *rps, u8 val)
@@ -817,9 +827,9 @@ static int vlv_rps_set(struct intel_rps *rps, u8 val)
struct drm_i915_private *i915 = rps_to_i915(rps);
int err;
- vlv_punit_get(i915);
- err = vlv_punit_write(i915, PUNIT_REG_GPU_FREQ_REQ, val);
- vlv_punit_put(i915);
+ vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT));
+ err = vlv_iosf_sb_write(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_REQ, val);
+ vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT));
GT_TRACE(rps_to_gt(rps), "set val:%x, freq:%d\n",
val, intel_gpu_freq(rps, val));
@@ -971,7 +981,7 @@ static int rps_set_boost_freq(struct intel_rps *rps, u32 val)
}
mutex_unlock(&rps->lock);
if (boost)
- schedule_work(&rps->work);
+ queue_work(rps_to_gt(rps)->i915->unordered_wq, &rps->work);
return 0;
}
@@ -996,6 +1006,10 @@ void intel_rps_dec_waiters(struct intel_rps *rps)
if (rps_uses_slpc(rps)) {
slpc = rps_to_slpc(rps);
+ /* Don't decrement num_waiters for req where increment was skipped */
+ if (slpc->power_profile == SLPC_POWER_PROFILES_POWER_SAVING)
+ return;
+
intel_guc_slpc_dec_waiters(slpc);
} else {
atomic_dec(&rps->num_waiters);
@@ -1009,6 +1023,10 @@ void intel_rps_boost(struct i915_request *rq)
if (i915_request_signaled(rq) || i915_request_has_waitboost(rq))
return;
+ /* Waitboost is not needed for contexts marked with a Freq hint */
+ if (test_bit(CONTEXT_LOW_LATENCY, &rq->context->flags))
+ return;
+
/* Serializes with i915_request_retire() */
if (!test_and_set_bit(I915_FENCE_FLAG_BOOST, &rq->fence.flags)) {
struct intel_rps *rps = &READ_ONCE(rq->engine)->gt->rps;
@@ -1016,14 +1034,23 @@ void intel_rps_boost(struct i915_request *rq)
if (rps_uses_slpc(rps)) {
slpc = rps_to_slpc(rps);
- if (slpc->min_freq_softlimit >= slpc->boost_freq)
+ /* Waitboost should not be done with power saving profile */
+ if (slpc->power_profile == SLPC_POWER_PROFILES_POWER_SAVING)
return;
/* Return if old value is non zero */
if (!atomic_fetch_inc(&slpc->num_waiters)) {
+ /*
+ * Skip queuing boost work if frequency is already boosted,
+ * but still increment num_waiters.
+ */
+ if (slpc->min_freq_softlimit >= slpc->boost_freq)
+ return;
+
GT_TRACE(rps_to_gt(rps), "boost fence:%llx:%llx\n",
rq->fence.context, rq->fence.seqno);
- schedule_work(&slpc->boost_work);
+ queue_work(rps_to_gt(rps)->i915->unordered_wq,
+ &slpc->boost_work);
}
return;
@@ -1039,7 +1066,7 @@ void intel_rps_boost(struct i915_request *rq)
rq->fence.context, rq->fence.seqno);
if (READ_ONCE(rps->cur_freq) < rps->boost_freq)
- schedule_work(&rps->work);
+ queue_work(rps_to_gt(rps)->i915->unordered_wq, &rps->work);
WRITE_ONCE(rps->boosts, rps->boosts + 1); /* debug only */
}
@@ -1081,11 +1108,7 @@ static u32 intel_rps_read_state_cap(struct intel_rps *rps)
struct drm_i915_private *i915 = rps_to_i915(rps);
struct intel_uncore *uncore = rps_to_uncore(rps);
- if (IS_PONTEVECCHIO(i915))
- return intel_uncore_read(uncore, PVC_RP_STATE_CAP);
- else if (IS_XEHPSDV(i915))
- return intel_uncore_read(uncore, XEHPSDV_RP_STATE_CAP);
- else if (IS_GEN9_LP(i915))
+ if (IS_GEN9_LP(i915))
return intel_uncore_read(uncore, BXT_RP_STATE_CAP);
else
return intel_uncore_read(uncore, GEN6_RP_STATE_CAP);
@@ -1156,7 +1179,7 @@ void gen6_rps_get_freq_caps(struct intel_rps *rps, struct intel_rps_freq_caps *c
{
struct drm_i915_private *i915 = rps_to_i915(rps);
- if (IS_METEORLAKE(i915))
+ if (GRAPHICS_VER_FULL(i915) >= IP_VER(12, 70))
return mtl_get_freq_caps(rps, caps);
else
return __gen6_rps_get_freq_caps(rps, caps);
@@ -1264,7 +1287,7 @@ static int chv_rps_max_freq(struct intel_rps *rps)
struct intel_gt *gt = rps_to_gt(rps);
u32 val;
- val = vlv_punit_read(i915, FB_GFX_FMAX_AT_VMAX_FUSE);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, FB_GFX_FMAX_AT_VMAX_FUSE);
switch (gt->info.sseu.eu_total) {
case 8:
@@ -1291,7 +1314,7 @@ static int chv_rps_rpe_freq(struct intel_rps *rps)
struct drm_i915_private *i915 = rps_to_i915(rps);
u32 val;
- val = vlv_punit_read(i915, PUNIT_GPU_DUTYCYCLE_REG);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_GPU_DUTYCYCLE_REG);
val >>= PUNIT_GPU_DUTYCYCLE_RPE_FREQ_SHIFT;
return val & PUNIT_GPU_DUTYCYCLE_RPE_FREQ_MASK;
@@ -1302,7 +1325,7 @@ static int chv_rps_guar_freq(struct intel_rps *rps)
struct drm_i915_private *i915 = rps_to_i915(rps);
u32 val;
- val = vlv_punit_read(i915, FB_GFX_FMAX_AT_VMAX_FUSE);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, FB_GFX_FMAX_AT_VMAX_FUSE);
return val & FB_GFX_FREQ_FUSE_MASK;
}
@@ -1312,7 +1335,7 @@ static u32 chv_rps_min_freq(struct intel_rps *rps)
struct drm_i915_private *i915 = rps_to_i915(rps);
u32 val;
- val = vlv_punit_read(i915, FB_GFX_FMIN_AT_VMIN_FUSE);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, FB_GFX_FMIN_AT_VMIN_FUSE);
val >>= FB_GFX_FMIN_AT_VMIN_FUSE_SHIFT;
return val & FB_GFX_FREQ_FUSE_MASK;
@@ -1346,14 +1369,14 @@ static bool chv_rps_enable(struct intel_rps *rps)
GEN6_PM_RP_DOWN_TIMEOUT);
/* Setting Fixed Bias */
- vlv_punit_get(i915);
+ vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT));
val = VLV_OVERRIDE_EN | VLV_SOC_TDP_EN | CHV_BIAS_CPU_50_SOC_50;
- vlv_punit_write(i915, VLV_TURBO_SOC_OVERRIDE, val);
+ vlv_iosf_sb_write(&i915->drm, VLV_IOSF_SB_PUNIT, VLV_TURBO_SOC_OVERRIDE, val);
- val = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_STS);
- vlv_punit_put(i915);
+ vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT));
/* RPS code assumes GPLL is used */
drm_WARN_ONCE(&i915->drm, (val & GPLLENABLE) == 0,
@@ -1371,7 +1394,7 @@ static int vlv_rps_guar_freq(struct intel_rps *rps)
struct drm_i915_private *i915 = rps_to_i915(rps);
u32 val, rp1;
- val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FREQ_FUSE);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_NC, IOSF_NC_FB_GFX_FREQ_FUSE);
rp1 = val & FB_GFX_FGUARANTEED_FREQ_FUSE_MASK;
rp1 >>= FB_GFX_FGUARANTEED_FREQ_FUSE_SHIFT;
@@ -1384,7 +1407,7 @@ static int vlv_rps_max_freq(struct intel_rps *rps)
struct drm_i915_private *i915 = rps_to_i915(rps);
u32 val, rp0;
- val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FREQ_FUSE);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_NC, IOSF_NC_FB_GFX_FREQ_FUSE);
rp0 = (val & FB_GFX_MAX_FREQ_FUSE_MASK) >> FB_GFX_MAX_FREQ_FUSE_SHIFT;
/* Clamp to max */
@@ -1398,9 +1421,9 @@ static int vlv_rps_rpe_freq(struct intel_rps *rps)
struct drm_i915_private *i915 = rps_to_i915(rps);
u32 val, rpe;
- val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FMAX_FUSE_LO);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_NC, IOSF_NC_FB_GFX_FMAX_FUSE_LO);
rpe = (val & FB_FMAX_VMIN_FREQ_LO_MASK) >> FB_FMAX_VMIN_FREQ_LO_SHIFT;
- val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FMAX_FUSE_HI);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_NC, IOSF_NC_FB_GFX_FMAX_FUSE_HI);
rpe |= (val & FB_FMAX_VMIN_FREQ_HI_MASK) << 5;
return rpe;
@@ -1411,7 +1434,7 @@ static int vlv_rps_min_freq(struct intel_rps *rps)
struct drm_i915_private *i915 = rps_to_i915(rps);
u32 val;
- val = vlv_punit_read(i915, PUNIT_REG_GPU_LFM) & 0xff;
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_LFM) & 0xff;
/*
* According to the BYT Punit GPU turbo HAS 1.1.6.3 the minimum value
* for the minimum frequency in GPLL mode is 0xc1. Contrary to this on
@@ -1447,15 +1470,15 @@ static bool vlv_rps_enable(struct intel_rps *rps)
/* WaGsvRC0ResidencyMethod:vlv */
rps->pm_events = GEN6_PM_RP_UP_EI_EXPIRED;
- vlv_punit_get(i915);
+ vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT));
/* Setting Fixed Bias */
val = VLV_OVERRIDE_EN | VLV_SOC_TDP_EN | VLV_BIAS_CPU_125_SOC_875;
- vlv_punit_write(i915, VLV_TURBO_SOC_OVERRIDE, val);
+ vlv_iosf_sb_write(&i915->drm, VLV_IOSF_SB_PUNIT, VLV_TURBO_SOC_OVERRIDE, val);
- val = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS);
+ val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_STS);
- vlv_punit_put(i915);
+ vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT));
/* RPS code assumes GPLL is used */
drm_WARN_ONCE(&i915->drm, (val & GPLLENABLE) == 0,
@@ -1555,10 +1578,12 @@ void intel_rps_enable(struct intel_rps *rps)
return;
GT_TRACE(rps_to_gt(rps),
- "min:%x, max:%x, freq:[%d, %d]\n",
+ "min:%x, max:%x, freq:[%d, %d], thresholds:[%u, %u]\n",
rps->min_freq, rps->max_freq,
intel_gpu_freq(rps, rps->min_freq),
- intel_gpu_freq(rps, rps->max_freq));
+ intel_gpu_freq(rps, rps->max_freq),
+ rps->power.up_threshold,
+ rps->power.down_threshold);
GEM_BUG_ON(rps->max_freq < rps->min_freq);
GEM_BUG_ON(rps->idle_freq > rps->max_freq);
@@ -1665,10 +1690,7 @@ static void vlv_init_gpll_ref_freq(struct intel_rps *rps)
{
struct drm_i915_private *i915 = rps_to_i915(rps);
- rps->gpll_ref_freq =
- vlv_get_cck_clock(i915, "GPLL ref",
- CCK_GPLL_CLOCK_CONTROL,
- i915->czclk_freq);
+ rps->gpll_ref_freq = vlv_clock_get_gpll(&i915->drm);
drm_dbg(&i915->drm, "GPLL reference freq: %d kHz\n",
rps->gpll_ref_freq);
@@ -1678,13 +1700,13 @@ static void vlv_rps_init(struct intel_rps *rps)
{
struct drm_i915_private *i915 = rps_to_i915(rps);
- vlv_iosf_sb_get(i915,
+ vlv_init_gpll_ref_freq(rps);
+
+ vlv_iosf_sb_get(&i915->drm,
BIT(VLV_IOSF_SB_PUNIT) |
BIT(VLV_IOSF_SB_NC) |
BIT(VLV_IOSF_SB_CCK));
- vlv_init_gpll_ref_freq(rps);
-
rps->max_freq = vlv_rps_max_freq(rps);
rps->rp0_freq = rps->max_freq;
drm_dbg(&i915->drm, "max GPU freq: %d MHz (%u)\n",
@@ -1702,7 +1724,7 @@ static void vlv_rps_init(struct intel_rps *rps)
drm_dbg(&i915->drm, "min GPU freq: %d MHz (%u)\n",
intel_gpu_freq(rps, rps->min_freq), rps->min_freq);
- vlv_iosf_sb_put(i915,
+ vlv_iosf_sb_put(&i915->drm,
BIT(VLV_IOSF_SB_PUNIT) |
BIT(VLV_IOSF_SB_NC) |
BIT(VLV_IOSF_SB_CCK));
@@ -1712,13 +1734,13 @@ static void chv_rps_init(struct intel_rps *rps)
{
struct drm_i915_private *i915 = rps_to_i915(rps);
- vlv_iosf_sb_get(i915,
+ vlv_init_gpll_ref_freq(rps);
+
+ vlv_iosf_sb_get(&i915->drm,
BIT(VLV_IOSF_SB_PUNIT) |
BIT(VLV_IOSF_SB_NC) |
BIT(VLV_IOSF_SB_CCK));
- vlv_init_gpll_ref_freq(rps);
-
rps->max_freq = chv_rps_max_freq(rps);
rps->rp0_freq = rps->max_freq;
drm_dbg(&i915->drm, "max GPU freq: %d MHz (%u)\n",
@@ -1736,7 +1758,7 @@ static void chv_rps_init(struct intel_rps *rps)
drm_dbg(&i915->drm, "min GPU freq: %d MHz (%u)\n",
intel_gpu_freq(rps, rps->min_freq), rps->min_freq);
- vlv_iosf_sb_put(i915,
+ vlv_iosf_sb_put(&i915->drm,
BIT(VLV_IOSF_SB_PUNIT) |
BIT(VLV_IOSF_SB_NC) |
BIT(VLV_IOSF_SB_CCK));
@@ -1755,6 +1777,7 @@ static void vlv_c0_read(struct intel_uncore *uncore, struct intel_rps_ei *ei)
static u32 vlv_wa_c0_ei(struct intel_rps *rps, u32 pm_iir)
{
+ struct drm_i915_private *i915 = rps_to_i915(rps);
struct intel_uncore *uncore = rps_to_uncore(rps);
const struct intel_rps_ei *prev = &rps->ei;
struct intel_rps_ei now;
@@ -1771,7 +1794,7 @@ static u32 vlv_wa_c0_ei(struct intel_rps *rps, u32 pm_iir)
time = ktime_us_delta(now.ktime, prev->ktime);
- time *= rps_to_i915(rps)->czclk_freq;
+ time *= vlv_clock_get_czclk(&i915->drm);
/* Workload can be split between render + media,
* e.g. SwapBuffers being blitted in X after being rendered in
@@ -1898,7 +1921,7 @@ void gen11_rps_irq_handler(struct intel_rps *rps, u32 pm_iir)
gen6_gt_pm_mask_irq(gt, events);
rps->pm_iir |= events;
- schedule_work(&rps->work);
+ queue_work(gt->i915->unordered_wq, &rps->work);
}
void gen6_rps_irq_handler(struct intel_rps *rps, u32 pm_iir)
@@ -1915,7 +1938,7 @@ void gen6_rps_irq_handler(struct intel_rps *rps, u32 pm_iir)
gen6_gt_pm_mask_irq(gt, events);
rps->pm_iir |= events;
- schedule_work(&rps->work);
+ queue_work(gt->i915->unordered_wq, &rps->work);
spin_unlock(gt->irq_lock);
}
@@ -2011,6 +2034,12 @@ void intel_rps_init(struct intel_rps *rps)
}
}
+ /* Set default thresholds in % */
+ rps->power.up_threshold = 95;
+ rps_to_gt(rps)->defaults.rps_up_threshold = rps->power.up_threshold;
+ rps->power.down_threshold = 85;
+ rps_to_gt(rps)->defaults.rps_down_threshold = rps->power.down_threshold;
+
/* Finally allow us to boost to max by default */
rps->boost_freq = rps->max_freq;
rps->idle_freq = rps->min_freq;
@@ -2095,9 +2124,9 @@ static u32 __read_cagf(struct intel_rps *rps, bool take_fw)
} else if (GRAPHICS_VER(i915) >= 12) {
r = GEN12_RPSTAT1;
} else if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915)) {
- vlv_punit_get(i915);
- freq = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS);
- vlv_punit_put(i915);
+ vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT));
+ freq = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_STS);
+ vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT));
} else if (GRAPHICS_VER(i915) >= 6) {
r = GEN6_RPSTAT1;
} else {
@@ -2565,6 +2594,58 @@ int intel_rps_set_min_frequency(struct intel_rps *rps, u32 val)
return set_min_freq(rps, val);
}
+u8 intel_rps_get_up_threshold(struct intel_rps *rps)
+{
+ return rps->power.up_threshold;
+}
+
+static int rps_set_threshold(struct intel_rps *rps, u8 *threshold, u8 val)
+{
+ int ret;
+
+ if (val > 100)
+ return -EINVAL;
+
+ ret = mutex_lock_interruptible(&rps->lock);
+ if (ret)
+ return ret;
+
+ if (*threshold == val)
+ goto out_unlock;
+
+ *threshold = val;
+
+ /* Force reset. */
+ rps->last_freq = -1;
+ mutex_lock(&rps->power.mutex);
+ rps->power.mode = -1;
+ mutex_unlock(&rps->power.mutex);
+
+ intel_rps_set(rps, clamp(rps->cur_freq,
+ rps->min_freq_softlimit,
+ rps->max_freq_softlimit));
+
+out_unlock:
+ mutex_unlock(&rps->lock);
+
+ return ret;
+}
+
+int intel_rps_set_up_threshold(struct intel_rps *rps, u8 threshold)
+{
+ return rps_set_threshold(rps, &rps->power.up_threshold, threshold);
+}
+
+u8 intel_rps_get_down_threshold(struct intel_rps *rps)
+{
+ return rps->power.down_threshold;
+}
+
+int intel_rps_set_down_threshold(struct intel_rps *rps, u8 threshold)
+{
+ return rps_set_threshold(rps, &rps->power.down_threshold, threshold);
+}
+
static void intel_rps_set_manual(struct intel_rps *rps, bool enable)
{
struct intel_uncore *uncore = rps_to_uncore(rps);
generated by cgit (git 2.34.1) at 2025-12-25 08:30:30 +0000