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.c61
1 files changed, 36 insertions, 25 deletions
diff --git a/drivers/gpu/drm/i915/gt/intel_rps.c b/drivers/gpu/drm/i915/gt/intel_rps.c
index b629eeb14002..ee5835c29c03 100644
--- a/drivers/gpu/drm/i915/gt/intel_rps.c
+++ b/drivers/gpu/drm/i915/gt/intel_rps.c
@@ -43,7 +43,7 @@ static u32 rps_pm_sanitize_mask(struct intel_rps *rps, u32 mask)
return mask & ~rps->pm_intrmsk_mbz;
}
-static inline void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
+static void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
{
intel_uncore_write_fw(uncore, reg, val);
}
@@ -400,7 +400,7 @@ static unsigned int gen5_invert_freq(struct intel_rps *rps,
return val;
}
-static bool gen5_rps_set(struct intel_rps *rps, u8 val)
+static int __gen5_rps_set(struct intel_rps *rps, u8 val)
{
struct intel_uncore *uncore = rps_to_uncore(rps);
u16 rgvswctl;
@@ -410,7 +410,7 @@ static bool gen5_rps_set(struct intel_rps *rps, u8 val)
rgvswctl = intel_uncore_read16(uncore, MEMSWCTL);
if (rgvswctl & MEMCTL_CMD_STS) {
DRM_DEBUG("gpu busy, RCS change rejected\n");
- return false; /* still busy with another command */
+ return -EBUSY; /* still busy with another command */
}
/* Invert the frequency bin into an ips delay */
@@ -426,7 +426,18 @@ static bool gen5_rps_set(struct intel_rps *rps, u8 val)
rgvswctl |= MEMCTL_CMD_STS;
intel_uncore_write16(uncore, MEMSWCTL, rgvswctl);
- return true;
+ return 0;
+}
+
+static int gen5_rps_set(struct intel_rps *rps, u8 val)
+{
+ int err;
+
+ spin_lock_irq(&mchdev_lock);
+ err = __gen5_rps_set(rps, val);
+ spin_unlock_irq(&mchdev_lock);
+
+ return err;
}
static unsigned long intel_pxfreq(u32 vidfreq)
@@ -557,7 +568,7 @@ static bool gen5_rps_enable(struct intel_rps *rps)
"stuck trying to change perf mode\n");
mdelay(1);
- gen5_rps_set(rps, rps->cur_freq);
+ __gen5_rps_set(rps, rps->cur_freq);
rps->ips.last_count1 = intel_uncore_read(uncore, DMIEC);
rps->ips.last_count1 += intel_uncore_read(uncore, DDREC);
@@ -599,7 +610,7 @@ static void gen5_rps_disable(struct intel_rps *rps)
intel_uncore_write(uncore, MEMINTRSTS, MEMINT_EVAL_CHG);
/* Go back to the starting frequency */
- gen5_rps_set(rps, rps->idle_freq);
+ __gen5_rps_set(rps, rps->idle_freq);
mdelay(1);
rgvswctl |= MEMCTL_CMD_STS;
intel_uncore_write(uncore, MEMSWCTL, rgvswctl);
@@ -797,20 +808,19 @@ static int rps_set(struct intel_rps *rps, u8 val, bool update)
struct drm_i915_private *i915 = rps_to_i915(rps);
int err;
- if (INTEL_GEN(i915) < 6)
- return 0;
-
if (val == rps->last_freq)
return 0;
if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915))
err = vlv_rps_set(rps, val);
- else
+ else if (INTEL_GEN(i915) >= 6)
err = gen6_rps_set(rps, val);
+ else
+ err = gen5_rps_set(rps, val);
if (err)
return err;
- if (update)
+ if (update && INTEL_GEN(i915) >= 6)
gen6_rps_set_thresholds(rps, val);
rps->last_freq = val;
@@ -852,6 +862,8 @@ void intel_rps_park(struct intel_rps *rps)
{
int adj;
+ GEM_BUG_ON(atomic_read(&rps->num_waiters));
+
if (!intel_rps_clear_active(rps))
return;
@@ -907,28 +919,27 @@ void intel_rps_park(struct intel_rps *rps)
void intel_rps_boost(struct i915_request *rq)
{
- struct intel_rps *rps = &READ_ONCE(rq->engine)->gt->rps;
- unsigned long flags;
-
- if (i915_request_signaled(rq) || !intel_rps_is_active(rps))
+ if (i915_request_signaled(rq) || i915_request_has_waitboost(rq))
return;
/* Serializes with i915_request_retire() */
- spin_lock_irqsave(&rq->lock, flags);
- if (!i915_request_has_waitboost(rq) &&
- !dma_fence_is_signaled_locked(&rq->fence)) {
- set_bit(I915_FENCE_FLAG_BOOST, &rq->fence.flags);
+ if (!test_and_set_bit(I915_FENCE_FLAG_BOOST, &rq->fence.flags)) {
+ struct intel_rps *rps = &READ_ONCE(rq->engine)->gt->rps;
+
+ if (atomic_fetch_inc(&rps->num_waiters))
+ return;
+
+ if (!intel_rps_is_active(rps))
+ return;
GT_TRACE(rps_to_gt(rps), "boost fence:%llx:%llx\n",
rq->fence.context, rq->fence.seqno);
- if (!atomic_fetch_inc(&rps->num_waiters) &&
- READ_ONCE(rps->cur_freq) < rps->boost_freq)
+ if (READ_ONCE(rps->cur_freq) < rps->boost_freq)
schedule_work(&rps->work);
- atomic_inc(&rps->boosts);
+ WRITE_ONCE(rps->boosts, rps->boosts + 1); /* debug only */
}
- spin_unlock_irqrestore(&rq->lock, flags);
}
int intel_rps_set(struct intel_rps *rps, u8 val)
@@ -1798,7 +1809,7 @@ void gen5_rps_irq_handler(struct intel_rps *rps)
rps->min_freq_softlimit,
rps->max_freq_softlimit);
- if (new_freq != rps->cur_freq && gen5_rps_set(rps, new_freq))
+ if (new_freq != rps->cur_freq && !__gen5_rps_set(rps, new_freq))
rps->cur_freq = new_freq;
spin_unlock(&mchdev_lock);
@@ -2109,7 +2120,7 @@ bool i915_gpu_turbo_disable(void)
spin_lock_irq(&mchdev_lock);
rps->max_freq_softlimit = rps->min_freq;
- ret = gen5_rps_set(&i915->gt.rps, rps->min_freq);
+ ret = !__gen5_rps_set(&i915->gt.rps, rps->min_freq);
spin_unlock_irq(&mchdev_lock);
drm_dev_put(&i915->drm);