From 8497376707be6d2ab30f6d6662f42ab0c15d2ba1 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 19 May 2020 14:20:46 +0100 Subject: [PATCH 01/16] drm/i915/gt: Incorporate the virtual engine into timeslicing It was quite the oversight to only factor in the normal queue to decide the timeslicing switch priority. By leaving out the next virtual request from the priority decision, we would not timeslice the current engine if there was an available virtual request. Testcase: igt/gem_exec_balancer/sliced Fixes: 3df2deed411e ("drm/i915/execlists: Enable timeslice on partial virtual engine dequeue") Signed-off-by: Chris Wilson Cc: Mika Kuoppala Cc: Tvrtko Ursulin Reviewed-by: Tvrtko Ursulin Link: https://patchwork.freedesktop.org/patch/msgid/20200519132046.22443-3-chris@chris-wilson.co.uk (cherry picked from commit 6ad249ba59badc7ff157d4db1f835748f0e2c9b6) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_lrc.c | 30 +++++++++++++++++++++++------ 1 file changed, 24 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 87e6c5bdd2dc..0b58e6c1e235 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1895,7 +1895,8 @@ static void defer_active(struct intel_engine_cs *engine) static bool need_timeslice(const struct intel_engine_cs *engine, - const struct i915_request *rq) + const struct i915_request *rq, + const struct rb_node *rb) { int hint; @@ -1903,6 +1904,24 @@ need_timeslice(const struct intel_engine_cs *engine, return false; hint = engine->execlists.queue_priority_hint; + + if (rb) { + const struct virtual_engine *ve = + rb_entry(rb, typeof(*ve), nodes[engine->id].rb); + const struct intel_engine_cs *inflight = + intel_context_inflight(&ve->context); + + if (!inflight || inflight == engine) { + struct i915_request *next; + + rcu_read_lock(); + next = READ_ONCE(ve->request); + if (next) + hint = max(hint, rq_prio(next)); + rcu_read_unlock(); + } + } + if (!list_is_last(&rq->sched.link, &engine->active.requests)) hint = max(hint, rq_prio(list_next_entry(rq, sched.link))); @@ -1977,10 +1996,9 @@ static void set_timeslice(struct intel_engine_cs *engine) set_timer_ms(&engine->execlists.timer, duration); } -static void start_timeslice(struct intel_engine_cs *engine) +static void start_timeslice(struct intel_engine_cs *engine, int prio) { struct intel_engine_execlists *execlists = &engine->execlists; - const int prio = queue_prio(execlists); unsigned long duration; if (!intel_engine_has_timeslices(engine)) @@ -2140,7 +2158,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine) __unwind_incomplete_requests(engine); last = NULL; - } else if (need_timeslice(engine, last) && + } else if (need_timeslice(engine, last, rb) && timeslice_expired(execlists, last)) { if (i915_request_completed(last)) { tasklet_hi_schedule(&execlists->tasklet); @@ -2188,7 +2206,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine) * Even if ELSP[1] is occupied and not worthy * of timeslices, our queue might be. */ - start_timeslice(engine); + start_timeslice(engine, queue_prio(execlists)); return; } } @@ -2223,7 +2241,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine) if (last && !can_merge_rq(last, rq)) { spin_unlock(&ve->base.active.lock); - start_timeslice(engine); + start_timeslice(engine, rq_prio(rq)); return; /* leave this for another sibling */ } From 54a9adc4608144c7a17530822246f3cfa9b15d76 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Wed, 27 May 2020 16:05:08 +0200 Subject: [PATCH 02/16] drm/i915/pmu: avoid an maybe-uninitialized warning Conditional spinlocks make it hard for gcc and for lockdep to follow the code flow. This one causes a warning with at least gcc-9 and higher: In file included from include/linux/irq.h:14, from drivers/gpu/drm/i915/i915_pmu.c:7: drivers/gpu/drm/i915/i915_pmu.c: In function 'i915_sample': include/linux/spinlock.h:289:3: error: 'flags' may be used uninitialized in this function [-Werror=maybe-uninitialized] 289 | _raw_spin_unlock_irqrestore(lock, flags); \ | ^~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/gpu/drm/i915/i915_pmu.c:288:17: note: 'flags' was declared here 288 | unsigned long flags; | ^~~~~ Split out the part between the locks into a separate function for readability and to let the compiler figure out what the logic actually is. Fixes: d79e1bd676f0 ("drm/i915/pmu: Only use exclusive mmio access for gen7") Signed-off-by: Arnd Bergmann Reviewed-by: Chris Wilson Signed-off-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20200527140526.1458215-1-arnd@arndb.de (cherry picked from commit 6ec81b82732e2b4a5ac0853fd33919ff1ca94238) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/i915_pmu.c | 84 ++++++++++++++++----------------- 1 file changed, 42 insertions(+), 42 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c index e991a707bdb7..962ded9ce73f 100644 --- a/drivers/gpu/drm/i915/i915_pmu.c +++ b/drivers/gpu/drm/i915/i915_pmu.c @@ -269,12 +269,48 @@ static bool exclusive_mmio_access(const struct drm_i915_private *i915) return IS_GEN(i915, 7); } +static void engine_sample(struct intel_engine_cs *engine, unsigned int period_ns) +{ + struct intel_engine_pmu *pmu = &engine->pmu; + bool busy; + u32 val; + + val = ENGINE_READ_FW(engine, RING_CTL); + if (val == 0) /* powerwell off => engine idle */ + return; + + if (val & RING_WAIT) + add_sample(&pmu->sample[I915_SAMPLE_WAIT], period_ns); + if (val & RING_WAIT_SEMAPHORE) + add_sample(&pmu->sample[I915_SAMPLE_SEMA], period_ns); + + /* No need to sample when busy stats are supported. */ + if (intel_engine_supports_stats(engine)) + return; + + /* + * While waiting on a semaphore or event, MI_MODE reports the + * ring as idle. However, previously using the seqno, and with + * execlists sampling, we account for the ring waiting as the + * engine being busy. Therefore, we record the sample as being + * busy if either waiting or !idle. + */ + busy = val & (RING_WAIT_SEMAPHORE | RING_WAIT); + if (!busy) { + val = ENGINE_READ_FW(engine, RING_MI_MODE); + busy = !(val & MODE_IDLE); + } + if (busy) + add_sample(&pmu->sample[I915_SAMPLE_BUSY], period_ns); +} + static void engines_sample(struct intel_gt *gt, unsigned int period_ns) { struct drm_i915_private *i915 = gt->i915; struct intel_engine_cs *engine; enum intel_engine_id id; + unsigned long flags; if ((i915->pmu.enable & ENGINE_SAMPLE_MASK) == 0) return; @@ -283,53 +319,17 @@ engines_sample(struct intel_gt *gt, unsigned int period_ns) return; for_each_engine(engine, gt, id) { - struct intel_engine_pmu *pmu = &engine->pmu; - spinlock_t *mmio_lock; - unsigned long flags; - bool busy; - u32 val; - if (!intel_engine_pm_get_if_awake(engine)) continue; - mmio_lock = NULL; - if (exclusive_mmio_access(i915)) - mmio_lock = &engine->uncore->lock; - - if (unlikely(mmio_lock)) - spin_lock_irqsave(mmio_lock, flags); - - val = ENGINE_READ_FW(engine, RING_CTL); - if (val == 0) /* powerwell off => engine idle */ - goto skip; - - if (val & RING_WAIT) - add_sample(&pmu->sample[I915_SAMPLE_WAIT], period_ns); - if (val & RING_WAIT_SEMAPHORE) - add_sample(&pmu->sample[I915_SAMPLE_SEMA], period_ns); - - /* No need to sample when busy stats are supported. */ - if (intel_engine_supports_stats(engine)) - goto skip; - - /* - * While waiting on a semaphore or event, MI_MODE reports the - * ring as idle. However, previously using the seqno, and with - * execlists sampling, we account for the ring waiting as the - * engine being busy. Therefore, we record the sample as being - * busy if either waiting or !idle. - */ - busy = val & (RING_WAIT_SEMAPHORE | RING_WAIT); - if (!busy) { - val = ENGINE_READ_FW(engine, RING_MI_MODE); - busy = !(val & MODE_IDLE); + if (exclusive_mmio_access(i915)) { + spin_lock_irqsave(&engine->uncore->lock, flags); + engine_sample(engine, period_ns); + spin_unlock_irqrestore(&engine->uncore->lock, flags); + } else { + engine_sample(engine, period_ns); } - if (busy) - add_sample(&pmu->sample[I915_SAMPLE_BUSY], period_ns); -skip: - if (unlikely(mmio_lock)) - spin_unlock_irqrestore(mmio_lock, flags); intel_engine_pm_put_async(engine); } } From 70cac501b576710b426f3e9f67fcd0a0558df72c Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Wed, 27 May 2020 16:05:09 +0200 Subject: [PATCH 03/16] drm/i915: work around false-positive maybe-uninitialized warning gcc-9 gets confused by the code flow in check_dirty_whitelist: drivers/gpu/drm/i915/gt/selftest_workarounds.c: In function 'check_dirty_whitelist': drivers/gpu/drm/i915/gt/selftest_workarounds.c:492:17: error: 'rsvd' may be used uninitialized in this function [-Werror=maybe-uninitialized] I could not figure out a good way to do this in a way that gcc understands better, so initialize the variable to zero, as last resort. Fixes: aee20aaed887 ("drm/i915: Implement read-only support in whitelist selftest") Signed-off-by: Arnd Bergmann Reviewed-by: Chris Wilson Signed-off-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20200527140526.1458215-2-arnd@arndb.de (cherry picked from commit cc649a9eafc1ef5c40db023084cb94422d08aa84) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/selftest_workarounds.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/i915/gt/selftest_workarounds.c b/drivers/gpu/drm/i915/gt/selftest_workarounds.c index 5ed323254ee1..32785463ec9e 100644 --- a/drivers/gpu/drm/i915/gt/selftest_workarounds.c +++ b/drivers/gpu/drm/i915/gt/selftest_workarounds.c @@ -623,6 +623,8 @@ err_request: err = -EINVAL; goto out_unpin; } + } else { + rsvd = 0; } expect = results[0]; From 3ffbe35321f4671c924a47bb50ca8087f87b2791 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 19 May 2020 07:31:14 +0100 Subject: [PATCH 04/16] drm/i915/selftests: Restore to default heartbeat Since we temporarily disable the heartbeat and restore back to the default value, we can use the stored defaults on the engine and avoid using a local. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20200519063123.20673-3-chris@chris-wilson.co.uk (cherry picked from commit 3a230a554dbbc6cd5016cf1b56ee77cfcd48c7d8) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/selftest_hangcheck.c | 25 +++---- drivers/gpu/drm/i915/gt/selftest_lrc.c | 67 +++++++------------ drivers/gpu/drm/i915/gt/selftest_rps.c | 69 ++++++++------------ drivers/gpu/drm/i915/gt/selftest_timeline.c | 15 ++--- 4 files changed, 67 insertions(+), 109 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c index 2b2efff6e19d..4aa4cc917d8b 100644 --- a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c +++ b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c @@ -310,22 +310,20 @@ static bool wait_until_running(struct hang *h, struct i915_request *rq) 1000)); } -static void engine_heartbeat_disable(struct intel_engine_cs *engine, - unsigned long *saved) +static void engine_heartbeat_disable(struct intel_engine_cs *engine) { - *saved = engine->props.heartbeat_interval_ms; engine->props.heartbeat_interval_ms = 0; intel_engine_pm_get(engine); intel_engine_park_heartbeat(engine); } -static void engine_heartbeat_enable(struct intel_engine_cs *engine, - unsigned long saved) +static void engine_heartbeat_enable(struct intel_engine_cs *engine) { intel_engine_pm_put(engine); - engine->props.heartbeat_interval_ms = saved; + engine->props.heartbeat_interval_ms = + engine->defaults.heartbeat_interval_ms; } static int igt_hang_sanitycheck(void *arg) @@ -473,7 +471,6 @@ static int igt_reset_nop_engine(void *arg) for_each_engine(engine, gt, id) { unsigned int reset_count, reset_engine_count, count; struct intel_context *ce; - unsigned long heartbeat; IGT_TIMEOUT(end_time); int err; @@ -485,7 +482,7 @@ static int igt_reset_nop_engine(void *arg) reset_engine_count = i915_reset_engine_count(global, engine); count = 0; - engine_heartbeat_disable(engine, &heartbeat); + engine_heartbeat_disable(engine); set_bit(I915_RESET_ENGINE + id, >->reset.flags); do { int i; @@ -529,7 +526,7 @@ static int igt_reset_nop_engine(void *arg) } } while (time_before(jiffies, end_time)); clear_bit(I915_RESET_ENGINE + id, >->reset.flags); - engine_heartbeat_enable(engine, heartbeat); + engine_heartbeat_enable(engine); pr_info("%s(%s): %d resets\n", __func__, engine->name, count); @@ -564,7 +561,6 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active) for_each_engine(engine, gt, id) { unsigned int reset_count, reset_engine_count; - unsigned long heartbeat; IGT_TIMEOUT(end_time); if (active && !intel_engine_can_store_dword(engine)) @@ -580,7 +576,7 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active) reset_count = i915_reset_count(global); reset_engine_count = i915_reset_engine_count(global, engine); - engine_heartbeat_disable(engine, &heartbeat); + engine_heartbeat_disable(engine); set_bit(I915_RESET_ENGINE + id, >->reset.flags); do { if (active) { @@ -632,7 +628,7 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active) } } while (time_before(jiffies, end_time)); clear_bit(I915_RESET_ENGINE + id, >->reset.flags); - engine_heartbeat_enable(engine, heartbeat); + engine_heartbeat_enable(engine); if (err) break; @@ -789,7 +785,6 @@ static int __igt_reset_engines(struct intel_gt *gt, struct active_engine threads[I915_NUM_ENGINES] = {}; unsigned long device = i915_reset_count(global); unsigned long count = 0, reported; - unsigned long heartbeat; IGT_TIMEOUT(end_time); if (flags & TEST_ACTIVE && @@ -832,7 +827,7 @@ static int __igt_reset_engines(struct intel_gt *gt, yield(); /* start all threads before we begin */ - engine_heartbeat_disable(engine, &heartbeat); + engine_heartbeat_disable(engine); set_bit(I915_RESET_ENGINE + id, >->reset.flags); do { struct i915_request *rq = NULL; @@ -906,7 +901,7 @@ static int __igt_reset_engines(struct intel_gt *gt, } } while (time_before(jiffies, end_time)); clear_bit(I915_RESET_ENGINE + id, >->reset.flags); - engine_heartbeat_enable(engine, heartbeat); + engine_heartbeat_enable(engine); pr_info("i915_reset_engine(%s:%s): %lu resets\n", engine->name, test_name, count); diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c index 824f99c4cc7c..ba97d1099e27 100644 --- a/drivers/gpu/drm/i915/gt/selftest_lrc.c +++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c @@ -51,22 +51,20 @@ static struct i915_vma *create_scratch(struct intel_gt *gt) return vma; } -static void engine_heartbeat_disable(struct intel_engine_cs *engine, - unsigned long *saved) +static void engine_heartbeat_disable(struct intel_engine_cs *engine) { - *saved = engine->props.heartbeat_interval_ms; engine->props.heartbeat_interval_ms = 0; intel_engine_pm_get(engine); intel_engine_park_heartbeat(engine); } -static void engine_heartbeat_enable(struct intel_engine_cs *engine, - unsigned long saved) +static void engine_heartbeat_enable(struct intel_engine_cs *engine) { intel_engine_pm_put(engine); - engine->props.heartbeat_interval_ms = saved; + engine->props.heartbeat_interval_ms = + engine->defaults.heartbeat_interval_ms; } static bool is_active(struct i915_request *rq) @@ -224,7 +222,6 @@ static int live_unlite_restore(struct intel_gt *gt, int prio) struct intel_context *ce[2] = {}; struct i915_request *rq[2]; struct igt_live_test t; - unsigned long saved; int n; if (prio && !intel_engine_has_preemption(engine)) @@ -237,7 +234,7 @@ static int live_unlite_restore(struct intel_gt *gt, int prio) err = -EIO; break; } - engine_heartbeat_disable(engine, &saved); + engine_heartbeat_disable(engine); for (n = 0; n < ARRAY_SIZE(ce); n++) { struct intel_context *tmp; @@ -345,7 +342,7 @@ err_ce: intel_context_put(ce[n]); } - engine_heartbeat_enable(engine, saved); + engine_heartbeat_enable(engine); if (igt_live_test_end(&t)) err = -EIO; if (err) @@ -466,7 +463,6 @@ static int live_hold_reset(void *arg) for_each_engine(engine, gt, id) { struct intel_context *ce; - unsigned long heartbeat; struct i915_request *rq; ce = intel_context_create(engine); @@ -475,7 +471,7 @@ static int live_hold_reset(void *arg) break; } - engine_heartbeat_disable(engine, &heartbeat); + engine_heartbeat_disable(engine); rq = igt_spinner_create_request(&spin, ce, MI_ARB_CHECK); if (IS_ERR(rq)) { @@ -535,7 +531,7 @@ static int live_hold_reset(void *arg) i915_request_put(rq); out: - engine_heartbeat_enable(engine, heartbeat); + engine_heartbeat_enable(engine); intel_context_put(ce); if (err) break; @@ -580,10 +576,9 @@ static int live_error_interrupt(void *arg) for_each_engine(engine, gt, id) { const struct error_phase *p; - unsigned long heartbeat; int err = 0; - engine_heartbeat_disable(engine, &heartbeat); + engine_heartbeat_disable(engine); for (p = phases; p->error[0] != GOOD; p++) { struct i915_request *client[ARRAY_SIZE(phases->error)]; @@ -682,7 +677,7 @@ out: } } - engine_heartbeat_enable(engine, heartbeat); + engine_heartbeat_enable(engine); if (err) { intel_gt_set_wedged(gt); return err; @@ -895,16 +890,14 @@ static int live_timeslice_preempt(void *arg) enum intel_engine_id id; for_each_engine(engine, gt, id) { - unsigned long saved; - if (!intel_engine_has_preemption(engine)) continue; memset(vaddr, 0, PAGE_SIZE); - engine_heartbeat_disable(engine, &saved); + engine_heartbeat_disable(engine); err = slice_semaphore_queue(engine, vma, count); - engine_heartbeat_enable(engine, saved); + engine_heartbeat_enable(engine); if (err) goto err_pin; @@ -1009,7 +1002,6 @@ static int live_timeslice_rewind(void *arg) enum { X = 1, Z, Y }; struct i915_request *rq[3] = {}; struct intel_context *ce; - unsigned long heartbeat; unsigned long timeslice; int i, err = 0; u32 *slot; @@ -1028,7 +1020,7 @@ static int live_timeslice_rewind(void *arg) * Expect execution/evaluation order XZY */ - engine_heartbeat_disable(engine, &heartbeat); + engine_heartbeat_disable(engine); timeslice = xchg(&engine->props.timeslice_duration_ms, 1); slot = memset32(engine->status_page.addr + 1000, 0, 4); @@ -1122,7 +1114,7 @@ err: wmb(); engine->props.timeslice_duration_ms = timeslice; - engine_heartbeat_enable(engine, heartbeat); + engine_heartbeat_enable(engine); for (i = 0; i < 3; i++) i915_request_put(rq[i]); if (igt_flush_test(gt->i915)) @@ -1202,12 +1194,11 @@ static int live_timeslice_queue(void *arg) .priority = I915_USER_PRIORITY(I915_PRIORITY_MAX), }; struct i915_request *rq, *nop; - unsigned long saved; if (!intel_engine_has_preemption(engine)) continue; - engine_heartbeat_disable(engine, &saved); + engine_heartbeat_disable(engine); memset(vaddr, 0, PAGE_SIZE); /* ELSP[0]: semaphore wait */ @@ -1284,7 +1275,7 @@ static int live_timeslice_queue(void *arg) err_rq: i915_request_put(rq); err_heartbeat: - engine_heartbeat_enable(engine, saved); + engine_heartbeat_enable(engine); if (err) break; } @@ -4153,7 +4144,6 @@ static int reset_virtual_engine(struct intel_gt *gt, { struct intel_engine_cs *engine; struct intel_context *ve; - unsigned long *heartbeat; struct igt_spinner spin; struct i915_request *rq; unsigned int n; @@ -4165,15 +4155,9 @@ static int reset_virtual_engine(struct intel_gt *gt, * descendents are not executed while the capture is in progress. */ - heartbeat = kmalloc_array(nsibling, sizeof(*heartbeat), GFP_KERNEL); - if (!heartbeat) + if (igt_spinner_init(&spin, gt)) return -ENOMEM; - if (igt_spinner_init(&spin, gt)) { - err = -ENOMEM; - goto out_free; - } - ve = intel_execlists_create_virtual(siblings, nsibling); if (IS_ERR(ve)) { err = PTR_ERR(ve); @@ -4181,7 +4165,7 @@ static int reset_virtual_engine(struct intel_gt *gt, } for (n = 0; n < nsibling; n++) - engine_heartbeat_disable(siblings[n], &heartbeat[n]); + engine_heartbeat_disable(siblings[n]); rq = igt_spinner_create_request(&spin, ve, MI_ARB_CHECK); if (IS_ERR(rq)) { @@ -4252,13 +4236,11 @@ out_rq: i915_request_put(rq); out_heartbeat: for (n = 0; n < nsibling; n++) - engine_heartbeat_enable(siblings[n], heartbeat[n]); + engine_heartbeat_enable(siblings[n]); intel_context_put(ve); out_spin: igt_spinner_fini(&spin); -out_free: - kfree(heartbeat); return err; } @@ -4932,9 +4914,7 @@ static int live_lrc_gpr(void *arg) return PTR_ERR(scratch); for_each_engine(engine, gt, id) { - unsigned long heartbeat; - - engine_heartbeat_disable(engine, &heartbeat); + engine_heartbeat_disable(engine); err = __live_lrc_gpr(engine, scratch, false); if (err) @@ -4945,7 +4925,7 @@ static int live_lrc_gpr(void *arg) goto err; err: - engine_heartbeat_enable(engine, heartbeat); + engine_heartbeat_enable(engine); if (igt_flush_test(gt->i915)) err = -EIO; if (err) @@ -5092,10 +5072,9 @@ static int live_lrc_timestamp(void *arg) */ for_each_engine(data.engine, gt, id) { - unsigned long heartbeat; int i, err = 0; - engine_heartbeat_disable(data.engine, &heartbeat); + engine_heartbeat_disable(data.engine); for (i = 0; i < ARRAY_SIZE(data.ce); i++) { struct intel_context *tmp; @@ -5128,7 +5107,7 @@ static int live_lrc_timestamp(void *arg) } err: - engine_heartbeat_enable(data.engine, heartbeat); + engine_heartbeat_enable(data.engine); for (i = 0; i < ARRAY_SIZE(data.ce); i++) { if (!data.ce[i]) break; diff --git a/drivers/gpu/drm/i915/gt/selftest_rps.c b/drivers/gpu/drm/i915/gt/selftest_rps.c index 6275d69aa9cc..5049c3dd08a6 100644 --- a/drivers/gpu/drm/i915/gt/selftest_rps.c +++ b/drivers/gpu/drm/i915/gt/selftest_rps.c @@ -20,24 +20,20 @@ /* Try to isolate the impact of cstates from determing frequency response */ #define CPU_LATENCY 0 /* -1 to disable pm_qos, 0 to disable cstates */ -static unsigned long engine_heartbeat_disable(struct intel_engine_cs *engine) +static void engine_heartbeat_disable(struct intel_engine_cs *engine) { - unsigned long old; - - old = fetch_and_zero(&engine->props.heartbeat_interval_ms); + engine->props.heartbeat_interval_ms = 0; intel_engine_pm_get(engine); intel_engine_park_heartbeat(engine); - - return old; } -static void engine_heartbeat_enable(struct intel_engine_cs *engine, - unsigned long saved) +static void engine_heartbeat_enable(struct intel_engine_cs *engine) { intel_engine_pm_put(engine); - engine->props.heartbeat_interval_ms = saved; + engine->props.heartbeat_interval_ms = + engine->defaults.heartbeat_interval_ms; } static void dummy_rps_work(struct work_struct *wrk) @@ -246,7 +242,6 @@ int live_rps_clock_interval(void *arg) intel_gt_check_clock_frequency(gt); for_each_engine(engine, gt, id) { - unsigned long saved_heartbeat; struct i915_request *rq; u32 cycles; u64 dt; @@ -254,13 +249,13 @@ int live_rps_clock_interval(void *arg) if (!intel_engine_can_store_dword(engine)) continue; - saved_heartbeat = engine_heartbeat_disable(engine); + engine_heartbeat_disable(engine); rq = igt_spinner_create_request(&spin, engine->kernel_context, MI_NOOP); if (IS_ERR(rq)) { - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); err = PTR_ERR(rq); break; } @@ -271,7 +266,7 @@ int live_rps_clock_interval(void *arg) pr_err("%s: RPS spinner did not start\n", engine->name); igt_spinner_end(&spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); intel_gt_set_wedged(engine->gt); err = -EIO; break; @@ -327,7 +322,7 @@ int live_rps_clock_interval(void *arg) intel_uncore_forcewake_put(gt->uncore, FORCEWAKE_ALL); igt_spinner_end(&spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); if (err == 0) { u64 time = intel_gt_pm_interval_to_ns(gt, cycles); @@ -405,7 +400,6 @@ int live_rps_control(void *arg) intel_gt_pm_get(gt); for_each_engine(engine, gt, id) { - unsigned long saved_heartbeat; struct i915_request *rq; ktime_t min_dt, max_dt; int f, limit; @@ -414,7 +408,7 @@ int live_rps_control(void *arg) if (!intel_engine_can_store_dword(engine)) continue; - saved_heartbeat = engine_heartbeat_disable(engine); + engine_heartbeat_disable(engine); rq = igt_spinner_create_request(&spin, engine->kernel_context, @@ -430,7 +424,7 @@ int live_rps_control(void *arg) pr_err("%s: RPS spinner did not start\n", engine->name); igt_spinner_end(&spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); intel_gt_set_wedged(engine->gt); err = -EIO; break; @@ -440,7 +434,7 @@ int live_rps_control(void *arg) pr_err("%s: could not set minimum frequency [%x], only %x!\n", engine->name, rps->min_freq, read_cagf(rps)); igt_spinner_end(&spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); show_pstate_limits(rps); err = -EINVAL; break; @@ -457,7 +451,7 @@ int live_rps_control(void *arg) pr_err("%s: could not restore minimum frequency [%x], only %x!\n", engine->name, rps->min_freq, read_cagf(rps)); igt_spinner_end(&spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); show_pstate_limits(rps); err = -EINVAL; break; @@ -472,7 +466,7 @@ int live_rps_control(void *arg) min_dt = ktime_sub(ktime_get(), min_dt); igt_spinner_end(&spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); pr_info("%s: range:[%x:%uMHz, %x:%uMHz] limit:[%x:%uMHz], %x:%x response %lluns:%lluns\n", engine->name, @@ -635,7 +629,6 @@ int live_rps_frequency_cs(void *arg) rps->work.func = dummy_rps_work; for_each_engine(engine, gt, id) { - unsigned long saved_heartbeat; struct i915_request *rq; struct i915_vma *vma; u32 *cancel, *cntr; @@ -644,14 +637,14 @@ int live_rps_frequency_cs(void *arg) int freq; } min, max; - saved_heartbeat = engine_heartbeat_disable(engine); + engine_heartbeat_disable(engine); vma = create_spin_counter(engine, engine->kernel_context->vm, false, &cancel, &cntr); if (IS_ERR(vma)) { err = PTR_ERR(vma); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); break; } @@ -732,7 +725,7 @@ err_vma: i915_vma_unpin(vma); i915_vma_put(vma); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); if (igt_flush_test(gt->i915)) err = -EIO; if (err) @@ -778,7 +771,6 @@ int live_rps_frequency_srm(void *arg) rps->work.func = dummy_rps_work; for_each_engine(engine, gt, id) { - unsigned long saved_heartbeat; struct i915_request *rq; struct i915_vma *vma; u32 *cancel, *cntr; @@ -787,14 +779,14 @@ int live_rps_frequency_srm(void *arg) int freq; } min, max; - saved_heartbeat = engine_heartbeat_disable(engine); + engine_heartbeat_disable(engine); vma = create_spin_counter(engine, engine->kernel_context->vm, true, &cancel, &cntr); if (IS_ERR(vma)) { err = PTR_ERR(vma); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); break; } @@ -874,7 +866,7 @@ err_vma: i915_vma_unpin(vma); i915_vma_put(vma); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); if (igt_flush_test(gt->i915)) err = -EIO; if (err) @@ -1066,16 +1058,14 @@ int live_rps_interrupt(void *arg) for_each_engine(engine, gt, id) { /* Keep the engine busy with a spinner; expect an UP! */ if (pm_events & GEN6_PM_RP_UP_THRESHOLD) { - unsigned long saved_heartbeat; - intel_gt_pm_wait_for_idle(engine->gt); GEM_BUG_ON(intel_rps_is_active(rps)); - saved_heartbeat = engine_heartbeat_disable(engine); + engine_heartbeat_disable(engine); err = __rps_up_interrupt(rps, engine, &spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); if (err) goto out; @@ -1084,15 +1074,13 @@ int live_rps_interrupt(void *arg) /* Keep the engine awake but idle and check for DOWN */ if (pm_events & GEN6_PM_RP_DOWN_THRESHOLD) { - unsigned long saved_heartbeat; - - saved_heartbeat = engine_heartbeat_disable(engine); + engine_heartbeat_disable(engine); intel_rc6_disable(>->rc6); err = __rps_down_interrupt(rps, engine); intel_rc6_enable(>->rc6); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); if (err) goto out; } @@ -1168,7 +1156,6 @@ int live_rps_power(void *arg) rps->work.func = dummy_rps_work; for_each_engine(engine, gt, id) { - unsigned long saved_heartbeat; struct i915_request *rq; struct { u64 power; @@ -1178,13 +1165,13 @@ int live_rps_power(void *arg) if (!intel_engine_can_store_dword(engine)) continue; - saved_heartbeat = engine_heartbeat_disable(engine); + engine_heartbeat_disable(engine); rq = igt_spinner_create_request(&spin, engine->kernel_context, MI_NOOP); if (IS_ERR(rq)) { - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); err = PTR_ERR(rq); break; } @@ -1195,7 +1182,7 @@ int live_rps_power(void *arg) pr_err("%s: RPS spinner did not start\n", engine->name); igt_spinner_end(&spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); intel_gt_set_wedged(engine->gt); err = -EIO; break; @@ -1208,7 +1195,7 @@ int live_rps_power(void *arg) min.power = measure_power_at(rps, &min.freq); igt_spinner_end(&spin); - engine_heartbeat_enable(engine, saved_heartbeat); + engine_heartbeat_enable(engine); pr_info("%s: min:%llumW @ %uMHz, max:%llumW @ %uMHz\n", engine->name, diff --git a/drivers/gpu/drm/i915/gt/selftest_timeline.c b/drivers/gpu/drm/i915/gt/selftest_timeline.c index c2578a0f2f14..ef1c35073dc0 100644 --- a/drivers/gpu/drm/i915/gt/selftest_timeline.c +++ b/drivers/gpu/drm/i915/gt/selftest_timeline.c @@ -751,22 +751,20 @@ out_free: return err; } -static void engine_heartbeat_disable(struct intel_engine_cs *engine, - unsigned long *saved) +static void engine_heartbeat_disable(struct intel_engine_cs *engine) { - *saved = engine->props.heartbeat_interval_ms; engine->props.heartbeat_interval_ms = 0; intel_engine_pm_get(engine); intel_engine_park_heartbeat(engine); } -static void engine_heartbeat_enable(struct intel_engine_cs *engine, - unsigned long saved) +static void engine_heartbeat_enable(struct intel_engine_cs *engine) { intel_engine_pm_put(engine); - engine->props.heartbeat_interval_ms = saved; + engine->props.heartbeat_interval_ms = + engine->defaults.heartbeat_interval_ms; } static int live_hwsp_rollover_kernel(void *arg) @@ -785,10 +783,9 @@ static int live_hwsp_rollover_kernel(void *arg) struct intel_context *ce = engine->kernel_context; struct intel_timeline *tl = ce->timeline; struct i915_request *rq[3] = {}; - unsigned long heartbeat; int i; - engine_heartbeat_disable(engine, &heartbeat); + engine_heartbeat_disable(engine); if (intel_gt_wait_for_idle(gt, HZ / 2)) { err = -EIO; goto out; @@ -839,7 +836,7 @@ static int live_hwsp_rollover_kernel(void *arg) out: for (i = 0; i < ARRAY_SIZE(rq); i++) i915_request_put(rq[i]); - engine_heartbeat_enable(engine, heartbeat); + engine_heartbeat_enable(engine); if (err) break; } From 4178b5a60cefd4721095cce974af7c142f2a6331 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 27 May 2020 17:24:18 +0100 Subject: [PATCH 05/16] drm/i915/gt: Prevent timeslicing into unpreemptable requests We have a I915_REQUEST_NOPREEMPT flag that we set when we must prevent the HW from preempting during the course of this request. We need to honour this flag and protect the HW even if we have a heartbeat request, or other maximum priority barrier, pending. As such, restrict the timeslicing check to avoid preempting into the topmost priority band, leaving the unpreemptable requests in blissful peace running uninterrupted on the HW. v2: Set the I915_PRIORITY_BARRIER to be less than I915_PRIORITY_UNPREEMPTABLE so that we never submit a request (heartbeat or barrier) that can legitimately preempt the current non-premptable request. Fixes: 2a98f4e65bba ("drm/i915: add infrastructure to hold off preemption on a request") Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Reviewed-by: Tvrtko Ursulin Link: https://patchwork.freedesktop.org/patch/msgid/20200527162418.24755-1-chris@chris-wilson.co.uk (cherry picked from commit b72f02d78e4f257761ed003444ae52083f962076) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_lrc.c | 1 + drivers/gpu/drm/i915/gt/selftest_lrc.c | 118 ++++++++++++++++++++- drivers/gpu/drm/i915/i915_priolist_types.h | 2 +- 3 files changed, 119 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 0b58e6c1e235..1858331e423e 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1925,6 +1925,7 @@ need_timeslice(const struct intel_engine_cs *engine, if (!list_is_last(&rq->sched.link, &engine->active.requests)) hint = max(hint, rq_prio(list_next_entry(rq, sched.link))); + GEM_BUG_ON(hint >= I915_PRIORITY_UNPREEMPTABLE); return hint >= effective_prio(rq); } diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c index ba97d1099e27..924bc01ef526 100644 --- a/drivers/gpu/drm/i915/gt/selftest_lrc.c +++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c @@ -823,7 +823,7 @@ slice_semaphore_queue(struct intel_engine_cs *outer, } } - err = release_queue(outer, vma, n, INT_MAX); + err = release_queue(outer, vma, n, I915_PRIORITY_BARRIER); if (err) goto out; @@ -1289,6 +1289,121 @@ err_obj: return err; } +static int live_timeslice_nopreempt(void *arg) +{ + struct intel_gt *gt = arg; + struct intel_engine_cs *engine; + enum intel_engine_id id; + struct igt_spinner spin; + int err = 0; + + /* + * We should not timeslice into a request that is marked with + * I915_REQUEST_NOPREEMPT. + */ + if (!IS_ACTIVE(CONFIG_DRM_I915_TIMESLICE_DURATION)) + return 0; + + if (igt_spinner_init(&spin, gt)) + return -ENOMEM; + + for_each_engine(engine, gt, id) { + struct intel_context *ce; + struct i915_request *rq; + unsigned long timeslice; + + if (!intel_engine_has_preemption(engine)) + continue; + + ce = intel_context_create(engine); + if (IS_ERR(ce)) { + err = PTR_ERR(ce); + break; + } + + engine_heartbeat_disable(engine); + timeslice = xchg(&engine->props.timeslice_duration_ms, 1); + + /* Create an unpreemptible spinner */ + + rq = igt_spinner_create_request(&spin, ce, MI_ARB_CHECK); + intel_context_put(ce); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + goto out_heartbeat; + } + + i915_request_get(rq); + i915_request_add(rq); + + if (!igt_wait_for_spinner(&spin, rq)) { + i915_request_put(rq); + err = -ETIME; + goto out_spin; + } + + set_bit(I915_FENCE_FLAG_NOPREEMPT, &rq->fence.flags); + i915_request_put(rq); + + /* Followed by a maximum priority barrier (heartbeat) */ + + ce = intel_context_create(engine); + if (IS_ERR(ce)) { + err = PTR_ERR(rq); + goto out_spin; + } + + rq = intel_context_create_request(ce); + intel_context_put(ce); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + goto out_spin; + } + + rq->sched.attr.priority = I915_PRIORITY_BARRIER; + i915_request_get(rq); + i915_request_add(rq); + + /* + * Wait until the barrier is in ELSP, and we know timeslicing + * will have been activated. + */ + if (wait_for_submit(engine, rq, HZ / 2)) { + i915_request_put(rq); + err = -ETIME; + goto out_spin; + } + + /* + * Since the ELSP[0] request is unpreemptible, it should not + * allow the maximum priority barrier through. Wait long + * enough to see if it is timesliced in by mistake. + */ + if (i915_request_wait(rq, 0, timeslice_threshold(engine)) >= 0) { + pr_err("%s: I915_PRIORITY_BARRIER request completed, bypassing no-preempt request\n", + engine->name); + err = -EINVAL; + } + i915_request_put(rq); + +out_spin: + igt_spinner_end(&spin); +out_heartbeat: + xchg(&engine->props.timeslice_duration_ms, timeslice); + engine_heartbeat_enable(engine); + if (err) + break; + + if (igt_flush_test(gt->i915)) { + err = -EIO; + break; + } + } + + igt_spinner_fini(&spin); + return err; +} + static int live_busywait_preempt(void *arg) { struct intel_gt *gt = arg; @@ -4296,6 +4411,7 @@ int intel_execlists_live_selftests(struct drm_i915_private *i915) SUBTEST(live_timeslice_preempt), SUBTEST(live_timeslice_rewind), SUBTEST(live_timeslice_queue), + SUBTEST(live_timeslice_nopreempt), SUBTEST(live_busywait_preempt), SUBTEST(live_preempt), SUBTEST(live_late_preempt), diff --git a/drivers/gpu/drm/i915/i915_priolist_types.h b/drivers/gpu/drm/i915/i915_priolist_types.h index 5003a71113cb..8aa7866ec6b6 100644 --- a/drivers/gpu/drm/i915/i915_priolist_types.h +++ b/drivers/gpu/drm/i915/i915_priolist_types.h @@ -42,7 +42,7 @@ enum { * active request. */ #define I915_PRIORITY_UNPREEMPTABLE INT_MAX -#define I915_PRIORITY_BARRIER INT_MAX +#define I915_PRIORITY_BARRIER (I915_PRIORITY_UNPREEMPTABLE - 1) struct i915_priolist { struct list_head requests[I915_PRIORITY_COUNT]; From a43555ac908c604f45ed98628805aec9355b9737 Mon Sep 17 00:00:00 2001 From: Khaled Almahallawy Date: Mon, 8 Jun 2020 13:45:37 -0700 Subject: [PATCH 06/16] drm/i915/tc: fix the reset of ln0 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Setting ln0 similar to ln1 Fixes: 3b51be4e4061b ("drm/i915/tc: Update DP_MODE programming") Cc: # v5.5+ Signed-off-by: Khaled Almahallawy Reviewed-by: José Roberto de Souza Signed-off-by: Imre Deak Link: https://patchwork.freedesktop.org/patch/msgid/20200608204537.28468-1-khaled.almahallawy@intel.com (cherry picked from commit 4f72a8ee819d57d7329d88f487a2fc9b45153177) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/display/intel_ddi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c index aa22465bb56e..f53de9664903 100644 --- a/drivers/gpu/drm/i915/display/intel_ddi.c +++ b/drivers/gpu/drm/i915/display/intel_ddi.c @@ -2987,7 +2987,7 @@ icl_program_mg_dp_mode(struct intel_digital_port *intel_dig_port, ln1 = intel_de_read(dev_priv, MG_DP_MODE(1, tc_port)); } - ln0 &= ~(MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X1_MODE); + ln0 &= ~(MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X2_MODE); ln1 &= ~(MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X2_MODE); /* DPPATC */ From 8ab3a3812aa90e488813e719308ffd807b865624 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 9 Jun 2020 16:17:23 +0100 Subject: [PATCH 07/16] drm/i915/gt: Incrementally check for rewinding In commit 5ba32c7be81e ("drm/i915/execlists: Always force a context reload when rewinding RING_TAIL"), we placed the check for rewinding a context on actually submitting the next request in that context. This was so that we only had to check once, and could do so with precision avoiding as many forced restores as possible. For example, to ensure that we can resubmit the same request a couple of times, we include a small wa_tail such that on the next submission, the ring->tail will appear to move forwards when resubmitting the same request. This is very common as it will happen for every lite-restore to fill the second port after a context switch. However, intel_ring_direction() is limited in precision to movements of upto half the ring size. The consequence being that if we tried to unwind many requests, we could exceed half the ring and flip the sense of the direction, so missing a force restore. As no request can be greater than half the ring (i.e. 2048 bytes in the smallest case), we can check for rollback incrementally. As we check against the tail that would be submitted, we do not lose any sensitivity and allow lite restores for the simple case. We still need to double check upon submitting the context, to allow for multiple preemptions and resubmissions. Fixes: 5ba32c7be81e ("drm/i915/execlists: Always force a context reload when rewinding RING_TAIL") Signed-off-by: Chris Wilson Cc: Mika Kuoppala Cc: # v5.4+ Reviewed-by: Bruce Chang Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20200609151723.12971-1-chris@chris-wilson.co.uk (cherry picked from commit e36ba817fa966f81fb1c8d16f3721b5a644b2fa9) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_engine_cs.c | 4 +- drivers/gpu/drm/i915/gt/intel_lrc.c | 21 +++- drivers/gpu/drm/i915/gt/intel_ring.c | 4 + drivers/gpu/drm/i915/gt/selftest_mocs.c | 18 ++- drivers/gpu/drm/i915/gt/selftest_ring.c | 110 ++++++++++++++++++ .../drm/i915/selftests/i915_mock_selftests.h | 1 + 6 files changed, 154 insertions(+), 4 deletions(-) create mode 100644 drivers/gpu/drm/i915/gt/selftest_ring.c diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c index da5b61085257..8691eb61e185 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c @@ -646,7 +646,7 @@ static int engine_setup_common(struct intel_engine_cs *engine) struct measure_breadcrumb { struct i915_request rq; struct intel_ring ring; - u32 cs[1024]; + u32 cs[2048]; }; static int measure_breadcrumb_dw(struct intel_context *ce) @@ -668,6 +668,8 @@ static int measure_breadcrumb_dw(struct intel_context *ce) frame->ring.vaddr = frame->cs; frame->ring.size = sizeof(frame->cs); + frame->ring.wrap = + BITS_PER_TYPE(frame->ring.size) - ilog2(frame->ring.size); frame->ring.effective_size = frame->ring.size; intel_ring_update_space(&frame->ring); frame->rq.ring = &frame->ring; diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 1858331e423e..7c3d8ef4a47c 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1134,6 +1134,13 @@ __unwind_incomplete_requests(struct intel_engine_cs *engine) list_move(&rq->sched.link, pl); set_bit(I915_FENCE_FLAG_PQUEUE, &rq->fence.flags); + /* Check in case we rollback so far we wrap [size/2] */ + if (intel_ring_direction(rq->ring, + intel_ring_wrap(rq->ring, + rq->tail), + rq->ring->tail) > 0) + rq->context->lrc.desc |= CTX_DESC_FORCE_RESTORE; + active = rq; } else { struct intel_engine_cs *owner = rq->context->engine; @@ -1498,8 +1505,9 @@ static u64 execlists_update_context(struct i915_request *rq) * HW has a tendency to ignore us rewinding the TAIL to the end of * an earlier request. */ + GEM_BUG_ON(ce->lrc_reg_state[CTX_RING_TAIL] != rq->ring->tail); + prev = rq->ring->tail; tail = intel_ring_set_tail(rq->ring, rq->tail); - prev = ce->lrc_reg_state[CTX_RING_TAIL]; if (unlikely(intel_ring_direction(rq->ring, tail, prev) <= 0)) desc |= CTX_DESC_FORCE_RESTORE; ce->lrc_reg_state[CTX_RING_TAIL] = tail; @@ -4758,6 +4766,14 @@ static int gen12_emit_flush(struct i915_request *request, u32 mode) return 0; } +static void assert_request_valid(struct i915_request *rq) +{ + struct intel_ring *ring __maybe_unused = rq->ring; + + /* Can we unwind this request without appearing to go forwards? */ + GEM_BUG_ON(intel_ring_direction(ring, rq->wa_tail, rq->head) <= 0); +} + /* * Reserve space for 2 NOOPs at the end of each request to be * used as a workaround for not being allowed to do lite @@ -4770,6 +4786,9 @@ static u32 *gen8_emit_wa_tail(struct i915_request *request, u32 *cs) *cs++ = MI_NOOP; request->wa_tail = intel_ring_offset(request, cs); + /* Check that entire request is less than half the ring */ + assert_request_valid(request); + return cs; } diff --git a/drivers/gpu/drm/i915/gt/intel_ring.c b/drivers/gpu/drm/i915/gt/intel_ring.c index 8cda1b7e17ba..bdb324167ef3 100644 --- a/drivers/gpu/drm/i915/gt/intel_ring.c +++ b/drivers/gpu/drm/i915/gt/intel_ring.c @@ -315,3 +315,7 @@ int intel_ring_cacheline_align(struct i915_request *rq) GEM_BUG_ON(rq->ring->emit & (CACHELINE_BYTES - 1)); return 0; } + +#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST) +#include "selftest_ring.c" +#endif diff --git a/drivers/gpu/drm/i915/gt/selftest_mocs.c b/drivers/gpu/drm/i915/gt/selftest_mocs.c index 8831ffee2061..63f87d8608c3 100644 --- a/drivers/gpu/drm/i915/gt/selftest_mocs.c +++ b/drivers/gpu/drm/i915/gt/selftest_mocs.c @@ -18,6 +18,20 @@ struct live_mocs { void *vaddr; }; +static struct intel_context *mocs_context_create(struct intel_engine_cs *engine) +{ + struct intel_context *ce; + + ce = intel_context_create(engine); + if (IS_ERR(ce)) + return ce; + + /* We build large requests to read the registers from the ring */ + ce->ring = __intel_context_ring_size(SZ_16K); + + return ce; +} + static int request_add_sync(struct i915_request *rq, int err) { i915_request_get(rq); @@ -301,7 +315,7 @@ static int live_mocs_clean(void *arg) for_each_engine(engine, gt, id) { struct intel_context *ce; - ce = intel_context_create(engine); + ce = mocs_context_create(engine); if (IS_ERR(ce)) { err = PTR_ERR(ce); break; @@ -395,7 +409,7 @@ static int live_mocs_reset(void *arg) for_each_engine(engine, gt, id) { struct intel_context *ce; - ce = intel_context_create(engine); + ce = mocs_context_create(engine); if (IS_ERR(ce)) { err = PTR_ERR(ce); break; diff --git a/drivers/gpu/drm/i915/gt/selftest_ring.c b/drivers/gpu/drm/i915/gt/selftest_ring.c new file mode 100644 index 000000000000..2a8c534dc125 --- /dev/null +++ b/drivers/gpu/drm/i915/gt/selftest_ring.c @@ -0,0 +1,110 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright © 2020 Intel Corporation + */ + +static struct intel_ring *mock_ring(unsigned long sz) +{ + struct intel_ring *ring; + + ring = kzalloc(sizeof(*ring) + sz, GFP_KERNEL); + if (!ring) + return NULL; + + kref_init(&ring->ref); + ring->size = sz; + ring->wrap = BITS_PER_TYPE(ring->size) - ilog2(sz); + ring->effective_size = sz; + ring->vaddr = (void *)(ring + 1); + atomic_set(&ring->pin_count, 1); + + intel_ring_update_space(ring); + + return ring; +} + +static void mock_ring_free(struct intel_ring *ring) +{ + kfree(ring); +} + +static int check_ring_direction(struct intel_ring *ring, + u32 next, u32 prev, + int expected) +{ + int result; + + result = intel_ring_direction(ring, next, prev); + if (result < 0) + result = -1; + else if (result > 0) + result = 1; + + if (result != expected) { + pr_err("intel_ring_direction(%u, %u):%d != %d\n", + next, prev, result, expected); + return -EINVAL; + } + + return 0; +} + +static int check_ring_step(struct intel_ring *ring, u32 x, u32 step) +{ + u32 prev = x, next = intel_ring_wrap(ring, x + step); + int err = 0; + + err |= check_ring_direction(ring, next, next, 0); + err |= check_ring_direction(ring, prev, prev, 0); + err |= check_ring_direction(ring, next, prev, 1); + err |= check_ring_direction(ring, prev, next, -1); + + return err; +} + +static int check_ring_offset(struct intel_ring *ring, u32 x, u32 step) +{ + int err = 0; + + err |= check_ring_step(ring, x, step); + err |= check_ring_step(ring, intel_ring_wrap(ring, x + 1), step); + err |= check_ring_step(ring, intel_ring_wrap(ring, x - 1), step); + + return err; +} + +static int igt_ring_direction(void *dummy) +{ + struct intel_ring *ring; + unsigned int half = 2048; + int step, err = 0; + + ring = mock_ring(2 * half); + if (!ring) + return -ENOMEM; + + GEM_BUG_ON(ring->size != 2 * half); + + /* Precision of wrap detection is limited to ring->size / 2 */ + for (step = 1; step < half; step <<= 1) { + err |= check_ring_offset(ring, 0, step); + err |= check_ring_offset(ring, half, step); + } + err |= check_ring_step(ring, 0, half - 64); + + /* And check unwrapped handling for good measure */ + err |= check_ring_offset(ring, 0, 2 * half + 64); + err |= check_ring_offset(ring, 3 * half, 1); + + mock_ring_free(ring); + return err; +} + +int intel_ring_mock_selftests(void) +{ + static const struct i915_subtest tests[] = { + SUBTEST(igt_ring_direction), + }; + + return i915_subtests(tests, NULL); +} diff --git a/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h b/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h index 6a2be7d0dd95..6090ce35226b 100644 --- a/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h +++ b/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h @@ -21,6 +21,7 @@ selftest(fence, i915_sw_fence_mock_selftests) selftest(scatterlist, scatterlist_mock_selftests) selftest(syncmap, i915_syncmap_mock_selftests) selftest(uncore, intel_uncore_mock_selftests) +selftest(ring, intel_ring_mock_selftests) selftest(engine, intel_engine_cs_mock_selftests) selftest(timelines, intel_timeline_mock_selftests) selftest(requests, i915_request_mock_selftests) From 898e4e57adaeeefcfcf23f5d83360e513b69172f Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Wed, 10 Jun 2020 01:06:16 +0300 Subject: [PATCH 08/16] drm/i915/icl: Disable DIP on MST ports with the transcoder clock still on According to BSpec the Data Island Packet should be disabled after disabling the transcoder, but before the transcoder clock select is set to none. On an ICL RVP, daisy-chained MST config not following this leads to a hang with the following MCE when disabling the output: [ 870.948739] mce: [Hardware Error]: CPU 0: Machine Check Exception: 5 Bank 6: ba00000011000402 [ 871.019212] mce: [Hardware Error]: RIP !INEXACT! 10: {poll_idle+0x92/0xb0} [ 871.019212] mce: [Hardware Error]: TSC 135a261fe61 [ 871.019212] mce: [Hardware Error]: PROCESSOR 0:706e5 TIME 1591739604 SOCKET 0 APIC 0 microcode 20 [ 871.019212] mce: [Hardware Error]: Run the above through 'mcelog --ascii' [ 871.019212] mce: [Hardware Error]: Machine check: Processor context corrupt [ 871.019212] Kernel panic - not syncing: Fatal machine check [ 871.019212] Kernel Offset: disabled Bspec: 4287 Fixes: fa37a213275c ("drm/i915: Stop sending DP SDPs on ddi disable") Cc: Gwan-gyeong Mun Cc: Uma Shankar Signed-off-by: Imre Deak Reviewed-by: Uma Shankar Link: https://patchwork.freedesktop.org/patch/msgid/20200609220616.6015-1-imre.deak@intel.com (cherry picked from commit c980216dd224c52b5c70172753c209b653d84958) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/display/intel_ddi.c | 4 +++- drivers/gpu/drm/i915/display/intel_dp_mst.c | 8 ++++++++ 2 files changed, 11 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c index f53de9664903..f92c5ed3f5d2 100644 --- a/drivers/gpu/drm/i915/display/intel_ddi.c +++ b/drivers/gpu/drm/i915/display/intel_ddi.c @@ -3472,7 +3472,9 @@ static void intel_ddi_post_disable_dp(struct intel_atomic_state *state, INTEL_OUTPUT_DP_MST); enum phy phy = intel_port_to_phy(dev_priv, encoder->port); - intel_dp_set_infoframes(encoder, false, old_crtc_state, old_conn_state); + if (!is_mst) + intel_dp_set_infoframes(encoder, false, + old_crtc_state, old_conn_state); /* * Power down sink before disabling the port, otherwise we end diff --git a/drivers/gpu/drm/i915/display/intel_dp_mst.c b/drivers/gpu/drm/i915/display/intel_dp_mst.c index d18b406f2a7d..f29e51ce489c 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/display/intel_dp_mst.c @@ -397,6 +397,14 @@ static void intel_mst_post_disable_dp(struct intel_atomic_state *state, */ drm_dp_send_power_updown_phy(&intel_dp->mst_mgr, connector->port, false); + + /* + * BSpec 4287: disable DIP after the transcoder is disabled and before + * the transcoder clock select is set to none. + */ + if (last_mst_stream) + intel_dp_set_infoframes(&intel_dig_port->base, false, + old_crtc_state, NULL); /* * From TGL spec: "If multi-stream slave transcoder: Configure * Transcoder Clock Select to direct no clock to the transcoder" From ef50fa9bd17d13d0611e39e13b37bbd3e1ea50bf Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 11 Jun 2020 10:30:15 +0100 Subject: [PATCH 09/16] drm/i915/gt: Move hsw GT workarounds from init_clock_gating to workarounds Rescue the GT workarounds from being buried inside init_clock_gating so that we remember to apply them after a GT reset, and that they are included in our verification that the workarounds are applied. v2: Leave HSW_SCRATCH to set an explicit value, not or in our disable bit. Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2011 Signed-off-by: Chris Wilson Cc: Mika Kuoppala Reviewed-by: Mika Kuoppala Cc: stable@vger.kernel.org Link: https://patchwork.freedesktop.org/patch/msgid/20200611093015.11370-1-chris@chris-wilson.co.uk (cherry picked from commit f93ec5fb563779bda4501890b1854526de58e0f1) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_workarounds.c | 48 +++++++++++++++++++++ drivers/gpu/drm/i915/intel_pm.c | 39 +---------------- 2 files changed, 50 insertions(+), 37 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index 90a2b9e399b0..b5d0fbe3b211 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -178,6 +178,12 @@ wa_write_or(struct i915_wa_list *wal, i915_reg_t reg, u32 set) wa_write_masked_or(wal, reg, set, set); } +static void +wa_write_clr(struct i915_wa_list *wal, i915_reg_t reg, u32 clr) +{ + wa_write_masked_or(wal, reg, clr, 0); +} + static void wa_masked_en(struct i915_wa_list *wal, i915_reg_t reg, u32 val) { @@ -686,6 +692,46 @@ int intel_engine_emit_ctx_wa(struct i915_request *rq) return 0; } +static void +hsw_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) +{ + /* L3 caching of data atomics doesn't work -- disable it. */ + wa_write(wal, HSW_SCRATCH1, HSW_SCRATCH1_L3_DATA_ATOMICS_DISABLE); + + wa_add(wal, + HSW_ROW_CHICKEN3, 0, + _MASKED_BIT_ENABLE(HSW_ROW_CHICKEN3_L3_GLOBAL_ATOMICS_DISABLE), + 0 /* XXX does this reg exist? */); + + /* WaVSRefCountFullforceMissDisable:hsw */ + wa_write_clr(wal, GEN7_FF_THREAD_MODE, GEN7_FF_VS_REF_CNT_FFME); + + wa_masked_dis(wal, + CACHE_MODE_0_GEN7, + /* WaDisable_RenderCache_OperationalFlush:hsw */ + RC_OP_FLUSH_ENABLE | + /* enable HiZ Raw Stall Optimization */ + HIZ_RAW_STALL_OPT_DISABLE); + + /* WaDisable4x2SubspanOptimization:hsw */ + wa_masked_en(wal, CACHE_MODE_1, PIXEL_SUBSPAN_COLLECT_OPT_DISABLE); + + /* + * BSpec recommends 8x4 when MSAA is used, + * however in practice 16x4 seems fastest. + * + * Note that PS/WM thread counts depend on the WIZ hashing + * disable bit, which we don't touch here, but it's good + * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). + */ + wa_add(wal, GEN7_GT_MODE, 0, + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4), + GEN6_WIZ_HASHING_16x4); + + /* WaSampleCChickenBitEnable:hsw */ + wa_masked_en(wal, HALF_SLICE_CHICKEN3, HSW_SAMPLE_C_PERFORMANCE); +} + static void gen9_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) { @@ -963,6 +1009,8 @@ gt_init_workarounds(struct drm_i915_private *i915, struct i915_wa_list *wal) bxt_gt_workarounds_init(i915, wal); else if (IS_SKYLAKE(i915)) skl_gt_workarounds_init(i915, wal); + else if (IS_HASWELL(i915)) + hsw_gt_workarounds_init(i915, wal); else if (INTEL_GEN(i915) <= 8) return; else diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 696491d71a1d..0b7a4c5f179d 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -7230,45 +7230,10 @@ static void bdw_init_clock_gating(struct drm_i915_private *dev_priv) static void hsw_init_clock_gating(struct drm_i915_private *dev_priv) { - /* L3 caching of data atomics doesn't work -- disable it. */ - I915_WRITE(HSW_SCRATCH1, HSW_SCRATCH1_L3_DATA_ATOMICS_DISABLE); - I915_WRITE(HSW_ROW_CHICKEN3, - _MASKED_BIT_ENABLE(HSW_ROW_CHICKEN3_L3_GLOBAL_ATOMICS_DISABLE)); - /* This is required by WaCatErrorRejectionIssue:hsw */ I915_WRITE(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG, - I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) | - GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB); - - /* WaVSRefCountFullforceMissDisable:hsw */ - I915_WRITE(GEN7_FF_THREAD_MODE, - I915_READ(GEN7_FF_THREAD_MODE) & ~GEN7_FF_VS_REF_CNT_FFME); - - /* WaDisable_RenderCache_OperationalFlush:hsw */ - I915_WRITE(CACHE_MODE_0_GEN7, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); - - /* enable HiZ Raw Stall Optimization */ - I915_WRITE(CACHE_MODE_0_GEN7, - _MASKED_BIT_DISABLE(HIZ_RAW_STALL_OPT_DISABLE)); - - /* WaDisable4x2SubspanOptimization:hsw */ - I915_WRITE(CACHE_MODE_1, - _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE)); - - /* - * BSpec recommends 8x4 when MSAA is used, - * however in practice 16x4 seems fastest. - * - * Note that PS/WM thread counts depend on the WIZ hashing - * disable bit, which we don't touch here, but it's good - * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). - */ - I915_WRITE(GEN7_GT_MODE, - _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); - - /* WaSampleCChickenBitEnable:hsw */ - I915_WRITE(HALF_SLICE_CHICKEN3, - _MASKED_BIT_ENABLE(HSW_SAMPLE_C_PERFORMANCE)); + I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) | + GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB); /* WaSwitchSolVfFArbitrationPriority:hsw */ I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL); From 7237b190add0794bd95979018a23eda698f2705d Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 11 Jun 2020 09:01:36 +0100 Subject: [PATCH 10/16] drm/i915/gt: Move ivb GT workarounds from init_clock_gating to workarounds Rescue the GT workarounds from being buried inside init_clock_gating so that we remember to apply them after a GT reset, and that they are included in our verification that the workarounds are applied. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Cc: stable@vger.kernel.org Link: https://patchwork.freedesktop.org/patch/msgid/20200611080140.30228-2-chris@chris-wilson.co.uk (cherry picked from commit 19f1f627b33385a2f0855cbc7d33d86d7f4a1e78) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_workarounds.c | 62 +++++++++++++++++++++ drivers/gpu/drm/i915/i915_reg.h | 2 +- drivers/gpu/drm/i915/intel_pm.c | 48 ---------------- 3 files changed, 63 insertions(+), 49 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index b5d0fbe3b211..d9b72679ca82 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -692,6 +692,66 @@ int intel_engine_emit_ctx_wa(struct i915_request *rq) return 0; } +static void +ivb_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) +{ + /* WaDisableEarlyCull:ivb */ + wa_masked_en(wal, _3D_CHICKEN3, _3D_CHICKEN_SF_DISABLE_OBJEND_CULL); + + /* WaDisablePSDDualDispatchEnable:ivb */ + if (IS_IVB_GT1(i915)) + wa_masked_en(wal, + GEN7_HALF_SLICE_CHICKEN1, + GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE); + + /* WaDisable_RenderCache_OperationalFlush:ivb */ + wa_masked_dis(wal, CACHE_MODE_0_GEN7, RC_OP_FLUSH_ENABLE); + + /* Apply the WaDisableRHWOOptimizationForRenderHang:ivb workaround. */ + wa_masked_dis(wal, + GEN7_COMMON_SLICE_CHICKEN1, + GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC); + + /* WaApplyL3ControlAndL3ChickenMode:ivb */ + wa_write(wal, GEN7_L3CNTLREG1, GEN7_WA_FOR_GEN7_L3_CONTROL); + wa_write(wal, GEN7_L3_CHICKEN_MODE_REGISTER, GEN7_WA_L3_CHICKEN_MODE); + + /* WaForceL3Serialization:ivb */ + wa_write_clr(wal, GEN7_L3SQCREG4, L3SQ_URB_READ_CAM_MATCH_DISABLE); + + /* + * WaVSThreadDispatchOverride:ivb,vlv + * + * This actually overrides the dispatch + * mode for all thread types. + */ + wa_write_masked_or(wal, GEN7_FF_THREAD_MODE, + GEN7_FF_SCHED_MASK, + GEN7_FF_TS_SCHED_HW | + GEN7_FF_VS_SCHED_HW | + GEN7_FF_DS_SCHED_HW); + + if (0) { /* causes HiZ corruption on ivb:gt1 */ + /* enable HiZ Raw Stall Optimization */ + wa_masked_dis(wal, CACHE_MODE_0_GEN7, HIZ_RAW_STALL_OPT_DISABLE); + } + + /* WaDisable4x2SubspanOptimization:ivb */ + wa_masked_en(wal, CACHE_MODE_1, PIXEL_SUBSPAN_COLLECT_OPT_DISABLE); + + /* + * BSpec recommends 8x4 when MSAA is used, + * however in practice 16x4 seems fastest. + * + * Note that PS/WM thread counts depend on the WIZ hashing + * disable bit, which we don't touch here, but it's good + * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). + */ + wa_add(wal, GEN7_GT_MODE, 0, + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4), + GEN6_WIZ_HASHING_16x4); +} + static void hsw_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) { @@ -1011,6 +1071,8 @@ gt_init_workarounds(struct drm_i915_private *i915, struct i915_wa_list *wal) skl_gt_workarounds_init(i915, wal); else if (IS_HASWELL(i915)) hsw_gt_workarounds_init(i915, wal); + else if (IS_IVYBRIDGE(i915)) + ivb_gt_workarounds_init(i915, wal); else if (INTEL_GEN(i915) <= 8) return; else diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 7717581350bd..06cd1d28a176 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -7896,7 +7896,7 @@ enum { /* GEN7 chicken */ #define GEN7_COMMON_SLICE_CHICKEN1 _MMIO(0x7010) - #define GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC ((1 << 10) | (1 << 26)) + #define GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC (1 << 10) #define GEN9_RHWO_OPTIMIZATION_DISABLE (1 << 14) #define COMMON_SLICE_CHICKEN2 _MMIO(0x7014) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 0b7a4c5f179d..7236644af40b 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -7247,32 +7247,11 @@ static void ivb_init_clock_gating(struct drm_i915_private *dev_priv) I915_WRITE(ILK_DSPCLK_GATE_D, ILK_VRHUNIT_CLOCK_GATE_DISABLE); - /* WaDisableEarlyCull:ivb */ - I915_WRITE(_3D_CHICKEN3, - _MASKED_BIT_ENABLE(_3D_CHICKEN_SF_DISABLE_OBJEND_CULL)); - /* WaDisableBackToBackFlipFix:ivb */ I915_WRITE(IVB_CHICKEN3, CHICKEN3_DGMG_REQ_OUT_FIX_DISABLE | CHICKEN3_DGMG_DONE_FIX_DISABLE); - /* WaDisablePSDDualDispatchEnable:ivb */ - if (IS_IVB_GT1(dev_priv)) - I915_WRITE(GEN7_HALF_SLICE_CHICKEN1, - _MASKED_BIT_ENABLE(GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE)); - - /* WaDisable_RenderCache_OperationalFlush:ivb */ - I915_WRITE(CACHE_MODE_0_GEN7, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); - - /* Apply the WaDisableRHWOOptimizationForRenderHang:ivb workaround. */ - I915_WRITE(GEN7_COMMON_SLICE_CHICKEN1, - GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC); - - /* WaApplyL3ControlAndL3ChickenMode:ivb */ - I915_WRITE(GEN7_L3CNTLREG1, - GEN7_WA_FOR_GEN7_L3_CONTROL); - I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER, - GEN7_WA_L3_CHICKEN_MODE); if (IS_IVB_GT1(dev_priv)) I915_WRITE(GEN7_ROW_CHICKEN2, _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE)); @@ -7284,10 +7263,6 @@ static void ivb_init_clock_gating(struct drm_i915_private *dev_priv) _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE)); } - /* WaForceL3Serialization:ivb */ - I915_WRITE(GEN7_L3SQCREG4, I915_READ(GEN7_L3SQCREG4) & - ~L3SQ_URB_READ_CAM_MATCH_DISABLE); - /* * According to the spec, bit 13 (RCZUNIT) must be set on IVB. * This implements the WaDisableRCZUnitClockGating:ivb workaround. @@ -7302,29 +7277,6 @@ static void ivb_init_clock_gating(struct drm_i915_private *dev_priv) g4x_disable_trickle_feed(dev_priv); - gen7_setup_fixed_func_scheduler(dev_priv); - - if (0) { /* causes HiZ corruption on ivb:gt1 */ - /* enable HiZ Raw Stall Optimization */ - I915_WRITE(CACHE_MODE_0_GEN7, - _MASKED_BIT_DISABLE(HIZ_RAW_STALL_OPT_DISABLE)); - } - - /* WaDisable4x2SubspanOptimization:ivb */ - I915_WRITE(CACHE_MODE_1, - _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE)); - - /* - * BSpec recommends 8x4 when MSAA is used, - * however in practice 16x4 seems fastest. - * - * Note that PS/WM thread counts depend on the WIZ hashing - * disable bit, which we don't touch here, but it's good - * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). - */ - I915_WRITE(GEN7_GT_MODE, - _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); - snpcr = I915_READ(GEN6_MBCUNIT_SNPCR); snpcr &= ~GEN6_MBC_SNPCR_MASK; snpcr |= GEN6_MBC_SNPCR_MED; From 695a2b11649e99bbf15d278042247042c42b8728 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 11 Jun 2020 09:01:37 +0100 Subject: [PATCH 11/16] drm/i915/gt: Move vlv GT workarounds from init_clock_gating to workarounds Rescue the GT workarounds from being buried inside init_clock_gating so that we remember to apply them after a GT reset, and that they are included in our verification that the workarounds are applied. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Cc: stable@vger.kernel.org Link: https://patchwork.freedesktop.org/patch/msgid/20200611080140.30228-3-chris@chris-wilson.co.uk (cherry picked from commit 7331c356b6d2d8a01422cacab27478a1dba9fa2a) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_workarounds.c | 59 ++++++++++++++++++++ drivers/gpu/drm/i915/intel_pm.c | 61 --------------------- 2 files changed, 59 insertions(+), 61 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index d9b72679ca82..9923ff1a3982 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -752,6 +752,63 @@ ivb_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) GEN6_WIZ_HASHING_16x4); } +static void +vlv_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) +{ + /* WaDisableEarlyCull:vlv */ + wa_masked_en(wal, _3D_CHICKEN3, _3D_CHICKEN_SF_DISABLE_OBJEND_CULL); + + /* WaPsdDispatchEnable:vlv */ + /* WaDisablePSDDualDispatchEnable:vlv */ + wa_masked_en(wal, + GEN7_HALF_SLICE_CHICKEN1, + GEN7_MAX_PS_THREAD_DEP | + GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE); + + /* WaDisable_RenderCache_OperationalFlush:vlv */ + wa_masked_dis(wal, CACHE_MODE_0_GEN7, RC_OP_FLUSH_ENABLE); + + /* WaForceL3Serialization:vlv */ + wa_write_clr(wal, GEN7_L3SQCREG4, L3SQ_URB_READ_CAM_MATCH_DISABLE); + + /* + * WaVSThreadDispatchOverride:ivb,vlv + * + * This actually overrides the dispatch + * mode for all thread types. + */ + wa_write_masked_or(wal, + GEN7_FF_THREAD_MODE, + GEN7_FF_SCHED_MASK, + GEN7_FF_TS_SCHED_HW | + GEN7_FF_VS_SCHED_HW | + GEN7_FF_DS_SCHED_HW); + + /* + * BSpec says this must be set, even though + * WaDisable4x2SubspanOptimization isn't listed for VLV. + */ + wa_masked_en(wal, CACHE_MODE_1, PIXEL_SUBSPAN_COLLECT_OPT_DISABLE); + + /* + * BSpec recommends 8x4 when MSAA is used, + * however in practice 16x4 seems fastest. + * + * Note that PS/WM thread counts depend on the WIZ hashing + * disable bit, which we don't touch here, but it's good + * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). + */ + wa_add(wal, GEN7_GT_MODE, 0, + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4), + GEN6_WIZ_HASHING_16x4); + + /* + * WaIncreaseL3CreditsForVLVB0:vlv + * This is the hardware default actually. + */ + wa_write(wal, GEN7_L3SQCREG1, VLV_B0_WA_L3SQCREG1_VALUE); +} + static void hsw_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) { @@ -1071,6 +1128,8 @@ gt_init_workarounds(struct drm_i915_private *i915, struct i915_wa_list *wal) skl_gt_workarounds_init(i915, wal); else if (IS_HASWELL(i915)) hsw_gt_workarounds_init(i915, wal); + else if (IS_VALLEYVIEW(i915)) + vlv_gt_workarounds_init(i915, wal); else if (IS_IVYBRIDGE(i915)) ivb_gt_workarounds_init(i915, wal); else if (INTEL_GEN(i915) <= 8) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 7236644af40b..cea7923c9cd6 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -6986,24 +6986,6 @@ static void gen6_init_clock_gating(struct drm_i915_private *dev_priv) gen6_check_mch_setup(dev_priv); } -static void gen7_setup_fixed_func_scheduler(struct drm_i915_private *dev_priv) -{ - u32 reg = I915_READ(GEN7_FF_THREAD_MODE); - - /* - * WaVSThreadDispatchOverride:ivb,vlv - * - * This actually overrides the dispatch - * mode for all thread types. - */ - reg &= ~GEN7_FF_SCHED_MASK; - reg |= GEN7_FF_TS_SCHED_HW; - reg |= GEN7_FF_VS_SCHED_HW; - reg |= GEN7_FF_DS_SCHED_HW; - - I915_WRITE(GEN7_FF_THREAD_MODE, reg); -} - static void lpt_init_clock_gating(struct drm_i915_private *dev_priv) { /* @@ -7290,28 +7272,11 @@ static void ivb_init_clock_gating(struct drm_i915_private *dev_priv) static void vlv_init_clock_gating(struct drm_i915_private *dev_priv) { - /* WaDisableEarlyCull:vlv */ - I915_WRITE(_3D_CHICKEN3, - _MASKED_BIT_ENABLE(_3D_CHICKEN_SF_DISABLE_OBJEND_CULL)); - /* WaDisableBackToBackFlipFix:vlv */ I915_WRITE(IVB_CHICKEN3, CHICKEN3_DGMG_REQ_OUT_FIX_DISABLE | CHICKEN3_DGMG_DONE_FIX_DISABLE); - /* WaPsdDispatchEnable:vlv */ - /* WaDisablePSDDualDispatchEnable:vlv */ - I915_WRITE(GEN7_HALF_SLICE_CHICKEN1, - _MASKED_BIT_ENABLE(GEN7_MAX_PS_THREAD_DEP | - GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE)); - - /* WaDisable_RenderCache_OperationalFlush:vlv */ - I915_WRITE(CACHE_MODE_0_GEN7, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); - - /* WaForceL3Serialization:vlv */ - I915_WRITE(GEN7_L3SQCREG4, I915_READ(GEN7_L3SQCREG4) & - ~L3SQ_URB_READ_CAM_MATCH_DISABLE); - /* WaDisableDopClockGating:vlv */ I915_WRITE(GEN7_ROW_CHICKEN2, _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE)); @@ -7321,8 +7286,6 @@ static void vlv_init_clock_gating(struct drm_i915_private *dev_priv) I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) | GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB); - gen7_setup_fixed_func_scheduler(dev_priv); - /* * According to the spec, bit 13 (RCZUNIT) must be set on IVB. * This implements the WaDisableRCZUnitClockGating:vlv workaround. @@ -7336,30 +7299,6 @@ static void vlv_init_clock_gating(struct drm_i915_private *dev_priv) I915_WRITE(GEN7_UCGCTL4, I915_READ(GEN7_UCGCTL4) | GEN7_L3BANK2X_CLOCK_GATE_DISABLE); - /* - * BSpec says this must be set, even though - * WaDisable4x2SubspanOptimization isn't listed for VLV. - */ - I915_WRITE(CACHE_MODE_1, - _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE)); - - /* - * BSpec recommends 8x4 when MSAA is used, - * however in practice 16x4 seems fastest. - * - * Note that PS/WM thread counts depend on the WIZ hashing - * disable bit, which we don't touch here, but it's good - * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). - */ - I915_WRITE(GEN7_GT_MODE, - _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); - - /* - * WaIncreaseL3CreditsForVLVB0:vlv - * This is the hardware default actually. - */ - I915_WRITE(GEN7_L3SQCREG1, VLV_B0_WA_L3SQCREG1_VALUE); - /* * WaDisableVLVClockGating_VBIIssue:vlv * Disable clock gating on th GCFG unit to prevent a delay From fd2599bda5a989c3332f4956fd7760ec32bd51ee Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 11 Jun 2020 09:01:38 +0100 Subject: [PATCH 12/16] drm/i915/gt: Move snb GT workarounds from init_clock_gating to workarounds Rescue the GT workarounds from being buried inside init_clock_gating so that we remember to apply them after a GT reset, and that they are included in our verification that the workarounds are applied. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Cc: stable@vger.kernel.org Link: https://patchwork.freedesktop.org/patch/msgid/20200611080140.30228-4-chris@chris-wilson.co.uk (cherry picked from commit c3b93a943f2c9ee4a106db100a2fc3b2f126bfc5) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_workarounds.c | 41 +++++++++++++++++++++ drivers/gpu/drm/i915/intel_pm.c | 33 ----------------- 2 files changed, 41 insertions(+), 33 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index 9923ff1a3982..c2d57f65b147 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -692,6 +692,45 @@ int intel_engine_emit_ctx_wa(struct i915_request *rq) return 0; } +static void +snb_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) +{ + /* WaDisableHiZPlanesWhenMSAAEnabled:snb */ + wa_masked_en(wal, + _3D_CHICKEN, + _3D_CHICKEN_HIZ_PLANE_DISABLE_MSAA_4X_SNB); + + /* WaDisable_RenderCache_OperationalFlush:snb */ + wa_masked_dis(wal, CACHE_MODE_0, RC_OP_FLUSH_ENABLE); + + /* + * BSpec recommends 8x4 when MSAA is used, + * however in practice 16x4 seems fastest. + * + * Note that PS/WM thread counts depend on the WIZ hashing + * disable bit, which we don't touch here, but it's good + * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). + */ + wa_add(wal, + GEN6_GT_MODE, 0, + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4), + GEN6_WIZ_HASHING_16x4); + + wa_masked_dis(wal, CACHE_MODE_0, CM0_STC_EVICT_DISABLE_LRA_SNB); + + wa_masked_en(wal, + _3D_CHICKEN3, + /* WaStripsFansDisableFastClipPerformanceFix:snb */ + _3D_CHICKEN3_SF_DISABLE_FASTCLIP_CULL | + /* + * Bspec says: + * "This bit must be set if 3DSTATE_CLIP clip mode is set + * to normal and 3DSTATE_SF number of SF output attributes + * is more than 16." + */ + _3D_CHICKEN3_SF_DISABLE_PIPELINED_ATTR_FETCH); +} + static void ivb_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) { @@ -1132,6 +1171,8 @@ gt_init_workarounds(struct drm_i915_private *i915, struct i915_wa_list *wal) vlv_gt_workarounds_init(i915, wal); else if (IS_IVYBRIDGE(i915)) ivb_gt_workarounds_init(i915, wal); + else if (IS_GEN(i915, 6)) + snb_gt_workarounds_init(i915, wal); else if (INTEL_GEN(i915) <= 8) return; else diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index cea7923c9cd6..5db0ebe5eee0 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -6902,27 +6902,6 @@ static void gen6_init_clock_gating(struct drm_i915_private *dev_priv) I915_READ(ILK_DISPLAY_CHICKEN2) | ILK_ELPIN_409_SELECT); - /* WaDisableHiZPlanesWhenMSAAEnabled:snb */ - I915_WRITE(_3D_CHICKEN, - _MASKED_BIT_ENABLE(_3D_CHICKEN_HIZ_PLANE_DISABLE_MSAA_4X_SNB)); - - /* WaDisable_RenderCache_OperationalFlush:snb */ - I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); - - /* - * BSpec recoomends 8x4 when MSAA is used, - * however in practice 16x4 seems fastest. - * - * Note that PS/WM thread counts depend on the WIZ hashing - * disable bit, which we don't touch here, but it's good - * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). - */ - I915_WRITE(GEN6_GT_MODE, - _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); - - I915_WRITE(CACHE_MODE_0, - _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB)); - I915_WRITE(GEN6_UCGCTL1, I915_READ(GEN6_UCGCTL1) | GEN6_BLBUNIT_CLOCK_GATE_DISABLE | @@ -6945,18 +6924,6 @@ static void gen6_init_clock_gating(struct drm_i915_private *dev_priv) GEN6_RCPBUNIT_CLOCK_GATE_DISABLE | GEN6_RCCUNIT_CLOCK_GATE_DISABLE); - /* WaStripsFansDisableFastClipPerformanceFix:snb */ - I915_WRITE(_3D_CHICKEN3, - _MASKED_BIT_ENABLE(_3D_CHICKEN3_SF_DISABLE_FASTCLIP_CULL)); - - /* - * Bspec says: - * "This bit must be set if 3DSTATE_CLIP clip mode is set to normal and - * 3DSTATE_SF number of SF output attributes is more than 16." - */ - I915_WRITE(_3D_CHICKEN3, - _MASKED_BIT_ENABLE(_3D_CHICKEN3_SF_DISABLE_PIPELINED_ATTR_FETCH)); - /* * According to the spec the following bits should be * set in order to enable memory self-refresh and fbc: From eacf21040aa97fd1b3c6bb201bfd43820e1c49be Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 11 Jun 2020 09:01:39 +0100 Subject: [PATCH 13/16] drm/i915/gt: Move ilk GT workarounds from init_clock_gating to workarounds Rescue the GT workarounds from being buried inside init_clock_gating so that we remember to apply them after a GT reset, and that they are included in our verification that the workarounds are applied. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Cc: stable@vger.kernel.org Link: https://patchwork.freedesktop.org/patch/msgid/20200611080140.30228-5-chris@chris-wilson.co.uk (cherry picked from commit 806a45c0838d253e306a6384057e851b65d11099) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_workarounds.c | 14 ++++++++++++++ drivers/gpu/drm/i915/intel_pm.c | 10 ---------- 2 files changed, 14 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index c2d57f65b147..5ccfe36c2978 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -692,6 +692,18 @@ int intel_engine_emit_ctx_wa(struct i915_request *rq) return 0; } +static void +ilk_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) +{ + wa_masked_en(wal, _3D_CHICKEN2, _3D_CHICKEN2_WM_READ_PIPELINED); + + /* WaDisableRenderCachePipelinedFlush:ilk */ + wa_masked_en(wal, CACHE_MODE_0, CM0_PIPELINED_RENDER_FLUSH_DISABLE); + + /* WaDisable_RenderCache_OperationalFlush:ilk */ + wa_masked_dis(wal, CACHE_MODE_0, RC_OP_FLUSH_ENABLE); +} + static void snb_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) { @@ -1173,6 +1185,8 @@ gt_init_workarounds(struct drm_i915_private *i915, struct i915_wa_list *wal) ivb_gt_workarounds_init(i915, wal); else if (IS_GEN(i915, 6)) snb_gt_workarounds_init(i915, wal); + else if (IS_GEN(i915, 5)) + ilk_gt_workarounds_init(i915, wal); else if (INTEL_GEN(i915) <= 8) return; else diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 5db0ebe5eee0..7ebe4fa3a162 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -6830,16 +6830,6 @@ static void ilk_init_clock_gating(struct drm_i915_private *dev_priv) I915_WRITE(ILK_DISPLAY_CHICKEN2, I915_READ(ILK_DISPLAY_CHICKEN2) | ILK_ELPIN_409_SELECT); - I915_WRITE(_3D_CHICKEN2, - _3D_CHICKEN2_WM_READ_PIPELINED << 16 | - _3D_CHICKEN2_WM_READ_PIPELINED); - - /* WaDisableRenderCachePipelinedFlush:ilk */ - I915_WRITE(CACHE_MODE_0, - _MASKED_BIT_ENABLE(CM0_PIPELINED_RENDER_FLUSH_DISABLE)); - - /* WaDisable_RenderCache_OperationalFlush:ilk */ - I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); g4x_disable_trickle_feed(dev_priv); From 27582a9c917940bc71c0df0b8e022cbde8d735d2 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 11 Jun 2020 09:01:40 +0100 Subject: [PATCH 14/16] drm/i915/gt: Move gen4 GT workarounds from init_clock_gating to workarounds Rescue the GT workarounds from being buried inside init_clock_gating so that we remember to apply them after a GT reset, and that they are included in our verification that the workarounds are applied. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Cc: stable@vger.kernel.org Link: https://patchwork.freedesktop.org/patch/msgid/20200611080140.30228-6-chris@chris-wilson.co.uk (cherry picked from commit 2bcefd0d263ab4a72f0d61921ae6b0dc81606551) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/gt/intel_workarounds.c | 29 ++++++++++++++++----- drivers/gpu/drm/i915/intel_pm.c | 15 ----------- 2 files changed, 23 insertions(+), 21 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index 5ccfe36c2978..85d2bef51524 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -692,16 +692,29 @@ int intel_engine_emit_ctx_wa(struct i915_request *rq) return 0; } +static void +gen4_gt_workarounds_init(struct drm_i915_private *i915, + struct i915_wa_list *wal) +{ + /* WaDisable_RenderCache_OperationalFlush:gen4,ilk */ + wa_masked_dis(wal, CACHE_MODE_0, RC_OP_FLUSH_ENABLE); +} + +static void +g4x_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) +{ + gen4_gt_workarounds_init(i915, wal); + + /* WaDisableRenderCachePipelinedFlush:g4x,ilk */ + wa_masked_en(wal, CACHE_MODE_0, CM0_PIPELINED_RENDER_FLUSH_DISABLE); +} + static void ilk_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) { + g4x_gt_workarounds_init(i915, wal); + wa_masked_en(wal, _3D_CHICKEN2, _3D_CHICKEN2_WM_READ_PIPELINED); - - /* WaDisableRenderCachePipelinedFlush:ilk */ - wa_masked_en(wal, CACHE_MODE_0, CM0_PIPELINED_RENDER_FLUSH_DISABLE); - - /* WaDisable_RenderCache_OperationalFlush:ilk */ - wa_masked_dis(wal, CACHE_MODE_0, RC_OP_FLUSH_ENABLE); } static void @@ -1187,6 +1200,10 @@ gt_init_workarounds(struct drm_i915_private *i915, struct i915_wa_list *wal) snb_gt_workarounds_init(i915, wal); else if (IS_GEN(i915, 5)) ilk_gt_workarounds_init(i915, wal); + else if (IS_G4X(i915)) + g4x_gt_workarounds_init(i915, wal); + else if (IS_GEN(i915, 4)) + gen4_gt_workarounds_init(i915, wal); else if (INTEL_GEN(i915) <= 8) return; else diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 7ebe4fa3a162..07f663cd2d1c 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -7308,13 +7308,6 @@ static void g4x_init_clock_gating(struct drm_i915_private *dev_priv) dspclk_gate |= DSSUNIT_CLOCK_GATE_DISABLE; I915_WRITE(DSPCLK_GATE_D, dspclk_gate); - /* WaDisableRenderCachePipelinedFlush */ - I915_WRITE(CACHE_MODE_0, - _MASKED_BIT_ENABLE(CM0_PIPELINED_RENDER_FLUSH_DISABLE)); - - /* WaDisable_RenderCache_OperationalFlush:g4x */ - I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); - g4x_disable_trickle_feed(dev_priv); } @@ -7330,11 +7323,6 @@ static void i965gm_init_clock_gating(struct drm_i915_private *dev_priv) intel_uncore_write(uncore, MI_ARB_STATE, _MASKED_BIT_ENABLE(MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE)); - - /* WaDisable_RenderCache_OperationalFlush:gen4 */ - intel_uncore_write(uncore, - CACHE_MODE_0, - _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); } static void i965g_init_clock_gating(struct drm_i915_private *dev_priv) @@ -7347,9 +7335,6 @@ static void i965g_init_clock_gating(struct drm_i915_private *dev_priv) I915_WRITE(RENCLK_GATE_D2, 0); I915_WRITE(MI_ARB_STATE, _MASKED_BIT_ENABLE(MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE)); - - /* WaDisable_RenderCache_OperationalFlush:gen4 */ - I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); } static void gen3_init_clock_gating(struct drm_i915_private *dev_priv) From a3005c2edf7e8c3478880db1ca84028a2b6819bb Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Fri, 12 Jun 2020 15:17:31 +0300 Subject: [PATCH 15/16] drm/i915/icl+: Fix hotplug interrupt disabling after storm detection MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Atm, hotplug interrupts on TypeC ports are left enabled after detecting an interrupt storm, fix this. Reported-by: Kunal Joshi References: https://gitlab.freedesktop.org/drm/intel/-/issues/351 Bugzilla: https://gitlab.freedesktop.org/drm/intel/-/issues/1964 Cc: Kunal Joshi Cc: stable@vger.kernel.org Signed-off-by: Imre Deak Reviewed-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20200612121731.19596-1-imre.deak@intel.com (cherry picked from commit 587a87b9d7e94927edcdea018565bc1939381eb1) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/i915_irq.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 4dc601dffc08..284cf078135a 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -3125,6 +3125,7 @@ static void gen11_hpd_irq_setup(struct drm_i915_private *dev_priv) val = I915_READ(GEN11_DE_HPD_IMR); val &= ~hotplug_irqs; + val |= ~enabled_irqs & hotplug_irqs; I915_WRITE(GEN11_DE_HPD_IMR, val); POSTING_READ(GEN11_DE_HPD_IMR); From 8e68c6340d5833077b3753eabedab40755571383 Mon Sep 17 00:00:00 2001 From: Vandita Kulkarni Date: Fri, 12 Jun 2020 13:52:37 +0530 Subject: [PATCH 16/16] drm/i915/display: Fix the encoder type check For all ddi, encoder->type holds output type as ddi, assigning it to individual o/p types is no more valid. Fixes: 362bfb995b78 ("drm/i915/tgl: Add DKL PHY vswing table for HDMI") v2: Rebase, no functional change. Signed-off-by: Vandita Kulkarni Reviewed-by: Uma Shankar Signed-off-by: Uma Shankar Link: https://patchwork.freedesktop.org/patch/msgid/20200612082237.11886-1-vandita.kulkarni@intel.com (cherry picked from commit 94641eb6c69682884abbecf22fe5b7c185af6a06) Signed-off-by: Joonas Lahtinen --- drivers/gpu/drm/i915/display/intel_ddi.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c index f92c5ed3f5d2..0575a1eea2a1 100644 --- a/drivers/gpu/drm/i915/display/intel_ddi.c +++ b/drivers/gpu/drm/i915/display/intel_ddi.c @@ -2579,14 +2579,14 @@ static void icl_ddi_vswing_sequence(struct intel_encoder *encoder, static void tgl_dkl_phy_ddi_vswing_sequence(struct intel_encoder *encoder, int link_clock, - u32 level) + u32 level, enum intel_output_type type) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); enum tc_port tc_port = intel_port_to_tc(dev_priv, encoder->port); const struct tgl_dkl_phy_ddi_buf_trans *ddi_translations; u32 n_entries, val, ln, dpcnt_mask, dpcnt_val; - if (encoder->type == INTEL_OUTPUT_HDMI) { + if (type == INTEL_OUTPUT_HDMI) { n_entries = ARRAY_SIZE(tgl_dkl_phy_hdmi_ddi_trans); ddi_translations = tgl_dkl_phy_hdmi_ddi_trans; } else { @@ -2638,7 +2638,7 @@ static void tgl_ddi_vswing_sequence(struct intel_encoder *encoder, if (intel_phy_is_combo(dev_priv, phy)) icl_combo_phy_ddi_vswing_sequence(encoder, level, type); else - tgl_dkl_phy_ddi_vswing_sequence(encoder, link_clock, level); + tgl_dkl_phy_ddi_vswing_sequence(encoder, link_clock, level, type); } static u32 translate_signal_level(struct intel_dp *intel_dp, int signal_levels)