summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915/i915_request.c
AgeCommit message (Collapse)Author
2019-10-09drm/i915: Only enqueue already completed requestsChris Wilson
If we are asked to submit a completed request, just move it onto the active-list without modifying it's payload. If we try to emit the modified payload of a completed request, we risk racing with the ring->head update during retirement which may advance the head past our breadcrumb and so we generate a warning for the emission being behind the RING_HEAD. v2: Commentary for the sneaky, shared responsibility between functions. v3: Spelling mistakes and bonus assertion Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190923110056.15176-3-chris@chris-wilson.co.uk (cherry picked from commit c0bb487dc19fc45dbeede7dcf8f513df51a3cd33) Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2019-10-07drm/i915: Verify the engine after acquiring the active.lockChris Wilson
When using virtual engines, the rq->engine is not stable until we hold the engine->active.lock (as the virtual engine may be exchanged with the sibling). Since commit 22b7a426bbe1 ("drm/i915/execlists: Preempt-to-busy") we may retire a request concurrently with resubmitting it to HW, we need to be extra careful to verify we are holding the correct lock for the request's active list. This is similar to the issue we saw with rescheduling the virtual requests, see sched_lock_engine(). Or else: <4> [876.736126] list_add corruption. prev->next should be next (ffff8883f931a1f8), but was dead000000000100. (prev=ffff888361ffa610). <4> [876.736136] WARNING: CPU: 2 PID: 21 at lib/list_debug.c:28 __list_add_valid+0x4d/0x70 <4> [876.736137] Modules linked in: i915(+) amdgpu gpu_sched ttm vgem snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_codec_generic mei_hdcp x86_pkg_temp_thermal coretemp crct10dif_pclmul crc32_pclmul snd_intel_nhlt snd_hda_codec snd_hwdep snd_hda_core ghash_clmulni_intel e1000e cdc_ether usbnet mii snd_pcm ptp pps_core mei_me mei prime_numbers btusb btrtl btbcm btintel bluetooth ecdh_generic ecc [last unloaded: i915] <4> [876.736154] CPU: 2 PID: 21 Comm: ksoftirqd/2 Tainted: G U 5.3.0-CI-CI_DRM_6898+ #1 <4> [876.736156] Hardware name: Intel Corporation Ice Lake Client Platform/IceLake U DDR4 SODIMM PD RVP TLC, BIOS ICLSFWR1.R00.3183.A00.1905020411 05/02/2019 <4> [876.736157] RIP: 0010:__list_add_valid+0x4d/0x70 <4> [876.736159] Code: c3 48 89 d1 48 c7 c7 20 33 0e 82 48 89 c2 e8 4a 4a bc ff 0f 0b 31 c0 c3 48 89 c1 4c 89 c6 48 c7 c7 70 33 0e 82 e8 33 4a bc ff <0f> 0b 31 c0 c3 48 89 f2 4c 89 c1 48 89 fe 48 c7 c7 c0 33 0e 82 e8 <4> [876.736160] RSP: 0018:ffffc9000018bd30 EFLAGS: 00010082 <4> [876.736162] RAX: 0000000000000000 RBX: ffff888361ffc840 RCX: 0000000000000104 <4> [876.736163] RDX: 0000000080000104 RSI: 0000000000000000 RDI: 00000000ffffffff <4> [876.736164] RBP: ffffc9000018bd68 R08: 0000000000000000 R09: 0000000000000001 <4> [876.736165] R10: 00000000aed95de3 R11: 000000007fe927eb R12: ffff888361ffca10 <4> [876.736166] R13: ffff888361ffa610 R14: ffff888361ffc880 R15: ffff8883f931a1f8 <4> [876.736168] FS: 0000000000000000(0000) GS:ffff88849fd00000(0000) knlGS:0000000000000000 <4> [876.736169] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 <4> [876.736170] CR2: 00007f093a9173c0 CR3: 00000003bba08005 CR4: 0000000000760ee0 <4> [876.736171] PKRU: 55555554 <4> [876.736172] Call Trace: <4> [876.736226] __i915_request_submit+0x152/0x370 [i915] <4> [876.736263] __execlists_submission_tasklet+0x6da/0x1f50 [i915] <4> [876.736293] ? execlists_submission_tasklet+0x29/0x50 [i915] <4> [876.736321] execlists_submission_tasklet+0x34/0x50 [i915] <4> [876.736325] tasklet_action_common.isra.5+0x47/0xb0 <4> [876.736328] __do_softirq+0xd8/0x4ae <4> [876.736332] ? smpboot_thread_fn+0x23/0x280 <4> [876.736334] ? smpboot_thread_fn+0x6b/0x280 <4> [876.736336] run_ksoftirqd+0x2b/0x50 <4> [876.736338] smpboot_thread_fn+0x1d3/0x280 <4> [876.736341] ? sort_range+0x20/0x20 <4> [876.736343] kthread+0x119/0x130 <4> [876.736345] ? kthread_park+0xa0/0xa0 <4> [876.736347] ret_from_fork+0x24/0x50 <4> [876.736353] irq event stamp: 2290145 <4> [876.736356] hardirqs last enabled at (2290144): [<ffffffff8123cde8>] __slab_free+0x3e8/0x500 <4> [876.736358] hardirqs last disabled at (2290145): [<ffffffff819cfb4d>] _raw_spin_lock_irqsave+0xd/0x50 <4> [876.736360] softirqs last enabled at (2290114): [<ffffffff81c0033e>] __do_softirq+0x33e/0x4ae <4> [876.736361] softirqs last disabled at (2290119): [<ffffffff810b815b>] run_ksoftirqd+0x2b/0x50 <4> [876.736363] WARNING: CPU: 2 PID: 21 at lib/list_debug.c:28 __list_add_valid+0x4d/0x70 <4> [876.736364] ---[ end trace 3e58d6c7356c65bf ]--- <4> [876.736406] ------------[ cut here ]------------ <4> [876.736415] list_del corruption. prev->next should be ffff888361ffca10, but was ffff88840ac2c730 <4> [876.736421] WARNING: CPU: 2 PID: 5490 at lib/list_debug.c:53 __list_del_entry_valid+0x79/0x90 <4> [876.736422] Modules linked in: i915(+) amdgpu gpu_sched ttm vgem snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_codec_generic mei_hdcp x86_pkg_temp_thermal coretemp crct10dif_pclmul crc32_pclmul snd_intel_nhlt snd_hda_codec snd_hwdep snd_hda_core ghash_clmulni_intel e1000e cdc_ether usbnet mii snd_pcm ptp pps_core mei_me mei prime_numbers btusb btrtl btbcm btintel bluetooth ecdh_generic ecc [last unloaded: i915] <4> [876.736433] CPU: 2 PID: 5490 Comm: i915_selftest Tainted: G U W 5.3.0-CI-CI_DRM_6898+ #1 <4> [876.736435] Hardware name: Intel Corporation Ice Lake Client Platform/IceLake U DDR4 SODIMM PD RVP TLC, BIOS ICLSFWR1.R00.3183.A00.1905020411 05/02/2019 <4> [876.736436] RIP: 0010:__list_del_entry_valid+0x79/0x90 <4> [876.736438] Code: 0b 31 c0 c3 48 89 fe 48 c7 c7 30 34 0e 82 e8 ae 49 bc ff 0f 0b 31 c0 c3 48 89 f2 48 89 fe 48 c7 c7 68 34 0e 82 e8 97 49 bc ff <0f> 0b 31 c0 c3 48 c7 c7 a8 34 0e 82 e8 86 49 bc ff 0f 0b 31 c0 c3 <4> [876.736439] RSP: 0018:ffffc900003ef758 EFLAGS: 00010086 <4> [876.736440] RAX: 0000000000000000 RBX: ffff888361ffc840 RCX: 0000000000000002 <4> [876.736442] RDX: 0000000080000002 RSI: 0000000000000000 RDI: 00000000ffffffff <4> [876.736443] RBP: ffffc900003ef780 R08: 0000000000000000 R09: 0000000000000001 <4> [876.736444] R10: 000000001418e4b7 R11: 000000007f0ea93b R12: ffff888361ffcab8 <4> [876.736445] R13: ffff88843b6d0000 R14: 000000000000217c R15: 0000000000000001 <4> [876.736447] FS: 00007f4e6f255240(0000) GS:ffff88849fd00000(0000) knlGS:0000000000000000 <4> [876.736448] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 <4> [876.736449] CR2: 00007f093a9173c0 CR3: 00000003bba08005 CR4: 0000000000760ee0 <4> [876.736450] PKRU: 55555554 <4> [876.736451] Call Trace: <4> [876.736488] i915_request_retire+0x224/0x8e0 [i915] <4> [876.736521] i915_request_create+0x4b/0x1b0 [i915] <4> [876.736550] nop_virtual_engine+0x230/0x4d0 [i915] Fixes: 22b7a426bbe1 ("drm/i915/execlists: Preempt-to-busy") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111695 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Matthew Auld <matthew.auld@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190918145453.8800-1-chris@chris-wilson.co.uk (cherry picked from commit 37fa0de3c137d5f54f7e64f53495c9d501d42a4d) Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2019-09-06drm/i915: Hold irq-off for the entire fake lock periodChris Wilson
Sadly lockdep records when the irqs are re-enabled and then marks up the fake lock as being irq-unsafe. Our hand is forced and so we must mark up the entire fake lock critical section as irq-off. Hopefully this is the last tweak required! v2: Not quite, we need to mark the timeline spinlock as irqsafe. That was a genuine bug being hidden by the earlier lockdep splat. Fixes: d67739268cf0 ("drm/i915/gt: Mark up the nested engine-pm timeline lock as irqsafe") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190823132700.25286-2-chris@chris-wilson.co.uk (cherry picked from commit 6dcb85a0ad990455ae7c596e3fc966ad9c1ba9c5) Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2019-08-22Merge drm/drm-next into drm-intel-next-queuedRodrigo Vivi
We need the rename of reservation_object to dma_resv. The solution on this merge came from linux-next: From: Stephen Rothwell <sfr@canb.auug.org.au> Date: Wed, 14 Aug 2019 12:48:39 +1000 Subject: [PATCH] drm: fix up fallout from "dma-buf: rename reservation_object to dma_resv" Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> --- drivers/gpu/drm/i915/gt/intel_engine_pool.c | 8 ++++---- 3 files changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_engine_pool.c b/drivers/gpu/drm/i915/gt/intel_engine_pool.c index 03d90b49584a..4cd54c569911 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_pool.c +++ b/drivers/gpu/drm/i915/gt/intel_engine_pool.c @@ -43,12 +43,12 @@ static int pool_active(struct i915_active *ref) { struct intel_engine_pool_node *node = container_of(ref, typeof(*node), active); - struct reservation_object *resv = node->obj->base.resv; + struct dma_resv *resv = node->obj->base.resv; int err; - if (reservation_object_trylock(resv)) { - reservation_object_add_excl_fence(resv, NULL); - reservation_object_unlock(resv); + if (dma_resv_trylock(resv)) { + dma_resv_add_excl_fence(resv, NULL); + dma_resv_unlock(resv); } err = i915_gem_object_pin_pages(node->obj); which is a simplified version from a previous one which had: Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2019-08-21Merge tag 'drm-misc-next-2019-08-19' of ↵Dave Airlie
git://anongit.freedesktop.org/drm/drm-misc into drm-next drm-misc-next for 5.4: UAPI Changes: Cross-subsystem Changes: Core Changes: - dma-buf: add reservation_object_fences helper, relax reservation_object_add_shared_fence, remove reservation_object seq number (and then restored) - dma-fence: Shrinkage of the dma_fence structure, Merge dma_fence_signal and dma_fence_signal_locked, Store the timestamp in struct dma_fence in a union with cb_list Driver Changes: - More dt-bindings YAML conversions - More removal of drmP.h includes - dw-hdmi: Support get_eld and various i2s improvements - gm12u320: Few fixes - meson: Global cleanup - panfrost: Few refactors, Support for GPU heap allocations - sun4i: Support for DDC enable GPIO - New panels: TI nspire, NEC NL8048HL11, LG Philips LB035Q02, Sharp LS037V7DW01, Sony ACX565AKM, Toppoly TD028TTEC1 Toppoly TD043MTEA1 Signed-off-by: Dave Airlie <airlied@redhat.com> [airlied: fixup dma_resv rename fallout] From: Maxime Ripard <maxime.ripard@bootlin.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190819141923.7l2adietcr2pioct@flea
2019-08-20drm/i915: Serialize insertion into the file->mm.request_listChris Wilson
Currently, we remove the from per-file request list for throttling and retirement under a dedicated spinlock, but insertion is governed by struct_mutex. This needs to be the same lock so that the retirement/insertion of neighbouring requests (at the tail) doesn't break the list. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Matthew Auld <matthew.auld@intel.com> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190820080907.4665-1-chris@chris-wilson.co.uk
2019-08-19drm/i915: Use 0 for the unordered contextChris Wilson
Since commit 078dec3326e2 ("dma-buf: add dma_fence_get_stub") the 0 fence context became an impossible match as it is used for an always signaled fence. We can simplify our timeline tracking by knowing that 0 always means no match. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190819184404.24200-1-chris@chris-wilson.co.uk Link: https://patchwork.freedesktop.org/patch/msgid/20190819175109.5241-1-chris@chris-wilson.co.uk
2019-08-18drm/i915: Propagate fence errorsChris Wilson
Errors spread like wildfire, and must eventually be returned to the user. They need to be captured and passed along the flow of fences, infecting each in turn with the existing error, until finally they fall out of a user visible result. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190817232511.11391-1-chris@chris-wilson.co.uk
2019-08-16drm/i915/gt: Mark context->active_count as protected by timeline->mutexChris Wilson
We use timeline->mutex to protect modifications to context->active_count, and the associated enable/disable callbacks. Due to complications with engine-pm barrier there is a path where we used a "superlock" to provide serialised protect and so could not unconditionally assert with lockdep that it was always held. However, we can mark the mutex as taken (noting that we may be nested underneath ourselves) which means we can be reassured the right timeline->mutex is always treated as held and let lockdep roam free. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190816121000.8507-1-chris@chris-wilson.co.uk
2019-08-15drm/i915: Protect request retirement with timeline->mutexChris Wilson
Forgo the struct_mutex requirement for request retirement as we have been transitioning over to only using the timeline->mutex for controlling the lifetime of a request on that timeline. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190815205709.24285-4-chris@chris-wilson.co.uk
2019-08-15drm/i915: Move tasklet kicking to __i915_request_queue callerChris Wilson
Since __i915_request_queue() may be called from hardirq (timer) context, we cannot use local_bh_disable/enable at the lower level. As we do want to kick the tasklet to speed up initial submission or preemption for normal client submission, lift it to the normal process context callpath. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190815042031.27750-1-chris@chris-wilson.co.uk
2019-08-13drm/i915: Push the wakeref->count deferral to the backendChris Wilson
If the backend wishes to defer the wakeref parking, make it responsible for unlocking the wakeref (i.e. bumping the counter). This allows it to time the unlock much more carefully in case it happens to needs the wakeref to be active during its deferral. For instance, during engine parking we may choose to emit an idle barrier (a request). To do so, we borrow the engine->kernel_context timeline and to ensure exclusive access we keep the engine->wakeref.count as 0. However, to submit that request to HW may require a intel_engine_pm_get() (e.g. to keep the submission tasklet alive) and before we allow that we have to rewake our wakeref to avoid a recursive deadlock. <4> [257.742916] IRQs not enabled as expected <4> [257.742930] WARNING: CPU: 0 PID: 0 at kernel/softirq.c:169 __local_bh_enable_ip+0xa9/0x100 <4> [257.742936] Modules linked in: vgem snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_codec_generic i915 btusb btrtl btbcm btintel snd_hda_intel snd_intel_nhlt bluetooth snd_hda_codec coretemp snd_hwdep crct10dif_pclmul snd_hda_core crc32_pclmul ecdh_generic ecc ghash_clmulni_intel snd_pcm r8169 realtek lpc_ich prime_numbers i2c_hid <4> [257.742991] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G U W 5.3.0-rc3-g5d0a06cd532c-drmtip_340+ #1 <4> [257.742998] Hardware name: GIGABYTE GB-BXBT-1900/MZBAYAB-00, BIOS F6 02/17/2015 <4> [257.743008] RIP: 0010:__local_bh_enable_ip+0xa9/0x100 <4> [257.743017] Code: 37 5b 5d c3 8b 80 50 08 00 00 85 c0 75 a9 80 3d 0b be 25 01 00 75 a0 48 c7 c7 f3 0c 06 ac c6 05 fb bd 25 01 01 e8 77 84 ff ff <0f> 0b eb 89 48 89 ef e8 3b 41 06 00 eb 98 e8 e4 5c f4 ff 5b 5d c3 <4> [257.743025] RSP: 0018:ffffa78600003cb8 EFLAGS: 00010086 <4> [257.743035] RAX: 0000000000000000 RBX: 0000000000000200 RCX: 0000000000010302 <4> [257.743042] RDX: 0000000080010302 RSI: 0000000000000000 RDI: 00000000ffffffff <4> [257.743050] RBP: ffffffffc0494bb3 R08: 0000000000000000 R09: 0000000000000001 <4> [257.743058] R10: 0000000014c8f0e9 R11: 00000000fee2ff8e R12: ffffa23ba8c38008 <4> [257.743065] R13: ffffa23bacc579c0 R14: ffffa23bb7db0f60 R15: ffffa23b9cc8c430 <4> [257.743074] FS: 0000000000000000(0000) GS:ffffa23bbba00000(0000) knlGS:0000000000000000 <4> [257.743082] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 <4> [257.743089] CR2: 00007fe477b20778 CR3: 000000011f72a000 CR4: 00000000001006f0 <4> [257.743096] Call Trace: <4> [257.743104] <IRQ> <4> [257.743265] __i915_request_commit+0x240/0x5d0 [i915] <4> [257.743427] ? __i915_request_create+0x228/0x4c0 [i915] <4> [257.743584] __engine_park+0x64/0x250 [i915] <4> [257.743730] ____intel_wakeref_put_last+0x1c/0x70 [i915] <4> [257.743878] i915_sample+0x2ee/0x310 [i915] <4> [257.744030] ? i915_pmu_cpu_offline+0xb0/0xb0 [i915] <4> [257.744040] __hrtimer_run_queues+0x11e/0x4b0 <4> [257.744068] hrtimer_interrupt+0xea/0x250 <4> [257.744079] ? lockdep_hardirqs_off+0x79/0xd0 <4> [257.744101] smp_apic_timer_interrupt+0x96/0x280 <4> [257.744114] apic_timer_interrupt+0xf/0x20 <4> [257.744125] RIP: 0010:__do_softirq+0xb3/0x4ae v2: Keep the priority_hint assert v3: That assert was desperately trying to point out my bug. Sorry, little assert. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111378 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190813190705.23869-1-chris@chris-wilson.co.uk
2019-08-13dma-buf: rename reservation_object to dma_resvChristian König
Be more consistent with the naming of the other DMA-buf objects. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/323401/
2019-08-09drm/i915: Lift timeline into intel_contextChris Wilson
Move the timeline from being inside the intel_ring to intel_context itself. This saves much pointer dancing and makes the relations of the context to its timeline much clearer. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190809182518.20486-4-chris@chris-wilson.co.uk
2019-08-07drm/i915: avoid including intel_drv.h via i915_drv.h->i915_trace.hJani Nikula
Disentangle i915_drv.h from intel_drv.h, which gets included via i915_trace.h. This necessitates including i915_trace.h wherever it's needed. Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/ed82bf259d3b725a1a1a3c3e9d6fb5c08bc4d489.1565085691.git.jani.nikula@intel.com
2019-08-06drm/i915/gem: Make caps.scheduler staticChris Wilson
We do not notify userspace when the scheduler capabilities are changed (due to wedging the driver) and as such userspace will expect the caps to be static and unchanging. Make it so, and so we only need to compute our caps once during driver registration. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190806124300.24945-1-chris@chris-wilson.co.uk
2019-07-12drm/i915/gt: Use intel_gt as the primary object for handling resetsChris Wilson
Having taken the first step in encapsulating the functionality by moving the related files under gt/, the next step is to start encapsulating by passing around the relevant structs rather than the global drm_i915_private. In this step, we pass intel_gt to intel_reset.c Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190712192953.9187-1-chris@chris-wilson.co.uk
2019-07-09drm/i915: add infrastructure to hold off preemption on a requestLionel Landwerlin
We want to set this flag in the next commit on requests containing perf queries so that the result of the perf query can just be a delta of global counters, rather than doing post processing of the OA buffer. Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [ickle: add basic selftest for nopreempt] Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190709164227.25859-1-chris@chris-wilson.co.uk
2019-06-21drm/i915: Rename i915_timeline to intel_timeline and move under gtTvrtko Ursulin
Move all timeline code under gt and rename to intel_gt prefix. Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Suggested-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20190621070811.7006-32-tvrtko.ursulin@linux.intel.com
2019-06-20drm/i915/execlists: Preempt-to-busyChris Wilson
When using a global seqno, we required a precise stop-the-workd event to handle preemption and unwind the global seqno counter. To accomplish this, we would preempt to a special out-of-band context and wait for the machine to report that it was idle. Given an idle machine, we could very precisely see which requests had completed and which we needed to feed back into the run queue. However, now that we have scrapped the global seqno, we no longer need to precisely unwind the global counter and only track requests by their per-context seqno. This allows us to loosely unwind inflight requests while scheduling a preemption, with the enormous caveat that the requests we put back on the run queue are still _inflight_ (until the preemption request is complete). This makes request tracking much more messy, as at any point then we can see a completed request that we believe is not currently scheduled for execution. We also have to be careful not to rewind RING_TAIL past RING_HEAD on preempting to the running context, and for this we use a semaphore to prevent completion of the request before continuing. To accomplish this feat, we change how we track requests scheduled to the HW. Instead of appending our requests onto a single list as we submit, we track each submission to ELSP as its own block. Then upon receiving the CS preemption event, we promote the pending block to the inflight block (discarding what was previously being tracked). As normal CS completion events arrive, we then remove stale entries from the inflight tracker. v2: Be a tinge paranoid and ensure we flush the write into the HWS page for the GPU semaphore to pick in a timely fashion. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190620142052.19311-1-chris@chris-wilson.co.uk
2019-06-19drm/i915: Flush the execution-callbacks on retiringChris Wilson
In the unlikely case the request completes while we regard it as not even executing on the GPU (see the next patch!), we have to flush any pending execution callbacks at retirement and ensure that we do not add any more. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190618074153.16055-4-chris@chris-wilson.co.uk
2019-06-19drm/i915: Signal fence completion from i915_request_waitChris Wilson
With the upcoming change to automanaged i915_active, the intent is that whenever we wait on the set of active fences, they are signaled and collected. The requirement is that all successful returns from i915_request_wait() signal the fence, so fixup the one remaining path where we may return before the interrupt has been run. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190619112341.9082-3-chris@chris-wilson.co.uk
2019-06-19drm/i915: Stop passing I915_WAIT_LOCKED to i915_request_wait()Chris Wilson
Since commit eb8d0f5af4ec ("drm/i915: Remove GPU reset dependence on struct_mutex"), the I915_WAIT_LOCKED flags passed to i915_request_wait() has been defunct. Now go ahead and remove it from all callers. References: eb8d0f5af4ec ("drm/i915: Remove GPU reset dependence on struct_mutex") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190618074153.16055-3-chris@chris-wilson.co.uk
2019-06-19drm/i915: Make the semaphore saturation mask globalChris Wilson
The idea behind keeping the saturation mask local to a context backfired spectacularly. The premise with the local mask was that we would be more proactive in attempting to use semaphores after each time the context idled, and that all new contexts would attempt to use semaphores ignoring the current state of the system. This turns out to be horribly optimistic. If the system state is still oversaturated and the existing workloads have all stopped using semaphores, the new workloads would attempt to use semaphores and be deprioritised behind real work. The new contexts would not switch off using semaphores until their initial batch of low priority work had completed. Given sufficient backload load of equal user priority, this would completely starve the new work of any GPU time. To compensate, remove the local tracking in favour of keeping it as global state on the engine -- once the system is saturated and semaphores are disabled, everyone stops attempting to use semaphores until the system is idle again. One of the reason for preferring local context tracking was that it worked with virtual engines, so for switching to global state we could either do a complete check of all the virtual siblings or simply disable semaphores for those requests. This takes the simpler approach of disabling semaphores on virtual engines. The downside is that the decision that the engine is saturated is a local measure -- we are only checking whether or not this context was scheduled in a timely fashion, it may be legitimately delayed due to user priorities. We still have the same dilemma though, that we do not want to employ the semaphore poll unless it will be used. v2: Explain why we need to assume the worst wrt virtual engines. Fixes: ca6e56f654e7 ("drm/i915: Disable semaphore busywaits on saturated systems") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> Cc: Dmitry Ermilov <dmitry.ermilov@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190618074153.16055-8-chris@chris-wilson.co.uk
2019-06-18drm/i915: Use drm_gem_object.resvChris Wilson
Since commit 1ba627148ef5 ("drm: Add reservation_object to drm_gem_object"), struct drm_gem_object grew its own builtin reservation_object rendering our own private one bloat. Remove our redundant reservation_object and point into obj->base.resv instead. References: 1ba627148ef5 ("drm: Add reservation_object to drm_gem_object") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Matthew Auld <matthew.auld@intel.com> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190618125858.7295-1-chris@chris-wilson.co.uk
2019-06-14drm/i915: Replace engine->timeline with a plain listChris Wilson
To continue the onslaught of removing the assumption of a global execution ordering, another casualty is the engine->timeline. Without an actual timeline to track, it is overkill and we can replace it with a much less grand plain list. We still need a list of requests inflight, for the simple purpose of finding inflight requests (for retiring, resetting, preemption etc). Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190614164606.15633-3-chris@chris-wilson.co.uk
2019-06-14drm/i915: Stop retiring along engineChris Wilson
We no longer track the execution order along the engine and so no longer need to enforce ordering of retire along the engine. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190614164606.15633-2-chris@chris-wilson.co.uk
2019-06-14drm/i915: Keep contexts pinned until after the next kernel context switchChris Wilson
We need to keep the context image pinned in memory until after the GPU has finished writing into it. Since it continues to write as we signal the final breadcrumb, we need to keep it pinned until the request after it is complete. Currently we know the order in which requests execute on each engine, and so to remove that presumption we need to identify a request/context-switch we know must occur after our completion. Any request queued after the signal must imply a context switch, for simplicity we use a fresh request from the kernel context. The sequence of operations for keeping the context pinned until saved is: - On context activation, we preallocate a node for each physical engine the context may operate on. This is to avoid allocations during unpinning, which may be from inside FS_RECLAIM context (aka the shrinker) - On context deactivation on retirement of the last active request (which is before we know the context has been saved), we add the preallocated node onto a barrier list on each engine - On engine idling, we emit a switch to kernel context. When this switch completes, we know that all previous contexts must have been saved, and so on retiring this request we can finally unpin all the contexts that were marked as deactivated prior to the switch. We can enhance this in future by flushing all the idle contexts on a regular heartbeat pulse of a switch to kernel context, which will also be used to check for hung engines. v2: intel_context_active_acquire/_release Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190614164606.15633-1-chris@chris-wilson.co.uk
2019-06-14drm/i915: Refine i915_reset.lock_mapChris Wilson
We already use a mutex to serialise i915_reset() and wedging, so all we need it to link that into i915_request_wait() and we have our lock cycle detection. v2.5: Take error mutex for selftests Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190614071023.17929-3-chris@chris-wilson.co.uk
2019-06-14drm/i915: Execute signal callbacks from no-op i915_request_waitChris Wilson
If we enter i915_request_wait() with an already completed request, but unsignaled dma-fence, signal the fence before returning. This allows us to execute any of the signal callbacks at the earliest opportunity. v2: Also signal after busyspin success Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190614111053.25615-2-chris@chris-wilson.co.uk
2019-06-12drm/i915: Prevent lock-cycles between GPU waits and GPU resetsChris Wilson
We cannot allow ourselves to wait on the GPU while holding any lock as we may need to reset the GPU. While there is not an explicit lock between the two operations, lockdep cannot detect the dependency. So let's tell lockdep about the wait/reset dependency with an explicit lockmap. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190612085246.16374-1-chris@chris-wilson.co.uk
2019-06-10drm/i915: Allow interrupts when taking the timeline->mutexChris Wilson
Before we commit ourselves to writing commands into the ringbuffer and submitting the request, allow signals to interrupt acquisition of the timeline mutex. We allow ourselves to be interrupted at any time later if we need to block for space in the ring, anyway. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190610103610.19883-1-chris@chris-wilson.co.uk
2019-05-28drm/i915: Move more GEM objects under gem/Chris Wilson
Continuing the theme of separating out the GEM clutter. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190528092956.14910-8-chris@chris-wilson.co.uk
2019-05-22drm/i915: Extend execution fence to support a callbackChris Wilson
In the next patch, we will want to configure the slave request depending on which physical engine the master request is executed on. For this, we introduce a callback from the execute fence to convey this information. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190521211134.16117-8-chris@chris-wilson.co.uk
2019-05-22drm/i915: Apply an execution_mask to the virtual_engineChris Wilson
Allow the user to direct which physical engines of the virtual engine they wish to execute one, as sometimes it is necessary to override the load balancing algorithm. v2: Only kick the virtual engines on context-out if required Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190521211134.16117-7-chris@chris-wilson.co.uk
2019-05-17drm/i915: Downgrade NEWCLIENT to non-preemptiveChris Wilson
Commit 1413b2bc0717 ("drm/i915: Trim NEWCLIENT boosting") had the intended consequence of not allowing a sequence of work that merely crossed into a new engine the privilege to be promoted to NEWCLIENT status. It also had the unintended consequence of actually making NEWCLIENT effective on heavily oversubscribed transcode machines and impacting upon their throughput. If we consider a client packet composed of (rcsA, rcsB, vcs) and 30 of those clients, using the NEWCLIENT boost that will be scheduled as rcsA x 30, (rcsB, vcs) x 30 where as before it would have been (rcsA, rcsB, vcs) x 30 That is with NEWCLIENT only boosting the first request of each client, we would execute all rcsA requests prior to running on the vcs engines; acruing a lot of dead time as compared to the previous case where the vcs engine would be started in parallel to processing the second client. The previous patch has the effect of delaying submission until it is required by a third party (either the user with an explicit wait, or by another client/engine). We reduce the NEWCLIENT bump to a mere WAIT, which has the effect of removing its preemptive grant and reducing it to the same level as any other user interaction -- that it will not be promoted above the interengine dependencies, and so preventing NEWCLIENTS from starving other engines. This a large nerf to the rrul properties of the current NEWCLIENT, but it still does give prioritised submission to new requests from light workloads. References: b16c765122f9 ("drm/i915: Priority boost for new clients") Fixes: 1413b2bc0717 ("drm/i915: Trim NEWCLIENT boosting") # customer impact Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> Cc: Dmitry Ermilov <dmitry.ermilov@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190515130052.4475-4-chris@chris-wilson.co.uk
2019-05-17drm/i915: Bump signaler priority on adding a waiterChris Wilson
The handling of the no-preemption priority level imposes the restriction that we need to maintain the implied ordering even though preemption is disabled. Otherwise we may end up with an AB-BA deadlock across multiple engine due to a real preemption event reordering the no-preemption WAITs. To resolve this issue we currently promote all requests to WAIT on unsubmission, however this interferes with the timeslicing requirement that we do not apply any implicit promotion that will defeat the round-robin timeslice list. (If we automatically promote the active request it will go back to the head of the queue and not the tail!) So we need implicit promotion to prevent reordering around semaphores where we are not allowed to preempt, and we must avoid implicit promotion on unsubmission. So instead of at unsubmit, if we apply that implicit promotion on adding the dependency, we avoid the semaphore deadlock and we also reduce the gains made by the promotion for user space waiting. Furthermore, by keeping the earlier dependencies at a higher level, we reduce the search space for timeslicing without altering runtime scheduling too badly (no dependencies at all will be assigned a higher priority for rrul). v2: Limit the bump to external edges (as originally intended) i.e. between contexts and out to the user. Testcase: igt/gem_concurrent_blit Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190515130052.4475-3-chris@chris-wilson.co.uk
2019-05-17drm/i915: Truly bump ready tasks ahead of busywaitsChris Wilson
In commit b7404c7ecb38 ("drm/i915: Bump ready tasks ahead of busywaits"), I tried cutting a corner in order to not install a signal for each of our dependencies, and only listened to requests on which we were intending to busywait. The compromise that was made was that instead of then being able to promote the request with a full NOSEMAPHORE like its non-busywaiting brethren, as we had not ensured we had cleared the semaphore chain, we settled for only using the NEWCLIENT boost. With an over saturated system with multiple NEWCLIENTS in flight at any time, this was found to be an inadequate promotion and left us with a much poorer scheduling order than prior to using semaphores. The outcome of this patch, is that all requests have NOSEMAPHORE priority when they have no dependencies and are ready to run and not busywait, restoring the pre-semaphore ordering on saturated systems. We can demonstrate the effect of poor scheduling order by oversaturating the system using gem_wsim on a system with multiple vcs engines (i.e running the same workloads across more clients than required for peak throughput, e.g. media_load_balance_17i7.wsim -c4 -b context): x v5.1 (normalized) + tip * fix +------------------------------------------------------------------------+ | x | | x | | x | | x | | %x | | %%x | | %%x | | %%x | | %%x | | %%x | | %%x | | %%x | | %%x | | %%x | | %%x | | %#x | | %#x | | %#x | | %#x | | %#x | | + %#xx | | + %#xx | | + %%#xx | | + %%#xx | | + %%#xx | | + %%#xx | | + %%##x | | +++ %%##x | | +++ %%##x | | +++ %%##x | | ++++ %%##x | | ++++ %%##x | | ++++ %%##xx | | ++++ %###xx | | ++++ %###xx | | ++++ %###xx | | ++++ %###xx | | ++++ + %#O#xx | | ++++ + %#O#xx | | ++++++ + %#O#xx | | ++++++++++ %OOOxxx| | ++++++++++ + %#OOO#xx| | + ++++++++++++ ++ +++++ + ++ @@OOOO#xx| | |A_| | ||__________M_______A____________________| | | |A_| | +------------------------------------------------------------------------+ N Min Max Median Avg Stddev x 120 0.99456 1.00628 0.999985 1.0001545 0.0024387139 + 120 0.873021 1.00037 0.884134 0.90148752 0.039190862 Difference at 99.5% confidence -0.098667 +/- 0.0110762 -9.86517% +/- 1.10745% (Student's t, pooled s = 0.0277657) % 120 0.990207 1.00165 0.9970265 0.99699748 0.0021024 Difference at 99.5% confidence -0.003157 +/- 0.000908245 -0.315651% +/- 0.0908105% (Student's t, pooled s = 0.00227678) Fixes: b7404c7ecb38 ("drm/i915: Bump ready tasks ahead of busywaits") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> Cc: Dmitry Ermilov <dmitry.ermilov@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190515130052.4475-2-chris@chris-wilson.co.uk
2019-05-17drm/i915: Mark semaphores as complete on unsubmit out if payload was startedChris Wilson
Avoid charging us for the presumed busywait if the request was preempted after successfully using semaphores to reduce inter-engine latency. v2: Bump the priority to reflect the lack of semaphores now required. References: ca6e56f654e7 ("drm/i915: Disable semaphore busywaits on saturated systems") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190515130052.4475-1-chris@chris-wilson.co.uk
2019-05-08drm/i915: Seal races between async GPU cancellation, retirement and signalingChris Wilson
Currently there is an underlying assumption that i915_request_unsubmit() is synchronous wrt the GPU -- that is the request is no longer in flight as we remove it. In the near future that may change, and this may upset our signaling as we can process an interrupt for that request while it is no longer in flight. CPU0 CPU1 intel_engine_breadcrumbs_irq (queue request completion) i915_request_cancel_signaling ... ... i915_request_enable_signaling dma_fence_signal Hence in the time it took us to drop the lock to signal the request, a preemption event may have occurred and re-queued the request. In the process, that request would have seen I915_FENCE_FLAG_SIGNAL clear and so reused the rq->signal_link that was in use on CPU0, leading to bad pointer chasing in intel_engine_breadcrumbs_irq. A related issue was that if someone started listening for a signal on a completed but no longer in-flight request, we missed the opportunity to immediately signal that request. Furthermore, as intel_contexts may be immediately released during request retirement, in order to be entirely sure that intel_engine_breadcrumbs_irq may no longer dereference the intel_context (ce->signals and ce->signal_link), we must wait for irq spinlock. In order to prevent the race, we use a bit in the fence.flags to signal the transfer onto the signal list inside intel_engine_breadcrumbs_irq. For simplicity, we use the DMA_FENCE_FLAG_SIGNALED_BIT as it then quickly signals to any outside observer that the fence is indeed signaled. v2: Sketch out potential dma-fence API for manual signaling v3: And the test_and_set_bit() Fixes: 52c0fdb25c7c ("drm/i915: Replace global breadcrumbs with per-context interrupt tracking") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190508112452.18942-1-chris@chris-wilson.co.uk
2019-05-07drm/i915: Only reschedule the submission tasklet if preemption is possibleChris Wilson
If we couple the scheduler more tightly with the execlists policy, we can apply the preemption policy to the question of whether we need to kick the tasklet at all for this priority bump. v2: Rephrase it as a core i915 policy and not an execlists foible. v3: Pull the kick together. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190507122544.12698-1-chris@chris-wilson.co.uk
2019-05-07drm/i915: Acquire the signaler's timeline HWSP lastChris Wilson
Acquiring the signaler's timeline takes an active reference to their HWSP that we would like to avoid if possible, so take it after performing all of our allocations required to set up the fencing. The acquisition also provides the final check that the target has not already signaled allowing us to avoid the semaphore at the last moment. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190503140239.32668-1-chris@chris-wilson.co.uk
2019-05-04drm/i915: Disable semaphore busywaits on saturated systemsChris Wilson
Asking the GPU to busywait on a memory address, perhaps not unexpectedly in hindsight for a shared system, leads to bus contention that affects CPU programs trying to concurrently access memory. This can manifest as a drop in transcode throughput on highly over-saturated workloads. The only clue offered by perf, is that the bus-cycles (perf stat -e bus-cycles) jumped by 50% when enabling semaphores. This corresponds with extra CPU active cycles being attributed to intel_idle's mwait. This patch introduces a heuristic to try and detect when more than one client is submitting to the GPU pushing it into an oversaturated state. As we already keep track of when the semaphores are signaled, we can inspect their state on submitting the busywait batch and if we planned to use a semaphore but were too late, conclude that the GPU is overloaded and not try to use semaphores in future requests. In practice, this means we optimistically try to use semaphores for the first frame of a transcode job split over multiple engines, and fail if there are multiple clients active and continue not to use semaphores for the subsequent frames in the sequence. Periodically, we try to optimistically switch semaphores back on whenever the client waits to catch up with the transcode results. With 1 client, on Broxton J3455, with the relative fps normalized by %cpu: x no semaphores + drm-tip * patched +------------------------------------------------------------------------+ | * | | *+ | | **+ | | **+ x | | x * +**+ x | | x x * * +***x xx | | x x * * *+***x *x | | x x* + * * *****x *x x | | + x xx+x* + *** * ********* x * | | + x xx+x* * *** +** ********* xx * | | * + ++++* + x*x****+*+* ***+*************+x* * | |*+ +** *+ + +* + *++****** *xxx**********x***+*****************+*++ *| | |__________A_____M_____| | | |_______________A____M_________| | | |____________A___M________| | +------------------------------------------------------------------------+ N Min Max Median Avg Stddev x 120 2.60475 3.50941 3.31123 3.2143953 0.21117399 + 120 2.3826 3.57077 3.25101 3.1414161 0.28146407 Difference at 95.0% confidence -0.0729792 +/- 0.0629585 -2.27039% +/- 1.95864% (Student's t, pooled s = 0.248814) * 120 2.35536 3.66713 3.2849 3.2059917 0.24618565 No difference proven at 95.0% confidence With 10 clients over-saturating the pipeline: x no semaphores + drm-tip * patched +------------------------------------------------------------------------+ | ++ ** | | ++ ** | | ++ ** | | ++ ** | | ++ xx *** | | ++ xx *** | | ++ xxx*** | | ++ xxx*** | | +++ xxx*** | | +++ xx**** | | +++ xx**** | | +++ xx**** | | +++ xx**** | | ++++ xx**** | | +++++ xx**** | | +++++ x x****** | | ++++++ xxx******* | | ++++++ xxx******* | | ++++++ xxx******* | | ++++++ xx******** | | ++++++ xxxx******** | | ++++++ xxxx******** | | ++++++++ xxxxx********* | |+ + + + ++++++++ xxx*xx**********x* *| | |__A__| | | |__AM__| | | |__A_| | +------------------------------------------------------------------------+ N Min Max Median Avg Stddev x 120 2.47855 2.8972 2.72376 2.7193402 0.074604933 + 120 1.17367 1.77459 1.71977 1.6966782 0.085850697 Difference at 95.0% confidence -1.02266 +/- 0.0203502 -37.607% +/- 0.748352% (Student's t, pooled s = 0.0804246) * 120 2.57868 3.00821 2.80142 2.7923878 0.058646477 Difference at 95.0% confidence 0.0730476 +/- 0.0169791 2.68622% +/- 0.624383% (Student's t, pooled s = 0.0671018) Indicating that we've recovered the regression from enabling semaphores on this saturated setup, with a hint towards an overall improvement. Very similar, but of smaller magnitude, results are observed on both Skylake(gt2) and Kabylake(gt4). This may be due to the reduced impact of bus-cycles, where we see a 50% hit on Broxton, it is only 10% on the big core, in this particular test. One observation to make here is that for a greedy client trying to maximise its own throughput, using semaphores is the right choice. It is only the holistic system-wide view that semaphores of one client impacts another and reduces the overall throughput where we would choose to disable semaphores. The most noticeable negactive impact this has is on the no-op microbenchmarks, which are also very notable for having no cpu bus load. In particular, this increases the runtime and energy consumption of gem_exec_whisper. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> Cc: Dmitry Ermilov <dmitry.ermilov@intel.com> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190504070707.30902-1-chris@chris-wilson.co.uk
2019-05-03drm/i915: Delay semaphore submission until the start of the signalerChris Wilson
Currently we submit the semaphore busywait as soon as the signaler is submitted to HW. However, we may submit the signaler as the tail of a batch of requests, and even not as the first context in the HW list, i.e. the busywait may start spinning far in advance of the signaler even starting. If we wait until the request before the signaler is completed before submitting the busywait, we prevent the busywait from starting too early, if the signaler is not first in submission port. To handle the case where the signaler is at the start of the second (or later) submission port, we will need to delay the execution callback until we know the context is promoted to port0. A challenge for later. Fixes: e88619646971 ("drm/i915: Use HW semaphores for inter-engine synchroni sation on gen8+") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190501114541.10077-9-chris@chris-wilson.co.uk
2019-04-26drm/i915: Move i915_request_alloc into selftests/Chris Wilson
Having transitioned GEM over to using intel_context as its primary means of tracking the GEM context and engine combined and using i915_request_create(), we can move the older i915_request_alloc() helper function into selftests/ where the remaining users are confined. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190426163336.15906-9-chris@chris-wilson.co.uk
2019-04-26drm/i915: Switch back to an array of logical per-engine HW contextsChris Wilson
We switched to a tree of per-engine HW context to accommodate the introduction of virtual engines. However, we plan to also support multiple instances of the same engine within the GEM context, defeating our use of the engine as a key to looking up the HW context. Just allocate a logical per-engine instance and always use an index into the ctx->engines[]. Later on, this ctx->engines[] may be replaced by a user specified map. v2: Add for_each_gem_engine() helper to iterator within the engines lock v3: intel_context_create_request() helper v4: s/unsigned long/unsigned int/ 4 billion engines is quite enough. v5: Push iterator locking to caller Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190426163336.15906-7-chris@chris-wilson.co.uk
2019-04-26drm/i915: Export intel_context_instance()Chris Wilson
We want to pass in a intel_context into intel_context_pin() and that requires us to first be able to lookup the intel_context! Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190426163336.15906-2-chris@chris-wilson.co.uk
2019-04-25drm/i915: Explicitly pin the logical context for execbufChris Wilson
In order to separate the reservation phase of building a request from its emission phase, we need to pull some of the request alloc activities from deep inside i915_request to the surface, GEM_EXECBUFFER. v2: Be frivolous, use a local drm_i915_private. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190425050143.811-1-chris@chris-wilson.co.uk
2019-04-24drm/i915: Invert the GEM wakeref hierarchyChris Wilson
In the current scheme, on submitting a request we take a single global GEM wakeref, which trickles down to wake up all GT power domains. This is undesirable as we would like to be able to localise our power management to the available power domains and to remove the global GEM operations from the heart of the driver. (The intent there is to push global GEM decisions to the boundary as used by the GEM user interface.) Now during request construction, each request is responsible via its logical context to acquire a wakeref on each power domain it intends to utilize. Currently, each request takes a wakeref on the engine(s) and the engines themselves take a chipset wakeref. This gives us a transition on each engine which we can extend if we want to insert more powermangement control (such as soft rc6). The global GEM operations that currently require a struct_mutex are reduced to listening to pm events from the chipset GT wakeref. As we reduce the struct_mutex requirement, these listeners should evaporate. Perhaps the biggest immediate change is that this removes the struct_mutex requirement around GT power management, allowing us greater flexibility in request construction. Another important knock-on effect, is that by tracking engine usage, we can insert a switch back to the kernel context on that engine immediately, avoiding any extra delay or inserting global synchronisation barriers. This makes tracking when an engine and its associated contexts are idle much easier -- important for when we forgo our assumed execution ordering and need idle barriers to unpin used contexts. In the process, it means we remove a large chunk of code whose only purpose was to switch back to the kernel context. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Imre Deak <imre.deak@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190424200717.1686-5-chris@chris-wilson.co.uk
2019-04-24drm/i915: Pass intel_context to i915_request_create()Chris Wilson
Start acquiring the logical intel_context and using that as our primary means for request allocation. This is the initial step to allow us to avoid requiring struct_mutex for request allocation along the perma-pinned kernel context, but it also provides a foundation for breaking up the complex request allocation to handle different scenarios inside execbuf. For the purpose of emitting a request from inside retirement (see the next patch for engine power management), we also need to lift control over the timeline mutex to the caller. v2: Note that the request carries the active reference upon construction. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190424200717.1686-4-chris@chris-wilson.co.uk
generated by cgit (git 2.34.1) at 2024-05-26 15:07:11 +0000