<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head> <body text="#000000" bgcolor="#FFFFFF"> <div class="moz-cite-prefix">Am 05.02.2018 um 09:22 schrieb Chunming Zhou: </div> <blockquote type="cite" cite="mid:d2ec14e8-d4cd-a312-5e23-ef2e2198d6ea@amd.com"> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> On 2018年01月31日 18:54, Christian König wrote: <blockquote type="cite" cite="mid:ff405b17-7f7c-3a9e-030b-28dcfc256546@amd.com"> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> <div class="moz-cite-prefix"> <blockquote type="cite">So I think preventing validation on same place is a simpler way: process B bo's place is fpfn~lpfn, it will only try to evict LRU BOs in that range, while eviction, we just prevent those validation to this range(fpfn~lpfn), if out of this range, the allocation/validation still can be go on. Any negative? </blockquote> That won't work either. The most common use of fpfn~lpfn range is to limit a BO to visible VRAM, the other use cases are to fullfil hardware limitations. So blocking this would result in blocking all normal GTT and VRAM allocations, adding a mutex to validate would have the same effect. </div> </blockquote> No, different effect, mutex will make the every allocation serialized. My approach only effects busy case, that is said, only when space is used up, the allocation is serialized in that range, otherwise still in parallel. </blockquote> That would still not allow for concurrent evictions, not as worse as completely blocking concurrent validation but still not a good idea as far as I can see. Going to give it a try today to use the ww_mutex owner to detect eviction of already reserved BOs. Maybe that can be used instead, Christian. <blockquote type="cite" cite="mid:d2ec14e8-d4cd-a312-5e23-ef2e2198d6ea@amd.com"> Regards, David Zhou <blockquote type="cite" cite="mid:ff405b17-7f7c-3a9e-030b-28dcfc256546@amd.com"> <div class="moz-cite-prefix"> Regards, Christian. Am 31.01.2018 um 11:30 schrieb Chunming Zhou: </div> <blockquote type="cite" cite="mid:cdb7726b-a82d-494c-a98c-ca0100f323cc@amd.com"> <div class="moz-cite-prefix">On 2018年01月26日 22:35, Christian König wrote: </div> <blockquote type="cite" cite="mid:a6d78bba-bbe9-769b-f9d3-665cdd8c04da@gmail.com"> <div class="moz-cite-prefix">I just realized that a change I'm thinking about for a while would solve your problem as well, but keep concurrent allocation possible. See ttm_mem_evict_first() unlocks the BO after evicting it: <blockquote type="cite"> ttm_bo_del_from_lru(bo); spin_unlock(&glob->lru_lock); ret = ttm_bo_evict(bo, ctx); if (locked) { ttm_bo_unreserve(bo); <-------- here } else { spin_lock(&glob->lru_lock); ttm_bo_add_to_lru(bo); spin_unlock(&glob->lru_lock); } kref_put(&bo->list_kref, ttm_bo_release_list); </blockquote> The effect is that in your example process C can not only beat process B once, but many many times because we run into a ping/pong situation where B evicts resources while C moves them back in. </div> </blockquote> For ping/pong case, I want to disable busy placement for allocation period, only enable it for cs bo validation. <blockquote type="cite" cite="mid:a6d78bba-bbe9-769b-f9d3-665cdd8c04da@gmail.com"> <div class="moz-cite-prefix"> For a while now I'm thinking about dropping those reservations only after the original allocation succeeded. The effect would be that process C can still beat process B initially, but sooner or process B would evict some resources from process C as well and then it can succeed with its allocation. </div> </blockquote> If it is from process C cs validation, process B still need evict the resource only after process C command submission completion. <blockquote type="cite" cite="mid:a6d78bba-bbe9-769b-f9d3-665cdd8c04da@gmail.com"> <div class="moz-cite-prefix"> The problem is for this approach to work we need to core change to the ww_mutexes to be able to handle this efficiently. </div> </blockquote> Yes, ww_mutex doesn't support this net lock, which easily deadlock without ticket and class. So I think preventing validation on same place is a simpler way: process B bo's place is fpfn~lpfn, it will only try to evict LRU BOs in that range, while eviction, we just prevent those validation to this range(fpfn~lpfn), if out of this range, the allocation/validation still can be go on. Any negative? Regards, David Zhou <blockquote type="cite" cite="mid:a6d78bba-bbe9-769b-f9d3-665cdd8c04da@gmail.com"> <div class="moz-cite-prefix"> Regards, Christian. Am 26.01.2018 um 14:59 schrieb Christian König: </div> <blockquote type="cite" cite="mid:e3ed22b7-ca1a-09d0-47a0-204affef780f@amd.com"> <div class="moz-cite-prefix">I know, but this has the same effect. You prevent concurrent allocation from happening. What we could do is to pipeline reusing of deleted memory as well, this makes it less likely to cause the problem you are seeing because the evicting processes doesn't need to block for deleted BOs any more. But that other processes can grab memory during eviction is intentional. Otherwise greedy processes would completely dominate command submission. Regards, Christian. Am 26.01.2018 um 14:50 schrieb Zhou, David(ChunMing): </div> <blockquote type="cite" cite="mid:smartisan1516974623698"> <meta content="text/html; charset=utf-8"> <div name="smartisanmessageid" id="smartisan1516974623698">I don't want to prevent all, my new approach is to prevent the later allocation is trying and ahead of front to get the memory space that the front made from eviction. 发自坚果 Pro <style type="text/css">  </style> <div class="quote"> <div style="margin:0 0px; font-size:105%">Christian K鰊ig <a class="moz-txt-link-rfc2396E" href="mailto:ckoenig.leichtzumerken@gmail.com" moz-do-not-send="true"><ckoenig.leichtzumerken@gmail.com></a> 于 2018年1月26日下午9:24写道：</div> </div> </div> <div> <div class="moz-cite-prefix">Yes, exactly that's the problem. See when you want to prevent a process B from allocating the memory process A has evicted, you need to prevent all concurrent allocation. And we don't do that because it causes a major performance drop. Regards, Christian. Am 26.01.2018 um 14:21 schrieb Zhou, David(ChunMing): </div> <blockquote type="cite"> <meta content="text/html; charset=utf-8"> <div name="smartisanmessageid" id="smartisan1516972869817">You patch will prevent concurrent allocation, and will result in allocation performance drop much. 发自坚果 Pro <style type="text/css">  </style> <div class="quote"> <div style="margin:0 0px; font-size:105%">Christian K鰊ig <a class="moz-txt-link-rfc2396E" href="mailto:ckoenig.leichtzumerken@gmail.com" moz-do-not-send="true"><ckoenig.leichtzumerken@gmail.com></a> 于 2018年1月26日下午9:04写道：</div> </div> </div> <div> <div class="moz-cite-prefix">Attached is what you actually want to do cleanly implemented. But as I said this is a NO-GO. Regards, Christian. Am 26.01.2018 um 13:43 schrieb Christian König: </div> <blockquote type="cite"> <div class="moz-cite-prefix"> <blockquote type="cite">After my investigation, this issue should be detect of TTM design self, which breaks scheduling balance. </blockquote> Yeah, but again. This is indented design we can't change easily. Regards, Christian. Am 26.01.2018 um 13:36 schrieb Zhou, David(ChunMing): </div> <blockquote type="cite"> <meta name="Generator" content="Microsoft Exchange Server"> <style>  </style> <div> <div name="x_smartisanmessageid" id="x_smartisan1516970199102">I am off work, so reply mail by phone, the format could not be text. back to topic itself: the problem indeed happen on amdgpu driver, someone reports me that application runs with two instances, the performance are different. I also reproduced the issue with unit test(bo_eviction_test). They always think our scheduler isn't working as expected. After my investigation, this issue should be detect of TTM design self, which breaks scheduling balance. Further, if we run containers for our gpu, container A could run high score, container B runs low score with same benchmark. So this is bug that we need fix. Regards, David Zhou 发自坚果 Pro <style type="text/css">  </style> <div class="x_quote"> <div style="margin:0 0px; font-size:105%">Christian K鰊ig <a class="moz-txt-link-rfc2396E" href="mailto:ckoenig.leichtzumerken@gmail.com" moz-do-not-send="true"><ckoenig.leichtzumerken@gmail.com></a> 于 2018年1月26日下午6:31写道：</div> </div> </div> </div> <div class="PlainText">Am 26.01.2018 um 11:22 schrieb Chunming Zhou: > there is a scheduling balance issue about get node like: > a. process A allocates full memory and use it for submission. > b. process B tries to allocates memory, will wait for process A BO idle in eviction. > c. process A completes the job, process B eviction will put process A BO node, > but in the meantime, process C is comming to allocate BO, whill directly get node successfully, and do submission, > process B will again wait for process C BO idle. > d. repeat the above setps, process B could be delayed much more. > > later allocation must not be ahead of front in same place. Again NAK to the whole approach. At least with amdgpu the problem you described above never occurs because evictions are pipelined operations. We could only block for deleted regions to become free. But independent of that incoming memory requests while we make room for eviction are intended to be served first. Changing that is certainly a no-go cause that would favor memory hungry applications over small clients. Regards, Christian. > > Change-Id: I3daa892e50f82226c552cc008a29e55894a98f18 > Signed-off-by: Chunming Zhou <a class="moz-txt-link-rfc2396E" href="mailto:david1.zhou@amd.com" moz-do-not-send="true"> <david1.zhou@amd.com></a> > --- > drivers/gpu/drm/ttm/ttm_bo.c | 69 +++++++++++++++++++++++++++++++++++++++-- > include/drm/ttm/ttm_bo_api.h | 7 +++++ > include/drm/ttm/ttm_bo_driver.h | 7 +++++ > 3 files changed, 80 insertions(+), 3 deletions(-) > > diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c > index d33a6bb742a1..558ec2cf465d 100644 > --- a/drivers/gpu/drm/ttm/ttm_bo.c > +++ b/drivers/gpu/drm/ttm/ttm_bo.c > @@ -841,6 +841,58 @@ static int ttm_bo_add_move_fence(struct ttm_buffer_object *bo, > return 0; > } > > +static void ttm_man_init_waiter(struct ttm_bo_waiter *waiter, > + struct ttm_buffer_object *bo, > + const struct ttm_place *place) > +{ > + waiter->tbo = bo; > + memcpy((void *)&waiter->place, (void *)place, sizeof(*place)); > + INIT_LIST_HEAD(&waiter->list); > +} > + > +static void ttm_man_add_waiter(struct ttm_mem_type_manager *man, > + struct ttm_bo_waiter *waiter) > +{ > + if (!waiter) > + return; > + spin_lock(&man->wait_lock); > + list_add_tail(&waiter->list, &man->waiter_list); > + spin_unlock(&man->wait_lock); > +} > + > +static void ttm_man_del_waiter(struct ttm_mem_type_manager *man, > + struct ttm_bo_waiter *waiter) > +{ > + if (!waiter) > + return; > + spin_lock(&man->wait_lock); > + if (!list_empty(&waiter->list)) > + list_del(&waiter->list); > + spin_unlock(&man->wait_lock); > + kfree(waiter); > +} > + > +int ttm_man_check_bo(struct ttm_mem_type_manager *man, > + struct ttm_buffer_object *bo, > + const struct ttm_place *place) > +{ > + struct ttm_bo_waiter *waiter, *tmp; > + > + spin_lock(&man->wait_lock); > + list_for_each_entry_safe(waiter, tmp, &man->waiter_list, list) { > + if ((bo != waiter->tbo) && > + ((place->fpfn >= waiter->place.fpfn && > + place->fpfn <= waiter->place.lpfn) || > + (place->lpfn <= waiter->place.lpfn && place->lpfn >= > + waiter->place.fpfn))) > + goto later_bo; > + } > + spin_unlock(&man->wait_lock); > + return true; > +later_bo: > + spin_unlock(&man->wait_lock); > + return false; > +} > /** > * Repeatedly evict memory from the LRU for @mem_type until we create enough > * space, or we've evicted everything and there isn't enough space. > @@ -853,17 +905,26 @@ static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo, > { > struct ttm_bo_device *bdev = bo->bdev; > struct ttm_mem_type_manager *man = &bdev->man[mem_type]; > + struct ttm_bo_waiter waiter; > int ret; > > + ttm_man_init_waiter(&waiter, bo, place); > + ttm_man_add_waiter(man, &waiter); > do { > ret = (*man->func->get_node)(man, bo, place, mem); > - if (unlikely(ret != 0)) > + if (unlikely(ret != 0)) { > + ttm_man_del_waiter(man, &waiter); > return ret; > - if (mem->mm_node) > + } > + if (mem->mm_node) { > + ttm_man_del_waiter(man, &waiter); > break; > + } > ret = ttm_mem_evict_first(bdev, mem_type, place, ctx); > - if (unlikely(ret != 0)) > + if (unlikely(ret != 0)) { > + ttm_man_del_waiter(man, &waiter); > return ret; > + } > } while (1); > mem->mem_type = mem_type; > return ttm_bo_add_move_fence(bo, man, mem); > @@ -1450,6 +1511,8 @@ int ttm_bo_init_mm(struct ttm_bo_device *bdev, unsigned type, > man->use_io_reserve_lru = false; > mutex_init(&man->io_reserve_mutex); > spin_lock_init(&man->move_lock); > + spin_lock_init(&man->wait_lock); > + INIT_LIST_HEAD(&man->waiter_list); > INIT_LIST_HEAD(&man->io_reserve_lru); > > ret = bdev->driver->init_mem_type(bdev, type, man); > diff --git a/include/drm/ttm/ttm_bo_api.h b/include/drm/ttm/ttm_bo_api.h > index 2cd025c2abe7..0fce4dbd02e7 100644 > --- a/include/drm/ttm/ttm_bo_api.h > +++ b/include/drm/ttm/ttm_bo_api.h > @@ -40,6 +40,7 @@ > #include <linux/mm.h> > #include <linux/bitmap.h> > #include <linux/reservation.h> > +#include <drm/ttm/ttm_placement.h> > > struct ttm_bo_device; > > @@ -232,6 +233,12 @@ struct ttm_buffer_object { > struct mutex wu_mutex; > }; > > +struct ttm_bo_waiter { > + struct ttm_buffer_object *tbo; > + struct ttm_place place; > + struct list_head list; > +}; > + > /** > * struct ttm_bo_kmap_obj > * > diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h > index 9b417eb2df20..dc6b8b4c9e06 100644 > --- a/include/drm/ttm/ttm_bo_driver.h > +++ b/include/drm/ttm/ttm_bo_driver.h > @@ -293,6 +293,10 @@ struct ttm_mem_type_manager { > bool io_reserve_fastpath; > spinlock_t move_lock; > > + /* waiters in list */ > + spinlock_t wait_lock; > + struct list_head waiter_list; > + > /* > * Protected by @io_reserve_mutex: > */ > @@ -748,6 +752,9 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo, > struct ttm_mem_reg *mem, > struct ttm_operation_ctx *ctx); > > +int ttm_man_check_bo(struct ttm_mem_type_manager *man, > + struct ttm_buffer_object *bo, > + const struct ttm_place *place); > void ttm_bo_mem_put(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); > void ttm_bo_mem_put_locked(struct ttm_buffer_object *bo, > struct ttm_mem_reg *mem); </div> <fieldset class="mimeAttachmentHeader"></fieldset> <pre>_______________________________________________ amd-gfx mailing list <a class="moz-txt-link-abbreviated" href="mailto:amd-gfx@lists.freedesktop.org" moz-do-not-send="true">amd-gfx@lists.freedesktop.org</a> <a class="moz-txt-link-freetext" href="https://lists.freedesktop.org/mailman/listinfo/amd-gfx" moz-do-not-send="true">https://lists.freedesktop.org/mailman/listinfo/amd-gfx</a> </pre> </blockquote> </blockquote> </div> <fieldset class="mimeAttachmentHeader"></fieldset> <pre>_______________________________________________ dri-devel mailing list <a class="moz-txt-link-abbreviated" href="mailto:dri-devel@lists.freedesktop.org" moz-do-not-send="true">dri-devel@lists.freedesktop.org</a> <a class="moz-txt-link-freetext" href="https://lists.freedesktop.org/mailman/listinfo/dri-devel" moz-do-not-send="true">https://lists.freedesktop.org/mailman/listinfo/dri-devel</a> </pre> </blockquote> </div> </blockquote> <fieldset class="mimeAttachmentHeader"></fieldset> <pre wrap="">_______________________________________________ amd-gfx mailing list <a class="moz-txt-link-abbreviated" href="mailto:amd-gfx@lists.freedesktop.org" moz-do-not-send="true">amd-gfx@lists.freedesktop.org</a> <a class="moz-txt-link-freetext" href="https://lists.freedesktop.org/mailman/listinfo/amd-gfx" moz-do-not-send="true">https://lists.freedesktop.org/mailman/listinfo/amd-gfx</a> </pre> </blockquote> </blockquote> </blockquote> </blockquote> <fieldset class="mimeAttachmentHeader"></fieldset> <pre wrap="">_______________________________________________ amd-gfx mailing list <a class="moz-txt-link-abbreviated" href="mailto:amd-gfx@lists.freedesktop.org">amd-gfx@lists.freedesktop.org</a> <a class="moz-txt-link-freetext" href="https://lists.freedesktop.org/mailman/listinfo/amd-gfx">https://lists.freedesktop.org/mailman/listinfo/amd-gfx</a> </pre> </blockquote> </body> </html>