From 8e94a46c1770884166b31adc99eba7da65a446a7 Mon Sep 17 00:00:00 2001 From: Mario Kleiner Date: Wed, 9 Nov 2016 02:25:15 +0100 Subject: drm/amdgpu: Attach exclusive fence to prime exported bo's. (v5) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit External clients which import our bo's wait only for exclusive dmabuf-fences, not on shared ones, ditto for bo's which we import from external providers and write to. Therefore attach exclusive fences on prime shared buffers if our exported buffer gets imported by an external client, or if we import a buffer from an external exporter. See discussion in thread: https://lists.freedesktop.org/archives/dri-devel/2016-October/122370.html Prime export tested on Intel iGPU + AMD Tonga dGPU as DRI3/Present Prime render offload, and with the Tonga standalone as primary gpu. v2: Add a wait for all shared fences before prime export, as suggested by Christian Koenig. v3: - Mark buffer prime_exported in amdgpu_gem_prime_pin, so we only use the exclusive fence when exporting a bo to external clients like a separate iGPU, but not when exporting/importing from/to ourselves as part of regular DRI3 fd passing. - Propagate failure of reservation_object_wait_rcu back to caller. v4: - Switch to a prime_shared_count counter instead of a flag, which gets in/decremented on prime_pin/unpin, so we can switch back to shared fences if all clients detach from our exported bo. - Also switch to exclusive fence for prime imported bo's. v5: - Drop lret, instead use int ret -> long ret, as proposed by Christian. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=95472 Tested-by: Mike Lothian (v1) Signed-off-by: Mario Kleiner Reviewed-by: Christian König . Cc: Christian König Cc: Michel Dänzer Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c | 20 +++++++++++++++++++- 3 files changed, 21 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 039b57e4644c..496f72b134eb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -459,6 +459,7 @@ struct amdgpu_bo { u64 metadata_flags; void *metadata; u32 metadata_size; + unsigned prime_shared_count; /* list of all virtual address to which this bo * is associated to */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c index 651115dcce12..c02db01f6583 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c @@ -132,7 +132,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev, entry->priority = min(info[i].bo_priority, AMDGPU_BO_LIST_MAX_PRIORITY); entry->tv.bo = &entry->robj->tbo; - entry->tv.shared = true; + entry->tv.shared = !entry->robj->prime_shared_count; if (entry->robj->prefered_domains == AMDGPU_GEM_DOMAIN_GDS) gds_obj = entry->robj; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c index 7700dc22f243..3826d5aea0a6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c @@ -74,20 +74,36 @@ amdgpu_gem_prime_import_sg_table(struct drm_device *dev, if (ret) return ERR_PTR(ret); + bo->prime_shared_count = 1; return &bo->gem_base; } int amdgpu_gem_prime_pin(struct drm_gem_object *obj) { struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); - int ret = 0; + long ret = 0; ret = amdgpu_bo_reserve(bo, false); if (unlikely(ret != 0)) return ret; + /* + * Wait for all shared fences to complete before we switch to future + * use of exclusive fence on this prime shared bo. + */ + ret = reservation_object_wait_timeout_rcu(bo->tbo.resv, true, false, + MAX_SCHEDULE_TIMEOUT); + if (unlikely(ret < 0)) { + DRM_DEBUG_PRIME("Fence wait failed: %li\n", ret); + amdgpu_bo_unreserve(bo); + return ret; + } + /* pin buffer into GTT */ ret = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL); + if (likely(ret == 0)) + bo->prime_shared_count++; + amdgpu_bo_unreserve(bo); return ret; } @@ -102,6 +118,8 @@ void amdgpu_gem_prime_unpin(struct drm_gem_object *obj) return; amdgpu_bo_unpin(bo); + if (bo->prime_shared_count) + bo->prime_shared_count--; amdgpu_bo_unreserve(bo); } -- cgit v1.2.3 From cb434658a8ff151c221a9ac1d44fb6788100cd0d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 15 Nov 2016 11:39:08 -0500 Subject: drm/amdgpu/powerplay: drop a redundant NULL check Left over from an earlier rev of the patch. Acked-by: Colin Ian King Cc: Dan Carpenter Cc: Colin King Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c index b0c929dd8beb..13f2b705ea49 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c @@ -1469,8 +1469,6 @@ static int smu7_get_evv_voltages(struct pp_hwmgr *hwmgr) table_info->vddgfx_lookup_table, vv_id, &sclk)) { if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_ClockStretcher)) { - if (table_info == NULL) - return -EINVAL; sclk_table = table_info->vdd_dep_on_sclk; for (j = 1; j < sclk_table->count; j++) { -- cgit v1.2.3 From 1da2c326e43b0834105993d13610647337bbad67 Mon Sep 17 00:00:00 2001 From: Monk Liu Date: Fri, 11 Nov 2016 11:24:29 +0800 Subject: drm/amdgpu:fix vpost_needed routine 1,cleanup description/comments 2,for FIJI & passthrough, force post when smc fw version below 22.15 3,for other cases, follow regular rules Signed-off-by: Monk Liu Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 27 +++++++-------------------- 1 file changed, 7 insertions(+), 20 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 7ca07e7b25c1..3161d77bf299 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -658,12 +658,10 @@ static bool amdgpu_vpost_needed(struct amdgpu_device *adev) return false; if (amdgpu_passthrough(adev)) { - /* for FIJI: In whole GPU pass-through virtualization case - * old smc fw won't clear some registers (e.g. MEM_SIZE, BIOS_SCRATCH) - * so amdgpu_card_posted return false and driver will incorrectly skip vPost. - * but if we force vPost do in pass-through case, the driver reload will hang. - * whether doing vPost depends on amdgpu_card_posted if smc version is above - * 00160e00 for FIJI. + /* for FIJI: In whole GPU pass-through virtualization case, after VM reboot + * some old smc fw still need driver do vPost otherwise gpu hang, while + * those smc fw version above 22.15 doesn't have this flaw, so we force + * vpost executed for smc version below 22.15 */ if (adev->asic_type == CHIP_FIJI) { int err; @@ -674,22 +672,11 @@ static bool amdgpu_vpost_needed(struct amdgpu_device *adev) return true; fw_ver = *((uint32_t *)adev->pm.fw->data + 69); - if (fw_ver >= 0x00160e00) - return !amdgpu_card_posted(adev); + if (fw_ver < 0x00160e00) + return true; } - } else { - /* in bare-metal case, amdgpu_card_posted return false - * after system reboot/boot, and return true if driver - * reloaded. - * we shouldn't do vPost after driver reload otherwise GPU - * could hang. - */ - if (amdgpu_card_posted(adev)) - return false; } - - /* we assume vPost is neede for all other cases */ - return true; + return !amdgpu_card_posted(adev); } /** -- cgit v1.2.3