drm/amdgpu: Make amdgpu_bo_reserve use uninterruptible waits for cleanup
A message that is not in the archive (sent off-list or not yet received): [email protected]
A message that is not in the archive (sent off-list or not yet received): mailman.1551.1493369856.21890.amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
Xie, AlexBin
Reviewed-by Christian KönigAm 28.04.2017 um 10:47 schrieb Michel Dänzer:
Quoted 7 lines
From: Michel Dänzer <[email protected]>
Some of these paths probably cannot be interrupted by a signal anyway. Those that can would fail to clean up things if they actually got interrupted.
Signed-off-by Michel Dänzer <[email protected]>
Alex X: Just a reminder: amdgpu_unpin_work_func is called by work queue. Signal is blocked already. un-interruptible waiting might slow thing down very slightly.
Quoted 399 lines
--- drivers/gpu/drm/amd/amdgpu/amdgpu_display.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 2 +- drivers/gpu/drm/amd/amdgpu/dce_v10_0.c | 6 +++--- drivers/gpu/drm/amd/amdgpu/dce_v11_0.c | 6 +++--- drivers/gpu/drm/amd/amdgpu/dce_v6_0.c | 6 +++--- drivers/gpu/drm/amd/amdgpu/dce_v8_0.c | 6 +++--- drivers/gpu/drm/amd/amdgpu/dce_virtual.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c | 6 +++--- drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c | 10 +++++----- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 8 ++++---- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 8 ++++---- 16 files changed, 36 insertions(+), 36 deletions(-)drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +1
-1@@ -123,7 +123,7 @@ static void amdgpu_unpin_work_func(struct work_struct *__work)int r;/* unpin of the old buffer */ - r =amdgpu_bo_reserve (work->old_abo, false); + r =amdgpu_bo_reserve (work->old_abo, true);if (likely (r ==0 )) { r =amdgpu_bo_unpin (work->old_abo);if (unlikely (r !=0 )) {drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c +1
-1@@ -113,7 +113,7 @@ static void amdgpufb_destroy_pinned_object(struct drm_gem_object *gobj)struct amdgpu_bo *abo =gem_to_amdgpu_bo (gobj);int ret; - ret =amdgpu_bo_reserve (abo, false); + ret =amdgpu_bo_reserve (abo, true);if (likely (ret ==0 )) {amdgpu_bo_kunmap (abo);amdgpu_bo_unpin (abo);drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c +1
-1@@ -183,7 +183,7 @@ void amdgpu_gart_table_vram_unpin(struct amdgpu_device *adev)if (adev->gart.robj ==NULL ) {return ; } - r =amdgpu_bo_reserve (adev->gart.robj, false); + r =amdgpu_bo_reserve (adev->gart.robj, true);if (likely (r ==0 )) {amdgpu_bo_kunmap (adev->gart.robj);amdgpu_bo_unpin (adev->gart.robj);drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +1
-1@@ -819,7 +819,7 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,if (amdgpu_sriov_vf (adev)) {/* TODO: how to handle reserve failure */ -BUG_ON (amdgpu_bo_reserve (adev->virt.csa_obj, false)); +BUG_ON (amdgpu_bo_reserve (adev->virt.csa_obj, true));amdgpu_vm_bo_rmv (adev, fpriv->vm.csa_bo_va); fpriv->vm.csa_bo_va =NULL ;amdgpu_bo_unreserve (adev->virt.csa_obj);drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c +1
-1@@ -113,7 +113,7 @@ void amdgpu_gem_prime_unpin(struct drm_gem_object *obj)struct amdgpu_bo *bo =gem_to_amdgpu_bo (obj);int ret =0 ; - ret =amdgpu_bo_reserve (bo, false); + ret =amdgpu_bo_reserve (bo, true);if (unlikely (ret !=0 ))return ;drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c +1
-1@@ -130,7 +130,7 @@ int amdgpu_sa_bo_manager_suspend(struct amdgpu_device *adev,return -EINVAL ; } - r =amdgpu_bo_reserve (sa_manager->bo, false); + r =amdgpu_bo_reserve (sa_manager->bo, true);if (!r) {amdgpu_bo_kunmap (sa_manager->bo);amdgpu_bo_unpin (sa_manager->bo);drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +1
-1@@ -1198,7 +1198,7 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev)return ;amdgpu_ttm_debugfs_fini (adev);if (adev->stollen_vga_memory) { - r =amdgpu_bo_reserve (adev->stollen_vga_memory, false); + r =amdgpu_bo_reserve (adev->stollen_vga_memory, true);if (r ==0 ) {amdgpu_bo_unpin (adev->stollen_vga_memory);amdgpu_bo_unreserve (adev->stollen_vga_memory);drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +3
-3@@ -2231,7 +2231,7 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc,if (!atomic && fb && fb != crtc->primary->fb) { amdgpu_fb =to_amdgpu_framebuffer (fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r !=0 ))return r;amdgpu_bo_unpin (abo);@@ -2590,7 +2590,7 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc, unpin:if (amdgpu_crtc->cursor_bo) {struct amdgpu_bo *aobj =gem_to_amdgpu_bo (amdgpu_crtc->cursor_bo); - ret =amdgpu_bo_reserve (aobj, false); + ret =amdgpu_bo_reserve (aobj, true);if (likely (ret ==0 )) {amdgpu_bo_unpin (aobj);amdgpu_bo_unreserve (aobj);@@ -2720,7 +2720,7 @@ static void dce_v10_0_crtc_disable(struct drm_crtc *crtc) amdgpu_fb =to_amdgpu_framebuffer (crtc->primary->fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r))DRM_ERROR ("failed to reserve abo before unpin\n" );else {drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +3
-3@@ -2215,7 +2215,7 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc,if (!atomic && fb && fb != crtc->primary->fb) { amdgpu_fb =to_amdgpu_framebuffer (fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r !=0 ))return r;amdgpu_bo_unpin (abo);@@ -2610,7 +2610,7 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc, unpin:if (amdgpu_crtc->cursor_bo) {struct amdgpu_bo *aobj =gem_to_amdgpu_bo (amdgpu_crtc->cursor_bo); - ret =amdgpu_bo_reserve (aobj, false); + ret =amdgpu_bo_reserve (aobj, true);if (likely (ret ==0 )) {amdgpu_bo_unpin (aobj);amdgpu_bo_unreserve (aobj);@@ -2740,7 +2740,7 @@ static void dce_v11_0_crtc_disable(struct drm_crtc *crtc) amdgpu_fb =to_amdgpu_framebuffer (crtc->primary->fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r))DRM_ERROR ("failed to reserve abo before unpin\n" );else {drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +3
-3@@ -2150,7 +2150,7 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc,if (!atomic && fb && fb != crtc->primary->fb) { amdgpu_fb =to_amdgpu_framebuffer (fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r !=0 ))return r;amdgpu_bo_unpin (abo);@@ -2467,7 +2467,7 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc, unpin:if (amdgpu_crtc->cursor_bo) {struct amdgpu_bo *aobj =gem_to_amdgpu_bo (amdgpu_crtc->cursor_bo); - ret =amdgpu_bo_reserve (aobj, false); + ret =amdgpu_bo_reserve (aobj, true);if (likely (ret ==0 )) {amdgpu_bo_unpin (aobj);amdgpu_bo_unreserve (aobj);@@ -2592,7 +2592,7 @@ static void dce_v6_0_crtc_disable(struct drm_crtc *crtc) amdgpu_fb =to_amdgpu_framebuffer (crtc->primary->fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r))DRM_ERROR ("failed to reserve abo before unpin\n" );else {drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +3
-3@@ -2090,7 +2090,7 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc,if (!atomic && fb && fb != crtc->primary->fb) { amdgpu_fb =to_amdgpu_framebuffer (fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r !=0 ))return r;amdgpu_bo_unpin (abo);@@ -2441,7 +2441,7 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc, unpin:if (amdgpu_crtc->cursor_bo) {struct amdgpu_bo *aobj =gem_to_amdgpu_bo (amdgpu_crtc->cursor_bo); - ret =amdgpu_bo_reserve (aobj, false); + ret =amdgpu_bo_reserve (aobj, true);if (likely (ret ==0 )) {amdgpu_bo_unpin (aobj);amdgpu_bo_unreserve (aobj);@@ -2571,7 +2571,7 @@ static void dce_v8_0_crtc_disable(struct drm_crtc *crtc) amdgpu_fb =to_amdgpu_framebuffer (crtc->primary->fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r))DRM_ERROR ("failed to reserve abo before unpin\n" );else {drivers/gpu/drm/amd/amdgpu/dce_virtual.c +1
-1@@ -247,7 +247,7 @@ static void dce_virtual_crtc_disable(struct drm_crtc *crtc) amdgpu_fb =to_amdgpu_framebuffer (crtc->primary->fb); abo =gem_to_amdgpu_bo (amdgpu_fb->obj); - r =amdgpu_bo_reserve (abo, false); + r =amdgpu_bo_reserve (abo, true);if (unlikely (r))DRM_ERROR ("failed to reserve abo before unpin\n" );else {drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +3
-3@@ -2437,7 +2437,7 @@ static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev)int r;if (adev->gfx.rlc.save_restore_obj) { - r =amdgpu_bo_reserve (adev->gfx.rlc.save_restore_obj, false); + r =amdgpu_bo_reserve (adev->gfx.rlc.save_restore_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve RLC sr bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.rlc.save_restore_obj);@@ -2448,7 +2448,7 @@ static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev) }if (adev->gfx.rlc.clear_state_obj) { - r =amdgpu_bo_reserve (adev->gfx.rlc.clear_state_obj, false); + r =amdgpu_bo_reserve (adev->gfx.rlc.clear_state_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve RLC c bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.rlc.clear_state_obj);@@ -2459,7 +2459,7 @@ static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev) }if (adev->gfx.rlc.cp_table_obj) { - r =amdgpu_bo_reserve (adev->gfx.rlc.cp_table_obj, false); + r =amdgpu_bo_reserve (adev->gfx.rlc.cp_table_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve RLC cp table bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.rlc.cp_table_obj);drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +5
-5@@ -2792,7 +2792,7 @@ static void gfx_v7_0_cp_compute_fini(struct amdgpu_device *adev)struct amdgpu_ring *ring = &adev->gfx.compute_ring[i];if (ring->mqd_obj) { - r =amdgpu_bo_reserve (ring->mqd_obj, false); + r =amdgpu_bo_reserve (ring->mqd_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve MQD bo failed\n" , r);@@ -2810,7 +2810,7 @@ static void gfx_v7_0_mec_fini(struct amdgpu_device *adev)int r;if (adev->gfx.mec.hpd_eop_obj) { - r =amdgpu_bo_reserve (adev->gfx.mec.hpd_eop_obj, false); + r =amdgpu_bo_reserve (adev->gfx.mec.hpd_eop_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve HPD EOP bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.mec.hpd_eop_obj);@@ -3359,7 +3359,7 @@ static void gfx_v7_0_rlc_fini(struct amdgpu_device *adev)/* save restore block */ if (adev->gfx.rlc.save_restore_obj) { - r =amdgpu_bo_reserve (adev->gfx.rlc.save_restore_obj, false); + r =amdgpu_bo_reserve (adev->gfx.rlc.save_restore_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve RLC sr bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.rlc.save_restore_obj);@@ -3371,7 +3371,7 @@ static void gfx_v7_0_rlc_fini(struct amdgpu_device *adev)/* clear state block */ if (adev->gfx.rlc.clear_state_obj) { - r =amdgpu_bo_reserve (adev->gfx.rlc.clear_state_obj, false); + r =amdgpu_bo_reserve (adev->gfx.rlc.clear_state_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve RLC c bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.rlc.clear_state_obj);@@ -3383,7 +3383,7 @@ static void gfx_v7_0_rlc_fini(struct amdgpu_device *adev)/* clear state block */ if (adev->gfx.rlc.cp_table_obj) { - r =amdgpu_bo_reserve (adev->gfx.rlc.cp_table_obj, false); + r =amdgpu_bo_reserve (adev->gfx.rlc.cp_table_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve RLC cp table bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.rlc.cp_table_obj);drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +4
-4@@ -1244,7 +1244,7 @@ static void gfx_v8_0_rlc_fini(struct amdgpu_device *adev)/* clear state block */ if (adev->gfx.rlc.clear_state_obj) { - r =amdgpu_bo_reserve (adev->gfx.rlc.clear_state_obj, false); + r =amdgpu_bo_reserve (adev->gfx.rlc.clear_state_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve RLC cbs bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.rlc.clear_state_obj);@@ -1255,7 +1255,7 @@ static void gfx_v8_0_rlc_fini(struct amdgpu_device *adev)/* jump table block */ if (adev->gfx.rlc.cp_table_obj) { - r =amdgpu_bo_reserve (adev->gfx.rlc.cp_table_obj, false); + r =amdgpu_bo_reserve (adev->gfx.rlc.cp_table_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve RLC cp table bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.rlc.cp_table_obj);@@ -1368,7 +1368,7 @@ static void gfx_v8_0_mec_fini(struct amdgpu_device *adev)int r;if (adev->gfx.mec.hpd_eop_obj) { - r =amdgpu_bo_reserve (adev->gfx.mec.hpd_eop_obj, false); + r =amdgpu_bo_reserve (adev->gfx.mec.hpd_eop_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve HPD EOP bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.mec.hpd_eop_obj);@@ -1496,7 +1496,7 @@ static int gfx_v8_0_kiq_init(struct amdgpu_device *adev)memset (hpd,0 ,MEC_HPD_SIZE ); - r =amdgpu_bo_reserve (kiq->eop_obj, false); + r =amdgpu_bo_reserve (kiq->eop_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve kiq eop bo failed\n" , r);amdgpu_bo_kunmap (kiq->eop_obj);drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +3
-3@@ -453,7 +453,7 @@ static void gfx_v9_0_mec_fini(struct amdgpu_device *adev)int r;if (adev->gfx.mec.hpd_eop_obj) { - r =amdgpu_bo_reserve (adev->gfx.mec.hpd_eop_obj, false); + r =amdgpu_bo_reserve (adev->gfx.mec.hpd_eop_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve HPD EOP bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.mec.hpd_eop_obj);@@ -463,7 +463,7 @@ static void gfx_v9_0_mec_fini(struct amdgpu_device *adev) adev->gfx.mec.hpd_eop_obj =NULL ; }if (adev->gfx.mec.mec_fw_obj) { - r =amdgpu_bo_reserve (adev->gfx.mec.mec_fw_obj, false); + r =amdgpu_bo_reserve (adev->gfx.mec.mec_fw_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve mec firmware bo failed\n" , r);amdgpu_bo_unpin (adev->gfx.mec.mec_fw_obj);@@ -599,7 +599,7 @@ static int gfx_v9_0_kiq_init(struct amdgpu_device *adev)memset (hpd,0 ,MEC_HPD_SIZE ); - r =amdgpu_bo_reserve (kiq->eop_obj, false); + r =amdgpu_bo_reserve (kiq->eop_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve kiq eop bo failed\n" , r);amdgpu_bo_kunmap (kiq->eop_obj);
drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c static void gfx_v9_0_cp_compute_fini(struct amdgpu_device *adev) · line 1792
Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +1
-1@@ -1786,7 +1786,7 @@ static void gfx_v9_0_cp_compute_fini(struct amdgpu_device *adev)struct amdgpu_ring *ring = &adev->gfx.compute_ring[i];if (ring->mqd_obj) { - r =amdgpu_bo_reserve (ring->mqd_obj, false); + r =amdgpu_bo_reserve (ring->mqd_obj, true);if (unlikely (r !=0 ))dev_warn (adev->dev,"(%d) reserve MQD bo failed\n" , r);
------------------------------
Subject: Digest Footer
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
amd-gfx Info Page - freedesktop.org<https://lists.freedesktop.org/mailman/listinfo/amd-gfx>
lists.freedesktop.org
Subscribing to amd-gfx: Subscribe to amd-gfx by filling out the following form. Use of all freedesktop.org lists is subject to our Code of ...
------------------------------
End of amd-gfx Digest, Vol 11, Issue 407
****************************************
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): MWHPR1201MB0045C097101B0EDA13F28BD9F2130-3iK1xFAIwjq45V35fqe+hGrFom/aUZj6nBOFsp37pqbUKgpGm//BTAC/[email protected]
Quoted 7 lines · AlexBin
Quoted 7 lines
Am 28.04.2017 um 10:47 schrieb Michel Dänzer:
Quoted 6 lines
From: Michel Dänzer <[email protected]>
Some of these paths probably cannot be interrupted by a signal anyway. Those that can would fail to clean up things if they actually got interrupted.
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Signed-off-by Michel Dänzer <[email protected]>Reviewed-by Christian König <[email protected]>Alex X: Just a reminder:
amdgpu_unpin_work_funcis called by work queue. Signal is blocked already. un-interruptible waiting might slow thing down very slightly.
How so?
Signature · 6 lines
Earthling Michel Dänzer | http://www.amd.com Libre software enthusiast | Mesa and X developer _______________________________________________ amd-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): 73b82af4-7903-512e-fe61-f05ebf18381f-otUistvHUpPR7s880joybQ@public.gmane.org
Quoted 7 lines · AlexBin
Quoted 7 lines
Am 28.04.2017 um 10:47 schrieb Michel Dänzer:
Quoted 6 lines
From: Michel Dänzer <[email protected]>
Some of these paths probably cannot be interrupted by a signal anyway. Those that can would fail to clean up things if they actually got interrupted.
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Signed-off-by Michel Dänzer <[email protected]>Reviewed-by Christian König <[email protected]>Alex X: Just a reminder:
amdgpu_unpin_work_funcis called by work queue. Signal is blocked already. un-interruptible waiting might slow thing down very slightly.
How so?
Alex X: I said "might". If you think it is not slower, it is fine for me.
My real concern is that the signals are blocked for work queue already. We don't need this change to avoid unnecessary risk. In theory, I am not against this change.
--
Earthling Michel Dänzer | http://www.amd.com
Libre software enthusiast | Mesa and X developer
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): MWHPR1201MB004532EF4E87E5503BEAC217F2140-3iK1xFAIwjq45V35fqe+hGrFom/aUZj6nBOFsp37pqbUKgpGm//BTAC/[email protected]
Am 01.05.2017 um 02:13 schrieb Xie, AlexBin:
Quoted 18 lines · AlexBin
Quoted 8 lines
Quoted 8 lines
Am 28.04.2017 um 10:47 schrieb Michel Dänzer:
Quoted 6 lines
From: Michel Dänzer <[email protected]>
Some of these paths probably cannot be interrupted by a signal anyway. Those that can would fail to clean up things if they actually got interrupted.
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Signed-off-by Michel Dänzer <[email protected]>Reviewed-by Christian König <[email protected]>Alex X: Just a reminder:
amdgpu_unpin_work_funcis called by work queue. Signal is blocked already. un-interruptible waiting might slow thing down very slightly.
Quoted 5 lines
How so?
Alex X: I said "might". If you think it is not slower, it is fine for me.
My real concern is that the signals are blocked for work queue already. We don't need this change to avoid unnecessary risk. In theory, I am not against this change.
I have to agree with that. For waits from work queues it is usually best to implicitly use un-interruptible waits for documentation purposes.
In other words even when we are sure that the code can't receive a signal we should write it conservatively.
Christian.
Quoted 10 lines
Signature · 9 lines
-- Earthling Michel Dänzer | http://www.amd.com <http://www.amd.com> Libre software enthusiast | Mesa and X developer _______________________________________________ amd-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/amd-gfx
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): 064b2ea6-1b57-0c4d-0818-6eb017887e55-ANTagKRnAhcb1SvskN2V4Q@public.gmane.org
I just realised that there are some stress test that sends huge amount of signals. I used those tests in the pass 2 years...
-Alex Bin
________________________________
From: Christian König <[email protected]>
Sent: Monday, May 1, 2017 10:28 AM
To: Xie, AlexBin; Michel Dänzer
Cc: [email protected]
Subject: Re: [PATCH] drm/amdgpu: Make amdgpu_bo_reserve use uninterruptible waits for cleanup
Am 01.05.2017 um 02:13 schrieb Xie, AlexBin:
Quoted 7 lines · AlexBin
Quoted 7 lines
Am 28.04.2017 um 10:47 schrieb Michel Dänzer:
Quoted 6 lines
From: Michel Dänzer <[email protected]><mailto:[email protected]>
Some of these paths probably cannot be interrupted by a signal anyway. Those that can would fail to clean up things if they actually got interrupted.
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Alex X: Just a reminder:
amdgpu_unpin_work_funcis called by work queue. Signal is blocked already. un-interruptible waiting might slow thing down very slightly.
How so?
Alex X: I said "might". If you think it is not slower, it is fine for me.
My real concern is that the signals are blocked for work queue already. We don't need this change to avoid unnecessary risk. In theory, I am not against this change.
I have to agree with that. For waits from work queues it is usually best to implicitly use un-interruptible waits for documentation purposes.
In other words even when we are sure that the code can't receive a signal we should write it conservatively.
Christian.
--
Earthling Michel Dänzer | http://www.amd.com
Libre software enthusiast | Mesa and X developer
_______________________________________________
amd-gfx mailing list [email protected]<mailto:[email protected]>
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx