dri-devel March 2014

dri-devel@lists.freedesktop.org

144 participants
926 discussions

[PULL] vmwgfx-fixes-3.14
by Thomas Hellstrom 02 Mar '14

02 Mar '14

Dave, A couple of minor fixes. The following changes since commit b0447888dc29f4fb91c3b3b02e3f1f0bfc6e1222: MAINTAINERS: update drm git tree entry (2014-02-27 14:49:55 +1000) are available in the git repository at: git://people.freedesktop.org/~thomash/linux tags/vmwgfx-fixes-3.14-2014-03-02 for you to fetch changes up to 6950e23e54b1c94abf8c60c1ddfac79133d41de2: drm/vmwgfx: avoid null pointer dereference at failure paths (2014-03-02 09:49:59 +0100) ---------------------------------------------------------------- Pull request of 2014-03-02 ---------------------------------------------------------------- Alexey Khoroshilov (1): drm/vmwgfx: avoid null pointer dereference at failure paths Thomas Hellstrom (2): drm/vmwgfx: Remove some unused surface formats drm/vmwgfx: Make sure backing mobs are cleared when allocated. Update driver date. drivers/gpu/drm/vmwgfx/svga3d_reg.h | 7 +----- drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 2 +- drivers/gpu/drm/vmwgfx/vmwgfx_mob.c | 35 ++++++++++++++++-------------- drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 3 +-- 4 files changed, 22 insertions(+), 25 deletions(-)

1 0

[PATCH 0/6] Radeon memory management improvements v3
by Marek Olšák 01 Mar '14

01 Mar '14

While updating "[PATCH 5/6] drm/radeon: validate relocations in the order determined by userspace" based on feedback, which should have been a harmless change, I discovered that the performance dropped. The problem was that list_add/move from one list to another reversed the list of relocations and the only thing which restored performance was to sort the relocation list with a stable sort. (preserve ordering of buffers whose priorities are equal). The only explanation I've come up with is that the buffers which appear sooner in the list are likely to be used more often by an IB than buffers which appear later in the list. For example, a texture used in all draw commands will most probably be somewhere at the beginning. However, a texture which appears at the end is probably only used in a few last draw commands. Also, the stats are changed to atomic64_t as discusssed before. Marek

1 6

3.14-rc1 ring/display regression in "low" power_profile (drm/radeon/pm: move pm handling into the asic specific code)
by Rafał Miłecki 01 Mar '14

01 Mar '14

01:00.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Blackcomb [Radeon HD 6970M/6990M] [1002:6720] I don't use dpm, but a profile method. Right after booting (without touching power_profile) I can suspend & resume as many times as I want (tested with ~50 s&r routines). Starting with: commit 6c7bccea390853bdec5b76fe31fc50f3b36f75d5 Author: Alex Deucher <alexander.deucher(a)amd.com> Date: Wed Dec 18 14:07:14 2013 -0500 drm/radeon/pm: move pm handling into the asic specific code We need more control over the ordering of dpm init with respect to the rest of the asic. Specifically, the SMC has to be initialized before the rlc and cg/pg. The pm code currently initializes late in the driver, but we need it to happen much earlier so move pm handling into the asic specific callbacks. This makes dpm more reliable and makes clockgating work properly on CIK parts and should help on SI parts as well. Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> I can't successfully resume using "low" power_profile. My testing looks like this: boot s & r (result: GOOD) s & r (result: GOOD) s & r (result: GOOD) echo "low" > /sys/class/drm/card0/device/power_profile s & r (result: BAD) BAD means display corruption [0] and: [ 80.244475] [drm:r600_ring_test] *ERROR* radeon: ring 0 test failed (scratch(0x8504)=0xCAFEDEAD) [ 80.244475] [drm:evergreen_resume] *ERROR* evergreen startup failed on resume [0] http://files.zajec.net/barts-low-resume-corruption.jpeg -- Rafał

2 3

[PATCH] drm/vmwgfx: avoid null pointer dereference at failure paths
by Alexey Khoroshilov 01 Mar '14

01 Mar '14

vmw_takedown_otable_base() and vmw_mob_unbind() check for potential vmw_fifo_reserve() failure and print error message, but then immediately dereference NULL pointer. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by: Alexey Khoroshilov <khoroshilov(a)ispras.ru> --- drivers/gpu/drm/vmwgfx/vmwgfx_mob.c | 35 +++++++++++++++++++---------------- 1 file changed, 19 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c index d4a5a19cb8c3..04a64b8cd3cd 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c @@ -188,18 +188,20 @@ static void vmw_takedown_otable_base(struct vmw_private *dev_priv, bo = otable->page_table->pt_bo; cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd)); - if (unlikely(cmd == NULL)) - DRM_ERROR("Failed reserving FIFO space for OTable setup.\n"); - - memset(cmd, 0, sizeof(*cmd)); - cmd->header.id = SVGA_3D_CMD_SET_OTABLE_BASE; - cmd->header.size = sizeof(cmd->body); - cmd->body.type = type; - cmd->body.baseAddress = 0; - cmd->body.sizeInBytes = 0; - cmd->body.validSizeInBytes = 0; - cmd->body.ptDepth = SVGA3D_MOBFMT_INVALID; - vmw_fifo_commit(dev_priv, sizeof(*cmd)); + if (unlikely(cmd == NULL)) { + DRM_ERROR("Failed reserving FIFO space for OTable " + "takedown.\n"); + } else { + memset(cmd, 0, sizeof(*cmd)); + cmd->header.id = SVGA_3D_CMD_SET_OTABLE_BASE; + cmd->header.size = sizeof(cmd->body); + cmd->body.type = type; + cmd->body.baseAddress = 0; + cmd->body.sizeInBytes = 0; + cmd->body.validSizeInBytes = 0; + cmd->body.ptDepth = SVGA3D_MOBFMT_INVALID; + vmw_fifo_commit(dev_priv, sizeof(*cmd)); + } if (bo) { int ret; @@ -562,11 +564,12 @@ void vmw_mob_unbind(struct vmw_private *dev_priv, if (unlikely(cmd == NULL)) { DRM_ERROR("Failed reserving FIFO space for Memory " "Object unbinding.\n"); + } else { + cmd->header.id = SVGA_3D_CMD_DESTROY_GB_MOB; + cmd->header.size = sizeof(cmd->body); + cmd->body.mobid = mob->id; + vmw_fifo_commit(dev_priv, sizeof(*cmd)); } - cmd->header.id = SVGA_3D_CMD_DESTROY_GB_MOB; - cmd->header.size = sizeof(cmd->body); - cmd->body.mobid = mob->id; - vmw_fifo_commit(dev_priv, sizeof(*cmd)); if (bo) { vmw_fence_single_bo(bo, NULL); ttm_bo_unreserve(bo); -- 1.8.3.2

2 1

[Bug 44499] New: r280 and xbmc - choppy menu and video playback
by bugzilla-daemon＠freedesktop.org 01 Mar '14

01 Mar '14

https://bugs.freedesktop.org/show_bug.cgi?id=44499 Bug #: 44499 Summary: r280 and xbmc - choppy menu and video playback Classification: Unclassified Product: Mesa Version: git Platform: x86 (IA32) OS/Version: Linux (All) Status: NEW Severity: normal Priority: medium Component: Drivers/DRI/r200 AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: smoki00790(a)gmail.com After mesa commit b2596c36c8f73e8bb7a0b1679b491662aeb2f9d9 i can experianced choppy/micro/freezes in video playback and also going thorugh menu in xbmc and other video players using GL based output. This can be also seen in many GL games menus like in ppracer/etracer, pingus (from git it now have GL option), etc. Have 9d21b5dd265959b28ab1e4103fb381074ec51a08 currently and same problem. DDX from current git, xorg 1.10.4, kernel 3.1.6. card:1002:5960:17af:2020 kms. Debian. Color tilling enabled and 75 fps at 1024x768 in UT4.1 with correct rendering 50 in OpenArena. Drawing tests in gtkperf are first slow later faster. -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 18

[PATCH] drm/radeon: fix VCE suspend/resume
by Christian König 01 Mar '14

01 Mar '14

From: Christian König <christian.koenig(a)amd.com> Signed-off-by: Christian König <christian.koenig(a)amd.com> --- drivers/gpu/drm/radeon/radeon.h | 1 - drivers/gpu/drm/radeon/radeon_vce.c | 63 ++++++++++++++++++++----------------- 2 files changed, 34 insertions(+), 30 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 08933cd..ff4f816 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -1637,7 +1637,6 @@ int radeon_uvd_send_upll_ctlreq(struct radeon_device *rdev, struct radeon_vce { struct radeon_bo *vcpu_bo; - void *cpu_addr; uint64_t gpu_addr; unsigned fw_version; unsigned fb_version; diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c index d130432..39ec7d8 100644 --- a/drivers/gpu/drm/radeon/radeon_vce.c +++ b/drivers/gpu/drm/radeon/radeon_vce.c @@ -119,7 +119,7 @@ int radeon_vce_init(struct radeon_device *rdev) if (rdev->vce.fw_version != ((40 << 24) | (2 << 16) | (2 << 8))) return -EINVAL; - /* load firmware into VRAM */ + /* allocate firmware, stack and heap BO */ size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size) + RADEON_VCE_STACK_SIZE + RADEON_VCE_HEAP_SIZE; @@ -130,16 +130,21 @@ int radeon_vce_init(struct radeon_device *rdev) return r; } - r = radeon_vce_resume(rdev); - if (r) + r = radeon_bo_reserve(rdev->vce.vcpu_bo, false); + if (r) { + radeon_bo_unref(&rdev->vce.vcpu_bo); + dev_err(rdev->dev, "(%d) failed to reserve VCE bo\n", r); return r; + } - memset(rdev->vce.cpu_addr, 0, size); - memcpy(rdev->vce.cpu_addr, rdev->vce_fw->data, rdev->vce_fw->size); - - r = radeon_vce_suspend(rdev); - if (r) + r = radeon_bo_pin(rdev->vce.vcpu_bo, RADEON_GEM_DOMAIN_VRAM, + &rdev->vce.gpu_addr); + radeon_bo_unreserve(rdev->vce.vcpu_bo); + if (r) { + radeon_bo_unref(&rdev->vce.vcpu_bo); + dev_err(rdev->dev, "(%d) VCE bo pin failed\n", r); return r; + } for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) { atomic_set(&rdev->vce.handles[i], 0); @@ -158,8 +163,12 @@ int radeon_vce_init(struct radeon_device *rdev) */ void radeon_vce_fini(struct radeon_device *rdev) { - radeon_vce_suspend(rdev); + if (rdev->vce.vcpu_bo == NULL) + return; + radeon_bo_unref(&rdev->vce.vcpu_bo); + + release_firmware(rdev->vce_fw); } /** @@ -167,22 +176,23 @@ void radeon_vce_fini(struct radeon_device *rdev) * * @rdev: radeon_device pointer * - * TODO: Test VCE suspend/resume */ int radeon_vce_suspend(struct radeon_device *rdev) { - int r; + int i; if (rdev->vce.vcpu_bo == NULL) return 0; - r = radeon_bo_reserve(rdev->vce.vcpu_bo, false); - if (!r) { - radeon_bo_kunmap(rdev->vce.vcpu_bo); - radeon_bo_unpin(rdev->vce.vcpu_bo); - radeon_bo_unreserve(rdev->vce.vcpu_bo); - } - return r; + for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) + if (atomic_read(&rdev->vce.handles[i])) + break; + + if (i == RADEON_MAX_VCE_HANDLES) + return 0; + + /* TODO: suspending running encoding sessions isn't supported */ + return -EINVAL; } /** @@ -190,10 +200,10 @@ int radeon_vce_suspend(struct radeon_device *rdev) * * @rdev: radeon_device pointer * - * TODO: Test VCE suspend/resume */ int radeon_vce_resume(struct radeon_device *rdev) { + void *cpu_addr; int r; if (rdev->vce.vcpu_bo == NULL) @@ -201,26 +211,21 @@ int radeon_vce_resume(struct radeon_device *rdev) r = radeon_bo_reserve(rdev->vce.vcpu_bo, false); if (r) { - radeon_bo_unref(&rdev->vce.vcpu_bo); dev_err(rdev->dev, "(%d) failed to reserve VCE bo\n", r); return r; } - r = radeon_bo_pin(rdev->vce.vcpu_bo, RADEON_GEM_DOMAIN_VRAM, - &rdev->vce.gpu_addr); + r = radeon_bo_kmap(rdev->vce.vcpu_bo, &cpu_addr); if (r) { radeon_bo_unreserve(rdev->vce.vcpu_bo); - radeon_bo_unref(&rdev->vce.vcpu_bo); - dev_err(rdev->dev, "(%d) VCE bo pin failed\n", r); - return r; - } - - r = radeon_bo_kmap(rdev->vce.vcpu_bo, &rdev->vce.cpu_addr); - if (r) { dev_err(rdev->dev, "(%d) VCE map failed\n", r); return r; } + memcpy(cpu_addr, rdev->vce_fw->data, rdev->vce_fw->size); + + radeon_bo_kunmap(rdev->vce.vcpu_bo); + radeon_bo_unreserve(rdev->vce.vcpu_bo); return 0; -- 1.8.3.2

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel March 2014