dri-devel September 2013

dri-devel@lists.freedesktop.org

153 participants
824 discussions

[PATCH] drm/exynos: fix deadlock issue at mmap
by Inki Dae 24 Sep '13

24 Sep '13

This patch fixes the deadlock issue when another process requested mmap system call. This issue can incur the deadlock if another process sharing the same file called mmap system call between ->f_op pointer chaning and restoring. So this patch calls down_write(&mm->mmap_sem) before do_mmap_pgoff call and then up_write(&mm->mmap_sem) after do_mmap_pg_off call so that when another process called mmap system call, the process can wait for up_write() call until the ->f_op pointer is restored. Signed-off-by: Inki Dae <inki.dae(a)samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park(a)samsung.com> --- drivers/gpu/drm/exynos/exynos_drm_gem.c | 20 ++++++++++++++++++-- 1 file changed, 18 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c index 49f9cd2..779c2d7 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_gem.c +++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c @@ -13,6 +13,7 @@ #include <drm/drm_vma_manager.h> #include <linux/shmem_fs.h> +#include <linux/security.h> #include <drm/exynos_drm.h> #include "exynos_drm_drv.h" @@ -420,6 +421,7 @@ int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data, { struct drm_exynos_gem_mmap *args = data; struct drm_gem_object *obj; + struct mm_struct *mm = current->mm; unsigned long addr; if (!(dev->driver->driver_features & DRIVER_GEM)) { @@ -433,6 +435,8 @@ int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data, return -EINVAL; } + down_write(&mm->mmap_sem); + /* * We have to use gem object and its fops for specific mmaper, * but vm_mmap() can deliver only filp. So we have to change @@ -457,8 +461,17 @@ int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data, */ file_priv->filp->private_data = obj; - addr = vm_mmap(file_priv->filp, 0, args->size, - PROT_READ | PROT_WRITE, MAP_SHARED, 0); + addr = security_mmap_file(file_priv->filp, PROT_READ | PROT_WRITE, + MAP_SHARED); + if (!addr) { + unsigned long populate; + + addr = do_mmap_pgoff(file_priv->filp, 0, args->size, + PROT_READ | PROT_WRITE, + MAP_SHARED, 0, &populate); + if (populate) + mm_populate(addr, populate); + } drm_gem_object_unreference(obj); @@ -469,10 +482,13 @@ int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data, file_priv->filp->private_data = file_priv; } mutex_unlock(&dev->struct_mutex); + up_write(&mm->mmap_sem); + return (int)addr; } mutex_unlock(&dev->struct_mutex); + up_write(&mm->mmap_sem); args->mapped = addr; -- 1.7.9.5

1 1

[PATCH 0/7] DRM: Remove gem_init_object() and friends
by David Herrmann 23 Sep '13

23 Sep '13

Hi This series removes any access to gem_obj->driver_private from all drivers and then drops drm_gem_object_alloc() with its ->gem_init_object() driver callback. All this is only needed if a gem object is not embedded into the driver's bo. However, all drivers (except for nouveau, see patch #6) embed the objects. No need to allow dynamic allocations, anymore. Patches #1 to #5 are trivial and only remove dead code. Patch #6 converts nouveau to embed gem objects into nouveau_bo. I tested this on my nv50 and it runs fine. Patch #7 drops gem_init_object() from all drivers as it is unused and the callbacks are all empty (or BUG()). And the diff-stat looks pretty nice, too. Cheers David David Herrmann (7): drm/ast: remove unused driver_private access drm/mgag200: remove unused driver_private access drm/cirrus: remove unused driver_private access drm/qxl: remove unused object_pin/unpin() helpers drm/radeon: remove stale gem->driver_private access drm/nouveau: embed gem object in nouveau_bo drm: kill ->gem_init_object() and friends drivers/gpu/drm/ast/ast_drv.c | 1 - drivers/gpu/drm/ast/ast_drv.h | 1 - drivers/gpu/drm/ast/ast_main.c | 6 ----- drivers/gpu/drm/ast/ast_ttm.c | 1 - drivers/gpu/drm/cirrus/cirrus_drv.c | 1 - drivers/gpu/drm/cirrus/cirrus_drv.h | 1 - drivers/gpu/drm/cirrus/cirrus_main.c | 6 ----- drivers/gpu/drm/cirrus/cirrus_ttm.c | 1 - drivers/gpu/drm/drm_gem.c | 29 --------------------- drivers/gpu/drm/exynos/exynos_drm_drv.c | 1 - drivers/gpu/drm/exynos/exynos_drm_gem.c | 5 ---- drivers/gpu/drm/exynos/exynos_drm_gem.h | 3 --- drivers/gpu/drm/gma500/gem.c | 5 ---- drivers/gpu/drm/gma500/psb_drv.c | 1 - drivers/gpu/drm/gma500/psb_drv.h | 1 - drivers/gpu/drm/i915/i915_drv.c | 1 - drivers/gpu/drm/i915/i915_drv.h | 1 - drivers/gpu/drm/i915/i915_gem.c | 7 ------ drivers/gpu/drm/mgag200/mgag200_drv.c | 1 - drivers/gpu/drm/mgag200/mgag200_drv.h | 1 - drivers/gpu/drm/mgag200/mgag200_main.c | 6 ----- drivers/gpu/drm/mgag200/mgag200_ttm.c | 1 - drivers/gpu/drm/nouveau/nouveau_abi16.c | 4 +-- drivers/gpu/drm/nouveau/nouveau_bo.c | 2 +- drivers/gpu/drm/nouveau/nouveau_bo.h | 5 +++- drivers/gpu/drm/nouveau/nouveau_display.c | 10 ++++---- drivers/gpu/drm/nouveau/nouveau_drm.c | 1 - drivers/gpu/drm/nouveau/nouveau_fbcon.c | 2 +- drivers/gpu/drm/nouveau/nouveau_gem.c | 42 +++++++++++++------------------ drivers/gpu/drm/nouveau/nouveau_gem.h | 3 +-- drivers/gpu/drm/nouveau/nouveau_prime.c | 10 +++++--- drivers/gpu/drm/omapdrm/omap_drv.c | 1 - drivers/gpu/drm/omapdrm/omap_drv.h | 1 - drivers/gpu/drm/omapdrm/omap_gem.c | 5 ---- drivers/gpu/drm/qxl/qxl_drv.c | 1 - drivers/gpu/drm/qxl/qxl_drv.h | 4 --- drivers/gpu/drm/qxl/qxl_gem.c | 32 ----------------------- drivers/gpu/drm/qxl/qxl_object.c | 1 - drivers/gpu/drm/radeon/radeon_drv.c | 2 -- drivers/gpu/drm/radeon/radeon_gem.c | 7 ------ drivers/gpu/drm/radeon/radeon_object.c | 1 - drivers/gpu/drm/radeon/radeon_prime.c | 1 - drivers/gpu/drm/udl/udl_drv.c | 1 - drivers/gpu/drm/udl/udl_drv.h | 1 - drivers/gpu/drm/udl/udl_gem.c | 7 ------ include/drm/drmP.h | 5 ---- 46 files changed, 37 insertions(+), 193 deletions(-) -- 1.8.3.4

5 14

[Bug 69341] [radeonsi] KDE 4.11 is EXTREMELY slow with Raster QT backend
by bugzilla-daemon＠freedesktop.org 23 Sep '13

23 Sep '13

https://bugs.freedesktop.org/show_bug.cgi?id=69341 --- Comment #9 from darkbasic <darkbasic(a)linuxsystems.it> --- It goes to 100% with native backend, at least on SI. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 69341] [radeonsi] KDE 4.11 is EXTREMELY slow with Raster QT backend
by bugzilla-daemon＠freedesktop.org 23 Sep '13

23 Sep '13

https://bugs.freedesktop.org/show_bug.cgi?id=69341 --- Comment #8 from Denis M. (Phr33d0m) <god(a)politeia.in> --- (In reply to comment #7) > Yes that's the same corruption I have with the native backend. I suggest you > to have a look at top: native isn't slow, but aftware a few minutes the X > process starts eating the CPU ,that's the reason ;) In my case, X doesn't normally use much cpu (~10-15%) (well, really depends on your definition of 'much'), but there are moments (every 2-3 seconds) where it peaks at ~45-60% which causes everything to lag for a moment (for example if I move a window or even with glxgears)... And btw, thanks for the vblank tip, it helped a little. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 69341] [radeonsi] KDE 4.11 is EXTREMELY slow with Raster QT backend
by bugzilla-daemon＠freedesktop.org 23 Sep '13

23 Sep '13

https://bugs.freedesktop.org/show_bug.cgi?id=69341 --- Comment #7 from darkbasic <darkbasic(a)linuxsystems.it> --- Yes that's the same corruption I have with the native backend. I suggest you to have a look at top: native isn't slow, but aftware a few minutes the X process starts eating the CPU ,that's the reason ;) On the contrary Raster in EXTREMELY slow since the very first moment, but at least there is no corruption. Regarding your glxgears issue "vblank_mode=0 glxgears" will do the magic. -- You are receiving this mail because: You are the assignee for the bug.

1 0

HDMI stereo support v5
by Damien Lespiau 23 Sep '13

23 Sep '13

v4 was: http://lists.freedesktop.org/archives/dri-devel/2013-September/045340.html Changes from v4: - The kernel is now in charge of adjusting the stereo mode timings. - There is a per-connector opt-in boolean to expose stereo modes, letting people enable stereo for each driver/connector. - The series now depends on the latest bits from the drm-intel tree - Removed the 2 buffers check that was judged unnecessary - I realized that for frame packing modes (where you have to adjust the vdisplay to be 2 * vdisplay + vblank) with a NTSC clock I was reconstructing the clock without adjusting it by 1000 / 1001. Now that the timing computation is in the kernel, fixed it there. - I also checked that we correctly match against the 2D CEA mode for the AVI infoframes. The last bit to make it work was the introduction of crtc_clock to separate the original clock from the one computed by drm_mode_set_crtcinfo() - And finally, I booted with i915.fastboot=1 and discovered that the pipe_src register was getting clobbered, so fixed that as well

5 28

[PATCH] drm/radeon/uvd: lower msg&fb buffer requirements on UVD3
by Christian König 23 Sep '13

23 Sep '13

From: Christian König <christian.koenig(a)amd.com> Starting with UVD3 message and feedback buffers have their own 256MB segment, so no need to force them into VRAM any more. Signed-off-by: Christian König <christian.koenig(a)amd.com> --- drivers/gpu/drm/radeon/radeon_cs.c | 3 ++- drivers/gpu/drm/radeon/radeon_uvd.c | 3 +-- drivers/gpu/drm/radeon/uvd_v1_0.c | 4 ++-- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c index ac6ece6..1806f4d 100644 --- a/drivers/gpu/drm/radeon/radeon_cs.c +++ b/drivers/gpu/drm/radeon/radeon_cs.c @@ -85,8 +85,9 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) VRAM, also but everything into VRAM on AGP cards to avoid image corruptions */ if (p->ring == R600_RING_TYPE_UVD_INDEX && + p->rdev->family < CHIP_PALM && (i == 0 || p->rdev->flags & RADEON_IS_AGP)) { - /* TODO: is this still needed for NI+ ? */ + p->relocs[i].lobj.domain = RADEON_GEM_DOMAIN_VRAM; diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c index 1a01bbf..a0f1185 100644 --- a/drivers/gpu/drm/radeon/radeon_uvd.c +++ b/drivers/gpu/drm/radeon/radeon_uvd.c @@ -476,8 +476,7 @@ static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p, return -EINVAL; } - /* TODO: is this still necessary on NI+ ? */ - if ((cmd == 0 || cmd == 0x3) && + if (p->rdev->family < CHIP_PALM && (cmd == 0 || cmd == 0x3) && (start >> 28) != (p->rdev->uvd.gpu_addr >> 28)) { DRM_ERROR("msg/fb buffer %LX-%LX out of 256MB segment!\n", start, end); diff --git a/drivers/gpu/drm/radeon/uvd_v1_0.c b/drivers/gpu/drm/radeon/uvd_v1_0.c index 7266805..3100fa9 100644 --- a/drivers/gpu/drm/radeon/uvd_v1_0.c +++ b/drivers/gpu/drm/radeon/uvd_v1_0.c @@ -212,8 +212,8 @@ int uvd_v1_0_start(struct radeon_device *rdev) /* enable VCPU clock */ WREG32(UVD_VCPU_CNTL, 1 << 9); - /* enable UMC */ - WREG32_P(UVD_LMI_CTRL2, 0, ~(1 << 8)); + /* enable UMC and NC0 */ + WREG32_P(UVD_LMI_CTRL2, 1 << 13, ~((1 << 8) | (1 << 13))); /* boot up the VCPU */ WREG32(UVD_SOFT_RESET, 0); -- 1.8.1.2

2 1

[Bug 69341] [radeonsi] KDE 4.11 is EXTREMELY slow with Raster QT backend
by bugzilla-daemon＠freedesktop.org 23 Sep '13

23 Sep '13

https://bugs.freedesktop.org/show_bug.cgi?id=69341 --- Comment #6 from Denis M. (Phr33d0m) <god(a)politeia.in> --- Created attachment 86376 --> https://bugs.freedesktop.org/attachment.cgi?id=86376&action=edit color-corruption.png I also have this color corruption all over the place with Native/OpenGL backend and Glamor. As well as huge slowness/laggyness, glxgears at: $ ▶ glxgears Running synchronized to the vertical refresh. The framerate should be approximately the same as the monitor refresh rate. 237 frames in 5.0 seconds = 47.022 FPS 250 frames in 5.0 seconds = 49.853 FPS 234 frames in 5.0 seconds = 46.726 FPS 253 frames in 5.0 seconds = 50.592 FPS 234 frames in 5.0 seconds = 46.440 FPS 256 frames in 5.0 seconds = 51.165 FPS 236 frames in 5.0 seconds = 46.952 FPS On Radeon HD7870. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[PATCH 1/1] drm/radeon: Make r100_cp_ring_info() and radeon_ring_gfx() safe
by Alex Ivanov 23 Sep '13

23 Sep '13

Prevent NULL pointer dereference in case when radeon_ring_fini() did it's job. Reading of r100_cp_ring_info and radeon_ring_gfx debugfs entries will lead to a KP if ring buffer was deallocated, e.g. on failed ring test. Seen on PA-RISC machine having "radeon: ring test failed (scratch(0x8504)=0xCAFEDEAD)" issue. Signed-off-by: Alex Ivanov <gnidorah(a)p0n4ik.tk> --- diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index 2417571..413cdd1 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -2933,10 +2933,11 @@ static int r100_debugfs_cp_ring_info(struct seq_file *m, void *data) seq_printf(m, "CP_RB_RPTR 0x%08x\n", rdp); seq_printf(m, "%u free dwords in ring\n", ring->ring_free_dw); seq_printf(m, "%u dwords in ring\n", count); - for (j = 0; j <= count; j++) { - i = (rdp + j) & ring->ptr_mask; - seq_printf(m, "r[%04d]=0x%08x\n", i, ring->ring[i]); - } + if (ring->ready) + for (j = 0; j <= count; j++) { + i = (rdp + j) & ring->ptr_mask; + seq_printf(m, "r[%04d]=0x%08x\n", i, ring->ring[i]); + } return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index 46a25f0..a890756 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -839,10 +839,11 @@ static int radeon_debugfs_ring_info(struct seq_file *m, void *data) * packet that is the root issue */ i = (ring->rptr + ring->ptr_mask + 1 - 32) & ring->ptr_mask; - for (j = 0; j <= (count + 32); j++) { - seq_printf(m, "r[%5d]=0x%08x\n", i, ring->ring[i]); - i = (i + 1) & ring->ptr_mask; - } + if (ring->ready) + for (j = 0; j <= (count + 32); j++) { + seq_printf(m, "r[%5d]=0x%08x\n", i, ring->ring[i]); + i = (i + 1) & ring->ptr_mask; + } return 0; }

2 4

[git pull] drm fixes
by Dave Airlie 22 Sep '13

22 Sep '13

Hi Linus, some small fixes for msm and exynos, a regression revert affecting nouveau users with old userspace, intel pageflip deadlock and gpu hang fixes, hsw modesetting hangs, Dave. The following changes since commit 928c2f0c006bf7f381f58af2b2786d2a858ae311: drm/fb-helper: don't sleep for screen unblank when an oops is in progress (2013-09-19 11:54:34 +1000) are available in the git repository at: git://people.freedesktop.org/~airlied/linux drm-fixes for you to fetch changes up to 6ddf2ed6e00396883b3123032ccb4416205aac7c: Merge branch 'msm-fixes-3.12' of git://people.freedesktop.org/~robclark/linux into drm-fixes (2013-09-20 09:06:48 +1000) ---------------------------------------------------------------- Chris Wilson (1): drm/i915: Track pfit enable state separately from size Daniel Vetter (5): drm/i915: fix wait_for_pending_flips vs gpu hang deadlock drm/i915/sdvo: Fully translate sync flags in the dtd->mode conversion drm/i915/sdvo: Robustify the dtd<->drm_mode conversions drm/i915/dvo: set crtc timings again for panel fixed modes drm/i915: kill set_need_resched Dave Airlie (4): Revert "drm: mark context support as a legacy subsystem" Merge tag 'drm-intel-fixes-2013-09-19' of git://people.freedesktop.org/~danvet/drm-intel into drm-fixes Merge branch 'exynos-drm-fixes' of git://git.kernel.org/.../daeinki/drm-exynos into drm-fixes Merge branch 'msm-fixes-3.12' of git://people.freedesktop.org/~robclark/linux into drm-fixes Jani Nikula (1): drm/i915: do not update cursor in crtc mode set Rob Clark (6): drm/msm: handle read vs write fences drm/msm: hangcheck harder drm/msm: fix cmdstream size check drm/msm: return -EBUSY if bo still active drm/msm: workaround for missing irq drm/msm: drop unnecessary set_need_resched() Sachin Kamat (3): drm/exynos: Remove redundant OF dependency drm/exynos: Fix address space warning in exynos_drm_buf.c drm/exynos: Fix address space warnings in exynos_drm_fbdev.c Takashi Iwai (1): drm/i915: Use proper print format for debug prints Ville Syrjälä (1): drm/i915: Don't enable the cursor on a disable pipe Wei Yongjun (3): drm/msm: fix return value check in ERR_PTR() drm/msm: fix potential NULL pointer dereference drm/exynos: fix return value check in lowlevel_buffer_allocate() drivers/gpu/drm/drm_context.c | 73 ++++--------------------------- drivers/gpu/drm/drm_fops.c | 21 ++++++++- drivers/gpu/drm/drm_stub.c | 10 ++++- drivers/gpu/drm/exynos/Kconfig | 2 +- drivers/gpu/drm/exynos/exynos_drm_buf.c | 7 +-- drivers/gpu/drm/exynos/exynos_drm_fbdev.c | 5 ++- drivers/gpu/drm/i915/i915_gem.c | 11 ++--- drivers/gpu/drm/i915/i915_irq.c | 68 ++++++++++++++++++++++------ drivers/gpu/drm/i915/intel_ddi.c | 2 +- drivers/gpu/drm/i915/intel_display.c | 42 +++++++++--------- drivers/gpu/drm/i915/intel_drv.h | 1 + drivers/gpu/drm/i915/intel_dvo.c | 2 + drivers/gpu/drm/i915/intel_panel.c | 1 + drivers/gpu/drm/i915/intel_pm.c | 6 +-- drivers/gpu/drm/i915/intel_sdvo.c | 63 ++++++++++++++------------ drivers/gpu/drm/msm/adreno/adreno_gpu.c | 10 ++++- drivers/gpu/drm/msm/msm_drv.c | 50 ++++++++++++++------- drivers/gpu/drm/msm/msm_drv.h | 8 +++- drivers/gpu/drm/msm/msm_gem.c | 34 +++++++++----- drivers/gpu/drm/msm/msm_gem.h | 2 +- drivers/gpu/drm/msm/msm_gem_submit.c | 24 +++++----- drivers/gpu/drm/msm/msm_gpu.c | 24 ++++++++-- include/drm/drmP.h | 7 ++- 23 files changed, 276 insertions(+), 197 deletions(-)

1 0

← Newer
1
...
64
65
66
67
68
69
70
...
83
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel September 2013