dri-devel December 2015

dri-devel@lists.freedesktop.org

162 participants
1148 discussions

[PATCH v8 00/25] Color Management for DRM framework
by Shashank Sharma 03 Dec '15

03 Dec '15

This patch set adds Color Manager implementation in DRM layer. Color Manager is an extension in DRM framework to support color correction/enhancement. Various Hardware platforms can support several color correction capabilities. Color Manager provides abstraction of these capabilities and allows a user space UI agent to correct/enhance the display using the DRM property interface. How is this going to work? ========================== 1. This patch series adds a few new properties in DRM framework. These properties are: a. color_capabilities property (type blob) b. Color Transformation Matrix property for corrections like CSC (called CTM, type blob) c. Palette correction properties for corrections like gamma fixup (called palette_correction, type blob) 2. Also, this patch series adds few structures to indicate specifications of a property like size, no_of_samples for correction etc. 3. These properties are present in mode_config. 4. When the platform's display driver loads, it fills up the values of color_capabilities property using the standard structures (added in step 2). For example, Intel's I915 driver adds following color correction capabilities: a. gamma correction capability as palette correction property, with 257 correction coefficients and a max/min value b. csc correction capability as CTM correction property, with 3x3 transformation matrix values and max/min values 5. Now when userspace comes up, it queries the platform's color capabilities by doing a get_property() on color_capabilities DRM property 6. Reading the blob, the userspace understands the color capabilities of the platform. For example, userspace will understand it can support: a. palette_correction with 257 coefficients b. CSC correction with 3x3 = 9 values 7. To set color correction values, userspace: a. creates a blob using the create_blob_ioctl in standard palette_correction structure format, with the correction values b. calls the set_property_ioctl with the blob_id as value for the property 8. Driver refers to the blob, gets the correction values and applies the correction in HW. 9. To get currently applied color correction values, userspace: a. calls a get_property_ioctl on that color property b. gets the blob_id for the currently applied correction from DRM infrastructure c. gets the blob using get_blob_ioctl and hence the currently applied values That's all! :) About the patch series: ======================= The patch series first adds the color management support in DRM layer. Then it adds the color management framework in I915 layer. After that, it implements platform specific core color correction functios. Intel color manager registers color correction with DRM color manager in this way: - CSC transformation is registered as CTM DRM property - Gamma correction is registered as palette_after_ctm DRM property - Degamma correction is registered as palette_before_ctm DRM property Our thanks to all the reviewers who have given valuable comments in terms of design and implementation to our previous sets of patches. Special mention of thanks should go to Matt Roper for all his inputs/suggestions in implementation of this module, using DRM atomic CRTC commit path. V2: Worked on review comments from Matt, Jim, Thierry, Rob. V3: Worked on review comments from Matt, Jim, Rob: - Jim, Rob: ======== Re-arranged the whole patch series in the sequence of features, currently: First 5 patches add color management support in DRM layer Next 7 patches add Intel color management framework in I915 driver Next 5 patches add color correction for CHV (gamma, degamma and CSC) Next 2 patches enable color management, by attaching the properties to CRTC(Matt) Next 4 patches add color correction for BDW (gamma, degamma) - Matt: ===== Patch 3: Added refernce/unreference for blob Patch 7: return -EINVAL and added debug message Patch 8: check for valid blob, from create blob moved call to intel_crtc_attach_color_prop in the end of full implementation (CHV) Patch 9: DRM_ERROR->DRM_DEBUG for NULL blob case Patch 13: Added static for internal functions Patch 20-24: renamed gen9_* functions to bdw_* Added new variables in device_info structure num_samples_after_ctm and num_samples_before_ctm Added new function in patch 8 to load capabilities based on device_info across all platforms V4: Worked on review comments from Daniel, Matt, Rob, Jim - Rob, Jim: ========= Patch 15, 22: Prepare CSC coeff properly(chv, bdw). Patch 13, 20: Gamma max should be (1<<24) -1(chv, bdw). - Daniel, Matt: ============= Patch 2: Create separate properties to query color capabilities, not a single blob. Patch 4, 5, 10: Add set/get property interface in DRM layer, not in I915 layer. V5: Addressed review comments from Emil, Rob. -Emil: ====== Patch 3: Fixed changes from patch 2, which were merged in patch 3 Patch 4: Added static for drm_atomic_crtc_set_blob Patch 13: Removed unnecessary initialization for various functions Patch 14: Removed unnecessary initialization for variables Checked for overflow for length Dropped variable ret, using direct returns instead Patch 15: Fixed type of return value of function chv_prepare_csc_coeff Fixed the type of variable csc_s3_12_format Changed the looping approach to write csc coefficients Patch 16: Fixed typo in intel_drv.h (intel_color_manager.c) Patch 19: Made bdw_write_12_bit_gamma_precision function static Fixed alignment problems with debug message Patch 21: Fixed indentation of debug message Patch 22: Removed change which belonged to another patch Made function bdw_set_csc static -Rob: ====== Patch 19: Clipped -> Clamped SET_BITS in 10bit gamma was taking red_fraction variable for all channels, fixed that. Added check for MAX values in 12_bit_gamma function Added new function for 8_bit_gamma mode (legacy) with MAX value checks Patch 20: Fixed commit message (BDW degamma values was mentioned as 65 in it). Patch 21: Removed unused length variable Moved mode variable closer to a place where its being used Removed cast for correction_values DeGamma -> degamma V6: Addressed review comments from Emil, Daniel, Rob and Jim -Emil: ====== Patch 10: Put definition and prototype in the same patch for function intel_attach_color_properties_to_crtc Patch 19: Removed extra line before every case() of switch() Patch 13: Replace if..else in function chv_set_degamma with switch case to maintain consistency Patch 21: Replace if..else in function bdw_set_degamma with switch case to maintain consistency -Daniel/Rob: ============= Removed num_samples from overall palette implementation, use blob_size/sizeof(struct drm_r32g32b32) -Jim: ===== Patch 13: Replace if..else in function chv_set_degamma with switch case Patch 21: Replace if..else in function bdw_set_degamma with switch case V7: Worked on Gary's suggestion to optimize the commit calls, applying color correction only when there is a change in color property status (not every commit). Added patch 24 and 25 in series, for the same. V8: -Daniel: ======= Patch 1,2,4,8: Fixed make htmldoc warnings, for missing description of new variables added in drm_crtc_state and mode_config. -Rebase: ======= Changed few variables from u32->i915_reg_t to align with changes in I915_READ and I915_WRITE. Shashank Sharma (25): drm: Create Color Management DRM properties drm: Create Color Management query properties drm: Add color correction blobs in CRTC state drm: Add set property support for color manager drm: Add get property support for color manager drm: Add drm structures for palette color property drm: Add structure for CTM color property drm: Add color correction state flag drm/i915: Add set property interface for CRTC drm/i915: Create color management files drm/i915: Register color correction capabilities drm/i915: CHV: Load gamma color correction values drm/i915: CHV: Load degamma color correction values drm/i915: CHV: Pipe level Gamma correction drm/i915: CHV: Pipe level degamma correction drm/i915: CHV: Pipe level CSC correction drm/i915: Commit color correction to CRTC drm/i915: Attach color properties to CRTC drm/i915: BDW: Load gamma correction values drm/i915: BDW: Pipe level Gamma correction drm/i915: BDW: Load degamma correction values drm/i915: BDW: Pipe level degamma correction drm/i915: BDW: Pipe level CSC correction drm/i915: disable plane gamma drm/i915: Commit color correction only when needed drivers/gpu/drm/drm_atomic.c | 67 ++- drivers/gpu/drm/drm_atomic_helper.c | 9 + drivers/gpu/drm/drm_crtc.c | 32 ++ drivers/gpu/drm/i915/Makefile | 3 +- drivers/gpu/drm/i915/i915_drv.c | 17 + drivers/gpu/drm/i915/i915_drv.h | 2 + drivers/gpu/drm/i915/i915_reg.h | 58 +- drivers/gpu/drm/i915/intel_color_manager.c | 863 +++++++++++++++++++++++++++++ drivers/gpu/drm/i915/intel_color_manager.h | 111 ++++ drivers/gpu/drm/i915/intel_display.c | 6 +- drivers/gpu/drm/i915/intel_drv.h | 5 + drivers/gpu/drm/i915/intel_sprite.c | 7 +- include/drm/drm_crtc.h | 24 + include/uapi/drm/drm.h | 30 + 14 files changed, 1225 insertions(+), 9 deletions(-) create mode 100644 drivers/gpu/drm/i915/intel_color_manager.c create mode 100644 drivers/gpu/drm/i915/intel_color_manager.h -- 1.9.1

1 25

[PATCH] drm/atomic-helper: Reject attempts at re-stealing encoders
by Daniel Vetter 03 Dec '15

03 Dec '15

This can happen when we run out of encoders for a multi-crtc modeset, or also when userspace is silly and tries to clone multiple connectors that need the same encoder on the same crtc. Reported-and-Tested-and-Reviewed-by: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> Cc: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter(a)intel.com> --- drivers/gpu/drm/drm_atomic_helper.c | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index d6d80b30f751..aff616dc6caf 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -86,6 +86,27 @@ drm_atomic_helper_plane_changed(struct drm_atomic_state *state, } } +static bool +check_pending_encoder_assignment(struct drm_atomic_state *state, + struct drm_encoder *new_encoder, + struct drm_connector *new_connector) +{ + struct drm_connector *connector; + struct drm_connector_state *conn_state; + int i; + + for_each_connector_in_state(state, connector, conn_state, i) { + if (conn_state->best_encoder != new_encoder) + continue; + + /* encoder already assigned and we're trying to re-steal it! */ + if (connector->state->best_encoder != conn_state->best_encoder) + return false; + } + + return true; +} + static struct drm_crtc * get_current_crtc_for_encoder(struct drm_device *dev, struct drm_encoder *encoder) @@ -235,6 +256,13 @@ update_connector_routing(struct drm_atomic_state *state, int conn_idx) return 0; } + if (!check_pending_encoder_assignment(state, new_encoder, connector)) { + DRM_DEBUG_ATOMIC("Encoder for [CONNECTOR:%d:%s] already assigned\n", + connector->base.id, + connector->name); + return -EINVAL; + } + encoder_crtc = get_current_crtc_for_encoder(state->dev, new_encoder); -- 2.5.1

2 1

[PATCH] drm/atomic: Reject async flips on a disabled pipe
by Daniel Vetter 03 Dec '15

03 Dec '15

Two reasons for that: - This changes aligns atomic with the legacy page flip semantics as established by the i915 driver. - Asking for an async flip on a disabled pipe generally indicates a userspace bug. Worst case userspace will busy-loop rendering (since flips complete immediately). So better to catch this fast. Note that we can't do the same for synchronous commits since the legacy hooks allowed that. And there's probably piles of userspace code out there which relies on this. Cc: Daniel Stone <daniels(a)collabora.com> Signed-off-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> --- drivers/gpu/drm/drm_atomic.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index 791876e4657c..3c171e9e1cf5 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -1334,12 +1334,25 @@ EXPORT_SYMBOL(drm_atomic_commit); int drm_atomic_async_commit(struct drm_atomic_state *state) { struct drm_mode_config *config = &state->dev->mode_config; - int ret; + struct drm_crtc *crtc; + struct drm_crtc_state *crtc_state; + int i, ret; ret = drm_atomic_check_only(state); if (ret) return ret; + if (!state->allow_modeset) { + for_each_crtc_in_state(state, crtc, crtc_state, i) { + if (crtc_state->active) { + DRM_DEBUG_ATOMIC("[CRTC:%d] requires full modeset\n", + crtc->base.id); + return -EINVAL; + + } + } + } + DRM_DEBUG_ATOMIC("commiting %p asynchronously\n", state); return config->funcs->atomic_commit(state->dev, state, true); -- 2.5.1

1 0

[PATCH] drm/panel: Initialize nominal vrefresh from timings in simple panel driver
by Philipp Zabel 03 Dec '15

03 Dec '15

Panel timing descriptions using struct videomode don't have the vrefresh field, so calculate it instead using drm_mode_vrefresh. Signed-off-by: Philipp Zabel <p.zabel(a)pengutronix.de> --- drivers/gpu/drm/panel/panel-simple.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/panel/panel-simple.c b/drivers/gpu/drm/panel/panel-simple.c index f97b73e..cb7dd41 100644 --- a/drivers/gpu/drm/panel/panel-simple.c +++ b/drivers/gpu/drm/panel/panel-simple.c @@ -112,6 +112,7 @@ static int panel_simple_get_fixed_modes(struct panel_simple *panel) } drm_display_mode_from_videomode(&vm, mode); + mode->vrefresh = drm_mode_vrefresh(mode); drm_mode_set_name(mode); drm_mode_probed_add(connector, mode); -- 2.6.2

2 2

[RESEND 2/3] drm :fsl-dcu: Cleanup vblank interrupt mask and status setting code
by Dongsheng Wang 03 Dec '15

03 Dec '15

From: Jianwei Wang <jianwei.wang.chn(a)gmail.com> Switch update interrupt mask bit with regmap_update_bits, and clear interrupt status by writing 1 to relevant bit before setting mask in fsl_dcu_drm_irq_init function. Signed-off-by: Jianwei Wang <jianwei.wang.chn(a)gmail.com> Signed-off-by: Yi Meng <b56799(a)freescale.com> Signed-off-by: Wang Dongsheng <dongsheng.wang(a)freescale.com> diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c index 1930234..5c29ff7 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c @@ -38,21 +38,17 @@ static const struct regmap_config fsl_dcu_regmap_config = { static int fsl_dcu_drm_irq_init(struct drm_device *dev) { struct fsl_dcu_drm_device *fsl_dev = dev->dev_private; - unsigned int value; int ret; ret = drm_irq_install(dev, fsl_dev->irq); if (ret < 0) dev_err(dev->dev, "failed to install IRQ handler\n"); - ret = regmap_write(fsl_dev->regmap, DCU_INT_STATUS, 0); + ret = regmap_write(fsl_dev->regmap, DCU_INT_STATUS, 0xffffffff); if (ret) dev_err(dev->dev, "set DCU_INT_STATUS failed\n"); - ret = regmap_read(fsl_dev->regmap, DCU_INT_MASK, &value); - if (ret) - dev_err(dev->dev, "read DCU_INT_MASK failed\n"); - value &= DCU_INT_MASK_VBLANK; - ret = regmap_write(fsl_dev->regmap, DCU_INT_MASK, value); + ret = regmap_update_bits(fsl_dev->regmap, DCU_INT_MASK, + DCU_INT_MASK_VBLANK, ~DCU_INT_MASK_VBLANK); if (ret) dev_err(dev->dev, "set DCU_INT_MASK failed\n"); ret = regmap_write(fsl_dev->regmap, DCU_UPDATE_MODE, @@ -143,14 +139,10 @@ static irqreturn_t fsl_dcu_drm_irq(int irq, void *arg) static int fsl_dcu_drm_enable_vblank(struct drm_device *dev, unsigned int pipe) { struct fsl_dcu_drm_device *fsl_dev = dev->dev_private; - unsigned int value; int ret; - ret = regmap_read(fsl_dev->regmap, DCU_INT_MASK, &value); - if (ret) - dev_err(dev->dev, "read DCU_INT_MASK failed\n"); - value &= ~DCU_INT_MASK_VBLANK; - ret = regmap_write(fsl_dev->regmap, DCU_INT_MASK, value); + ret = regmap_update_bits(fsl_dev->regmap, DCU_INT_MASK, + DCU_INT_MASK_VBLANK, ~DCU_INT_MASK_VBLANK); if (ret) dev_err(dev->dev, "set DCU_INT_MASK failed\n"); return 0; @@ -160,14 +152,10 @@ static void fsl_dcu_drm_disable_vblank(struct drm_device *dev, unsigned int pipe) { struct fsl_dcu_drm_device *fsl_dev = dev->dev_private; - unsigned int value; int ret; - ret = regmap_read(fsl_dev->regmap, DCU_INT_MASK, &value); - if (ret) - dev_err(dev->dev, "read DCU_INT_MASK failed\n"); - value |= DCU_INT_MASK_VBLANK; - ret = regmap_write(fsl_dev->regmap, DCU_INT_MASK, value); + ret = regmap_update_bits(fsl_dev->regmap, DCU_INT_MASK, + DCU_INT_MASK_VBLANK, DCU_INT_MASK_VBLANK); if (ret) dev_err(dev->dev, "set DCU_INT_MASK failed\n"); } -- 2.1.0.27.g96db324

3 4

[Bug 92850] Segfault loading War Thunder
by bugzilla-daemon＠freedesktop.org 03 Dec '15

03 Dec '15

https://bugs.freedesktop.org/show_bug.cgi?id=92850 Bug ID: 92850 Summary: Segfault loading War Thunder Product: Mesa Version: git Hardware: x86-64 (AMD64) OS: Linux (All) Status: NEW Severity: critical Priority: medium Component: Drivers/Gallium/radeonsi Assignee: dri-devel(a)lists.freedesktop.org Reporter: bellamorte42(a)gmail.com QA Contact: dri-devel(a)lists.freedesktop.org After logging in the game will segfault almost immediately. LD_DEBUG=all, strace, and gdb of the core dump below. Apitrace did not yield any error messages. I've waited a few weeks to see if it would clear up on it's own but it hasn't yet. [0] 25515: symbol= 25515: binding file /usr/lib/libgcc_s.so.1 [0] to /usr/lib/libc.so.6 [0]: normal symbol `dl_iterate_phdr' [GLIBC_2.2.5] dl_iterate_phdr; lookup in file=/usr/lib/libgcc_s.so.1 [0] 25515: symbol=dl_iterate_phdr; lookup in file=/usr/lib/libc.so.6 [0] 25515: binding file /usr/lib/libgcc_s.so.1 [0] to /usr/lib/libc.so.6 [0]: normal symbol `dl_iterate_phdr' [GLIBC_2.2.5] 25515: symbol=modff; lookup in file=./aces [0] 25515: symbol=modff; lookup in file=/usr/lib/libpthread.so.0 [0] 25515: symbol=modff; lookup in file=/usr/lib/libresolv.so.2 [0] 25515: symbol=modff; lookup in file=/usr/lib/libdl.so.2 [0] 25515: symbol=modff; lookup in file=/usr/lib/librt.so.1 [0] 25515: symbol=modff; lookup in file=/usr/lib/libGL.so.1 [0] 25515: symbol=modff; lookup in file=/usr/lib/libX11.so.6 [0] 25515: symbol=modff; lookup in file=/usr/lib/libstdc++.so.6 [0] 25515: symbol=modff; lookup in file=/usr/lib/libm.so.6 [0] 25515: binding file ./aces [0] to /usr/lib/libm.so.6 [0]: normal symbol `modff' [GLIBC_2.2.5] war thunder: line 3: 25515 Segmentation fault (core dumped) MESA_GL_VERSION_OVERRIDE=4.1COMPAT LD_DEBUG=all ./aces ioctl(8, DRM_IOCTL_RADEON_GEM_VA, 0x7ffe41840040) = 0 ioctl(8, DRM_IOCTL_GEM_CLOSE, 0x7ffe41840030) = 0 ioctl(8, DRM_IOCTL_RADEON_GEM_BUSY, 0x7ffe4183ff40) = 0 futex(0x4990d0c, FUTEX_WAKE_OP_PRIVATE, 1, 1, 0x4990d08, {FUTEX_OP_SET, 0, FUTEX_OP_CMP_GT, 1}) = 1 futex(0x4990ce0, FUTEX_WAKE_PRIVATE, 1) = 1 poll([{fd=6, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=6, revents=POLLOUT}]) writev(6, [{"\224\1\22\0\10\0\0\1\v\0\0\1\270\7\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 72}], 1) = 72 select(10, [9], NULL, NULL, {0, 0}) = 0 (Timeout) select(10, [9], NULL, NULL, {0, 0}) = 0 (Timeout) ioctl(12, EVIOCGABS(ABS_X), {value=124, minimum=0, ...}) = 0 ioctl(12, EVIOCGABS(ABS_Y), {value=138, minimum=0, ...}) = 0 ioctl(12, EVIOCGABS(ABS_Z), {value=10, minimum=0, ...}) = 0 ioctl(12, EVIOCGABS(ABS_HAT0X), {value=0, minimum=4294967295, ...}) = 0 ioctl(12, EVIOCGABS(ABS_HAT0Y), {value=0, minimum=4294967295, ...}) = 0 ioctl(12, EVIOCGKEY(96), [ 0 ]) = 96 ioctl(18, EVIOCGABS(ABS_X), {value=0, minimum=0, ...}) = 0 ioctl(18, EVIOCGABS(ABS_Y), {value=0, minimum=0, ...}) = 0 ioctl(18, EVIOCGABS(ABS_RZ), {value=0, minimum=0, ...}) = 0 ioctl(18, EVIOCGABS(ABS_THROTTLE), {value=0, minimum=0, ...}) = 0 ioctl(18, EVIOCGABS(ABS_HAT0X), {value=0, minimum=4294967295, ...}) = 0 ioctl(18, EVIOCGABS(ABS_HAT0Y), {value=0, minimum=4294967295, ...}) = 0 ioctl(18, EVIOCGKEY(96), [ 0 ]) = 96 recvmsg(6, 0x7ffe41840000, 0) = -1 EAGAIN (Resource temporarily unavailable) select(10, [9], NULL, NULL, {0, 0}) = 0 (Timeout) select(10, [9], NULL, NULL, {0, 0}) = 0 (Timeout) ioctl(12, EVIOCGABS(ABS_X), {value=124, minimum=0, ...}) = 0 ioctl(12, EVIOCGABS(ABS_Y), {value=138, minimum=0, ...}) = 0 ioctl(12, EVIOCGABS(ABS_Z), {value=10, minimum=0, ...}) = 0 ioctl(12, EVIOCGABS(ABS_HAT0X), {value=0, minimum=4294967295, ...}) = 0 ioctl(12, EVIOCGABS(ABS_HAT0Y), {value=0, minimum=4294967295, ...}) = 0 ioctl(12, EVIOCGKEY(96), [ 0 ]) = 96 ioctl(18, EVIOCGABS(ABS_X), {value=0, minimum=0, ...}) = 0 ioctl(18, EVIOCGABS(ABS_Y), {value=0, minimum=0, ...}) = 0 ioctl(18, EVIOCGABS(ABS_RZ), {value=0, minimum=0, ...}) = 0 ioctl(18, EVIOCGABS(ABS_THROTTLE), {value=0, minimum=0, ...}) = 0 ioctl(18, EVIOCGABS(ABS_HAT0X), {value=0, minimum=4294967295, ...}) = 0 ioctl(18, EVIOCGABS(ABS_HAT0Y), {value=0, minimum=4294967295, ...}) = 0 ioctl(18, EVIOCGKEY(96), [ 0 ]) = 96 nanosleep({0, 33000000}, 0x7ffe41840250) = 0 futex(0x31ca380, FUTEX_WAIT_PRIVATE, 2, NULL <unfinished ...> +++ killed by SIGSEGV (core dumped) +++ war thunder: line 3: 3826 Segmentation fault (core dumped) MESA_GL_VERSION_OVERRIDE=4.1COMPAT strace ./aces warning: Could not load shared library symbols for linux-vdso.so.1. Do you need "set solib-search-path" or "set sysroot"? [Thread debugging using libthread_db enabled] Using host libthread_db library "/usr/lib/libthread_db.so.1". Core was generated by `./aces'. Program terminated with signal SIGSEGV, Segmentation fault. #0 0x00007f3851f5bde6 in ?? () from /usr/lib/xorg/modules/dri/radeonsi_dri.so [Current thread is 1 (Thread 0x7f384eccc700 (LWP 3923))] -- You are receiving this mail because: You are the assignee for the bug.

1 28

Issue with lock when switching graphics (nouveau)
by York Possemiers 03 Dec '15

03 Dec '15

So having conferred with Karol Herbst in #nouveau on freenode, when attempting to run $ DRI_PRIME=1 glxinfo, the following stacktrace turns up in dmesg https://gist.githubusercontent.com/karolherbst/cd8292689959417a0111/raw/d58… Running kernel 4.4.0-0.rc3.git1.1.fc24.x86_64 (Fedora Rawhide) I'm not really sure where to direct this, I was advised that I should post this here. If any more information is required, I am happy to provide.

1 0

[PATCH v5 1/3] drm: Implement drm_modeset_lock_all_ctx()
by Thierry Reding 03 Dec '15

03 Dec '15

From: Thierry Reding <treding(a)nvidia.com> This function is like drm_modeset_lock_all(), but it takes the lock acquisition context as a parameter rather than storing it in the DRM device's mode_config structure. Implement drm_modeset_{,un}lock_all() in terms of the new function for better code reuse, and add a note to the kerneldoc that new code should use the new functions. v2: improve kerneldoc v4: rename drm_modeset_lock_all_crtcs() to drm_modeset_lock_all_ctx() and take mode_config's .connection_mutex instead of .mutex lock to avoid lock inversion (Daniel Vetter), use drm_modeset_drop_locks() which is now the equivalent of drm_modeset_unlock_all_ctx() v5: do not take the dev->mode_config.connection_mutex in drm_atomic_legacy_backoff() since drm_modeset_lock_all_ctx() already keeps it, enhance kerneldoc for drm_modeset_lock_all_ctx() (Daniel Vetter) Signed-off-by: Thierry Reding <treding(a)nvidia.com> --- drivers/gpu/drm/drm_atomic.c | 7 +-- drivers/gpu/drm/drm_modeset_lock.c | 89 +++++++++++++++++++++++++------------- include/drm/drm_modeset_lock.h | 4 +- 3 files changed, 63 insertions(+), 37 deletions(-) diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index 55b4debad79b..ef5f7663a718 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -1188,12 +1188,7 @@ void drm_atomic_legacy_backoff(struct drm_atomic_state *state) retry: drm_modeset_backoff(state->acquire_ctx); - ret = drm_modeset_lock(&state->dev->mode_config.connection_mutex, - state->acquire_ctx); - if (ret) - goto retry; - ret = drm_modeset_lock_all_crtcs(state->dev, - state->acquire_ctx); + ret = drm_modeset_lock_all_ctx(state->dev, state->acquire_ctx); if (ret) goto retry; } diff --git a/drivers/gpu/drm/drm_modeset_lock.c b/drivers/gpu/drm/drm_modeset_lock.c index 6675b1428410..c2f5971146ba 100644 --- a/drivers/gpu/drm/drm_modeset_lock.c +++ b/drivers/gpu/drm/drm_modeset_lock.c @@ -57,11 +57,18 @@ /** * drm_modeset_lock_all - take all modeset locks - * @dev: drm device + * @dev: DRM device * * This function takes all modeset locks, suitable where a more fine-grained - * scheme isn't (yet) implemented. Locks must be dropped with - * drm_modeset_unlock_all. + * scheme isn't (yet) implemented. Locks must be dropped by calling the + * drm_modeset_unlock_all() function. + * + * This function is deprecated. It allocates a lock acquisition context and + * stores it in the DRM device's ->mode_config. This facilitate conversion of + * existing code because it removes the need to manually deal with the + * acquisition context, but it is also brittle because the context is global + * and care must be taken not to nest calls. New code should use the + * drm_modeset_lock_all_ctx() function and pass in the context explicitly. */ void drm_modeset_lock_all(struct drm_device *dev) { @@ -78,39 +85,43 @@ void drm_modeset_lock_all(struct drm_device *dev) drm_modeset_acquire_init(ctx, 0); retry: - ret = drm_modeset_lock(&config->connection_mutex, ctx); - if (ret) - goto fail; - ret = drm_modeset_lock_all_crtcs(dev, ctx); - if (ret) - goto fail; + ret = drm_modeset_lock_all_ctx(dev, ctx); + if (ret < 0) { + if (ret == -EDEADLK) { + drm_modeset_backoff(ctx); + goto retry; + } + + drm_modeset_acquire_fini(ctx); + kfree(ctx); + return; + } WARN_ON(config->acquire_ctx); - /* now we hold the locks, so now that it is safe, stash the - * ctx for drm_modeset_unlock_all(): + /* + * We hold the locks now, so it is safe to stash the acquisition + * context for drm_modeset_unlock_all(). */ config->acquire_ctx = ctx; drm_warn_on_modeset_not_all_locked(dev); - - return; - -fail: - if (ret == -EDEADLK) { - drm_modeset_backoff(ctx); - goto retry; - } - - kfree(ctx); } EXPORT_SYMBOL(drm_modeset_lock_all); /** * drm_modeset_unlock_all - drop all modeset locks - * @dev: device + * @dev: DRM device * - * This function drop all modeset locks taken by drm_modeset_lock_all. + * This function drops all modeset locks taken by a previous call to the + * drm_modeset_lock_all() function. + * + * This function is deprecated. It uses the lock acquisition context stored + * in the DRM device's ->mode_config. This facilitates conversion of existing + * code because it removes the need to manually deal with the acquisition + * context, but it is also brittle because the context is global and care must + * be taken not to nest calls. New code should pass the acquisition context + * directly to the drm_modeset_drop_locks() function. */ void drm_modeset_unlock_all(struct drm_device *dev) { @@ -431,14 +442,34 @@ void drm_modeset_unlock(struct drm_modeset_lock *lock) } EXPORT_SYMBOL(drm_modeset_unlock); -/* In some legacy codepaths it's convenient to just grab all the crtc and plane - * related locks. */ -int drm_modeset_lock_all_crtcs(struct drm_device *dev, - struct drm_modeset_acquire_ctx *ctx) +/** + * drm_modeset_lock_all_ctx - take all modeset locks + * @dev: DRM device + * @ctx: lock acquisition context + * + * This function takes all modeset locks, suitable where a more fine-grained + * scheme isn't (yet) implemented. + * + * Unlike drm_modeset_lock_all(), it doesn't take the dev->mode_config.mutex + * since that lock isn't required for modeset state changes. Callers which + * need to grab that lock too need to do so outside of the acquire context + * @ctx. + * + * Locks acquired with this function should be released by calling the + * drm_modeset_drop_locks() function on @ctx. + * + * Returns: 0 on success or a negative error-code on failure. + */ +int drm_modeset_lock_all_ctx(struct drm_device *dev, + struct drm_modeset_acquire_ctx *ctx) { struct drm_crtc *crtc; struct drm_plane *plane; - int ret = 0; + int ret; + + ret = drm_modeset_lock(&dev->mode_config.connection_mutex, ctx); + if (ret) + return ret; drm_for_each_crtc(crtc, dev) { ret = drm_modeset_lock(&crtc->mutex, ctx); @@ -454,4 +485,4 @@ int drm_modeset_lock_all_crtcs(struct drm_device *dev, return 0; } -EXPORT_SYMBOL(drm_modeset_lock_all_crtcs); +EXPORT_SYMBOL(drm_modeset_lock_all_ctx); diff --git a/include/drm/drm_modeset_lock.h b/include/drm/drm_modeset_lock.h index 94938d89347c..c5576fbcb909 100644 --- a/include/drm/drm_modeset_lock.h +++ b/include/drm/drm_modeset_lock.h @@ -138,7 +138,7 @@ void drm_warn_on_modeset_not_all_locked(struct drm_device *dev); struct drm_modeset_acquire_ctx * drm_modeset_legacy_acquire_ctx(struct drm_crtc *crtc); -int drm_modeset_lock_all_crtcs(struct drm_device *dev, - struct drm_modeset_acquire_ctx *ctx); +int drm_modeset_lock_all_ctx(struct drm_device *dev, + struct drm_modeset_acquire_ctx *ctx); #endif /* DRM_MODESET_LOCK_H_ */ -- 2.5.0

3 4

[Bug 92982] slow switching to KD_GRAPHICS (KMS?)
by bugzilla-daemon＠freedesktop.org 02 Dec '15

02 Dec '15

https://bugs.freedesktop.org/show_bug.cgi?id=92982 --- Comment #11 from royvandervegt(a)gmail.com --- It turns out the problems are caused by DPM. Without DPM (using radeon.dpm=0) everything works as it should. I have tried tweaking the DPM parameters using https://wiki.archlinux.org/index.php/ATI#Dynamic_power_management, but changing dpm_state to `performance` and power_dpm_force_performance_level to `high` doesn't solve the issues (in fact, changing these parameters seems to have no effect at all?). Obviously disabling DPM isn't ideal, so it would be nice if there was a workaround/fix. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[PATCH] radeon and amdgpu drm-fixes-4.4
by Alex Deucher 02 Dec '15

02 Dec '15

Hi Dave, radeon and amdgpu fixes for 4.4. The following changes since commit 8c14f72b57570d71afda03c5c51b06f0b4af4367: Merge branch 'linux-4.4' of git://anongit.freedesktop.org/git/nouveau/linux-2.6 into drm-fixes (2015-11-28 06:50:34 +1000) are available in the git repository at: git://people.freedesktop.org/~agd5f/linux drm-fixes-4.4 for you to fetch changes up to 07df04dfcfe51a45239d056d43cbc09702262c82: drm/amdgpu: fix race condition in amd_sched_entity_push_job (2015-12-02 15:04:04 -0500) ---------------------------------------------------------------- Christian König (2): drm/amdgpu: fix userptr flags check drm/amdgpu: fix VM page table reference counting Chunming Zhou (1): drm/amdgpu: add err check for pin userptr Nicolai Hähnle (1): drm/amdgpu: fix race condition in amd_sched_entity_push_job Pavel Machek (1): add blacklist for thinkpad T40p drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 5 +++-- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 10 +++++++--- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 5 +++++ drivers/gpu/drm/amd/scheduler/gpu_scheduler.c | 5 +++-- drivers/gpu/drm/radeon/radeon_agp.c | 3 +++ 7 files changed, 23 insertions(+), 7 deletions(-)

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel December 2015