dri-devel September 2017

dri-devel@lists.freedesktop.org

210 participants
1526 discussions

[PATCH v8 0/3] Add support Rockchip Soc LVDS
by Sandy Huang 11 Sep '17

11 Sep '17

this patches add support Rockchip RK3288 LVDS support, based on patches from Mark yao[0] and Heiko Stuebner[1]. [0]: https://github.com/RockchipOpensourceCommunity/popmetal-kernel-3.14 [1]: http://lists.infradead.org/pipermail/linux-rockchip/2015-April/002830.html Changes: - Update rockchip_lvds_encoder_helper_funcs to atomic API - Add bridge function for RGB/LVDS convert to other output type - Fix some unreasonable define - Adapter to the latest rockchip DRM driver framework - Add pinctrl for RGB output type - Reseved some define for rockchip next Soc Sandy Huang (3): dt-bindings: display: Add Document for Rockchip Soc LVDS ARM: dts: Add LVDS info for rk3288 drm/rockchip: Add support for Rockchip Soc LVDS .../bindings/display/rockchip/rockchip-lvds.txt | 99 ++++ arch/arm/boot/dts/rk3288.dtsi | 52 ++ drivers/gpu/drm/rockchip/Kconfig | 8 + drivers/gpu/drm/rockchip/Makefile | 1 + drivers/gpu/drm/rockchip/rockchip_drm_drv.c | 2 + drivers/gpu/drm/rockchip/rockchip_drm_drv.h | 1 + drivers/gpu/drm/rockchip/rockchip_lvds.c | 582 +++++++++++++++++++++ drivers/gpu/drm/rockchip/rockchip_lvds.h | 114 ++++ 8 files changed, 859 insertions(+) create mode 100644 Documentation/devicetree/bindings/display/rockchip/rockchip-lvds.txt create mode 100644 drivers/gpu/drm/rockchip/rockchip_lvds.c create mode 100644 drivers/gpu/drm/rockchip/rockchip_lvds.h -- 2.7.4

5 9

[PATCH] qxl: fix primary surface handling
by Gerd Hoffmann 11 Sep '17

11 Sep '17

The atomic conversion of the qxl driver didn't got the primary surface handling completely right. It works in the common simple cases, but fails for example when changing the display resolution using xrandr or in multihead setups. The rules are simple: There is one primary surface. Before defining a new one you have to destroy the old one. This patch makes qxl_primary_atomic_update() destroy the primary surface before defining a new one. It fixes is_primary flag updates. It adds is_primary checks so we don't try to update the primary surface in case it already has the state we want it being in. Signed-off-by: Gerd Hoffmann <kraxel(a)redhat.com> --- drivers/gpu/drm/qxl/qxl_display.c | 36 ++++++++++++++++++++---------------- 1 file changed, 20 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c index 14c5613b43..e1dd05423e 100644 --- a/drivers/gpu/drm/qxl/qxl_display.c +++ b/drivers/gpu/drm/qxl/qxl_display.c @@ -509,23 +509,25 @@ static void qxl_primary_atomic_update(struct drm_plane *plane, .y2 = qfb->base.height }; - if (!old_state->fb) { - qxl_io_log(qdev, - "create primary fb: %dx%d,%d,%d\n", - bo->surf.width, bo->surf.height, - bo->surf.stride, bo->surf.format); - - qxl_io_create_primary(qdev, 0, bo); - bo->is_primary = true; - return; - - } else { + if (old_state->fb) { qfb_old = to_qxl_framebuffer(old_state->fb); bo_old = gem_to_qxl_bo(qfb_old->obj); + } else { + bo_old = NULL; + } + + if (bo == bo_old) + return; + + if (bo_old && bo_old->is_primary) { + qxl_io_destroy_primary(qdev); bo_old->is_primary = false; } - bo->is_primary = true; + if (!bo->is_primary) { + qxl_io_create_primary(qdev, 0, bo); + bo->is_primary = true; + } qxl_draw_dirty_fb(qdev, qfb, bo, 0, 0, &norect, 1, 1); } @@ -534,13 +536,15 @@ static void qxl_primary_atomic_disable(struct drm_plane *plane, { struct qxl_device *qdev = plane->dev->dev_private; - if (old_state->fb) - { struct qxl_framebuffer *qfb = + if (old_state->fb) { + struct qxl_framebuffer *qfb = to_qxl_framebuffer(old_state->fb); struct qxl_bo *bo = gem_to_qxl_bo(qfb->obj); - qxl_io_destroy_primary(qdev); - bo->is_primary = false; + if (bo->is_primary) { + qxl_io_destroy_primary(qdev); + bo->is_primary = false; + } } } -- 2.9.3

3 2

[PATCH 1/2] dma-buf: make reservation_object_copy_fences rcu save
by Christian König 11 Sep '17

11 Sep '17

From: Christian König <christian.koenig(a)amd.com> Stop requiring that the src reservation object is locked for this operation. Signed-off-by: Christian König <christian.koenig(a)amd.com> --- drivers/dma-buf/reservation.c | 56 ++++++++++++++++++++++++++++++++----------- 1 file changed, 42 insertions(+), 14 deletions(-) diff --git a/drivers/dma-buf/reservation.c b/drivers/dma-buf/reservation.c index dec3a81..b44d9d7 100644 --- a/drivers/dma-buf/reservation.c +++ b/drivers/dma-buf/reservation.c @@ -266,8 +266,7 @@ EXPORT_SYMBOL(reservation_object_add_excl_fence); * @dst: the destination reservation object * @src: the source reservation object * -* Copy all fences from src to dst. Both src->lock as well as dst-lock must be -* held. +* Copy all fences from src to dst. dst-lock must be held. */ int reservation_object_copy_fences(struct reservation_object *dst, struct reservation_object *src) @@ -277,33 +276,62 @@ int reservation_object_copy_fences(struct reservation_object *dst, size_t size; unsigned i; - src_list = reservation_object_get_list(src); + rcu_read_lock(); + src_list = rcu_dereference(src->fence); +retry: if (src_list) { - size = offsetof(typeof(*src_list), - shared[src_list->shared_count]); + unsigned shared_count = src_list->shared_count; + + size = offsetof(typeof(*src_list), shared[shared_count]); + rcu_read_unlock(); + dst_list = kmalloc(size, GFP_KERNEL); if (!dst_list) return -ENOMEM; - dst_list->shared_count = src_list->shared_count; - dst_list->shared_max = src_list->shared_count; - for (i = 0; i < src_list->shared_count; ++i) - dst_list->shared[i] = - dma_fence_get(src_list->shared[i]); + rcu_read_lock(); + src_list = rcu_dereference(src->fence); + if (!src_list || src_list->shared_count > shared_count) { + kfree(dst_list); + goto retry; + } + + dst_list->shared_count = 0; + dst_list->shared_max = shared_count; + for (i = 0; i < src_list->shared_count; ++i) { + struct dma_fence *fence; + + fence = rcu_dereference(src_list->shared[i]); + if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, + &fence->flags)) + continue; + + if (!dma_fence_get_rcu(fence)) { + kfree(dst_list); + src_list = rcu_dereference(src->fence); + goto retry; + } + + if (dma_fence_is_signaled(fence)) { + dma_fence_put(fence); + continue; + } + + dst_list->shared[dst_list->shared_count++] = fence; + } } else { dst_list = NULL; } + new = dma_fence_get_rcu_safe(&src->fence_excl); + rcu_read_unlock(); + kfree(dst->staged); dst->staged = NULL; src_list = reservation_object_get_list(dst); - old = reservation_object_get_excl(dst); - new = reservation_object_get_excl(src); - - dma_fence_get(new); preempt_disable(); write_seqcount_begin(&dst->seq); -- 2.7.4

4 14

[PATCH v3 0/2] add Silicon Image SiI9234 driver
by Maciej Purski 11 Sep '17

11 Sep '17

Hi everyone, These patches are a continuation of work by Tomasz Stanislawski on sii9324 driver, which was described in th following letter: https://lists.freedesktop.org/archives/dri-devel/2014-April/057481.html The main differences from the previous code are: * driver moved to /gpu/drm/bridge/ and integrated with drm/bridge subsystem * added filtering-out unsupported display modes * changed gpio interface to up-to-date * changed interrupt handling * improve code style * add hdmi and sii9324 to exynos4412-trats2 device tree All comments are welcome. Regards, Maciej Purski Changes in v3: - change sii9234 device tree node name - use defines from dt-bindings/pinctrl/samsung.h Changes in v2: - regulators used in driver now model all physical regulators used by the device - substitute some of the magic values with macros - improve coding style - improve error handling in sii9234_probe() - fix commit message Maciej Purski (2): drm/bridge: add Silicon Image SiI9234 driver ARM: dts: exynos: Add HDMI and Sil9234 to Trats2 board .../devicetree/bindings/display/bridge/sii9234.txt | 34 + arch/arm/boot/dts/exynos4412-trats2.dts | 111 +++ drivers/gpu/drm/bridge/Kconfig | 8 + drivers/gpu/drm/bridge/Makefile | 1 + drivers/gpu/drm/bridge/sii9234.c | 993 +++++++++++++++++++++ 5 files changed, 1147 insertions(+) create mode 100644 Documentation/devicetree/bindings/display/bridge/sii9234.txt create mode 100644 drivers/gpu/drm/bridge/sii9234.c -- 2.7.4

3 7

[PATCH] drm/exynos/hdmi: add 85.5MHz pixel clock for v14 HDMI PHY
by Tobias Jakobi 11 Sep '17

11 Sep '17

From: Daniel Drake <drake(a)endlessm.com> Configuration details from Samsung. This enables 1366x768@60Hz, which also needs the 257px timing hack to work around a mixer limitation. Signed-off-by: Daniel Drake <drake(a)endlessm.com> Signed-off-by: Tobias Jakobi <tjakobi(a)math.uni-bielefeld.de> --- drivers/gpu/drm/exynos/exynos_hdmi.c | 9 +++++++++ drivers/gpu/drm/exynos/exynos_mixer.c | 1 + 2 files changed, 10 insertions(+) diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c index 2b14cc7def6f..adc69331dafe 100644 --- a/drivers/gpu/drm/exynos/exynos_hdmi.c +++ b/drivers/gpu/drm/exynos/exynos_hdmi.c @@ -298,6 +298,15 @@ static const struct hdmiphy_config hdmiphy_v14_configs[] = { }, }, { + .pixel_clock = 85500000, + .conf = { + 0x01, 0xd1, 0x24, 0x11, 0x40, 0x40, 0xd0, 0x08, + 0x84, 0xa0, 0xd6, 0xd8, 0x45, 0xa0, 0xac, 0x80, + 0x08, 0x80, 0x11, 0x04, 0x02, 0x22, 0x44, 0x86, + 0x54, 0x90, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80, + }, + }, + { .pixel_clock = 106500000, .conf = { 0x01, 0xd1, 0x2c, 0x12, 0x40, 0x0c, 0x09, 0x08, diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c index 96c3f61f080e..1dd65e261385 100644 --- a/drivers/gpu/drm/exynos/exynos_mixer.c +++ b/drivers/gpu/drm/exynos/exynos_mixer.c @@ -1127,6 +1127,7 @@ static int mixer_atomic_check(struct exynos_drm_crtc *crtc, /* Check against some specific resolutions. */ if ((w == 1024 && h == 768) || + (w == 1366 && h == 768) || (w == 1280 && h == 1024)) return 0; -- 2.13.5

1 0

[PATCH] drm/i915: Fix an error handling in 'intel_framebuffer_init()'
by Christophe JAILLET 11 Sep '17

11 Sep '17

We should go through the error handling path to decrease the 'framebuffer_references' as done everywhere else in this function. Fixes: 2e2adb05736c ("drm/i915: Add render decompression support") Signed-off-by: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> --- drivers/gpu/drm/i915/intel_display.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 7cd392f2cd94..478fa449003f 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -14077,7 +14077,7 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb, if (mode_cmd->handles[i] != mode_cmd->handles[0]) { DRM_DEBUG_KMS("bad plane %d handle\n", i); - return -EINVAL; + goto err; } stride_alignment = intel_fb_stride_alignment(fb, i); -- 2.11.0

2 1

[Bug 102616] [CI][SNB,HSW,APL,KBL] igt@gem_eio@in-flight - Failed assertion: __vgem_fence_signal(fd, fence) == 0
by bugzilla-daemon＠freedesktop.org 11 Sep '17

11 Sep '17

https://bugs.freedesktop.org/show_bug.cgi?id=102616 --- Comment #6 from Chris Wilson <chris(a)chris-wilson.co.uk> --- Related igt commit 4f082c35d2df545f81d202ae1a08463f6c123552 (upstream/master) Author: Chris Wilson <chris(a)chris-wilson.co.uk> Date: Fri Sep 8 13:48:05 2017 +0100 igt/gem_eio: Install an exithandler to unwedge the device after failure Under normal conditions, we try to repair the damage we inflict to the GPU, but if we fail we don't. Make sure that if the test does die, we do try to restore normal operation by using an atexit handler. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102616 Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler(a)intel.com> -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 102369] [CI][SNB] Failed assertion: !kmsg_check("*ERROR*") in igt@tools_test@check_dmesg
by bugzilla-daemon＠freedesktop.org 11 Sep '17

11 Sep '17

https://bugs.freedesktop.org/show_bug.cgi?id=102369 Martin Peres <martin.peres(a)free.fr> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED --- Comment #3 from Martin Peres <martin.peres(a)free.fr> --- Thanks Chris!We wanted to run the others bug not inflight, once the new test stops killing the kernel, we should be ok to run all of gem_eio (or right now if killing the kernel with an oops and -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 102369] [CI][SNB] Failed assertion: !kmsg_check("*ERROR*") in igt@tools_test@check_dmesg
by bugzilla-daemon＠freedesktop.org 11 Sep '17

11 Sep '17

1 0

[Bug 102655] [CI][HSW] igt@gem_flink_race@flink_close - Failed assertion: obj_count == 0
by bugzilla-daemon＠freedesktop.org 11 Sep '17

11 Sep '17

https://bugs.freedesktop.org/show_bug.cgi?id=102655 --- Comment #2 from Chris Wilson <chris(a)chris-wilson.co.uk> --- Given all the floating references held on objects, I'm not sure if we can do a "stable_obj_count" without the loop and sleeping. I'm not sure of all the places that we do have such objects in a timer, so getting the interval right or adding a flush control is hard. E.g. one thing we are not flushing before counting are the kms workqueues which hold a reference on the old object until finished. -- You are receiving this mail because: You are the assignee for the bug.

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel September 2017