dri-devel January 2016

dri-devel@lists.freedesktop.org

194 participants
1200 discussions

[PULL] drm-amdkfd-fixes
by Oded Gabbay 28 Jan '16

28 Jan '16

Hi Dave, Just two small static analyzer code fixes for 4.5 Thanks, Oded The following changes since commit 92e963f50fc74041b5e9e744c330dca48e04f08d: Linux 4.5-rc1 (2016-01-24 13:06:47 -0800) are available in the git repository at: git://people.freedesktop.org/~gabbayo/linux tags/drm-amdkfd-fixes-2016-01-28 for you to fetch changes up to 642f0f2a1563b5ab6f5de896f602177ac1f06652: drm/amdkfd: Remove unnecessary cast in kfree (2016-01-28 14:40:11 +0200) ---------------------------------------------------------------- Amitoj Kaur Chawla (1): drm/amdkfd: Remove unnecessary cast in kfree Colin Ian King (1): drm/amdgpu: fix non-ANSI declaration of amdgpu_amdkfd_gfx_*_get_functions() drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c | 2 +- drivers/gpu/drm/amd/amdkfd/kfd_process.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)

1 0

[PATCH] gpu: drm: amd: amdkfd: Remove unnecessary cast in kfree
by Amitoj Kaur Chawla 28 Jan '16

28 Jan '16

Remove an unnecassary cast in the argument to kfree. Found using Coccinelle. The semantic patch used to find this is as follows: //<smpl> @@ type T; expression *f; @@ - kfree((T *)(f)); + kfree(f); //</smpl> Signed-off-by: Amitoj Kaur Chawla <amitoj1606(a)gmail.com> --- drivers/gpu/drm/amd/amdkfd/kfd_process.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 9be0070..a902ae0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -194,7 +194,7 @@ static void kfd_process_wq_release(struct work_struct *work) kfree(p); - kfree((void *)work); + kfree(work); } static void kfd_process_destroy_delayed(struct rcu_head *rcu) -- 1.9.1

2 2

[PATCH] drm/amdgpu: fix non-ANSI declaration of amdgpu_amdkfd_gfx_*_get_functions()
by Colin King 28 Jan '16

28 Jan '16

From: Colin Ian King <colin.king(a)canonical.com> amdgpu_amdkfd_gfx_7_get_functions and amdgpu_amdkfd_gfx_8_0_get_functions have no parameters, so use the normal void parameter convention to make them match their prototypes in the header file drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h Signed-off-by: Colin Ian King <colin.king(a)canonical.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c index 0e13763..362bedc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c @@ -154,7 +154,7 @@ static const struct kfd2kgd_calls kfd2kgd = { .get_fw_version = get_fw_version }; -struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions() +struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void) { return (struct kfd2kgd_calls *)&kfd2kgd; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c index 79fa5c7..04b744d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c @@ -115,7 +115,7 @@ static const struct kfd2kgd_calls kfd2kgd = { .get_fw_version = get_fw_version }; -struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions() +struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void) { return (struct kfd2kgd_calls *)&kfd2kgd; } -- 2.7.0.rc3

2 1

[PATCH 1/3] vc4: Add the DRM header file.
by Eric Anholt 28 Jan '16

28 Jan '16

I'll build some libdrm C code soon, but for lets libdrm users use vc4 ioctls. Produced from headers_install of 1df59b8497f47495e873c23abd6d3d290c730505 (drm-next) in the kernel. Signed-off-by: Eric Anholt <eric(a)anholt.net> --- Makefile.sources | 1 + include/drm/vc4_drm.h | 279 ++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 280 insertions(+) create mode 100644 include/drm/vc4_drm.h diff --git a/Makefile.sources b/Makefile.sources index a77f48d..1a1f0fe 100644 --- a/Makefile.sources +++ b/Makefile.sources @@ -32,6 +32,7 @@ LIBDRM_INCLUDE_H_FILES := \ include/drm/savage_drm.h \ include/drm/sis_drm.h \ include/drm/tegra_drm.h \ + include/drm/vc4_drm.h \ include/drm/via_drm.h LIBDRM_INCLUDE_VMWGFX_H_FILES := \ diff --git a/include/drm/vc4_drm.h b/include/drm/vc4_drm.h new file mode 100644 index 0000000..da3caa0 --- /dev/null +++ b/include/drm/vc4_drm.h @@ -0,0 +1,279 @@ +/* + * Copyright © 2014-2015 Broadcom + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef _VC4_DRM_H_ +#define _VC4_DRM_H_ + +#include "drm.h" + +#define DRM_VC4_SUBMIT_CL 0x00 +#define DRM_VC4_WAIT_SEQNO 0x01 +#define DRM_VC4_WAIT_BO 0x02 +#define DRM_VC4_CREATE_BO 0x03 +#define DRM_VC4_MMAP_BO 0x04 +#define DRM_VC4_CREATE_SHADER_BO 0x05 +#define DRM_VC4_GET_HANG_STATE 0x06 + +#define DRM_IOCTL_VC4_SUBMIT_CL DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_SUBMIT_CL, struct drm_vc4_submit_cl) +#define DRM_IOCTL_VC4_WAIT_SEQNO DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_WAIT_SEQNO, struct drm_vc4_wait_seqno) +#define DRM_IOCTL_VC4_WAIT_BO DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_WAIT_BO, struct drm_vc4_wait_bo) +#define DRM_IOCTL_VC4_CREATE_BO DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_CREATE_BO, struct drm_vc4_create_bo) +#define DRM_IOCTL_VC4_MMAP_BO DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_MMAP_BO, struct drm_vc4_mmap_bo) +#define DRM_IOCTL_VC4_CREATE_SHADER_BO DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_CREATE_SHADER_BO, struct drm_vc4_create_shader_bo) +#define DRM_IOCTL_VC4_GET_HANG_STATE DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_GET_HANG_STATE, struct drm_vc4_get_hang_state) + +struct drm_vc4_submit_rcl_surface { + __u32 hindex; /* Handle index, or ~0 if not present. */ + __u32 offset; /* Offset to start of buffer. */ + /* + * Bits for either render config (color_write) or load/store packet. + * Bits should all be 0 for MSAA load/stores. + */ + __u16 bits; + +#define VC4_SUBMIT_RCL_SURFACE_READ_IS_FULL_RES (1 << 0) + __u16 flags; +}; + +/** + * struct drm_vc4_submit_cl - ioctl argument for submitting commands to the 3D + * engine. + * + * Drivers typically use GPU BOs to store batchbuffers / command lists and + * their associated state. However, because the VC4 lacks an MMU, we have to + * do validation of memory accesses by the GPU commands. If we were to store + * our commands in BOs, we'd need to do uncached readback from them to do the + * validation process, which is too expensive. Instead, userspace accumulates + * commands and associated state in plain memory, then the kernel copies the + * data to its own address space, and then validates and stores it in a GPU + * BO. + */ +struct drm_vc4_submit_cl { + /* Pointer to the binner command list. + * + * This is the first set of commands executed, which runs the + * coordinate shader to determine where primitives land on the screen, + * then writes out the state updates and draw calls necessary per tile + * to the tile allocation BO. + */ + __u64 bin_cl; + + /* Pointer to the shader records. + * + * Shader records are the structures read by the hardware that contain + * pointers to uniforms, shaders, and vertex attributes. The + * reference to the shader record has enough information to determine + * how many pointers are necessary (fixed number for shaders/uniforms, + * and an attribute count), so those BO indices into bo_handles are + * just stored as __u32s before each shader record passed in. + */ + __u64 shader_rec; + + /* Pointer to uniform data and texture handles for the textures + * referenced by the shader. + * + * For each shader state record, there is a set of uniform data in the + * order referenced by the record (FS, VS, then CS). Each set of + * uniform data has a __u32 index into bo_handles per texture + * sample operation, in the order the QPU_W_TMUn_S writes appear in + * the program. Following the texture BO handle indices is the actual + * uniform data. + * + * The individual uniform state blocks don't have sizes passed in, + * because the kernel has to determine the sizes anyway during shader + * code validation. + */ + __u64 uniforms; + __u64 bo_handles; + + /* Size in bytes of the binner command list. */ + __u32 bin_cl_size; + /* Size in bytes of the set of shader records. */ + __u32 shader_rec_size; + /* Number of shader records. + * + * This could just be computed from the contents of shader_records and + * the address bits of references to them from the bin CL, but it + * keeps the kernel from having to resize some allocations it makes. + */ + __u32 shader_rec_count; + /* Size in bytes of the uniform state. */ + __u32 uniforms_size; + + /* Number of BO handles passed in (size is that times 4). */ + __u32 bo_handle_count; + + /* RCL setup: */ + __u16 width; + __u16 height; + __u8 min_x_tile; + __u8 min_y_tile; + __u8 max_x_tile; + __u8 max_y_tile; + struct drm_vc4_submit_rcl_surface color_read; + struct drm_vc4_submit_rcl_surface color_write; + struct drm_vc4_submit_rcl_surface zs_read; + struct drm_vc4_submit_rcl_surface zs_write; + struct drm_vc4_submit_rcl_surface msaa_color_write; + struct drm_vc4_submit_rcl_surface msaa_zs_write; + __u32 clear_color[2]; + __u32 clear_z; + __u8 clear_s; + + __u32 pad:24; + +#define VC4_SUBMIT_CL_USE_CLEAR_COLOR (1 << 0) + __u32 flags; + + /* Returned value of the seqno of this render job (for the + * wait ioctl). + */ + __u64 seqno; +}; + +/** + * struct drm_vc4_wait_seqno - ioctl argument for waiting for + * DRM_VC4_SUBMIT_CL completion using its returned seqno. + * + * timeout_ns is the timeout in nanoseconds, where "0" means "don't + * block, just return the status." + */ +struct drm_vc4_wait_seqno { + __u64 seqno; + __u64 timeout_ns; +}; + +/** + * struct drm_vc4_wait_bo - ioctl argument for waiting for + * completion of the last DRM_VC4_SUBMIT_CL on a BO. + * + * This is useful for cases where multiple processes might be + * rendering to a BO and you want to wait for all rendering to be + * completed. + */ +struct drm_vc4_wait_bo { + __u32 handle; + __u32 pad; + __u64 timeout_ns; +}; + +/** + * struct drm_vc4_create_bo - ioctl argument for creating VC4 BOs. + * + * There are currently no values for the flags argument, but it may be + * used in a future extension. + */ +struct drm_vc4_create_bo { + __u32 size; + __u32 flags; + /** Returned GEM handle for the BO. */ + __u32 handle; + __u32 pad; +}; + +/** + * struct drm_vc4_mmap_bo - ioctl argument for mapping VC4 BOs. + * + * This doesn't actually perform an mmap. Instead, it returns the + * offset you need to use in an mmap on the DRM device node. This + * means that tools like valgrind end up knowing about the mapped + * memory. + * + * There are currently no values for the flags argument, but it may be + * used in a future extension. + */ +struct drm_vc4_mmap_bo { + /** Handle for the object being mapped. */ + __u32 handle; + __u32 flags; + /** offset into the drm node to use for subsequent mmap call. */ + __u64 offset; +}; + +/** + * struct drm_vc4_create_shader_bo - ioctl argument for creating VC4 + * shader BOs. + * + * Since allowing a shader to be overwritten while it's also being + * executed from would allow privlege escalation, shaders must be + * created using this ioctl, and they can't be mmapped later. + */ +struct drm_vc4_create_shader_bo { + /* Size of the data argument. */ + __u32 size; + /* Flags, currently must be 0. */ + __u32 flags; + + /* Pointer to the data. */ + __u64 data; + + /** Returned GEM handle for the BO. */ + __u32 handle; + /* Pad, must be 0. */ + __u32 pad; +}; + +struct drm_vc4_get_hang_state_bo { + __u32 handle; + __u32 paddr; + __u32 size; + __u32 pad; +}; + +/** + * struct drm_vc4_hang_state - ioctl argument for collecting state + * from a GPU hang for analysis. +*/ +struct drm_vc4_get_hang_state { + /** Pointer to array of struct drm_vc4_get_hang_state_bo. */ + __u64 bo; + /** + * On input, the size of the bo array. Output is the number + * of bos to be returned. + */ + __u32 bo_count; + + __u32 start_bin, start_render; + + __u32 ct0ca, ct0ea; + __u32 ct1ca, ct1ea; + __u32 ct0cs, ct1cs; + __u32 ct0ra0, ct1ra0; + + __u32 bpca, bpcs; + __u32 bpoa, bpos; + + __u32 vpmbase; + + __u32 dbge; + __u32 fdbgo; + __u32 fdbgb; + __u32 fdbgr; + __u32 fdbgs; + __u32 errstat; + + /* Pad that we may save more registers into in the future. */ + __u32 pad[16]; +}; + +#endif /* _VC4_DRM_H_ */ -- 2.6.4

4 11

[PATCH 01/15] drm/vblank: Use drm_event_reserve_init
by Daniel Vetter 28 Jan '16

28 Jan '16

Well we can't use that directly since that code must hold dev->event_lock already. Extract an _unlocked version. Embarrassingly I've totally forgotten about this patch and any kind of event-based vblank wait totally blew up, killing the kernel. Cc: Alex Deucher <alexander.deucher(a)amd.com> Cc: Daniel Stone <daniels(a)collabora.com> Cc: Laurent Pinchart <laurent.pinchart(a)ideasonboard.com> Signed-off-by: Daniel Vetter <daniel.vetter(a)intel.com> --- drivers/gpu/drm/drm_fops.c | 64 +++++++++++++++++++++++++++++++++++----------- drivers/gpu/drm/drm_irq.c | 10 +++----- include/drm/drmP.h | 4 +++ 3 files changed, 56 insertions(+), 22 deletions(-) diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c index e13501e3606e..eb6a02f78697 100644 --- a/drivers/gpu/drm/drm_fops.c +++ b/drivers/gpu/drm/drm_fops.c @@ -678,7 +678,7 @@ unsigned int drm_poll(struct file *filp, struct poll_table_struct *wait) EXPORT_SYMBOL(drm_poll); /** - * drm_event_reserve_init - init a DRM event and reserve space for it + * drm_event_reserve_init_locked - init a DRM event and reserve space for it * @dev: DRM device * @file_priv: DRM file private data * @p: tracking structure for the pending event @@ -694,24 +694,20 @@ EXPORT_SYMBOL(drm_poll); * If callers embedded @p into a larger structure it must be allocated with * kmalloc and @p must be the first member element. * + * This is the locked version of drm_event_reserve_init() for callers which + * already hold dev->event_lock. + * * RETURNS: * * 0 on success or a negative error code on failure. */ -int drm_event_reserve_init(struct drm_device *dev, - struct drm_file *file_priv, - struct drm_pending_event *p, - struct drm_event *e) +int drm_event_reserve_init_locked(struct drm_device *dev, + struct drm_file *file_priv, + struct drm_pending_event *p, + struct drm_event *e) { - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&dev->event_lock, flags); - - if (file_priv->event_space < e->length) { - ret = -ENOMEM; - goto out; - } + if (file_priv->event_space < e->length) + return -ENOMEM; file_priv->event_space -= e->length; @@ -721,8 +717,46 @@ int drm_event_reserve_init(struct drm_device *dev, /* we *could* pass this in as arg, but everyone uses kfree: */ p->destroy = (void (*) (struct drm_pending_event *)) kfree; -out: + return 0; +} +EXPORT_SYMBOL(drm_event_reserve_init_locked); + +/** + * drm_event_reserve_init - init a DRM event and reserve space for it + * @dev: DRM device + * @file_priv: DRM file private data + * @p: tracking structure for the pending event + * @e: actual event data to deliver to userspace + * + * This function prepares the passed in event for eventual delivery. If the event + * doesn't get delivered (because the IOCTL fails later on, before queuing up + * anything) then the even must be cancelled and freed using + * drm_event_cancel_free(). Successfully initialized events should be sent out + * using drm_send_event() or drm_send_event_locked() to signal completion of the + * asynchronous event to userspace. + * + * If callers embedded @p into a larger structure it must be allocated with + * kmalloc and @p must be the first member element. + * + * Callers which already hold dev->event_lock should use + * drm_event_reserve_init() instead. + * + * RETURNS: + * + * 0 on success or a negative error code on failure. + */ +int drm_event_reserve_init(struct drm_device *dev, + struct drm_file *file_priv, + struct drm_pending_event *p, + struct drm_event *e) +{ + unsigned long flags; + int ret; + + spin_lock_irqsave(&dev->event_lock, flags); + ret = drm_event_reserve_init_locked(dev, file_priv, p, e); spin_unlock_irqrestore(&dev->event_lock, flags); + return ret; } EXPORT_SYMBOL(drm_event_reserve_init); diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c index 4ec8bca643ac..aa602199cb5d 100644 --- a/drivers/gpu/drm/drm_irq.c +++ b/drivers/gpu/drm/drm_irq.c @@ -1598,9 +1598,6 @@ static int drm_queue_vblank_event(struct drm_device *dev, unsigned int pipe, e->event.base.type = DRM_EVENT_VBLANK; e->event.base.length = sizeof(e->event); e->event.user_data = vblwait->request.signal; - e->base.event = &e->event.base; - e->base.file_priv = file_priv; - e->base.destroy = (void (*) (struct drm_pending_event *)) kfree; spin_lock_irqsave(&dev->event_lock, flags); @@ -1616,12 +1613,11 @@ static int drm_queue_vblank_event(struct drm_device *dev, unsigned int pipe, goto err_unlock; } - if (file_priv->event_space < sizeof(e->event)) { - ret = -EBUSY; + ret = drm_event_reserve_init_locked(dev, file_priv, &e->base, &e->event); + + if (ret) goto err_unlock; - } - file_priv->event_space -= sizeof(e->event); seq = drm_vblank_count_and_time(dev, pipe, &now); if ((vblwait->request.type & _DRM_VBLANK_NEXTONMISS) && diff --git a/include/drm/drmP.h b/include/drm/drmP.h index 306ef32ec086..1b71852d0a55 100644 --- a/include/drm/drmP.h +++ b/include/drm/drmP.h @@ -926,6 +926,10 @@ ssize_t drm_read(struct file *filp, char __user *buffer, int drm_release(struct inode *inode, struct file *filp); int drm_new_set_master(struct drm_device *dev, struct drm_file *fpriv); unsigned int drm_poll(struct file *filp, struct poll_table_struct *wait); +int drm_event_reserve_init_locked(struct drm_device *dev, + struct drm_file *file_priv, + struct drm_pending_event *p, + struct drm_event *e); int drm_event_reserve_init(struct drm_device *dev, struct drm_file *file_priv, struct drm_pending_event *p, -- 2.7.0.rc3

2 16

[patch] drm/vmwgfx: fix a NULL dereference
by Dan Carpenter 28 Jan '16

28 Jan '16

We dereference "eaction->event" inside the call to drm_send_event_locked() so should hold off on setting it to NULL until afterward. Fixes: fb740cf2492c ("drm: Create drm_send_event helpers") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c index e0edf14..37c305b 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c @@ -880,8 +880,8 @@ static void vmw_event_fence_action_seq_passed(struct vmw_fence_action *action) } list_del_init(&eaction->fpriv_head); - eaction->event = NULL; drm_send_event_locked(dev, eaction->event); + eaction->event = NULL; spin_unlock_irqrestore(&dev->event_lock, irq_flags); }

2 1

[RESEND,V2] drm: fsl-dcu: Fix no fb check bug
by Meng Yi 28 Jan '16

28 Jan '16

For state->fb or state->crtc may be NULL in fsl_dcu_drm_plane_atomic_check function, if so, return 0. Signed-off-by: Meng Yi <meng.yi(a)nxp.com> Signed-off-by: Jianwei Wang <jianwei.wang.chn(a)gmail.com> --- change in v2: -Add state->crtc check -return 0 when state->fb or state->crtc is NULL, instead of -EINVAL Adviced by Daniel Stone drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c index 4b13cf9..8965580 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c @@ -41,6 +41,9 @@ static int fsl_dcu_drm_plane_atomic_check(struct drm_plane *plane, { struct drm_framebuffer *fb = state->fb; + if (!state->fb || !state->crtc) + return 0; + switch (fb->pixel_format) { case DRM_FORMAT_RGB565: case DRM_FORMAT_RGB888: @@ -85,9 +88,6 @@ static void fsl_dcu_drm_plane_atomic_update(struct drm_plane *plane, unsigned int alpha, bpp; int index, ret; - if (!fb) - return; - index = fsl_dcu_drm_plane_index(plane); if (index < 0) return; -- 2.1.0.27.g96db324

3 5

[PATCH] MAINTAINERS: update for Freescale DCU DRM driver
by Stefan Agner 28 Jan '16

28 Jan '16

Promote myself as new maintainer of the Freescale DCU DRM driver. Signed-off-by: Stefan Agner <stefan(a)agner.ch> --- This has been previously discussed privately. The original driver author and maintainer Jianwei does not work for Freescale anymore and can not carve out spare time to maintain the driver. As I started to make use of the driver on the Freescale Vybrid SoC and already posted several fixes and improvements, I am willing to maintain the driver. Thanks Jianwei for your work on that driver! Dave, I hope you are OK with that? I plan to send a pull request with the patches which have been on the ml lately, specifically the "Fix no fb check bug" and my fixes and enhancements patchset, along with this maintainer change, does that sound reasonable? There is one more pending patchset for this drivers, the TCON and Vybrid support patchset, but this is probably not ready yet. I also plan to post another patchset which enables multiple overlay planes and a cursor plane. -- Stefan MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index e9caa4b..3f0aea3 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3678,7 +3678,7 @@ F: include/drm/exynos* F: include/uapi/drm/exynos* DRM DRIVERS FOR FREESCALE DCU -M: Jianwei Wang <jianwei.wang.chn(a)gmail.com> +M: Stefan Agner <stefan(a)agner.ch> M: Alison Wang <alison.wang(a)freescale.com> L: dri-devel(a)lists.freedesktop.org S: Supported -- 2.6.4

3 3

[PATCH resend] kmstest: fix link error due to library ordering
by Reinette Chatre 27 Jan '16

27 Jan '16

Recent commit below introduced a new library to be linked (libutil.la): commit 0caf58a6cb82327a3f6a53f05dea8e02f1412a05 Author: Stefan Agner <stefan(a)agner.ch> Date: Sat Dec 19 21:52:58 2015 -0800 kmstest: Use util_open() Use the new util_open() helper instead of open-coding the method for finding a usable device. While at it, make the command-line interface more consistent with that of modetest by adding the -D and -M options. Signed-off-by: Stefan Agner <stefan(a)agner.ch> v2: correctly use util_open() - swap device, module Signed-off-by: Emil Velikov <emil.l.velikov(a)gmail.com> The ordering of the libraries in kmstest_LDADD is causing the following error on my build system: ld: ../../tests/util/.libs/libutil.a(libutil_la-kms.o): undefined reference to symbol 'drmOpen' ../../.libs/libdrm.so: error adding symbols: DSO missing from command line collect2: error: ld returned 1 exit status Makefile:436: recipe for target 'kmstest' failed make[3]: *** [kmstest] Error 1 The above error is resolved by switching the linking order. Cc: Stefan Agner <stefan(a)agner.ch> Cc: Kristian Høgsberg Kristensen <kristian.h.kristensen(a)intel.com> Signed-off-by: Reinette Chatre <reinette.chatre(a)intel.com> --- This is a resend since first attempt was made when I was not subscribed to mailing list and it only made it to cc list, not the mailing list. tests/kmstest/Makefile.am | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/kmstest/Makefile.am b/tests/kmstest/Makefile.am index 100662e..9083620 100644 --- a/tests/kmstest/Makefile.am +++ b/tests/kmstest/Makefile.am @@ -18,8 +18,8 @@ kmstest_SOURCES = \ kmstest_LDADD = \ $(top_builddir)/libdrm.la \ - $(top_builddir)/libkms/libkms.la \ - $(top_builddir)/tests/util/libutil.la + $(top_builddir)/tests/util/libutil.la \ + $(top_builddir)/libkms/libkms.la run: kmstest ./kmstest -- 2.4.3

4 4

[PATCH libdrm] Update/fix the {copy,commit}-headers targets
by Emil Velikov 27 Jan '16

27 Jan '16

As some headers do not reside in include/drm we need to tweak our rules, and exclude headers that shouldn't be distributed [XXX: clarify why ?]. To avoid the extra magic of diving into the kernel tree running `make headers_install', just sed out the only reason why we need it - __user. Cc: Alex Deucher <alexander.deucher(a)amd.com> Cc: Michel Dänzer <michel.daenzer(a)amd.com> Cc: Ben Skeggs <bskeggs(a)redhat.com> Cc: Dave Airlie <airlied(a)redhat.com> Cc: Daniel Stone <daniels(a)collabora.com> Cc: Daniel Vetter <daniel.vetter(a)ffwll.ch> Cc: Inki Dae <inki.dae(a)samsung.com> Cc: Rob Clark <robclark(a)freedesktop.org> Cc: Tomi Valkeinen <tomi.valkeinen(a)ti.com> Cc: Daniel Kurtz <djkurtz(a)chromium.org> Signed-off-by: Emil Velikov <emil.l.velikov(a)gmail.com> --- Gents, As one runs `make copy-headers' we get a massive diff stat (+1500/-800) and a handful of issues gets pointed out. Please let me know of your prefered solution (regardless if one will get to it soon) and if we should consider it a blocker (B) or not (N). Thanks Emil - (N) Header license miss-match - omap, msm, exynos. Update the kernel ones ? - (N) Broken compat ioctls - exynos (and the UMS drivers) - using unsigned int as opposed to _u32/64. Considering they're 32bit only, we can get away with 'breaking' the ABI and using the proper ones ? - (N, keep local for now) C++ compat - libdrm has a hack/workaround (virtual is a keyword in C++), which I'd like us to upstream plus some extern C wrappers. - (?) Missing UMS symbols - see _DRM_GEM - (?) Non C89 compliant (see DRM_DRAWABLE_CLIPRECTS) - do we still want/need that ? - (B) Using include <drm/...> as opposed to include "..." - drm.h, nouveau_drm.h. Should be fixed in kernel. - (N) ABI 'break' + drm - struct drm_mode_get_connector extra pad + tegra - struct drm_tegra_gem_mmap extra pad - (B) API break + drm - missing DRM_MODE_OBJECT_* + nouveau - missing (gs)etparam - both structs and macros. everything else is fine/unused. + radeon - RADEON_TILING_R600_NO_SCANOUT, CIK_TILE_MODE_COLOR_2D* and CIK_TILE_MODE_DEPTH_STENCIL_2D_TILESPLIT_* - quick grep shows no users + omap - struct drm_omap_get_base, DRM_OMAP_GET_BASE + IOCTL - (N) (unneeded?) API additions - nouveau's DRM_IOCTL_NOUVEAU_GEM_* - (N) __KERNEL__ condiditionals. Is it really an issue - sure if looks a bit spurious but that's about it. Makefile.am | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/Makefile.am b/Makefile.am index ca41508..6c71d3a 100644 --- a/Makefile.am +++ b/Makefile.am @@ -126,7 +126,14 @@ endif copy-headers : cp -r $(kernel_source)/include/uapi/drm/*.h $(top_srcdir)/include/drm/ + sed -i "s/__user //g" $(top_srcdir)/include/drm/*.h + mv $(top_srcdir)/include/drm/exynos_drm.h $(top_srcdir)/exynos/ + mv $(top_srcdir)/include/drm/msm_drm.h $(top_srcdir)/freedreno/msm/ + mv $(top_srcdir)/include/drm/omap_drm.h $(top_srcdir)/omap/ + rm $(top_srcdir)/include/drm/armada_drm.h + rm $(top_srcdir)/include/drm/etnaviv_drm.h + rm $(top_srcdir)/include/drm/i810_drm.h + rm $(top_srcdir)/include/drm/vc4_drm.h commit-headers : copy-headers - git add include/drm/*.h git commit -am "Copy headers from kernel $$(GIT_DIR=$(kernel_source)/.git git describe)" -- 2.6.2

3 4

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel January 2016