OpenCloudOS-Kernel/include/drm/drmP.h

1070 lines
33 KiB
C
Raw Normal View History

/*
* Internal Header for the Direct Rendering Manager
*
* Copyright 1999 Precision Insight, Inc., Cedar Park, Texas.
* Copyright 2000 VA Linux Systems, Inc., Sunnyvale, California.
* Copyright (c) 2009-2010, Code Aurora Forum.
* All rights reserved.
*
* Author: Rickard E. (Rik) Faith <faith@valinux.com>
* Author: Gareth Hughes <gareth@valinux.com>
*
* 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
* VA LINUX SYSTEMS AND/OR ITS SUPPLIERS 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 _DRM_P_H_
#define _DRM_P_H_
#include <linux/agp_backend.h>
#include <linux/cdev.h>
#include <linux/dma-mapping.h>
#include <linux/file.h>
#include <linux/fs.h>
#include <linux/highmem.h>
#include <linux/idr.h>
#include <linux/init.h>
#include <linux/io.h>
#include <linux/jiffies.h>
#include <linux/kernel.h>
#include <linux/kref.h>
#include <linux/miscdevice.h>
#include <linux/mm.h>
#include <linux/mutex.h>
#include <linux/pci.h>
#include <linux/platform_device.h>
#include <linux/poll.h>
#include <linux/ratelimit.h>
#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/types.h>
#include <linux/vmalloc.h>
#include <linux/workqueue.h>
#include <linux/fence.h>
#include <asm/mman.h>
#include <asm/pgalloc.h>
#include <asm/uaccess.h>
#include <uapi/drm/drm.h>
#include <uapi/drm/drm_mode.h>
#include <drm/drm_agpsupport.h>
#include <drm/drm_crtc.h>
#include <drm/drm_fourcc.h>
#include <drm/drm_global.h>
#include <drm/drm_hashtab.h>
#include <drm/drm_mem_util.h>
#include <drm/drm_mm.h>
#include <drm/drm_os_linux.h>
#include <drm/drm_sarea.h>
#include <drm/drm_vma_manager.h>
struct module;
struct drm_file;
struct drm_device;
struct drm_agp_head;
struct drm_local_map;
struct drm_device_dma;
struct drm_dma_handle;
struct drm_gem_object;
struct drm_master;
struct drm_vblank_crtc;
struct device_node;
struct videomode;
struct reservation_object;
struct dma_buf_attachment;
/*
* The following categories are defined:
*
* CORE: Used in the generic drm code: drm_ioctl.c, drm_mm.c, drm_memory.c, ...
* This is the category used by the DRM_DEBUG() macro.
*
* DRIVER: Used in the vendor specific part of the driver: i915, radeon, ...
* This is the category used by the DRM_DEBUG_DRIVER() macro.
*
* KMS: used in the modesetting code.
* This is the category used by the DRM_DEBUG_KMS() macro.
*
* PRIME: used in the prime code.
* This is the category used by the DRM_DEBUG_PRIME() macro.
*
* ATOMIC: used in the atomic code.
* This is the category used by the DRM_DEBUG_ATOMIC() macro.
*
* VBL: used for verbose debug message in the vblank code
* This is the category used by the DRM_DEBUG_VBL() macro.
*
* Enabling verbose debug messages is done through the drm.debug parameter,
* each category being enabled by a bit.
*
* drm.debug=0x1 will enable CORE messages
* drm.debug=0x2 will enable DRIVER messages
* drm.debug=0x3 will enable CORE and DRIVER messages
* ...
* drm.debug=0x3f will enable all messages
*
* An interesting feature is that it's possible to enable verbose logging at
* run-time by echoing the debug value in its sysfs node:
* # echo 0xf > /sys/module/drm/parameters/debug
*/
#define DRM_UT_CORE 0x01
#define DRM_UT_DRIVER 0x02
#define DRM_UT_KMS 0x04
#define DRM_UT_PRIME 0x08
#define DRM_UT_ATOMIC 0x10
#define DRM_UT_VBL 0x20
2009-06-02 14:09:47 +08:00
extern __printf(2, 3)
void drm_ut_debug_printk(const char *function_name,
const char *format, ...);
extern __printf(1, 2)
void drm_err(const char *format, ...);
/***********************************************************************/
/** \name DRM template customization defaults */
/*@{*/
/* driver capabilities and requirements mask */
#define DRIVER_USE_AGP 0x1
#define DRIVER_LEGACY 0x2
#define DRIVER_PCI_DMA 0x8
#define DRIVER_SG 0x10
#define DRIVER_HAVE_DMA 0x20
#define DRIVER_HAVE_IRQ 0x40
#define DRIVER_IRQ_SHARED 0x80
#define DRIVER_GEM 0x1000
#define DRIVER_MODESET 0x2000
#define DRIVER_PRIME 0x4000
#define DRIVER_RENDER 0x8000
#define DRIVER_ATOMIC 0x10000
#define DRIVER_KMS_LEGACY_CONTEXT 0x20000
/***********************************************************************/
/** \name Macros to make printk easier */
/*@{*/
/**
* Error output.
*
* \param fmt printf() like format string.
* \param arg arguments
*/
#define DRM_ERROR(fmt, ...) \
drm_err(fmt, ##__VA_ARGS__)
/**
* Rate limited error output. Like DRM_ERROR() but won't flood the log.
*
* \param fmt printf() like format string.
* \param arg arguments
*/
#define DRM_ERROR_RATELIMITED(fmt, ...) \
({ \
static DEFINE_RATELIMIT_STATE(_rs, \
DEFAULT_RATELIMIT_INTERVAL, \
DEFAULT_RATELIMIT_BURST); \
\
if (__ratelimit(&_rs)) \
drm_err(fmt, ##__VA_ARGS__); \
})
#define DRM_INFO(fmt, ...) \
printk(KERN_INFO "[" DRM_NAME "] " fmt, ##__VA_ARGS__)
#define DRM_INFO_ONCE(fmt, ...) \
printk_once(KERN_INFO "[" DRM_NAME "] " fmt, ##__VA_ARGS__)
/**
* Debug output.
*
* \param fmt printf() like format string.
* \param arg arguments
*/
2009-06-02 14:09:47 +08:00
#define DRM_DEBUG(fmt, args...) \
do { \
if (unlikely(drm_debug & DRM_UT_CORE)) \
drm_ut_debug_printk(__func__, fmt, ##args); \
2009-06-02 14:09:47 +08:00
} while (0)
#define DRM_DEBUG_DRIVER(fmt, args...) \
2009-06-02 14:09:47 +08:00
do { \
if (unlikely(drm_debug & DRM_UT_DRIVER)) \
drm_ut_debug_printk(__func__, fmt, ##args); \
2009-06-02 14:09:47 +08:00
} while (0)
#define DRM_DEBUG_KMS(fmt, args...) \
2009-06-02 14:09:47 +08:00
do { \
if (unlikely(drm_debug & DRM_UT_KMS)) \
drm_ut_debug_printk(__func__, fmt, ##args); \
2009-06-02 14:09:47 +08:00
} while (0)
#define DRM_DEBUG_PRIME(fmt, args...) \
do { \
if (unlikely(drm_debug & DRM_UT_PRIME)) \
drm_ut_debug_printk(__func__, fmt, ##args); \
} while (0)
#define DRM_DEBUG_ATOMIC(fmt, args...) \
do { \
if (unlikely(drm_debug & DRM_UT_ATOMIC)) \
drm_ut_debug_printk(__func__, fmt, ##args); \
} while (0)
#define DRM_DEBUG_VBL(fmt, args...) \
do { \
if (unlikely(drm_debug & DRM_UT_VBL)) \
drm_ut_debug_printk(__func__, fmt, ##args); \
} while (0)
#define _DRM_DEFINE_DEBUG_RATELIMITED(level, fmt, args...) \
do { \
if (unlikely(drm_debug & DRM_UT_ ## level)) { \
static DEFINE_RATELIMIT_STATE( \
_rs, \
DEFAULT_RATELIMIT_INTERVAL, \
DEFAULT_RATELIMIT_BURST); \
\
if (__ratelimit(&_rs)) { \
drm_ut_debug_printk(__func__, fmt, \
##args); \
} \
} \
} while (0)
/**
* Rate limited debug output. Like DRM_DEBUG() but won't flood the log.
*
* \param fmt printf() like format string.
* \param arg arguments
*/
#define DRM_DEBUG_RATELIMITED(fmt, args...) \
_DRM_DEFINE_DEBUG_RATELIMITED(CORE, fmt, ##args)
#define DRM_DEBUG_DRIVER_RATELIMITED(fmt, args...) \
_DRM_DEFINE_DEBUG_RATELIMITED(DRIVER, fmt, ##args)
#define DRM_DEBUG_KMS_RATELIMITED(fmt, args...) \
_DRM_DEFINE_DEBUG_RATELIMITED(KMS, fmt, ##args)
#define DRM_DEBUG_PRIME_RATELIMITED(fmt, args...) \
_DRM_DEFINE_DEBUG_RATELIMITED(PRIME, fmt, ##args)
/*@}*/
/***********************************************************************/
/** \name Internal types and structures */
/*@{*/
#define DRM_IF_VERSION(maj, min) (maj << 16 | min)
/**
* Ioctl function type.
*
* \param inode device inode.
* \param file_priv DRM file private pointer.
* \param cmd command.
* \param arg argument.
*/
typedef int drm_ioctl_t(struct drm_device *dev, void *data,
struct drm_file *file_priv);
drm: 32/64-bit DRM ioctl compatibility patch The patch is against a 2.6.11 kernel tree. I am running this with a 32-bit X server (compiled up from X.org CVS as of a couple of weeks ago) and 32-bit DRI libraries and clients. All the userland stuff is identical to what I am using under a 32-bit kernel on my G4 powerbook (which is a 32-bit machine of course). I haven't tried compiling up a 64-bit X server or clients yet. In the compatibility routines I have assumed that the kernel can safely access user addresses after set_fs(KERNEL_DS). That is, where an ioctl argument structure contains pointers to other structures, and those other structures are already compatible between the 32-bit and 64-bit ABIs (i.e. they only contain things like chars, shorts or ints), I just check the address with access_ok() and then pass it through to the 64-bit ioctl code. I believe this approach may not work on sparc64, but it does work on ppc64 and x86_64 at least. One tricky area which may need to be revisited is the question of how to handle the handles which we pass back to userspace to identify mappings. These handles are generated in the ADDMAP ioctl and then passed in as the offset value to mmap. However, offset values for mmap seem to be generated in other ways as well, particularly for AGP mappings. The approach I have ended up with is to generate a fake 32-bit handle only for _DRM_SHM mappings. The handles for other mappings (AGP, REG, FB) are physical addresses which are already limited to 32 bits, and generating fake handles for them created all sorts of problems in the mmap/nopage code. This patch has been updated to use the new compatibility ioctls. From: Paul Mackerras <paulus@samba.org> Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 19:29:18 +08:00
typedef int drm_ioctl_compat_t(struct file *filp, unsigned int cmd,
unsigned long arg);
#define DRM_IOCTL_NR(n) _IOC_NR(n)
#define DRM_MAJOR 226
#define DRM_AUTH 0x1
#define DRM_MASTER 0x2
#define DRM_ROOT_ONLY 0x4
#define DRM_CONTROL_ALLOW 0x8
#define DRM_UNLOCKED 0x10
drm: implement experimental render nodes Render nodes provide an API for userspace to use non-privileged GPU commands without any running DRM-Master. It is useful for offscreen rendering, GPGPU clients, and normal render clients which do not perform modesetting. Compared to legacy clients, render clients no longer need any authentication to perform client ioctls. Instead, user-space controls render/client access to GPUs via filesystem access-modes on the render-node. Once a render-node was opened, a client has full access to the client/render operations on the GPU. However, no modesetting or ioctls that affect global state are allowed on render nodes. To prevent privilege-escalation, drivers must explicitly state that they support render nodes. They must mark their render-only ioctls as DRM_RENDER_ALLOW so render clients can use them. Furthermore, they must support clients without any attached master. If filesystem access-modes are not enough for fine-grained access control to render nodes (very unlikely, considering the versaitlity of FS-ACLs), you may still fall-back to fd-passing from server to client (which allows arbitrary access-control). However, note that revoking access is currently impossible and unlikely to get implemented. Note: Render clients no longer have any associated DRM-Master as they are supposed to be independent of any server state. DRM core highly depends on file_priv->master to be non-NULL for modesetting/ctx/etc. commands. Therefore, drivers must be very careful to not require DRM-Master if they support DRIVER_RENDER. So far render-nodes are protected by "drm_rnodes". As long as this module-parameter is not set to 1, a driver will not create render nodes. This allows us to experiment with the API a bit before we stabilize it. v2: drop insecure GEM_FLINK to force use of dmabuf Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-08-26 00:29:00 +08:00
#define DRM_RENDER_ALLOW 0x20
struct drm_ioctl_desc {
unsigned int cmd;
int flags;
drm_ioctl_t *func;
drm: Use names of ioctls in debug traces The intention here is to make the output of dmesg with full verbosity a bit easier for a human to parse. This commit transforms: [drm:drm_ioctl], pid=699, cmd=0x6458, nr=0x58, dev 0xe200, auth=1 [drm:drm_ioctl], pid=699, cmd=0xc010645b, nr=0x5b, dev 0xe200, auth=1 [drm:drm_ioctl], pid=699, cmd=0xc0106461, nr=0x61, dev 0xe200, auth=1 [drm:drm_ioctl], pid=699, cmd=0xc01c64ae, nr=0xae, dev 0xe200, auth=1 [drm:drm_mode_addfb], [FB:32] [drm:drm_ioctl], pid=699, cmd=0xc0106464, nr=0x64, dev 0xe200, auth=1 [drm:drm_vm_open_locked], 0x7fd9302fe000,0x00a00000 [drm:drm_ioctl], pid=699, cmd=0x400c645f, nr=0x5f, dev 0xe200, auth=1 [drm:drm_ioctl], pid=699, cmd=0xc00464af, nr=0xaf, dev 0xe200, auth=1 [drm:intel_crtc_set_config], [CRTC:3] [NOFB] into: [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_THROTTLE [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_CREATE [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_SET_TILING [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, IOCTL_MODE_ADDFB [drm:drm_mode_addfb], [FB:32] [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_MMAP_GTT [drm:drm_vm_open_locked], 0x7fd9302fe000,0x00a00000 [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_SET_DOMAIN [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, DRM_IOCTL_MODE_RMFB [drm:intel_crtc_set_config], [CRTC:3] [NOFB] v2: drm_ioctls is now a constant (Ville Syrjälä) Signed-off-by: Chris Cummins <christopher.e.cummins@intel.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-05-09 21:20:40 +08:00
const char *name;
};
/**
* Creates a driver or general drm_ioctl_desc array entry for the given
* ioctl, for use by drm_ioctl().
*/
#define DRM_IOCTL_DEF_DRV(ioctl, _func, _flags) \
[DRM_IOCTL_NR(DRM_IOCTL_##ioctl) - DRM_COMMAND_BASE] = { \
.cmd = DRM_IOCTL_##ioctl, \
.func = _func, \
.flags = _flags, \
.name = #ioctl \
}
/* Event queued up for userspace to read */
struct drm_pending_event {
struct completion *completion;
struct drm_event *event;
struct fence *fence;
struct list_head link;
struct list_head pending_link;
struct drm_file *file_priv;
pid_t pid; /* pid of requester, no guarantee it's valid by the time
we deliver the event, for tracing only */
};
/* initial implementaton using a linked list - todo hashtab */
struct drm_prime_file_private {
struct list_head head;
struct mutex lock;
};
/** File private data */
struct drm_file {
unsigned authenticated :1;
/* true when the client has asked us to expose stereo 3D mode flags */
unsigned stereo_allowed :1;
/*
* true if client understands CRTC primary planes and cursor planes
* in the plane list
*/
unsigned universal_planes:1;
/* true if client understands atomic properties */
unsigned atomic:1;
/*
* This client is the creator of @master.
* Protected by struct drm_device::master_mutex.
*/
unsigned is_master:1;
struct pid *pid;
kuid_t uid;
drm_magic_t magic;
struct list_head lhead;
struct drm_minor *minor;
unsigned long lock_count;
/** Mapping of mm object handles to object pointers. */
struct idr object_idr;
/** Lock for synchronization of access to object_idr. */
spinlock_t table_lock;
struct file *filp;
void *driver_priv;
struct drm_master *master; /* master this node is currently associated with
2016-06-21 16:54:12 +08:00
N.B. not always dev->master */
drm: revamp locking around fb creation/destruction Well, at least step 1. The goal here is that framebuffer objects can survive outside of the mode_config lock, with just a reference held as protection. The first step to get there is to introduce a special fb_lock which protects fb lookup, creation and destruction, to make them appear atomic. This new fb_lock can nest within the mode_config lock. But the idea is (once the reference counting part is completed) that we only quickly take that fb_lock to lookup a framebuffer and grab a reference, without any other locks involved. vmwgfx is the only driver which does framebuffer lookups itself, also wrap those calls to drm_mode_object_find with the new lock. Also protect the fb_list walking in i915 and omapdrm with the new lock. As a slight complication there's also the list of user-created fbs attached to the file private. The problem now is that at fclose() time we need to walk that list, eventually do a modeset call to remove the fb from active usage (and are required to be able to take the mode_config lock), but in the end we need to grab the new fb_lock to remove the fb from the list. The easiest solution is to add another mutex to protect this per-file list. Currently that new fbs_lock nests within the modeset locks and so appears redudant. But later patches will switch around this sequence so that taking the modeset locks in the fb destruction path is optional in the fastpath. Ultimately the goal is that addfb and rmfb do not require the mode_config lock, since otherwise they have the potential to introduce stalls in the pageflip sequence of a compositor (if the compositor e.g. switches to a fullscreen client or if it enables a plane). But that requires a few more steps and hoops to jump through. Note that framebuffer creation/destruction is now double-protected - once by the fb_lock and in parts by the idr_lock. The later would be unnecessariy if framebuffers would have their own idr allocator. But that's material for another patch (series). v2: Properly initialize the fb->filp_head list in _init, otherwise the newly added WARN to check whether the fb isn't on a fpriv list any more will fail for driver-private objects. v3: Fixup two error-case unlock bugs spotted by Richard Wilbur. Reviewed-by: Rob Clark <rob@ti.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2012-12-11 04:19:18 +08:00
/**
* fbs - List of framebuffers associated with this file.
*
* Protected by fbs_lock. Note that the fbs list holds a reference on
* the fb object to prevent it from untimely disappearing.
*/
struct list_head fbs;
drm: revamp locking around fb creation/destruction Well, at least step 1. The goal here is that framebuffer objects can survive outside of the mode_config lock, with just a reference held as protection. The first step to get there is to introduce a special fb_lock which protects fb lookup, creation and destruction, to make them appear atomic. This new fb_lock can nest within the mode_config lock. But the idea is (once the reference counting part is completed) that we only quickly take that fb_lock to lookup a framebuffer and grab a reference, without any other locks involved. vmwgfx is the only driver which does framebuffer lookups itself, also wrap those calls to drm_mode_object_find with the new lock. Also protect the fb_list walking in i915 and omapdrm with the new lock. As a slight complication there's also the list of user-created fbs attached to the file private. The problem now is that at fclose() time we need to walk that list, eventually do a modeset call to remove the fb from active usage (and are required to be able to take the mode_config lock), but in the end we need to grab the new fb_lock to remove the fb from the list. The easiest solution is to add another mutex to protect this per-file list. Currently that new fbs_lock nests within the modeset locks and so appears redudant. But later patches will switch around this sequence so that taking the modeset locks in the fb destruction path is optional in the fastpath. Ultimately the goal is that addfb and rmfb do not require the mode_config lock, since otherwise they have the potential to introduce stalls in the pageflip sequence of a compositor (if the compositor e.g. switches to a fullscreen client or if it enables a plane). But that requires a few more steps and hoops to jump through. Note that framebuffer creation/destruction is now double-protected - once by the fb_lock and in parts by the idr_lock. The later would be unnecessariy if framebuffers would have their own idr allocator. But that's material for another patch (series). v2: Properly initialize the fb->filp_head list in _init, otherwise the newly added WARN to check whether the fb isn't on a fpriv list any more will fail for driver-private objects. v3: Fixup two error-case unlock bugs spotted by Richard Wilbur. Reviewed-by: Rob Clark <rob@ti.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2012-12-11 04:19:18 +08:00
struct mutex fbs_lock;
/** User-created blob properties; this retains a reference on the
* property. */
struct list_head blobs;
wait_queue_head_t event_wait;
struct list_head pending_event_list;
struct list_head event_list;
int event_space;
struct mutex event_read_lock;
struct drm_prime_file_private prime;
};
/**
* Lock data.
*/
struct drm_lock_data {
struct drm_hw_lock *hw_lock; /**< Hardware lock */
/** Private of lock holder's file (NULL=kernel) */
struct drm_file *file_priv;
wait_queue_head_t lock_queue; /**< Queue of blocked processes */
unsigned long lock_time; /**< Time of last lock in jiffies */
spinlock_t spinlock;
uint32_t kernel_waiters;
uint32_t user_waiters;
int idle_has_lock;
};
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
/* Flags and return codes for get_vblank_timestamp() driver function. */
#define DRM_CALLED_FROM_VBLIRQ 1
#define DRM_VBLANKTIME_SCANOUTPOS_METHOD (1 << 0)
#define DRM_VBLANKTIME_IN_VBLANK (1 << 1)
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
/* get_scanout_position() return flags */
#define DRM_SCANOUTPOS_VALID (1 << 0)
#define DRM_SCANOUTPOS_IN_VBLANK (1 << 1)
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
#define DRM_SCANOUTPOS_ACCURATE (1 << 2)
/**
* DRM driver structure. This structure represent the common code for
* a family of cards. There will one drm_device for each card present
* in this family
*/
struct drm_driver {
int (*load) (struct drm_device *, unsigned long flags);
int (*firstopen) (struct drm_device *);
int (*open) (struct drm_device *, struct drm_file *);
void (*preclose) (struct drm_device *, struct drm_file *file_priv);
void (*postclose) (struct drm_device *, struct drm_file *);
void (*lastclose) (struct drm_device *);
int (*unload) (struct drm_device *);
int (*dma_ioctl) (struct drm_device *dev, void *data, struct drm_file *file_priv);
int (*dma_quiescent) (struct drm_device *);
int (*context_dtor) (struct drm_device *dev, int context);
int (*set_busid)(struct drm_device *dev, struct drm_master *master);
/**
* get_vblank_counter - get raw hardware vblank counter
* @dev: DRM device
* @pipe: counter to fetch
*
* Driver callback for fetching a raw hardware vblank counter for @crtc.
* If a device doesn't have a hardware counter, the driver can simply
* use drm_vblank_no_hw_counter() function. The DRM core will account for
* missed vblank events while interrupts where disabled based on system
* timestamps.
*
* Wraparound handling and loss of events due to modesetting is dealt
* with in the DRM core code.
*
* RETURNS
* Raw vblank counter value.
*/
u32 (*get_vblank_counter) (struct drm_device *dev, unsigned int pipe);
/**
* enable_vblank - enable vblank interrupt events
* @dev: DRM device
* @pipe: which irq to enable
*
* Enable vblank interrupts for @crtc. If the device doesn't have
* a hardware vblank counter, the driver should use the
* drm_vblank_no_hw_counter() function that keeps a virtual counter.
*
* RETURNS
* Zero on success, appropriate errno if the given @crtc's vblank
* interrupt cannot be enabled.
*/
int (*enable_vblank) (struct drm_device *dev, unsigned int pipe);
/**
* disable_vblank - disable vblank interrupt events
* @dev: DRM device
* @pipe: which irq to enable
*
* Disable vblank interrupts for @crtc. If the device doesn't have
* a hardware vblank counter, the driver should use the
* drm_vblank_no_hw_counter() function that keeps a virtual counter.
*/
void (*disable_vblank) (struct drm_device *dev, unsigned int pipe);
/**
* Called by \c drm_device_is_agp. Typically used to determine if a
* card is really attached to AGP or not.
*
* \param dev DRM device handle
*
* \returns
* One of three values is returned depending on whether or not the
* card is absolutely \b not AGP (return of 0), absolutely \b is AGP
* (return of 1), or may or may not be AGP (return of 2).
*/
int (*device_is_agp) (struct drm_device *dev);
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
/**
* Called by vblank timestamping code.
*
drm: Push latency sensitive bits of vblank scanoutpos timestamping into kms drivers. A change in locking of some kms drivers (currently intel-kms) make the old approach too inaccurate and also incompatible with the PREEMPT_RT realtime kernel patchset. The driver->get_scanout_position() method of intel-kms now needs to aquire a spinlock, which clashes badly with the former preempt_disable() calls in the drm, and it also introduces larger delays and timing uncertainty on a contended lock than acceptable. This patch changes the prototype of driver->get_scanout_position() to require/allow kms drivers to perform the ktime_get() system time queries which go along with actual scanout position readout in a way that provides maximum precision and to return those timestamps to the drm. kms drivers implementations of get_scanout_position() are asked to implement timestamping and scanoutpos readout in a way that is as precise as possible and compatible with preempt_disable() on a PREMPT_RT kernel. A driver should follow this pattern in get_scanout_position() for precision and compatibility: spin_lock...(...); preempt_disable_rt(); // On a PREEMPT_RT kernel, otherwise omit. if (stime) *stime = ktime_get(); ... Minimum amount of MMIO register reads to get scanout position ... ... no taking of locks allowed here! ... if (etime) *etime = ktime_get(); preempt_enable_rt(); // On PREEMPT_RT kernel, otherwise omit. spin_unlock...(...); v2: Fix formatting of new multi-line code comments. Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-10-30 12:13:06 +08:00
* Return the current display scanout position from a crtc, and an
* optional accurate ktime_get timestamp of when position was measured.
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
*
* \param dev DRM device.
* \param pipe Id of the crtc to query.
* \param flags Flags from the caller (DRM_CALLED_FROM_VBLIRQ or 0).
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
* \param *vpos Target location for current vertical scanout position.
* \param *hpos Target location for current horizontal scanout position.
drm: Push latency sensitive bits of vblank scanoutpos timestamping into kms drivers. A change in locking of some kms drivers (currently intel-kms) make the old approach too inaccurate and also incompatible with the PREEMPT_RT realtime kernel patchset. The driver->get_scanout_position() method of intel-kms now needs to aquire a spinlock, which clashes badly with the former preempt_disable() calls in the drm, and it also introduces larger delays and timing uncertainty on a contended lock than acceptable. This patch changes the prototype of driver->get_scanout_position() to require/allow kms drivers to perform the ktime_get() system time queries which go along with actual scanout position readout in a way that provides maximum precision and to return those timestamps to the drm. kms drivers implementations of get_scanout_position() are asked to implement timestamping and scanoutpos readout in a way that is as precise as possible and compatible with preempt_disable() on a PREMPT_RT kernel. A driver should follow this pattern in get_scanout_position() for precision and compatibility: spin_lock...(...); preempt_disable_rt(); // On a PREEMPT_RT kernel, otherwise omit. if (stime) *stime = ktime_get(); ... Minimum amount of MMIO register reads to get scanout position ... ... no taking of locks allowed here! ... if (etime) *etime = ktime_get(); preempt_enable_rt(); // On PREEMPT_RT kernel, otherwise omit. spin_unlock...(...); v2: Fix formatting of new multi-line code comments. Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-10-30 12:13:06 +08:00
* \param *stime Target location for timestamp taken immediately before
* scanout position query. Can be NULL to skip timestamp.
* \param *etime Target location for timestamp taken immediately after
* scanout position query. Can be NULL to skip timestamp.
* \param mode Current display timings.
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
*
* Returns vpos as a positive number while in active scanout area.
* Returns vpos as a negative number inside vblank, counting the number
* of scanlines to go until end of vblank, e.g., -1 means "one scanline
* until start of active scanout / end of vblank."
*
* \return Flags, or'ed together as follows:
*
* DRM_SCANOUTPOS_VALID = Query successful.
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
* DRM_SCANOUTPOS_INVBL = Inside vblank.
* DRM_SCANOUTPOS_ACCURATE = Returned position is accurate. A lack of
* this flag means that returned position may be offset by a constant
* but unknown small number of scanlines wrt. real scanout position.
*
*/
int (*get_scanout_position) (struct drm_device *dev, unsigned int pipe,
unsigned int flags, int *vpos, int *hpos,
ktime_t *stime, ktime_t *etime,
const struct drm_display_mode *mode);
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
/**
* Called by \c drm_get_last_vbltimestamp. Should return a precise
* timestamp when the most recent VBLANK interval ended or will end.
*
* Specifically, the timestamp in @vblank_time should correspond as
* closely as possible to the time when the first video scanline of
* the video frame after the end of VBLANK will start scanning out,
* the time immediately after end of the VBLANK interval. If the
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
* @crtc is currently inside VBLANK, this will be a time in the future.
* If the @crtc is currently scanning out a frame, this will be the
* past start time of the current scanout. This is meant to adhere
* to the OpenML OML_sync_control extension specification.
*
* \param dev dev DRM device handle.
* \param pipe crtc for which timestamp should be returned.
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
* \param *max_error Maximum allowable timestamp error in nanoseconds.
* Implementation should strive to provide timestamp
* with an error of at most *max_error nanoseconds.
* Returns true upper bound on error for timestamp.
* \param *vblank_time Target location for returned vblank timestamp.
* \param flags 0 = Defaults, no special treatment needed.
* \param DRM_CALLED_FROM_VBLIRQ = Function is called from vblank
* irq handler. Some drivers need to apply some workarounds
* for gpu-specific vblank irq quirks if flag is set.
*
* \returns
* Zero if timestamping isn't supported in current display mode or a
* negative number on failure. A positive status code on success,
* which describes how the vblank_time timestamp was computed.
*/
int (*get_vblank_timestamp) (struct drm_device *dev, unsigned int pipe,
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
int *max_error,
struct timeval *vblank_time,
unsigned flags);
/* these have to be filled in */
irqreturn_t(*irq_handler) (int irq, void *arg);
void (*irq_preinstall) (struct drm_device *dev);
int (*irq_postinstall) (struct drm_device *dev);
void (*irq_uninstall) (struct drm_device *dev);
/* Master routines */
int (*master_create)(struct drm_device *dev, struct drm_master *master);
void (*master_destroy)(struct drm_device *dev, struct drm_master *master);
/**
* master_set is called whenever the minor master is set.
* master_drop is called whenever the minor master is dropped.
*/
int (*master_set)(struct drm_device *dev, struct drm_file *file_priv,
bool from_open);
void (*master_drop)(struct drm_device *dev, struct drm_file *file_priv);
int (*debugfs_init)(struct drm_minor *minor);
void (*debugfs_cleanup)(struct drm_minor *minor);
/**
* @gem_free_object: deconstructor for drm_gem_objects
*
* This is deprecated and should not be used by new drivers. Use
* @gem_free_object_unlocked instead.
*/
void (*gem_free_object) (struct drm_gem_object *obj);
/**
* @gem_free_object_unlocked: deconstructor for drm_gem_objects
*
* This is for drivers which are not encumbered with dev->struct_mutex
* legacy locking schemes. Use this hook instead of @gem_free_object.
*/
void (*gem_free_object_unlocked) (struct drm_gem_object *obj);
int (*gem_open_object) (struct drm_gem_object *, struct drm_file *);
void (*gem_close_object) (struct drm_gem_object *, struct drm_file *);
/**
* Hook for allocating the GEM object struct, for use by core
* helpers.
*/
struct drm_gem_object *(*gem_create_object)(struct drm_device *dev,
size_t size);
/* prime: */
/* export handle -> fd (see drm_gem_prime_handle_to_fd() helper) */
int (*prime_handle_to_fd)(struct drm_device *dev, struct drm_file *file_priv,
uint32_t handle, uint32_t flags, int *prime_fd);
/* import fd -> handle (see drm_gem_prime_fd_to_handle() helper) */
int (*prime_fd_to_handle)(struct drm_device *dev, struct drm_file *file_priv,
int prime_fd, uint32_t *handle);
/* export GEM -> dmabuf */
struct dma_buf * (*gem_prime_export)(struct drm_device *dev,
struct drm_gem_object *obj, int flags);
/* import dmabuf -> GEM */
struct drm_gem_object * (*gem_prime_import)(struct drm_device *dev,
struct dma_buf *dma_buf);
drm: add prime helpers Instead of reimplementing all of the dma_buf functionality in every driver, create helpers drm_prime_import and drm_prime_export that implement them in terms of new, lower-level hook functions: gem_prime_pin: callback when a buffer is created, used to pin buffers into GTT gem_prime_get_sg_table: convert a drm_gem_object to an sg_table for export gem_prime_import_sg_table: convert an sg_table into a drm_gem_object gem_prime_vmap, gem_prime_vunmap: map and unmap an object These hooks are optional; drivers can opt in by using drm_gem_prime_import and drm_gem_prime_export as the .gem_prime_import and .gem_prime_export fields of struct drm_driver. v2: - Drop .begin_cpu_access. None of the drivers this code replaces implemented it. Having it here was a leftover from when I was trying to include i915 in this rework. - Use mutex_lock instead of mutex_lock_interruptible, as these three drivers did. This patch series shouldn't change that behavior. - Rename helpers to gem_prime_get_sg_table and gem_prime_import_sg_table. Rename struct sg_table* variables to 'sgt' for clarity. - Update drm.tmpl for these new hooks. v3: - Pass the vaddr down to the driver. This lets drivers that just call vunmap on the pointer avoid having to store the pointer in their GEM private structures. - Move documentation into a /** DOC */ comment in drm_prime.c and include it in drm.tmpl with a !P line. I tried to use !F lines to include documentation of the individual functions from drmP.h, but the docproc / kernel-doc scripts barf on that file, so hopefully this is good enough for now. - apply refcount fix from commit be8a42ae60addd8b6092535c11b42d099d6470ec ("drm/prime: drop reference on imported dma-buf come from gem") Signed-off-by: Aaron Plattner <aplattner@nvidia.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: David Airlie <airlied@linux.ie> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-01-16 04:47:42 +08:00
/* low-level interface used by drm_gem_prime_{import,export} */
int (*gem_prime_pin)(struct drm_gem_object *obj);
void (*gem_prime_unpin)(struct drm_gem_object *obj);
struct reservation_object * (*gem_prime_res_obj)(
struct drm_gem_object *obj);
drm: add prime helpers Instead of reimplementing all of the dma_buf functionality in every driver, create helpers drm_prime_import and drm_prime_export that implement them in terms of new, lower-level hook functions: gem_prime_pin: callback when a buffer is created, used to pin buffers into GTT gem_prime_get_sg_table: convert a drm_gem_object to an sg_table for export gem_prime_import_sg_table: convert an sg_table into a drm_gem_object gem_prime_vmap, gem_prime_vunmap: map and unmap an object These hooks are optional; drivers can opt in by using drm_gem_prime_import and drm_gem_prime_export as the .gem_prime_import and .gem_prime_export fields of struct drm_driver. v2: - Drop .begin_cpu_access. None of the drivers this code replaces implemented it. Having it here was a leftover from when I was trying to include i915 in this rework. - Use mutex_lock instead of mutex_lock_interruptible, as these three drivers did. This patch series shouldn't change that behavior. - Rename helpers to gem_prime_get_sg_table and gem_prime_import_sg_table. Rename struct sg_table* variables to 'sgt' for clarity. - Update drm.tmpl for these new hooks. v3: - Pass the vaddr down to the driver. This lets drivers that just call vunmap on the pointer avoid having to store the pointer in their GEM private structures. - Move documentation into a /** DOC */ comment in drm_prime.c and include it in drm.tmpl with a !P line. I tried to use !F lines to include documentation of the individual functions from drmP.h, but the docproc / kernel-doc scripts barf on that file, so hopefully this is good enough for now. - apply refcount fix from commit be8a42ae60addd8b6092535c11b42d099d6470ec ("drm/prime: drop reference on imported dma-buf come from gem") Signed-off-by: Aaron Plattner <aplattner@nvidia.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: David Airlie <airlied@linux.ie> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-01-16 04:47:42 +08:00
struct sg_table *(*gem_prime_get_sg_table)(struct drm_gem_object *obj);
struct drm_gem_object *(*gem_prime_import_sg_table)(
struct drm_device *dev,
struct dma_buf_attachment *attach,
drm: add prime helpers Instead of reimplementing all of the dma_buf functionality in every driver, create helpers drm_prime_import and drm_prime_export that implement them in terms of new, lower-level hook functions: gem_prime_pin: callback when a buffer is created, used to pin buffers into GTT gem_prime_get_sg_table: convert a drm_gem_object to an sg_table for export gem_prime_import_sg_table: convert an sg_table into a drm_gem_object gem_prime_vmap, gem_prime_vunmap: map and unmap an object These hooks are optional; drivers can opt in by using drm_gem_prime_import and drm_gem_prime_export as the .gem_prime_import and .gem_prime_export fields of struct drm_driver. v2: - Drop .begin_cpu_access. None of the drivers this code replaces implemented it. Having it here was a leftover from when I was trying to include i915 in this rework. - Use mutex_lock instead of mutex_lock_interruptible, as these three drivers did. This patch series shouldn't change that behavior. - Rename helpers to gem_prime_get_sg_table and gem_prime_import_sg_table. Rename struct sg_table* variables to 'sgt' for clarity. - Update drm.tmpl for these new hooks. v3: - Pass the vaddr down to the driver. This lets drivers that just call vunmap on the pointer avoid having to store the pointer in their GEM private structures. - Move documentation into a /** DOC */ comment in drm_prime.c and include it in drm.tmpl with a !P line. I tried to use !F lines to include documentation of the individual functions from drmP.h, but the docproc / kernel-doc scripts barf on that file, so hopefully this is good enough for now. - apply refcount fix from commit be8a42ae60addd8b6092535c11b42d099d6470ec ("drm/prime: drop reference on imported dma-buf come from gem") Signed-off-by: Aaron Plattner <aplattner@nvidia.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: David Airlie <airlied@linux.ie> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-01-16 04:47:42 +08:00
struct sg_table *sgt);
void *(*gem_prime_vmap)(struct drm_gem_object *obj);
void (*gem_prime_vunmap)(struct drm_gem_object *obj, void *vaddr);
int (*gem_prime_mmap)(struct drm_gem_object *obj,
struct vm_area_struct *vma);
/* vga arb irq handler */
void (*vgaarb_irq)(struct drm_device *dev, bool state);
/* dumb alloc support */
int (*dumb_create)(struct drm_file *file_priv,
struct drm_device *dev,
struct drm_mode_create_dumb *args);
int (*dumb_map_offset)(struct drm_file *file_priv,
struct drm_device *dev, uint32_t handle,
uint64_t *offset);
int (*dumb_destroy)(struct drm_file *file_priv,
struct drm_device *dev,
uint32_t handle);
/* Driver private ops for this object */
const struct vm_operations_struct *gem_vm_ops;
int major;
int minor;
int patchlevel;
char *name;
char *desc;
char *date;
u32 driver_features;
int dev_priv_size;
const struct drm_ioctl_desc *ioctls;
int num_ioctls;
const struct file_operations *fops;
/* List of devices hanging off this driver with stealth attach. */
struct list_head legacy_dev_list;
};
enum drm_minor_type {
DRM_MINOR_PRIMARY,
DRM_MINOR_CONTROL,
DRM_MINOR_RENDER,
DRM_MINOR_CNT,
};
/**
* Info file list entry. This structure represents a debugfs or proc file to
* be created by the drm core
*/
struct drm_info_list {
const char *name; /** file name */
int (*show)(struct seq_file*, void*); /** show callback */
u32 driver_features; /**< Required driver features for this entry */
void *data;
};
/**
* debugfs node structure. This structure represents a debugfs file.
*/
struct drm_info_node {
struct list_head list;
struct drm_minor *minor;
const struct drm_info_list *info_ent;
struct dentry *dent;
};
/**
* DRM minor structure. This structure represents a drm minor number.
*/
struct drm_minor {
int index; /**< Minor device number */
int type; /**< Control or render */
struct device *kdev; /**< Linux device */
struct drm_device *dev;
struct dentry *debugfs_root;
struct list_head debugfs_list;
struct mutex debugfs_lock; /* Protects debugfs_list. */
};
/**
* DRM device structure. This structure represent a complete card that
* may contain multiple heads.
*/
struct drm_device {
struct list_head legacy_dev_list;/**< list of devices per driver for stealth attach cleanup */
int if_version; /**< Highest interface version set */
/** \name Lifetime Management */
/*@{ */
struct kref ref; /**< Object ref-count */
struct device *dev; /**< Device structure of bus-device */
struct drm_driver *driver; /**< DRM driver managing the device */
void *dev_private; /**< DRM driver private data */
struct drm_minor *control; /**< Control node */
struct drm_minor *primary; /**< Primary node */
struct drm_minor *render; /**< Render node */
2016-06-21 16:54:12 +08:00
/* currently active master for this device. Protected by master_mutex */
struct drm_master *master;
atomic_t unplugged; /**< Flag whether dev is dead */
struct inode *anon_inode; /**< inode for private address-space */
char *unique; /**< unique name of the device */
/*@} */
/** \name Locks */
/*@{ */
struct mutex struct_mutex; /**< For others */
struct mutex master_mutex; /**< For drm_minor::master and drm_file::is_master */
/*@} */
/** \name Usage Counters */
/*@{ */
drm: Rip out totally bogus vga_switcheroo->can_switch locking So I just wanted to add a new field to struct drm_device and accidentally stumbled over something. According to comments dev->open_count is protected by dev->count_lock, but that's totally not the case. It's protected by drm_global_mutex. Unfortunately the vga switcheroo callbacks took this comment at face value. The problem is that we can't just take the drm_global_mutex because: - It would lead to a locking inversion with the driver load/unload paths. - It wouldn't actually protect anything, for that we'd need to wrap the entire vga switcheroo code in the drm_global_mutex. And I'm not sure whether that would actually solve anything. What we probably want is a try_to_grab_switcheroo reference kind of thing which is used in the driver's ->open callback. Then we could move all that ->can_switch madness into the vga switcheroo core where it really belongs. But since that would amount to real work take the easy way out and just add a comment. It's definitely not going to make anything worse since doing switcheroo state changes while restarting X just isn't recommended. Even though the delayed switching code does exactly that. v2: - Simplify the ->can_switch implementations more (Thierry) - Fix comment about the dev->open_count locking (Thierry) Cc: Thierry Reding <treding@nvidia.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> (v1) Reviewed-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-11-04 03:46:34 +08:00
int open_count; /**< Outstanding files open, protected by drm_global_mutex. */
spinlock_t buf_lock; /**< For drm_device::buf_use and a few other things. */
int buf_use; /**< Buffers in use -- cannot alloc */
atomic_t buf_alloc; /**< Buffer allocation in progress */
/*@} */
struct mutex filelist_mutex;
struct list_head filelist;
/** \name Memory management */
/*@{ */
struct list_head maplist; /**< Linked list of regions */
struct drm_open_hash map_hash; /**< User token hash table for maps */
/** \name Context handle management */
/*@{ */
struct list_head ctxlist; /**< Linked list of context handles */
struct mutex ctxlist_mutex; /**< For ctxlist */
struct idr ctx_idr;
struct list_head vmalist; /**< List of vmas (for debugging) */
/*@} */
/** \name DMA support */
/*@{ */
struct drm_device_dma *dma; /**< Optional pointer for DMA support */
/*@} */
/** \name Context support */
/*@{ */
__volatile__ long context_flag; /**< Context swapping flag */
int last_context; /**< Last current context */
/*@} */
/** \name VBLANK IRQ support */
/*@{ */
bool irq_enabled;
int irq;
/*
* If true, vblank interrupt will be disabled immediately when the
* refcount drops to zero, as opposed to via the vblank disable
* timer.
* This can be set to true it the hardware has a working vblank
* counter and the driver uses drm_vblank_on() and drm_vblank_off()
* appropriately.
*/
bool vblank_disable_immediate;
/* array of size num_crtcs */
struct drm_vblank_crtc *vblank;
drm/vblank: Add support for precise vblank timestamping. The DRI2 swap & sync implementation needs precise vblank counts and precise timestamps corresponding to those vblank counts. For conformance to the OpenML OML_sync_control extension specification the DRM timestamp associated with a vblank count should correspond to the start of video scanout of the first scanline of the video frame following the vblank interval for that vblank count. Therefore we need to carry around precise timestamps for vblanks. Currently the DRM and KMS drivers generate timestamps ad-hoc via do_gettimeofday() in some places. The resulting timestamps are sometimes not very precise due to interrupt handling delays, they don't conform to OML_sync_control and some are wrong, as they aren't taken synchronized to the vblank. This patch implements support inside the drm core for precise and robust timestamping. It consists of the following interrelated pieces. 1. Vblank timestamp caching: A per-crtc ringbuffer stores the most recent vblank timestamps corresponding to vblank counts. The ringbuffer can be read out lock-free via the accessor function: struct timeval timestamp; vblankcount = drm_vblank_count_and_time(dev, crtcid, &timestamp). The function returns the current vblank count and the corresponding timestamp for start of video scanout following the vblank interval. It can be used anywhere between enclosing drm_vblank_get(dev, crtcid) and drm_vblank_put(dev,crtcid) statements. It is used inside the drmWaitVblank ioctl and in the vblank event queueing and handling. It should be used by kms drivers for timestamping of bufferswap completion. The timestamp ringbuffer is reinitialized each time vblank irq's get reenabled in drm_vblank_get()/ drm_update_vblank_count(). It is invalidated when vblank irq's get disabled. The ringbuffer is updated inside drm_handle_vblank() at each vblank irq. 2. Calculation of precise vblank timestamps: drm_get_last_vbltimestamp() is used to compute the timestamp for the end of the most recent vblank (if inside active scanout), or the expected end of the current vblank interval (if called inside a vblank interval). The function calls into a new optional kms driver entry point dev->driver->get_vblank_timestamp() which is supposed to provide the precise timestamp. If a kms driver doesn't implement the entry point or if the call fails, a simple do_gettimeofday() timestamp is returned as crude approximation of the true vblank time. A new drm module parameter drm.timestamp_precision_usec allows to disable high precision timestamps (if set to zero) or to specify the maximum acceptable error in the timestamps in microseconds. Kms drivers could implement their get_vblank_timestamp() function in a gpu specific way, as long as returned timestamps conform to OML_sync_control, e.g., by use of gpu specific hardware timestamps. Optionally, kms drivers can simply wrap and use the new utility function drm_calc_vbltimestamp_from_scanoutpos(). This function calls a new optional kms driver function dev->driver->get_scanout_position() which returns the current horizontal and vertical video scanout position of the crtc. The scanout position together with the drm_display_timing of the current video mode is used to calculate elapsed time relative to start of active scanout for the current video frame. This elapsed time is subtracted from the current do_gettimeofday() time to get the timestamp corresponding to start of video scanout. Currently non-interlaced, non-doublescan video modes, with or without panel scaling are handled correctly. Interlaced/ doublescan modes are tbd in a future patch. 3. Filtering of redundant vblank irq's and removal of some race-conditions in the vblank irq enable/disable path: Some gpu's (e.g., Radeon R500/R600) send spurious vblank irq's outside the vblank if vblank irq's get reenabled. These get detected by use of the vblank timestamps and filtered out to avoid miscounting of vblanks. Some race-conditions between the vblank irq enable/disable functions, the vblank irq handler and the gpu itself (updating its hardware vblank counter in the "wrong" moment) are fixed inside vblank_disable_and_save() and drm_update_vblank_count() by use of the vblank timestamps and a new spinlock dev->vblank_time_lock. The time until vblank irq disable is now configurable via a new drm module parameter drm.vblankoffdelay to allow experimentation with timeouts that are much shorter than the current 5 seconds and should allow longer vblank off periods for better power savings. Followup patches will use these new functions to implement precise timestamping for the intel and radeon kms drivers. Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-10-23 10:20:23 +08:00
spinlock_t vblank_time_lock; /**< Protects vblank count and time updates during vblank enable/disable */
spinlock_t vbl_lock;
u32 max_vblank_count; /**< size of vblank counter register */
/**
* List of events
*/
struct list_head vblank_event_list;
spinlock_t event_lock;
/*@} */
struct drm_agp_head *agp; /**< AGP data */
struct pci_dev *pdev; /**< PCI device structure */
#ifdef __alpha__
struct pci_controller *hose;
#endif
struct platform_device *platformdev; /**< Platform device struture */
struct virtio_device *virtdev;
struct drm_sg_mem *sg; /**< Scatter gather memory */
unsigned int num_crtcs; /**< Number of CRTCs on this device */
struct {
int context;
struct drm_hw_lock *lock;
} sigdata;
struct drm_local_map *agp_buffer_map;
unsigned int agp_buffer_token;
struct drm_mode_config mode_config; /**< Current mode config */
/** \name GEM information */
/*@{ */
struct mutex object_name_lock;
struct idr object_name_idr;
struct drm_vma_offset_manager *vma_offset_manager;
/*@} */
int switch_power_state;
};
#include <drm/drm_irq.h>
#define DRM_SWITCH_POWER_ON 0
#define DRM_SWITCH_POWER_OFF 1
#define DRM_SWITCH_POWER_CHANGING 2
#define DRM_SWITCH_POWER_DYNAMIC_OFF 3
static __inline__ int drm_core_check_feature(struct drm_device *dev,
int feature)
{
return ((dev->driver->driver_features & feature) ? 1 : 0);
}
static inline void drm_device_set_unplugged(struct drm_device *dev)
{
smp_wmb();
atomic_set(&dev->unplugged, 1);
}
static inline int drm_device_is_unplugged(struct drm_device *dev)
{
int ret = atomic_read(&dev->unplugged);
smp_rmb();
return ret;
}
static inline bool drm_is_render_client(const struct drm_file *file_priv)
drm: implement experimental render nodes Render nodes provide an API for userspace to use non-privileged GPU commands without any running DRM-Master. It is useful for offscreen rendering, GPGPU clients, and normal render clients which do not perform modesetting. Compared to legacy clients, render clients no longer need any authentication to perform client ioctls. Instead, user-space controls render/client access to GPUs via filesystem access-modes on the render-node. Once a render-node was opened, a client has full access to the client/render operations on the GPU. However, no modesetting or ioctls that affect global state are allowed on render nodes. To prevent privilege-escalation, drivers must explicitly state that they support render nodes. They must mark their render-only ioctls as DRM_RENDER_ALLOW so render clients can use them. Furthermore, they must support clients without any attached master. If filesystem access-modes are not enough for fine-grained access control to render nodes (very unlikely, considering the versaitlity of FS-ACLs), you may still fall-back to fd-passing from server to client (which allows arbitrary access-control). However, note that revoking access is currently impossible and unlikely to get implemented. Note: Render clients no longer have any associated DRM-Master as they are supposed to be independent of any server state. DRM core highly depends on file_priv->master to be non-NULL for modesetting/ctx/etc. commands. Therefore, drivers must be very careful to not require DRM-Master if they support DRIVER_RENDER. So far render-nodes are protected by "drm_rnodes". As long as this module-parameter is not set to 1, a driver will not create render nodes. This allows us to experiment with the API a bit before we stabilize it. v2: drop insecure GEM_FLINK to force use of dmabuf Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-08-26 00:29:00 +08:00
{
return file_priv->minor->type == DRM_MINOR_RENDER;
}
static inline bool drm_is_control_client(const struct drm_file *file_priv)
{
return file_priv->minor->type == DRM_MINOR_CONTROL;
}
static inline bool drm_is_primary_client(const struct drm_file *file_priv)
{
return file_priv->minor->type == DRM_MINOR_PRIMARY;
}
/******************************************************************/
/** \name Internal function definitions */
/*@{*/
/* Driver support (drm_drv.h) */
extern int drm_ioctl_permit(u32 flags, struct drm_file *file_priv);
extern long drm_ioctl(struct file *filp,
unsigned int cmd, unsigned long arg);
extern long drm_compat_ioctl(struct file *filp,
unsigned int cmd, unsigned long arg);
extern bool drm_ioctl_flags(unsigned int nr, unsigned int *flags);
/* File Operations (drm_fops.c) */
int drm_open(struct inode *inode, struct file *filp);
ssize_t drm_read(struct file *filp, char __user *buffer,
size_t count, loff_t *offset);
int drm_release(struct inode *inode, struct file *filp);
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,
struct drm_event *e);
void drm_event_cancel_free(struct drm_device *dev,
struct drm_pending_event *p);
void drm_send_event_locked(struct drm_device *dev, struct drm_pending_event *e);
void drm_send_event(struct drm_device *dev, struct drm_pending_event *e);
/* Misc. IOCTL support (drm_ioctl.c) */
int drm_noop(struct drm_device *dev, void *data,
struct drm_file *file_priv);
int drm_invalid_op(struct drm_device *dev, void *data,
struct drm_file *file_priv);
/* Cache management (drm_cache.c) */
void drm_clflush_pages(struct page *pages[], unsigned long num_pages);
void drm_clflush_sg(struct sg_table *st);
void drm_clflush_virt_range(void *addr, unsigned long length);
/*
* These are exported to drivers so that they can implement fencing using
* DMA quiscent + idle. DMA quiescent usually requires the hardware lock.
*/
/* Modesetting support */
extern void drm_vblank_pre_modeset(struct drm_device *dev, unsigned int pipe);
extern void drm_vblank_post_modeset(struct drm_device *dev, unsigned int pipe);
/* drm_drv.c */
void drm_put_dev(struct drm_device *dev);
void drm_unplug_dev(struct drm_device *dev);
extern unsigned int drm_debug;
/* Debugfs support */
#if defined(CONFIG_DEBUG_FS)
extern int drm_debugfs_create_files(const struct drm_info_list *files,
int count, struct dentry *root,
struct drm_minor *minor);
extern int drm_debugfs_remove_files(const struct drm_info_list *files,
int count, struct drm_minor *minor);
#else
static inline int drm_debugfs_create_files(const struct drm_info_list *files,
int count, struct dentry *root,
struct drm_minor *minor)
{
return 0;
}
static inline int drm_debugfs_remove_files(const struct drm_info_list *files,
int count, struct drm_minor *minor)
{
return 0;
}
#endif
drm: add prime helpers Instead of reimplementing all of the dma_buf functionality in every driver, create helpers drm_prime_import and drm_prime_export that implement them in terms of new, lower-level hook functions: gem_prime_pin: callback when a buffer is created, used to pin buffers into GTT gem_prime_get_sg_table: convert a drm_gem_object to an sg_table for export gem_prime_import_sg_table: convert an sg_table into a drm_gem_object gem_prime_vmap, gem_prime_vunmap: map and unmap an object These hooks are optional; drivers can opt in by using drm_gem_prime_import and drm_gem_prime_export as the .gem_prime_import and .gem_prime_export fields of struct drm_driver. v2: - Drop .begin_cpu_access. None of the drivers this code replaces implemented it. Having it here was a leftover from when I was trying to include i915 in this rework. - Use mutex_lock instead of mutex_lock_interruptible, as these three drivers did. This patch series shouldn't change that behavior. - Rename helpers to gem_prime_get_sg_table and gem_prime_import_sg_table. Rename struct sg_table* variables to 'sgt' for clarity. - Update drm.tmpl for these new hooks. v3: - Pass the vaddr down to the driver. This lets drivers that just call vunmap on the pointer avoid having to store the pointer in their GEM private structures. - Move documentation into a /** DOC */ comment in drm_prime.c and include it in drm.tmpl with a !P line. I tried to use !F lines to include documentation of the individual functions from drmP.h, but the docproc / kernel-doc scripts barf on that file, so hopefully this is good enough for now. - apply refcount fix from commit be8a42ae60addd8b6092535c11b42d099d6470ec ("drm/prime: drop reference on imported dma-buf come from gem") Signed-off-by: Aaron Plattner <aplattner@nvidia.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: David Airlie <airlied@linux.ie> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-01-16 04:47:42 +08:00
extern struct dma_buf *drm_gem_prime_export(struct drm_device *dev,
struct drm_gem_object *obj, int flags);
extern int drm_gem_prime_handle_to_fd(struct drm_device *dev,
struct drm_file *file_priv, uint32_t handle, uint32_t flags,
int *prime_fd);
drm: add prime helpers Instead of reimplementing all of the dma_buf functionality in every driver, create helpers drm_prime_import and drm_prime_export that implement them in terms of new, lower-level hook functions: gem_prime_pin: callback when a buffer is created, used to pin buffers into GTT gem_prime_get_sg_table: convert a drm_gem_object to an sg_table for export gem_prime_import_sg_table: convert an sg_table into a drm_gem_object gem_prime_vmap, gem_prime_vunmap: map and unmap an object These hooks are optional; drivers can opt in by using drm_gem_prime_import and drm_gem_prime_export as the .gem_prime_import and .gem_prime_export fields of struct drm_driver. v2: - Drop .begin_cpu_access. None of the drivers this code replaces implemented it. Having it here was a leftover from when I was trying to include i915 in this rework. - Use mutex_lock instead of mutex_lock_interruptible, as these three drivers did. This patch series shouldn't change that behavior. - Rename helpers to gem_prime_get_sg_table and gem_prime_import_sg_table. Rename struct sg_table* variables to 'sgt' for clarity. - Update drm.tmpl for these new hooks. v3: - Pass the vaddr down to the driver. This lets drivers that just call vunmap on the pointer avoid having to store the pointer in their GEM private structures. - Move documentation into a /** DOC */ comment in drm_prime.c and include it in drm.tmpl with a !P line. I tried to use !F lines to include documentation of the individual functions from drmP.h, but the docproc / kernel-doc scripts barf on that file, so hopefully this is good enough for now. - apply refcount fix from commit be8a42ae60addd8b6092535c11b42d099d6470ec ("drm/prime: drop reference on imported dma-buf come from gem") Signed-off-by: Aaron Plattner <aplattner@nvidia.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: David Airlie <airlied@linux.ie> Signed-off-by: Dave Airlie <airlied@redhat.com>
2013-01-16 04:47:42 +08:00
extern struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
struct dma_buf *dma_buf);
extern int drm_gem_prime_fd_to_handle(struct drm_device *dev,
struct drm_file *file_priv, int prime_fd, uint32_t *handle);
extern void drm_gem_dmabuf_release(struct dma_buf *dma_buf);
extern int drm_prime_sg_to_page_addr_arrays(struct sg_table *sgt, struct page **pages,
dma_addr_t *addrs, int max_pages);
extern struct sg_table *drm_prime_pages_to_sg(struct page **pages, unsigned int nr_pages);
extern void drm_prime_gem_destroy(struct drm_gem_object *obj, struct sg_table *sg);
extern struct drm_dma_handle *drm_pci_alloc(struct drm_device *dev, size_t size,
size_t align);
extern void drm_pci_free(struct drm_device *dev, struct drm_dma_handle * dmah);
/* sysfs support (drm_sysfs.c) */
extern void drm_sysfs_hotplug_event(struct drm_device *dev);
struct drm_device *drm_dev_alloc(struct drm_driver *driver,
struct device *parent);
int drm_dev_init(struct drm_device *dev,
struct drm_driver *driver,
struct device *parent);
void drm_dev_ref(struct drm_device *dev);
void drm_dev_unref(struct drm_device *dev);
int drm_dev_register(struct drm_device *dev, unsigned long flags);
void drm_dev_unregister(struct drm_device *dev);
struct drm_minor *drm_minor_acquire(unsigned int minor_id);
void drm_minor_release(struct drm_minor *minor);
/*@}*/
/* PCI section */
static __inline__ int drm_pci_device_is_agp(struct drm_device *dev)
{
if (dev->driver->device_is_agp != NULL) {
int err = (*dev->driver->device_is_agp) (dev);
if (err != 2) {
return err;
}
}
return pci_find_capability(dev->pdev, PCI_CAP_ID_AGP);
}
void drm_pci_agp_destroy(struct drm_device *dev);
extern int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver);
extern void drm_pci_exit(struct drm_driver *driver, struct pci_driver *pdriver);
#ifdef CONFIG_PCI
extern int drm_get_pci_dev(struct pci_dev *pdev,
const struct pci_device_id *ent,
struct drm_driver *driver);
extern int drm_pci_set_busid(struct drm_device *dev, struct drm_master *master);
#else
static inline int drm_get_pci_dev(struct pci_dev *pdev,
const struct pci_device_id *ent,
struct drm_driver *driver)
{
return -ENOSYS;
}
static inline int drm_pci_set_busid(struct drm_device *dev,
struct drm_master *master)
{
return -ENOSYS;
}
#endif
#define DRM_PCIE_SPEED_25 1
#define DRM_PCIE_SPEED_50 2
#define DRM_PCIE_SPEED_80 4
extern int drm_pcie_get_speed_cap_mask(struct drm_device *dev, u32 *speed_mask);
extern int drm_pcie_get_max_link_width(struct drm_device *dev, u32 *mlw);
/* platform section */
extern int drm_platform_init(struct drm_driver *driver, struct platform_device *platform_device);
/* returns true if currently okay to sleep */
static __inline__ bool drm_can_sleep(void)
{
if (in_atomic() || in_dbg_master() || irqs_disabled())
return false;
return true;
}
/* helper for handling conditionals in various for_each macros */
#define for_each_if(condition) if (!(condition)) {} else
#endif