2005-04-17 06:20:36 +08:00
|
|
|
/**
|
2005-09-25 12:28:13 +08:00
|
|
|
* \file drm_irq.c
|
2005-04-17 06:20:36 +08:00
|
|
|
* IRQ support
|
|
|
|
*
|
|
|
|
* \author Rickard E. (Rik) Faith <faith@valinux.com>
|
|
|
|
* \author Gareth Hughes <gareth@valinux.com>
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Created: Fri Mar 19 14:30:16 1999 by faith@valinux.com
|
|
|
|
*
|
|
|
|
* Copyright 1999, 2000 Precision Insight, Inc., Cedar Park, Texas.
|
|
|
|
* Copyright 2000 VA Linux Systems, Inc., Sunnyvale, California.
|
|
|
|
* All Rights Reserved.
|
|
|
|
*
|
|
|
|
* 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.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "drmP.h"
|
2010-07-02 07:47:31 +08:00
|
|
|
#include "drm_trace.h"
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
#include <linux/interrupt.h> /* For task queue support */
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
|
|
|
#include <linux/slab.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-09-21 12:33:58 +08:00
|
|
|
#include <linux/vgaarb.h>
|
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, ×tamp).
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
|
|
|
|
|
|
|
/* Access macro for slots in vblank timestamp ringbuffer. */
|
|
|
|
#define vblanktimestamp(dev, crtc, count) ( \
|
|
|
|
(dev)->_vblank_time[(crtc) * DRM_VBLANKTIME_RBSIZE + \
|
|
|
|
((count) % DRM_VBLANKTIME_RBSIZE)])
|
|
|
|
|
|
|
|
/* Retry timestamp calculation up to 3 times to satisfy
|
|
|
|
* drm_timestamp_precision before giving up.
|
|
|
|
*/
|
|
|
|
#define DRM_TIMESTAMP_MAXRETRIES 3
|
|
|
|
|
|
|
|
/* Threshold in nanoseconds for detection of redundant
|
|
|
|
* vblank irq in drm_handle_vblank(). 1 msec should be ok.
|
|
|
|
*/
|
|
|
|
#define DRM_REDUNDANT_VBLIRQ_THRESH_NS 1000000
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/**
|
|
|
|
* Get interrupt from bus id.
|
2005-09-25 12:28:13 +08:00
|
|
|
*
|
2005-04-17 06:20:36 +08:00
|
|
|
* \param inode device inode.
|
2007-08-25 18:23:09 +08:00
|
|
|
* \param file_priv DRM file private.
|
2005-04-17 06:20:36 +08:00
|
|
|
* \param cmd command.
|
|
|
|
* \param arg user argument, pointing to a drm_irq_busid structure.
|
|
|
|
* \return zero on success or a negative number on failure.
|
2005-09-25 12:28:13 +08:00
|
|
|
*
|
2005-04-17 06:20:36 +08:00
|
|
|
* Finds the PCI device with the specified bus id and gets its IRQ number.
|
|
|
|
* This IOCTL is deprecated, and will now return EINVAL for any busid not equal
|
|
|
|
* to that of the device that this DRM instance attached to.
|
|
|
|
*/
|
2007-09-03 10:06:45 +08:00
|
|
|
int drm_irq_by_busid(struct drm_device *dev, void *data,
|
|
|
|
struct drm_file *file_priv)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2007-09-03 10:06:45 +08:00
|
|
|
struct drm_irq_busid *p = data;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2010-12-15 01:16:38 +08:00
|
|
|
if (!dev->driver->bus->irq_by_busid)
|
2010-05-28 03:40:25 +08:00
|
|
|
return -EINVAL;
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
|
|
|
|
return -EINVAL;
|
|
|
|
|
2010-12-15 01:16:38 +08:00
|
|
|
return dev->driver->bus->irq_by_busid(dev, p);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
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, ×tamp).
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
|
|
|
/*
|
|
|
|
* Clear vblank timestamp buffer for a crtc.
|
|
|
|
*/
|
|
|
|
static void clear_vblank_timestamps(struct drm_device *dev, int crtc)
|
|
|
|
{
|
|
|
|
memset(&dev->_vblank_time[crtc * DRM_VBLANKTIME_RBSIZE], 0,
|
|
|
|
DRM_VBLANKTIME_RBSIZE * sizeof(struct timeval));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Disable vblank irq's on crtc, make sure that last vblank count
|
|
|
|
* of hardware and corresponding consistent software vblank counter
|
|
|
|
* are preserved, even if there are any spurious vblank irq's after
|
|
|
|
* disable.
|
|
|
|
*/
|
|
|
|
static void vblank_disable_and_save(struct drm_device *dev, int crtc)
|
|
|
|
{
|
|
|
|
unsigned long irqflags;
|
|
|
|
u32 vblcount;
|
|
|
|
s64 diff_ns;
|
|
|
|
int vblrc;
|
|
|
|
struct timeval tvblank;
|
|
|
|
|
|
|
|
/* Prevent vblank irq processing while disabling vblank irqs,
|
|
|
|
* so no updates of timestamps or count can happen after we've
|
|
|
|
* disabled. Needed to prevent races in case of delayed irq's.
|
|
|
|
* Disable preemption, so vblank_time_lock is held as short as
|
|
|
|
* possible, even under a kernel with PREEMPT_RT patches.
|
|
|
|
*/
|
|
|
|
preempt_disable();
|
|
|
|
spin_lock_irqsave(&dev->vblank_time_lock, irqflags);
|
|
|
|
|
|
|
|
dev->driver->disable_vblank(dev, crtc);
|
|
|
|
dev->vblank_enabled[crtc] = 0;
|
|
|
|
|
|
|
|
/* No further vblank irq's will be processed after
|
|
|
|
* this point. Get current hardware vblank count and
|
|
|
|
* vblank timestamp, repeat until they are consistent.
|
|
|
|
*
|
|
|
|
* FIXME: There is still a race condition here and in
|
|
|
|
* drm_update_vblank_count() which can cause off-by-one
|
|
|
|
* reinitialization of software vblank counter. If gpu
|
|
|
|
* vblank counter doesn't increment exactly at the leading
|
|
|
|
* edge of a vblank interval, then we can lose 1 count if
|
|
|
|
* we happen to execute between start of vblank and the
|
|
|
|
* delayed gpu counter increment.
|
|
|
|
*/
|
|
|
|
do {
|
|
|
|
dev->last_vblank[crtc] = dev->driver->get_vblank_counter(dev, crtc);
|
|
|
|
vblrc = drm_get_last_vbltimestamp(dev, crtc, &tvblank, 0);
|
|
|
|
} while (dev->last_vblank[crtc] != dev->driver->get_vblank_counter(dev, crtc));
|
|
|
|
|
|
|
|
/* Compute time difference to stored timestamp of last vblank
|
|
|
|
* as updated by last invocation of drm_handle_vblank() in vblank irq.
|
|
|
|
*/
|
|
|
|
vblcount = atomic_read(&dev->_vblank_count[crtc]);
|
|
|
|
diff_ns = timeval_to_ns(&tvblank) -
|
|
|
|
timeval_to_ns(&vblanktimestamp(dev, crtc, vblcount));
|
|
|
|
|
|
|
|
/* If there is at least 1 msec difference between the last stored
|
|
|
|
* timestamp and tvblank, then we are currently executing our
|
|
|
|
* disable inside a new vblank interval, the tvblank timestamp
|
|
|
|
* corresponds to this new vblank interval and the irq handler
|
|
|
|
* for this vblank didn't run yet and won't run due to our disable.
|
|
|
|
* Therefore we need to do the job of drm_handle_vblank() and
|
|
|
|
* increment the vblank counter by one to account for this vblank.
|
|
|
|
*
|
|
|
|
* Skip this step if there isn't any high precision timestamp
|
|
|
|
* available. In that case we can't account for this and just
|
|
|
|
* hope for the best.
|
|
|
|
*/
|
2011-02-21 12:42:01 +08:00
|
|
|
if ((vblrc > 0) && (abs64(diff_ns) > 1000000)) {
|
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, ×tamp).
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
|
|
|
atomic_inc(&dev->_vblank_count[crtc]);
|
2011-02-21 12:42:01 +08:00
|
|
|
smp_mb__after_atomic_inc();
|
|
|
|
}
|
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, ×tamp).
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
|
|
|
|
|
|
|
/* Invalidate all timestamps while vblank irq's are off. */
|
|
|
|
clear_vblank_timestamps(dev, crtc);
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
|
|
|
|
preempt_enable();
|
|
|
|
}
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
static void vblank_disable_fn(unsigned long arg)
|
|
|
|
{
|
|
|
|
struct drm_device *dev = (struct drm_device *)arg;
|
|
|
|
unsigned long irqflags;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (!dev->vblank_disable_allowed)
|
|
|
|
return;
|
|
|
|
|
|
|
|
for (i = 0; i < dev->num_crtcs; i++) {
|
|
|
|
spin_lock_irqsave(&dev->vbl_lock, irqflags);
|
|
|
|
if (atomic_read(&dev->vblank_refcount[i]) == 0 &&
|
|
|
|
dev->vblank_enabled[i]) {
|
|
|
|
DRM_DEBUG("disabling vblank on crtc %d\n", i);
|
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, ×tamp).
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
|
|
|
vblank_disable_and_save(dev, i);
|
2008-10-01 03:14:26 +08:00
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-11-19 01:30:25 +08:00
|
|
|
void drm_vblank_cleanup(struct drm_device *dev)
|
2008-10-01 03:14:26 +08:00
|
|
|
{
|
|
|
|
/* Bail if the driver didn't call drm_vblank_init() */
|
|
|
|
if (dev->num_crtcs == 0)
|
|
|
|
return;
|
|
|
|
|
|
|
|
del_timer(&dev->vblank_disable_timer);
|
|
|
|
|
|
|
|
vblank_disable_fn((unsigned long)dev);
|
|
|
|
|
2009-03-25 03:23:04 +08:00
|
|
|
kfree(dev->vbl_queue);
|
|
|
|
kfree(dev->_vblank_count);
|
|
|
|
kfree(dev->vblank_refcount);
|
|
|
|
kfree(dev->vblank_enabled);
|
|
|
|
kfree(dev->last_vblank);
|
|
|
|
kfree(dev->last_vblank_wait);
|
|
|
|
kfree(dev->vblank_inmodeset);
|
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, ×tamp).
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
|
|
|
kfree(dev->_vblank_time);
|
2008-10-01 03:14:26 +08:00
|
|
|
|
|
|
|
dev->num_crtcs = 0;
|
|
|
|
}
|
2010-01-07 22:39:13 +08:00
|
|
|
EXPORT_SYMBOL(drm_vblank_cleanup);
|
2008-10-01 03:14:26 +08:00
|
|
|
|
|
|
|
int drm_vblank_init(struct drm_device *dev, int num_crtcs)
|
|
|
|
{
|
|
|
|
int i, ret = -ENOMEM;
|
|
|
|
|
|
|
|
setup_timer(&dev->vblank_disable_timer, vblank_disable_fn,
|
|
|
|
(unsigned long)dev);
|
|
|
|
spin_lock_init(&dev->vbl_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, ×tamp).
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
|
|
|
spin_lock_init(&dev->vblank_time_lock);
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
dev->num_crtcs = num_crtcs;
|
|
|
|
|
2009-03-25 03:23:04 +08:00
|
|
|
dev->vbl_queue = kmalloc(sizeof(wait_queue_head_t) * num_crtcs,
|
|
|
|
GFP_KERNEL);
|
2008-10-01 03:14:26 +08:00
|
|
|
if (!dev->vbl_queue)
|
|
|
|
goto err;
|
|
|
|
|
2009-03-25 03:23:04 +08:00
|
|
|
dev->_vblank_count = kmalloc(sizeof(atomic_t) * num_crtcs, GFP_KERNEL);
|
2008-10-01 03:14:26 +08:00
|
|
|
if (!dev->_vblank_count)
|
|
|
|
goto err;
|
|
|
|
|
2009-03-25 03:23:04 +08:00
|
|
|
dev->vblank_refcount = kmalloc(sizeof(atomic_t) * num_crtcs,
|
|
|
|
GFP_KERNEL);
|
2008-10-01 03:14:26 +08:00
|
|
|
if (!dev->vblank_refcount)
|
|
|
|
goto err;
|
|
|
|
|
2009-03-25 03:23:04 +08:00
|
|
|
dev->vblank_enabled = kcalloc(num_crtcs, sizeof(int), GFP_KERNEL);
|
2008-10-01 03:14:26 +08:00
|
|
|
if (!dev->vblank_enabled)
|
|
|
|
goto err;
|
|
|
|
|
2009-03-25 03:23:04 +08:00
|
|
|
dev->last_vblank = kcalloc(num_crtcs, sizeof(u32), GFP_KERNEL);
|
2008-10-01 03:14:26 +08:00
|
|
|
if (!dev->last_vblank)
|
|
|
|
goto err;
|
|
|
|
|
2009-03-25 03:23:04 +08:00
|
|
|
dev->last_vblank_wait = kcalloc(num_crtcs, sizeof(u32), GFP_KERNEL);
|
2008-12-20 09:23:38 +08:00
|
|
|
if (!dev->last_vblank_wait)
|
|
|
|
goto err;
|
|
|
|
|
2009-03-25 03:23:04 +08:00
|
|
|
dev->vblank_inmodeset = kcalloc(num_crtcs, sizeof(int), GFP_KERNEL);
|
2008-10-01 03:14:26 +08:00
|
|
|
if (!dev->vblank_inmodeset)
|
|
|
|
goto err;
|
|
|
|
|
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, ×tamp).
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
|
|
|
dev->_vblank_time = kcalloc(num_crtcs * DRM_VBLANKTIME_RBSIZE,
|
|
|
|
sizeof(struct timeval), GFP_KERNEL);
|
|
|
|
if (!dev->_vblank_time)
|
|
|
|
goto err;
|
|
|
|
|
|
|
|
DRM_INFO("Supports vblank timestamp caching Rev 1 (10.10.2010).\n");
|
|
|
|
|
|
|
|
/* Driver specific high-precision vblank timestamping supported? */
|
|
|
|
if (dev->driver->get_vblank_timestamp)
|
|
|
|
DRM_INFO("Driver supports precise vblank timestamp query.\n");
|
|
|
|
else
|
|
|
|
DRM_INFO("No driver support for vblank timestamp query.\n");
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
/* Zero per-crtc vblank stuff */
|
|
|
|
for (i = 0; i < num_crtcs; i++) {
|
|
|
|
init_waitqueue_head(&dev->vbl_queue[i]);
|
|
|
|
atomic_set(&dev->_vblank_count[i], 0);
|
|
|
|
atomic_set(&dev->vblank_refcount[i], 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
dev->vblank_disable_allowed = 0;
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
err:
|
|
|
|
drm_vblank_cleanup(dev);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_vblank_init);
|
|
|
|
|
2009-09-21 12:33:58 +08:00
|
|
|
static void drm_irq_vgaarb_nokms(void *cookie, bool state)
|
|
|
|
{
|
|
|
|
struct drm_device *dev = cookie;
|
|
|
|
|
|
|
|
if (dev->driver->vgaarb_irq) {
|
|
|
|
dev->driver->vgaarb_irq(dev, state);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!dev->irq_enabled)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (state)
|
|
|
|
dev->driver->irq_uninstall(dev);
|
|
|
|
else {
|
|
|
|
dev->driver->irq_preinstall(dev);
|
|
|
|
dev->driver->irq_postinstall(dev);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/**
|
|
|
|
* Install IRQ handler.
|
|
|
|
*
|
|
|
|
* \param dev DRM device.
|
|
|
|
*
|
2008-10-01 03:14:26 +08:00
|
|
|
* Initializes the IRQ related data. Installs the handler, calling the driver
|
2005-04-17 06:20:36 +08:00
|
|
|
* \c drm_driver_irq_preinstall() and \c drm_driver_irq_postinstall() functions
|
|
|
|
* before and after the installation.
|
|
|
|
*/
|
2008-10-01 03:14:26 +08:00
|
|
|
int drm_irq_install(struct drm_device *dev)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2008-10-01 03:14:26 +08:00
|
|
|
int ret = 0;
|
2005-09-25 12:28:13 +08:00
|
|
|
unsigned long sh_flags = 0;
|
2009-06-02 14:50:35 +08:00
|
|
|
char *irqname;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
|
|
|
|
return -EINVAL;
|
|
|
|
|
2010-05-28 03:40:25 +08:00
|
|
|
if (drm_dev_to_irq(dev) == 0)
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EINVAL;
|
|
|
|
|
2006-02-02 16:37:46 +08:00
|
|
|
mutex_lock(&dev->struct_mutex);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* Driver must have been initialized */
|
2005-09-25 12:28:13 +08:00
|
|
|
if (!dev->dev_private) {
|
2006-02-02 16:37:46 +08:00
|
|
|
mutex_unlock(&dev->struct_mutex);
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2005-09-25 12:28:13 +08:00
|
|
|
if (dev->irq_enabled) {
|
2006-02-02 16:37:46 +08:00
|
|
|
mutex_unlock(&dev->struct_mutex);
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EBUSY;
|
|
|
|
}
|
|
|
|
dev->irq_enabled = 1;
|
2006-02-02 16:37:46 +08:00
|
|
|
mutex_unlock(&dev->struct_mutex);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2010-05-28 03:40:25 +08:00
|
|
|
DRM_DEBUG("irq=%d\n", drm_dev_to_irq(dev));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-09-25 12:28:13 +08:00
|
|
|
/* Before installing handler */
|
2005-04-17 06:20:36 +08:00
|
|
|
dev->driver->irq_preinstall(dev);
|
|
|
|
|
2005-09-25 12:28:13 +08:00
|
|
|
/* Install handler */
|
2005-04-17 06:20:36 +08:00
|
|
|
if (drm_core_check_feature(dev, DRIVER_IRQ_SHARED))
|
2006-07-02 10:29:34 +08:00
|
|
|
sh_flags = IRQF_SHARED;
|
2005-09-25 12:28:13 +08:00
|
|
|
|
2009-06-02 14:50:35 +08:00
|
|
|
if (dev->devname)
|
|
|
|
irqname = dev->devname;
|
|
|
|
else
|
|
|
|
irqname = dev->driver->name;
|
|
|
|
|
2008-09-16 06:00:33 +08:00
|
|
|
ret = request_irq(drm_dev_to_irq(dev), dev->driver->irq_handler,
|
2009-06-02 14:50:35 +08:00
|
|
|
sh_flags, irqname, dev);
|
2008-09-16 06:00:33 +08:00
|
|
|
|
2005-09-25 12:28:13 +08:00
|
|
|
if (ret < 0) {
|
2006-02-02 16:37:46 +08:00
|
|
|
mutex_lock(&dev->struct_mutex);
|
2005-04-17 06:20:36 +08:00
|
|
|
dev->irq_enabled = 0;
|
2006-02-02 16:37:46 +08:00
|
|
|
mutex_unlock(&dev->struct_mutex);
|
2005-04-17 06:20:36 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2009-09-21 12:33:58 +08:00
|
|
|
if (!drm_core_check_feature(dev, DRIVER_MODESET))
|
|
|
|
vga_client_register(dev->pdev, (void *)dev, drm_irq_vgaarb_nokms, NULL);
|
|
|
|
|
2005-09-25 12:28:13 +08:00
|
|
|
/* After installing handler */
|
2008-10-01 03:14:26 +08:00
|
|
|
ret = dev->driver->irq_postinstall(dev);
|
|
|
|
if (ret < 0) {
|
|
|
|
mutex_lock(&dev->struct_mutex);
|
|
|
|
dev->irq_enabled = 0;
|
|
|
|
mutex_unlock(&dev->struct_mutex);
|
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
return ret;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2008-10-01 03:14:26 +08:00
|
|
|
EXPORT_SYMBOL(drm_irq_install);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/**
|
|
|
|
* Uninstall the IRQ handler.
|
|
|
|
*
|
|
|
|
* \param dev DRM device.
|
|
|
|
*
|
|
|
|
* Calls the driver's \c drm_driver_irq_uninstall() function, and stops the irq.
|
|
|
|
*/
|
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, ×tamp).
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 drm_irq_uninstall(struct drm_device *dev)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2009-01-07 02:21:24 +08:00
|
|
|
unsigned long irqflags;
|
|
|
|
int irq_enabled, i;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
|
|
|
|
return -EINVAL;
|
|
|
|
|
2006-02-02 16:37:46 +08:00
|
|
|
mutex_lock(&dev->struct_mutex);
|
2005-04-17 06:20:36 +08:00
|
|
|
irq_enabled = dev->irq_enabled;
|
|
|
|
dev->irq_enabled = 0;
|
2006-02-02 16:37:46 +08:00
|
|
|
mutex_unlock(&dev->struct_mutex);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-01-07 02:21:24 +08:00
|
|
|
/*
|
|
|
|
* Wake up any waiters so they don't hang.
|
|
|
|
*/
|
|
|
|
spin_lock_irqsave(&dev->vbl_lock, irqflags);
|
|
|
|
for (i = 0; i < dev->num_crtcs; i++) {
|
|
|
|
DRM_WAKEUP(&dev->vbl_queue[i]);
|
|
|
|
dev->vblank_enabled[i] = 0;
|
2009-02-07 05:04:49 +08:00
|
|
|
dev->last_vblank[i] = dev->driver->get_vblank_counter(dev, i);
|
2009-01-07 02:21:24 +08:00
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
|
|
|
|
|
2005-09-25 12:28:13 +08:00
|
|
|
if (!irq_enabled)
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EINVAL;
|
|
|
|
|
2010-05-28 03:40:25 +08:00
|
|
|
DRM_DEBUG("irq=%d\n", drm_dev_to_irq(dev));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-09-21 12:33:58 +08:00
|
|
|
if (!drm_core_check_feature(dev, DRIVER_MODESET))
|
|
|
|
vga_client_register(dev->pdev, NULL, NULL, NULL);
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
dev->driver->irq_uninstall(dev);
|
|
|
|
|
2010-05-28 03:40:25 +08:00
|
|
|
free_irq(drm_dev_to_irq(dev), dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_irq_uninstall);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* IRQ control ioctl.
|
|
|
|
*
|
|
|
|
* \param inode device inode.
|
2007-08-25 18:23:09 +08:00
|
|
|
* \param file_priv DRM file private.
|
2005-04-17 06:20:36 +08:00
|
|
|
* \param cmd command.
|
|
|
|
* \param arg user argument, pointing to a drm_control structure.
|
|
|
|
* \return zero on success or a negative number on failure.
|
|
|
|
*
|
|
|
|
* Calls irq_install() or irq_uninstall() according to \p arg.
|
|
|
|
*/
|
2007-09-03 10:06:45 +08:00
|
|
|
int drm_control(struct drm_device *dev, void *data,
|
|
|
|
struct drm_file *file_priv)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2007-09-03 10:06:45 +08:00
|
|
|
struct drm_control *ctl = data;
|
2005-09-25 12:28:13 +08:00
|
|
|
|
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, ×tamp).
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
|
|
|
/* if we haven't irq we fallback for compatibility reasons -
|
|
|
|
* this used to be a separate function in drm_dma.h
|
|
|
|
*/
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
2007-09-03 10:06:45 +08:00
|
|
|
switch (ctl->func) {
|
2005-04-17 06:20:36 +08:00
|
|
|
case DRM_INST_HANDLER:
|
|
|
|
if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
|
|
|
|
return 0;
|
2008-11-08 06:05:41 +08:00
|
|
|
if (drm_core_check_feature(dev, DRIVER_MODESET))
|
|
|
|
return 0;
|
2005-04-17 06:20:36 +08:00
|
|
|
if (dev->if_version < DRM_IF_VERSION(1, 2) &&
|
2010-05-28 03:40:25 +08:00
|
|
|
ctl->irq != drm_dev_to_irq(dev))
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EINVAL;
|
2005-09-25 12:28:13 +08:00
|
|
|
return drm_irq_install(dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
case DRM_UNINST_HANDLER:
|
|
|
|
if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
|
|
|
|
return 0;
|
2008-11-08 06:05:41 +08:00
|
|
|
if (drm_core_check_feature(dev, DRIVER_MODESET))
|
|
|
|
return 0;
|
2005-09-25 12:28:13 +08:00
|
|
|
return drm_irq_uninstall(dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
default:
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
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, ×tamp).
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_calc_timestamping_constants - Calculate and
|
|
|
|
* store various constants which are later needed by
|
|
|
|
* vblank and swap-completion timestamping, e.g, by
|
|
|
|
* drm_calc_vbltimestamp_from_scanoutpos().
|
|
|
|
* They are derived from crtc's true scanout timing,
|
|
|
|
* so they take things like panel scaling or other
|
|
|
|
* adjustments into account.
|
|
|
|
*
|
|
|
|
* @crtc drm_crtc whose timestamp constants should be updated.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
void drm_calc_timestamping_constants(struct drm_crtc *crtc)
|
|
|
|
{
|
|
|
|
s64 linedur_ns = 0, pixeldur_ns = 0, framedur_ns = 0;
|
|
|
|
u64 dotclock;
|
|
|
|
|
|
|
|
/* Dot clock in Hz: */
|
|
|
|
dotclock = (u64) crtc->hwmode.clock * 1000;
|
|
|
|
|
2011-02-21 12:42:02 +08:00
|
|
|
/* Fields of interlaced scanout modes are only halve a frame duration.
|
|
|
|
* Double the dotclock to get halve the frame-/line-/pixelduration.
|
|
|
|
*/
|
|
|
|
if (crtc->hwmode.flags & DRM_MODE_FLAG_INTERLACE)
|
|
|
|
dotclock *= 2;
|
|
|
|
|
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, ×tamp).
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
|
|
|
/* Valid dotclock? */
|
|
|
|
if (dotclock > 0) {
|
|
|
|
/* Convert scanline length in pixels and video dot clock to
|
|
|
|
* line duration, frame duration and pixel duration in
|
|
|
|
* nanoseconds:
|
|
|
|
*/
|
|
|
|
pixeldur_ns = (s64) div64_u64(1000000000, dotclock);
|
|
|
|
linedur_ns = (s64) div64_u64(((u64) crtc->hwmode.crtc_htotal *
|
|
|
|
1000000000), dotclock);
|
|
|
|
framedur_ns = (s64) crtc->hwmode.crtc_vtotal * linedur_ns;
|
|
|
|
} else
|
|
|
|
DRM_ERROR("crtc %d: Can't calculate constants, dotclock = 0!\n",
|
|
|
|
crtc->base.id);
|
|
|
|
|
|
|
|
crtc->pixeldur_ns = pixeldur_ns;
|
|
|
|
crtc->linedur_ns = linedur_ns;
|
|
|
|
crtc->framedur_ns = framedur_ns;
|
|
|
|
|
|
|
|
DRM_DEBUG("crtc %d: hwmode: htotal %d, vtotal %d, vdisplay %d\n",
|
|
|
|
crtc->base.id, crtc->hwmode.crtc_htotal,
|
|
|
|
crtc->hwmode.crtc_vtotal, crtc->hwmode.crtc_vdisplay);
|
|
|
|
DRM_DEBUG("crtc %d: clock %d kHz framedur %d linedur %d, pixeldur %d\n",
|
|
|
|
crtc->base.id, (int) dotclock/1000, (int) framedur_ns,
|
|
|
|
(int) linedur_ns, (int) pixeldur_ns);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_calc_timestamping_constants);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* drm_calc_vbltimestamp_from_scanoutpos - helper routine for kms
|
|
|
|
* drivers. Implements calculation of exact vblank timestamps from
|
|
|
|
* given drm_display_mode timings and current video scanout position
|
|
|
|
* of a crtc. This can be called from within get_vblank_timestamp()
|
|
|
|
* implementation of a kms driver to implement the actual timestamping.
|
|
|
|
*
|
|
|
|
* Should return timestamps conforming to the OML_sync_control OpenML
|
|
|
|
* extension specification. The timestamp corresponds to the end of
|
|
|
|
* the vblank interval, aka start of scanout of topmost-leftmost display
|
|
|
|
* pixel in the following video frame.
|
|
|
|
*
|
|
|
|
* Requires support for optional dev->driver->get_scanout_position()
|
|
|
|
* in kms driver, plus a bit of setup code to provide a drm_display_mode
|
|
|
|
* that corresponds to the true scanout timing.
|
|
|
|
*
|
|
|
|
* The current implementation only handles standard video modes. It
|
|
|
|
* returns as no operation if a doublescan or interlaced video mode is
|
|
|
|
* active. Higher level code is expected to handle this.
|
|
|
|
*
|
|
|
|
* @dev: DRM device.
|
|
|
|
* @crtc: Which crtc's vblank timestamp to retrieve.
|
|
|
|
* @max_error: Desired maximum allowable error in timestamps (nanosecs).
|
|
|
|
* On return contains true maximum error of timestamp.
|
|
|
|
* @vblank_time: Pointer to struct timeval which should receive the timestamp.
|
|
|
|
* @flags: Flags to pass to driver:
|
|
|
|
* 0 = Default.
|
|
|
|
* DRM_CALLED_FROM_VBLIRQ = If function is called from vbl irq handler.
|
|
|
|
* @refcrtc: drm_crtc* of crtc which defines scanout timing.
|
|
|
|
*
|
|
|
|
* Returns negative value on error, failure or if not supported in current
|
|
|
|
* video mode:
|
|
|
|
*
|
|
|
|
* -EINVAL - Invalid crtc.
|
|
|
|
* -EAGAIN - Temporary unavailable, e.g., called before initial modeset.
|
|
|
|
* -ENOTSUPP - Function not supported in current display mode.
|
|
|
|
* -EIO - Failed, e.g., due to failed scanout position query.
|
|
|
|
*
|
|
|
|
* Returns or'ed positive status flags on success:
|
|
|
|
*
|
|
|
|
* DRM_VBLANKTIME_SCANOUTPOS_METHOD - Signal this method used for timestamping.
|
|
|
|
* DRM_VBLANKTIME_INVBL - Timestamp taken while scanout was in vblank interval.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
|
|
|
|
int *max_error,
|
|
|
|
struct timeval *vblank_time,
|
|
|
|
unsigned flags,
|
|
|
|
struct drm_crtc *refcrtc)
|
|
|
|
{
|
|
|
|
struct timeval stime, raw_time;
|
|
|
|
struct drm_display_mode *mode;
|
|
|
|
int vbl_status, vtotal, vdisplay;
|
|
|
|
int vpos, hpos, i;
|
|
|
|
s64 framedur_ns, linedur_ns, pixeldur_ns, delta_ns, duration_ns;
|
|
|
|
bool invbl;
|
|
|
|
|
|
|
|
if (crtc < 0 || crtc >= dev->num_crtcs) {
|
|
|
|
DRM_ERROR("Invalid crtc %d\n", crtc);
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Scanout position query not supported? Should not happen. */
|
|
|
|
if (!dev->driver->get_scanout_position) {
|
|
|
|
DRM_ERROR("Called from driver w/o get_scanout_position()!?\n");
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
mode = &refcrtc->hwmode;
|
|
|
|
vtotal = mode->crtc_vtotal;
|
|
|
|
vdisplay = mode->crtc_vdisplay;
|
|
|
|
|
|
|
|
/* Durations of frames, lines, pixels in nanoseconds. */
|
|
|
|
framedur_ns = refcrtc->framedur_ns;
|
|
|
|
linedur_ns = refcrtc->linedur_ns;
|
|
|
|
pixeldur_ns = refcrtc->pixeldur_ns;
|
|
|
|
|
|
|
|
/* If mode timing undefined, just return as no-op:
|
|
|
|
* Happens during initial modesetting of a crtc.
|
|
|
|
*/
|
|
|
|
if (vtotal <= 0 || vdisplay <= 0 || framedur_ns == 0) {
|
|
|
|
DRM_DEBUG("crtc %d: Noop due to uninitialized mode.\n", crtc);
|
|
|
|
return -EAGAIN;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Get current scanout position with system timestamp.
|
|
|
|
* Repeat query up to DRM_TIMESTAMP_MAXRETRIES times
|
|
|
|
* if single query takes longer than max_error nanoseconds.
|
|
|
|
*
|
|
|
|
* This guarantees a tight bound on maximum error if
|
|
|
|
* code gets preempted or delayed for some reason.
|
|
|
|
*/
|
|
|
|
for (i = 0; i < DRM_TIMESTAMP_MAXRETRIES; i++) {
|
|
|
|
/* Disable preemption to make it very likely to
|
|
|
|
* succeed in the first iteration even on PREEMPT_RT kernel.
|
|
|
|
*/
|
|
|
|
preempt_disable();
|
|
|
|
|
|
|
|
/* Get system timestamp before query. */
|
|
|
|
do_gettimeofday(&stime);
|
|
|
|
|
|
|
|
/* Get vertical and horizontal scanout pos. vpos, hpos. */
|
|
|
|
vbl_status = dev->driver->get_scanout_position(dev, crtc, &vpos, &hpos);
|
|
|
|
|
|
|
|
/* Get system timestamp after query. */
|
|
|
|
do_gettimeofday(&raw_time);
|
|
|
|
|
|
|
|
preempt_enable();
|
|
|
|
|
|
|
|
/* Return as no-op if scanout query unsupported or failed. */
|
|
|
|
if (!(vbl_status & DRM_SCANOUTPOS_VALID)) {
|
|
|
|
DRM_DEBUG("crtc %d : scanoutpos query failed [%d].\n",
|
|
|
|
crtc, vbl_status);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
duration_ns = timeval_to_ns(&raw_time) - timeval_to_ns(&stime);
|
|
|
|
|
|
|
|
/* Accept result with < max_error nsecs timing uncertainty. */
|
|
|
|
if (duration_ns <= (s64) *max_error)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Noisy system timing? */
|
|
|
|
if (i == DRM_TIMESTAMP_MAXRETRIES) {
|
|
|
|
DRM_DEBUG("crtc %d: Noisy timestamp %d us > %d us [%d reps].\n",
|
|
|
|
crtc, (int) duration_ns/1000, *max_error/1000, i);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Return upper bound of timestamp precision error. */
|
|
|
|
*max_error = (int) duration_ns;
|
|
|
|
|
|
|
|
/* Check if in vblank area:
|
|
|
|
* vpos is >=0 in video scanout area, but negative
|
|
|
|
* within vblank area, counting down the number of lines until
|
|
|
|
* start of scanout.
|
|
|
|
*/
|
|
|
|
invbl = vbl_status & DRM_SCANOUTPOS_INVBL;
|
|
|
|
|
|
|
|
/* Convert scanout position into elapsed time at raw_time query
|
|
|
|
* since start of scanout at first display scanline. delta_ns
|
|
|
|
* can be negative if start of scanout hasn't happened yet.
|
|
|
|
*/
|
|
|
|
delta_ns = (s64) vpos * linedur_ns + (s64) hpos * pixeldur_ns;
|
|
|
|
|
|
|
|
/* Is vpos outside nominal vblank area, but less than
|
|
|
|
* 1/100 of a frame height away from start of vblank?
|
|
|
|
* If so, assume this isn't a massively delayed vblank
|
|
|
|
* interrupt, but a vblank interrupt that fired a few
|
|
|
|
* microseconds before true start of vblank. Compensate
|
|
|
|
* by adding a full frame duration to the final timestamp.
|
|
|
|
* Happens, e.g., on ATI R500, R600.
|
|
|
|
*
|
|
|
|
* We only do this if DRM_CALLED_FROM_VBLIRQ.
|
|
|
|
*/
|
|
|
|
if ((flags & DRM_CALLED_FROM_VBLIRQ) && !invbl &&
|
|
|
|
((vdisplay - vpos) < vtotal / 100)) {
|
|
|
|
delta_ns = delta_ns - framedur_ns;
|
|
|
|
|
|
|
|
/* Signal this correction as "applied". */
|
|
|
|
vbl_status |= 0x8;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Subtract time delta from raw timestamp to get final
|
|
|
|
* vblank_time timestamp for end of vblank.
|
|
|
|
*/
|
|
|
|
*vblank_time = ns_to_timeval(timeval_to_ns(&raw_time) - delta_ns);
|
|
|
|
|
2011-04-18 11:35:52 +08:00
|
|
|
DRM_DEBUG("crtc %d : v %d p(%d,%d)@ %ld.%ld -> %ld.%ld [e %d us, %d rep]\n",
|
|
|
|
crtc, (int)vbl_status, hpos, vpos,
|
|
|
|
(long)raw_time.tv_sec, (long)raw_time.tv_usec,
|
|
|
|
(long)vblank_time->tv_sec, (long)vblank_time->tv_usec,
|
|
|
|
(int)duration_ns/1000, i);
|
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, ×tamp).
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
|
|
|
|
|
|
|
vbl_status = DRM_VBLANKTIME_SCANOUTPOS_METHOD;
|
|
|
|
if (invbl)
|
|
|
|
vbl_status |= DRM_VBLANKTIME_INVBL;
|
|
|
|
|
|
|
|
return vbl_status;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_calc_vbltimestamp_from_scanoutpos);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* drm_get_last_vbltimestamp - retrieve raw timestamp for the most recent
|
|
|
|
* vblank interval.
|
|
|
|
*
|
|
|
|
* @dev: DRM device
|
|
|
|
* @crtc: which crtc's vblank timestamp to retrieve
|
|
|
|
* @tvblank: Pointer to target struct timeval which should receive the timestamp
|
|
|
|
* @flags: Flags to pass to driver:
|
|
|
|
* 0 = Default.
|
|
|
|
* DRM_CALLED_FROM_VBLIRQ = If function is called from vbl irq handler.
|
|
|
|
*
|
|
|
|
* Fetches the system timestamp corresponding to the time of the most recent
|
|
|
|
* vblank interval on specified crtc. May call into kms-driver to
|
|
|
|
* compute the timestamp with a high-precision GPU specific method.
|
|
|
|
*
|
|
|
|
* Returns zero if timestamp originates from uncorrected do_gettimeofday()
|
|
|
|
* call, i.e., it isn't very precisely locked to the true vblank.
|
|
|
|
*
|
|
|
|
* Returns non-zero if timestamp is considered to be very precise.
|
|
|
|
*/
|
|
|
|
u32 drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
|
|
|
|
struct timeval *tvblank, unsigned flags)
|
|
|
|
{
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
/* Define requested maximum error on timestamps (nanoseconds). */
|
|
|
|
int max_error = (int) drm_timestamp_precision * 1000;
|
|
|
|
|
|
|
|
/* Query driver if possible and precision timestamping enabled. */
|
|
|
|
if (dev->driver->get_vblank_timestamp && (max_error > 0)) {
|
|
|
|
ret = dev->driver->get_vblank_timestamp(dev, crtc, &max_error,
|
|
|
|
tvblank, flags);
|
|
|
|
if (ret > 0)
|
|
|
|
return (u32) ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* GPU high precision timestamp query unsupported or failed.
|
|
|
|
* Return gettimeofday timestamp as best estimate.
|
|
|
|
*/
|
|
|
|
do_gettimeofday(tvblank);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_get_last_vbltimestamp);
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
/**
|
|
|
|
* drm_vblank_count - retrieve "cooked" vblank counter value
|
|
|
|
* @dev: DRM device
|
|
|
|
* @crtc: which counter to retrieve
|
|
|
|
*
|
|
|
|
* Fetches the "cooked" vblank count value that represents the number of
|
|
|
|
* vblank events since the system was booted, including lost events due to
|
|
|
|
* modesetting activity.
|
|
|
|
*/
|
|
|
|
u32 drm_vblank_count(struct drm_device *dev, int crtc)
|
|
|
|
{
|
|
|
|
return atomic_read(&dev->_vblank_count[crtc]);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_vblank_count);
|
|
|
|
|
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, ×tamp).
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_vblank_count_and_time - retrieve "cooked" vblank counter value
|
|
|
|
* and the system timestamp corresponding to that vblank counter value.
|
|
|
|
*
|
|
|
|
* @dev: DRM device
|
|
|
|
* @crtc: which counter to retrieve
|
|
|
|
* @vblanktime: Pointer to struct timeval to receive the vblank timestamp.
|
|
|
|
*
|
|
|
|
* Fetches the "cooked" vblank count value that represents the number of
|
|
|
|
* vblank events since the system was booted, including lost events due to
|
|
|
|
* modesetting activity. Returns corresponding system timestamp of the time
|
|
|
|
* of the vblank interval that corresponds to the current value vblank counter
|
|
|
|
* value.
|
|
|
|
*/
|
|
|
|
u32 drm_vblank_count_and_time(struct drm_device *dev, int crtc,
|
|
|
|
struct timeval *vblanktime)
|
|
|
|
{
|
|
|
|
u32 cur_vblank;
|
|
|
|
|
|
|
|
/* Read timestamp from slot of _vblank_time ringbuffer
|
|
|
|
* that corresponds to current vblank count. Retry if
|
|
|
|
* count has incremented during readout. This works like
|
|
|
|
* a seqlock.
|
|
|
|
*/
|
|
|
|
do {
|
|
|
|
cur_vblank = atomic_read(&dev->_vblank_count[crtc]);
|
|
|
|
*vblanktime = vblanktimestamp(dev, crtc, cur_vblank);
|
|
|
|
smp_rmb();
|
|
|
|
} while (cur_vblank != atomic_read(&dev->_vblank_count[crtc]));
|
|
|
|
|
|
|
|
return cur_vblank;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_vblank_count_and_time);
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
/**
|
|
|
|
* drm_update_vblank_count - update the master vblank counter
|
|
|
|
* @dev: DRM device
|
|
|
|
* @crtc: counter to update
|
|
|
|
*
|
|
|
|
* Call back into the driver to update the appropriate vblank counter
|
|
|
|
* (specified by @crtc). Deal with wraparound, if it occurred, and
|
|
|
|
* update the last read value so we can deal with wraparound on the next
|
|
|
|
* call if necessary.
|
|
|
|
*
|
|
|
|
* Only necessary when going from off->on, to account for frames we
|
|
|
|
* didn't get an interrupt for.
|
|
|
|
*
|
|
|
|
* Note: caller must hold dev->vbl_lock since this reads & writes
|
|
|
|
* device vblank fields.
|
|
|
|
*/
|
|
|
|
static void drm_update_vblank_count(struct drm_device *dev, int crtc)
|
|
|
|
{
|
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, ×tamp).
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
|
|
|
u32 cur_vblank, diff, tslot, rc;
|
|
|
|
struct timeval t_vblank;
|
2008-10-01 03:14:26 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Interrupts were disabled prior to this call, so deal with counter
|
|
|
|
* wrap if needed.
|
|
|
|
* NOTE! It's possible we lost a full dev->max_vblank_count events
|
|
|
|
* here if the register is small or we had vblank interrupts off for
|
|
|
|
* a long time.
|
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, ×tamp).
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
|
|
|
*
|
|
|
|
* We repeat the hardware vblank counter & timestamp query until
|
|
|
|
* we get consistent results. This to prevent races between gpu
|
|
|
|
* updating its hardware counter while we are retrieving the
|
|
|
|
* corresponding vblank timestamp.
|
2008-10-01 03:14:26 +08:00
|
|
|
*/
|
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, ×tamp).
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
|
|
|
do {
|
|
|
|
cur_vblank = dev->driver->get_vblank_counter(dev, crtc);
|
|
|
|
rc = drm_get_last_vbltimestamp(dev, crtc, &t_vblank, 0);
|
|
|
|
} while (cur_vblank != dev->driver->get_vblank_counter(dev, crtc));
|
|
|
|
|
|
|
|
/* Deal with counter wrap */
|
2008-10-01 03:14:26 +08:00
|
|
|
diff = cur_vblank - dev->last_vblank[crtc];
|
|
|
|
if (cur_vblank < dev->last_vblank[crtc]) {
|
|
|
|
diff += dev->max_vblank_count;
|
|
|
|
|
|
|
|
DRM_DEBUG("last_vblank[%d]=0x%x, cur_vblank=0x%x => diff=0x%x\n",
|
|
|
|
crtc, dev->last_vblank[crtc], cur_vblank, diff);
|
|
|
|
}
|
|
|
|
|
|
|
|
DRM_DEBUG("enabling vblank interrupts on crtc %d, missed %d\n",
|
|
|
|
crtc, diff);
|
|
|
|
|
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, ×tamp).
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
|
|
|
/* Reinitialize corresponding vblank timestamp if high-precision query
|
|
|
|
* available. Skip this step if query unsupported or failed. Will
|
|
|
|
* reinitialize delayed at next vblank interrupt in that case.
|
|
|
|
*/
|
|
|
|
if (rc) {
|
|
|
|
tslot = atomic_read(&dev->_vblank_count[crtc]) + diff;
|
|
|
|
vblanktimestamp(dev, crtc, tslot) = t_vblank;
|
|
|
|
}
|
|
|
|
|
2011-02-21 12:42:01 +08:00
|
|
|
smp_mb__before_atomic_inc();
|
2008-10-01 03:14:26 +08:00
|
|
|
atomic_add(diff, &dev->_vblank_count[crtc]);
|
2011-02-21 12:42:01 +08:00
|
|
|
smp_mb__after_atomic_inc();
|
2008-10-01 03:14:26 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* drm_vblank_get - get a reference count on vblank events
|
|
|
|
* @dev: DRM device
|
|
|
|
* @crtc: which CRTC to own
|
|
|
|
*
|
|
|
|
* Acquire a reference count on vblank events to avoid having them disabled
|
|
|
|
* while in use.
|
|
|
|
*
|
|
|
|
* RETURNS
|
|
|
|
* Zero on success, nonzero on failure.
|
|
|
|
*/
|
|
|
|
int drm_vblank_get(struct drm_device *dev, int crtc)
|
|
|
|
{
|
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, ×tamp).
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
|
|
|
unsigned long irqflags, irqflags2;
|
2008-10-01 03:14:26 +08:00
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&dev->vbl_lock, irqflags);
|
|
|
|
/* Going from 0->1 means we have to enable interrupts again */
|
2009-11-09 12:51:22 +08:00
|
|
|
if (atomic_add_return(1, &dev->vblank_refcount[crtc]) == 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, ×tamp).
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
|
|
|
/* Disable preemption while holding vblank_time_lock. Do
|
|
|
|
* it explicitely to guard against PREEMPT_RT kernel.
|
|
|
|
*/
|
|
|
|
preempt_disable();
|
|
|
|
spin_lock_irqsave(&dev->vblank_time_lock, irqflags2);
|
2009-11-09 12:51:22 +08:00
|
|
|
if (!dev->vblank_enabled[crtc]) {
|
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, ×tamp).
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
|
|
|
/* Enable vblank irqs under vblank_time_lock protection.
|
|
|
|
* All vblank count & timestamp updates are held off
|
|
|
|
* until we are done reinitializing master counter and
|
|
|
|
* timestamps. Filtercode in drm_handle_vblank() will
|
|
|
|
* prevent double-accounting of same vblank interval.
|
|
|
|
*/
|
2009-11-09 12:51:22 +08:00
|
|
|
ret = dev->driver->enable_vblank(dev, crtc);
|
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, ×tamp).
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_DEBUG("enabling vblank on crtc %d, ret: %d\n",
|
|
|
|
crtc, ret);
|
2009-11-09 12:51:22 +08:00
|
|
|
if (ret)
|
|
|
|
atomic_dec(&dev->vblank_refcount[crtc]);
|
|
|
|
else {
|
|
|
|
dev->vblank_enabled[crtc] = 1;
|
|
|
|
drm_update_vblank_count(dev, crtc);
|
|
|
|
}
|
|
|
|
}
|
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, ×tamp).
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
|
|
|
spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags2);
|
|
|
|
preempt_enable();
|
2009-11-09 12:51:22 +08:00
|
|
|
} else {
|
|
|
|
if (!dev->vblank_enabled[crtc]) {
|
2008-10-01 03:14:26 +08:00
|
|
|
atomic_dec(&dev->vblank_refcount[crtc]);
|
2009-11-09 12:51:22 +08:00
|
|
|
ret = -EINVAL;
|
2008-10-01 03:14:26 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_vblank_get);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* drm_vblank_put - give up ownership of vblank events
|
|
|
|
* @dev: DRM device
|
|
|
|
* @crtc: which counter to give up
|
|
|
|
*
|
|
|
|
* Release ownership of a given vblank counter, turning off interrupts
|
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, ×tamp).
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
|
|
|
* if possible. Disable interrupts after drm_vblank_offdelay milliseconds.
|
2008-10-01 03:14:26 +08:00
|
|
|
*/
|
|
|
|
void drm_vblank_put(struct drm_device *dev, int crtc)
|
|
|
|
{
|
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, ×tamp).
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
|
|
|
BUG_ON(atomic_read(&dev->vblank_refcount[crtc]) == 0);
|
2009-02-19 22:48:22 +08:00
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
/* Last user schedules interrupt disable */
|
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, ×tamp).
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
|
|
|
if (atomic_dec_and_test(&dev->vblank_refcount[crtc]) &&
|
|
|
|
(drm_vblank_offdelay > 0))
|
|
|
|
mod_timer(&dev->vblank_disable_timer,
|
|
|
|
jiffies + ((drm_vblank_offdelay * DRM_HZ)/1000));
|
2008-10-01 03:14:26 +08:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_vblank_put);
|
|
|
|
|
2009-11-09 12:51:22 +08:00
|
|
|
void drm_vblank_off(struct drm_device *dev, int crtc)
|
|
|
|
{
|
2011-04-27 14:10:57 +08:00
|
|
|
struct drm_pending_vblank_event *e, *t;
|
|
|
|
struct timeval now;
|
2009-11-09 12:51:22 +08:00
|
|
|
unsigned long irqflags;
|
2011-04-27 14:10:57 +08:00
|
|
|
unsigned int seq;
|
2009-11-09 12:51:22 +08:00
|
|
|
|
|
|
|
spin_lock_irqsave(&dev->vbl_lock, irqflags);
|
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, ×tamp).
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
|
|
|
vblank_disable_and_save(dev, crtc);
|
2009-11-09 12:51:22 +08:00
|
|
|
DRM_WAKEUP(&dev->vbl_queue[crtc]);
|
2011-04-27 14:10:57 +08:00
|
|
|
|
|
|
|
/* Send any queued vblank events, lest the natives grow disquiet */
|
|
|
|
seq = drm_vblank_count_and_time(dev, crtc, &now);
|
|
|
|
list_for_each_entry_safe(e, t, &dev->vblank_event_list, base.link) {
|
|
|
|
if (e->pipe != crtc)
|
|
|
|
continue;
|
|
|
|
DRM_DEBUG("Sending premature vblank event on disable: \
|
|
|
|
wanted %d, current %d\n",
|
|
|
|
e->event.sequence, seq);
|
|
|
|
|
|
|
|
e->event.sequence = seq;
|
|
|
|
e->event.tv_sec = now.tv_sec;
|
|
|
|
e->event.tv_usec = now.tv_usec;
|
|
|
|
drm_vblank_put(dev, e->pipe);
|
|
|
|
list_move_tail(&e->base.link, &e->base.file_priv->event_list);
|
|
|
|
wake_up_interruptible(&e->base.file_priv->event_wait);
|
|
|
|
trace_drm_vblank_event_delivered(e->base.pid, e->pipe,
|
|
|
|
e->event.sequence);
|
|
|
|
}
|
|
|
|
|
2009-11-09 12:51:22 +08:00
|
|
|
spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_vblank_off);
|
|
|
|
|
2008-11-08 06:05:41 +08:00
|
|
|
/**
|
|
|
|
* drm_vblank_pre_modeset - account for vblanks across mode sets
|
|
|
|
* @dev: DRM device
|
|
|
|
* @crtc: CRTC in question
|
|
|
|
* @post: post or pre mode set?
|
|
|
|
*
|
|
|
|
* Account for vblank events across mode setting events, which will likely
|
|
|
|
* reset the hardware frame counter.
|
|
|
|
*/
|
|
|
|
void drm_vblank_pre_modeset(struct drm_device *dev, int crtc)
|
|
|
|
{
|
2010-01-07 22:39:13 +08:00
|
|
|
/* vblank is not initialized (IRQ not installed ?) */
|
|
|
|
if (!dev->num_crtcs)
|
|
|
|
return;
|
2008-11-08 06:05:41 +08:00
|
|
|
/*
|
|
|
|
* To avoid all the problems that might happen if interrupts
|
|
|
|
* were enabled/disabled around or between these calls, we just
|
|
|
|
* have the kernel take a reference on the CRTC (just once though
|
|
|
|
* to avoid corrupting the count if multiple, mismatch calls occur),
|
|
|
|
* so that interrupts remain enabled in the interim.
|
|
|
|
*/
|
|
|
|
if (!dev->vblank_inmodeset[crtc]) {
|
2009-02-19 22:48:22 +08:00
|
|
|
dev->vblank_inmodeset[crtc] = 0x1;
|
|
|
|
if (drm_vblank_get(dev, crtc) == 0)
|
|
|
|
dev->vblank_inmodeset[crtc] |= 0x2;
|
2008-11-08 06:05:41 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_vblank_pre_modeset);
|
|
|
|
|
|
|
|
void drm_vblank_post_modeset(struct drm_device *dev, int crtc)
|
|
|
|
{
|
|
|
|
unsigned long irqflags;
|
|
|
|
|
|
|
|
if (dev->vblank_inmodeset[crtc]) {
|
|
|
|
spin_lock_irqsave(&dev->vbl_lock, irqflags);
|
|
|
|
dev->vblank_disable_allowed = 1;
|
|
|
|
spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
|
2009-02-19 22:48:22 +08:00
|
|
|
|
|
|
|
if (dev->vblank_inmodeset[crtc] & 0x2)
|
|
|
|
drm_vblank_put(dev, crtc);
|
|
|
|
|
|
|
|
dev->vblank_inmodeset[crtc] = 0;
|
2008-11-08 06:05:41 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_vblank_post_modeset);
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
/**
|
|
|
|
* drm_modeset_ctl - handle vblank event counter changes across mode switch
|
|
|
|
* @DRM_IOCTL_ARGS: standard ioctl arguments
|
|
|
|
*
|
|
|
|
* Applications should call the %_DRM_PRE_MODESET and %_DRM_POST_MODESET
|
|
|
|
* ioctls around modesetting so that any lost vblank events are accounted for.
|
|
|
|
*
|
|
|
|
* Generally the counter will reset across mode sets. If interrupts are
|
|
|
|
* enabled around this call, we don't have to do anything since the counter
|
|
|
|
* will have already been incremented.
|
|
|
|
*/
|
|
|
|
int drm_modeset_ctl(struct drm_device *dev, void *data,
|
|
|
|
struct drm_file *file_priv)
|
|
|
|
{
|
|
|
|
struct drm_modeset_ctl *modeset = data;
|
2011-02-24 06:35:06 +08:00
|
|
|
int ret = 0;
|
|
|
|
unsigned int crtc;
|
2008-10-01 03:14:26 +08:00
|
|
|
|
|
|
|
/* If drm_vblank_init() hasn't been called yet, just no-op */
|
|
|
|
if (!dev->num_crtcs)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
crtc = modeset->crtc;
|
|
|
|
if (crtc >= dev->num_crtcs) {
|
|
|
|
ret = -EINVAL;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (modeset->cmd) {
|
|
|
|
case _DRM_PRE_MODESET:
|
2008-11-08 06:05:41 +08:00
|
|
|
drm_vblank_pre_modeset(dev, crtc);
|
2008-10-01 03:14:26 +08:00
|
|
|
break;
|
|
|
|
case _DRM_POST_MODESET:
|
2008-11-08 06:05:41 +08:00
|
|
|
drm_vblank_post_modeset(dev, crtc);
|
2008-10-01 03:14:26 +08:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
ret = -EINVAL;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2009-09-12 02:33:34 +08:00
|
|
|
static int drm_queue_vblank_event(struct drm_device *dev, int pipe,
|
|
|
|
union drm_wait_vblank *vblwait,
|
|
|
|
struct drm_file *file_priv)
|
|
|
|
{
|
|
|
|
struct drm_pending_vblank_event *e;
|
|
|
|
struct timeval now;
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned int seq;
|
2010-12-02 03:41:31 +08:00
|
|
|
int ret;
|
2009-09-12 02:33:34 +08:00
|
|
|
|
|
|
|
e = kzalloc(sizeof *e, GFP_KERNEL);
|
2010-12-02 03:41:31 +08:00
|
|
|
if (e == NULL) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto err_put;
|
|
|
|
}
|
2009-09-12 02:33:34 +08:00
|
|
|
|
|
|
|
e->pipe = pipe;
|
2010-07-02 07:48:09 +08:00
|
|
|
e->base.pid = current->pid;
|
2009-09-12 02:33:34 +08:00
|
|
|
e->event.base.type = DRM_EVENT_VBLANK;
|
|
|
|
e->event.base.length = sizeof e->event;
|
|
|
|
e->event.user_data = vblwait->request.signal;
|
|
|
|
e->base.event = &e->event.base;
|
|
|
|
e->base.file_priv = file_priv;
|
|
|
|
e->base.destroy = (void (*) (struct drm_pending_event *)) kfree;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&dev->event_lock, flags);
|
|
|
|
|
|
|
|
if (file_priv->event_space < sizeof e->event) {
|
2010-12-02 03:41:31 +08:00
|
|
|
ret = -EBUSY;
|
|
|
|
goto err_unlock;
|
2009-09-12 02:33:34 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
file_priv->event_space -= sizeof e->event;
|
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, ×tamp).
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
|
|
|
seq = drm_vblank_count_and_time(dev, pipe, &now);
|
|
|
|
|
2009-09-12 02:33:34 +08:00
|
|
|
if ((vblwait->request.type & _DRM_VBLANK_NEXTONMISS) &&
|
|
|
|
(seq - vblwait->request.sequence) <= (1 << 23)) {
|
|
|
|
vblwait->request.sequence = seq + 1;
|
2009-11-10 16:21:25 +08:00
|
|
|
vblwait->reply.sequence = vblwait->request.sequence;
|
2009-09-12 02:33:34 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
DRM_DEBUG("event on vblank count %d, current %d, crtc %d\n",
|
|
|
|
vblwait->request.sequence, seq, pipe);
|
|
|
|
|
2010-07-02 07:48:09 +08:00
|
|
|
trace_drm_vblank_event_queued(current->pid, pipe,
|
|
|
|
vblwait->request.sequence);
|
|
|
|
|
2009-09-12 02:33:34 +08:00
|
|
|
e->event.sequence = vblwait->request.sequence;
|
|
|
|
if ((seq - vblwait->request.sequence) <= (1 << 23)) {
|
2010-12-10 23:00:19 +08:00
|
|
|
e->event.sequence = seq;
|
2009-09-12 02:33:34 +08:00
|
|
|
e->event.tv_sec = now.tv_sec;
|
|
|
|
e->event.tv_usec = now.tv_usec;
|
2010-12-09 13:35:40 +08:00
|
|
|
drm_vblank_put(dev, pipe);
|
2009-09-12 02:33:34 +08:00
|
|
|
list_add_tail(&e->base.link, &e->base.file_priv->event_list);
|
|
|
|
wake_up_interruptible(&e->base.file_priv->event_wait);
|
2010-12-10 23:00:19 +08:00
|
|
|
vblwait->reply.sequence = seq;
|
2010-07-02 07:48:09 +08:00
|
|
|
trace_drm_vblank_event_delivered(current->pid, pipe,
|
|
|
|
vblwait->request.sequence);
|
2009-09-12 02:33:34 +08:00
|
|
|
} else {
|
|
|
|
list_add_tail(&e->base.link, &dev->vblank_event_list);
|
2010-12-10 23:00:19 +08:00
|
|
|
vblwait->reply.sequence = vblwait->request.sequence;
|
2009-09-12 02:33:34 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&dev->event_lock, flags);
|
|
|
|
|
|
|
|
return 0;
|
2010-12-02 03:41:31 +08:00
|
|
|
|
|
|
|
err_unlock:
|
|
|
|
spin_unlock_irqrestore(&dev->event_lock, flags);
|
|
|
|
kfree(e);
|
|
|
|
err_put:
|
2010-12-09 13:35:40 +08:00
|
|
|
drm_vblank_put(dev, pipe);
|
2010-12-02 03:41:31 +08:00
|
|
|
return ret;
|
2009-09-12 02:33:34 +08:00
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/**
|
|
|
|
* Wait for VBLANK.
|
|
|
|
*
|
|
|
|
* \param inode device inode.
|
2007-08-25 18:23:09 +08:00
|
|
|
* \param file_priv DRM file private.
|
2005-04-17 06:20:36 +08:00
|
|
|
* \param cmd command.
|
|
|
|
* \param data user argument, pointing to a drm_wait_vblank structure.
|
|
|
|
* \return zero on success or a negative number on failure.
|
|
|
|
*
|
2009-01-28 13:19:41 +08:00
|
|
|
* This function enables the vblank interrupt on the pipe requested, then
|
|
|
|
* sleeps waiting for the requested sequence number to occur, and drops
|
|
|
|
* the vblank interrupt refcount afterwards. (vblank irq disable follows that
|
|
|
|
* after a timeout with no further vblank waits scheduled).
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2008-10-01 03:14:26 +08:00
|
|
|
int drm_wait_vblank(struct drm_device *dev, void *data,
|
|
|
|
struct drm_file *file_priv)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2007-09-03 10:06:45 +08:00
|
|
|
union drm_wait_vblank *vblwait = data;
|
2005-04-17 06:20:36 +08:00
|
|
|
int ret = 0;
|
2011-03-19 05:58:04 +08:00
|
|
|
unsigned int flags, seq, crtc, high_crtc;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2010-05-28 03:40:25 +08:00
|
|
|
if ((!drm_dev_to_irq(dev)) || (!dev->irq_enabled))
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EINVAL;
|
|
|
|
|
2009-01-28 13:19:41 +08:00
|
|
|
if (vblwait->request.type & _DRM_VBLANK_SIGNAL)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2007-09-03 10:06:45 +08:00
|
|
|
if (vblwait->request.type &
|
2011-03-19 05:58:04 +08:00
|
|
|
~(_DRM_VBLANK_TYPES_MASK | _DRM_VBLANK_FLAGS_MASK |
|
|
|
|
_DRM_VBLANK_HIGH_CRTC_MASK)) {
|
2006-10-24 20:24:38 +08:00
|
|
|
DRM_ERROR("Unsupported type value 0x%x, supported mask 0x%x\n",
|
2007-09-03 10:06:45 +08:00
|
|
|
vblwait->request.type,
|
2011-03-19 05:58:04 +08:00
|
|
|
(_DRM_VBLANK_TYPES_MASK | _DRM_VBLANK_FLAGS_MASK |
|
|
|
|
_DRM_VBLANK_HIGH_CRTC_MASK));
|
2006-10-24 20:24:38 +08:00
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2007-09-03 10:06:45 +08:00
|
|
|
flags = vblwait->request.type & _DRM_VBLANK_FLAGS_MASK;
|
2011-03-19 05:58:04 +08:00
|
|
|
high_crtc = (vblwait->request.type & _DRM_VBLANK_HIGH_CRTC_MASK);
|
|
|
|
if (high_crtc)
|
|
|
|
crtc = high_crtc >> _DRM_VBLANK_HIGH_CRTC_SHIFT;
|
|
|
|
else
|
|
|
|
crtc = flags & _DRM_VBLANK_SECONDARY ? 1 : 0;
|
2008-10-01 03:14:26 +08:00
|
|
|
if (crtc >= dev->num_crtcs)
|
2006-10-24 20:24:38 +08:00
|
|
|
return -EINVAL;
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
ret = drm_vblank_get(dev, crtc);
|
|
|
|
if (ret) {
|
2009-08-09 10:24:01 +08:00
|
|
|
DRM_DEBUG("failed to acquire vblank counter, %d\n", ret);
|
2008-10-01 03:14:26 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
seq = drm_vblank_count(dev, crtc);
|
2006-10-24 20:24:38 +08:00
|
|
|
|
2007-09-03 10:06:45 +08:00
|
|
|
switch (vblwait->request.type & _DRM_VBLANK_TYPES_MASK) {
|
2005-04-17 06:20:36 +08:00
|
|
|
case _DRM_VBLANK_RELATIVE:
|
2007-09-03 10:06:45 +08:00
|
|
|
vblwait->request.sequence += seq;
|
|
|
|
vblwait->request.type &= ~_DRM_VBLANK_RELATIVE;
|
2005-04-17 06:20:36 +08:00
|
|
|
case _DRM_VBLANK_ABSOLUTE:
|
|
|
|
break;
|
|
|
|
default:
|
2008-10-01 03:14:26 +08:00
|
|
|
ret = -EINVAL;
|
|
|
|
goto done;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2009-09-12 02:33:34 +08:00
|
|
|
if (flags & _DRM_VBLANK_EVENT)
|
|
|
|
return drm_queue_vblank_event(dev, crtc, vblwait, file_priv);
|
|
|
|
|
2006-10-24 21:34:18 +08:00
|
|
|
if ((flags & _DRM_VBLANK_NEXTONMISS) &&
|
2007-09-03 10:06:45 +08:00
|
|
|
(seq - vblwait->request.sequence) <= (1<<23)) {
|
|
|
|
vblwait->request.sequence = seq + 1;
|
2006-10-24 21:34:18 +08:00
|
|
|
}
|
|
|
|
|
2009-01-28 13:19:41 +08:00
|
|
|
DRM_DEBUG("waiting on vblank count %d, crtc %d\n",
|
|
|
|
vblwait->request.sequence, crtc);
|
|
|
|
dev->last_vblank_wait[crtc] = vblwait->request.sequence;
|
|
|
|
DRM_WAIT_ON(ret, dev->vbl_queue[crtc], 3 * DRM_HZ,
|
|
|
|
(((drm_vblank_count(dev, crtc) -
|
|
|
|
vblwait->request.sequence) <= (1 << 23)) ||
|
|
|
|
!dev->irq_enabled));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-01-28 13:19:41 +08:00
|
|
|
if (ret != -EINTR) {
|
|
|
|
struct timeval now;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
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, ×tamp).
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
|
|
|
vblwait->reply.sequence = drm_vblank_count_and_time(dev, crtc, &now);
|
2009-01-28 13:19:41 +08:00
|
|
|
vblwait->reply.tval_sec = now.tv_sec;
|
|
|
|
vblwait->reply.tval_usec = now.tv_usec;
|
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, ×tamp).
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
|
|
|
|
2009-01-28 13:19:41 +08:00
|
|
|
DRM_DEBUG("returning %d to client\n",
|
|
|
|
vblwait->reply.sequence);
|
2005-04-17 06:20:36 +08:00
|
|
|
} else {
|
2009-01-28 13:19:41 +08:00
|
|
|
DRM_DEBUG("vblank wait interrupted by signal\n");
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
done:
|
|
|
|
drm_vblank_put(dev, crtc);
|
2005-04-17 06:20:36 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2009-09-12 02:33:34 +08:00
|
|
|
void drm_handle_vblank_events(struct drm_device *dev, int crtc)
|
|
|
|
{
|
|
|
|
struct drm_pending_vblank_event *e, *t;
|
|
|
|
struct timeval now;
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned int seq;
|
|
|
|
|
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, ×tamp).
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
|
|
|
seq = drm_vblank_count_and_time(dev, crtc, &now);
|
2009-09-12 02:33:34 +08:00
|
|
|
|
|
|
|
spin_lock_irqsave(&dev->event_lock, flags);
|
|
|
|
|
|
|
|
list_for_each_entry_safe(e, t, &dev->vblank_event_list, base.link) {
|
|
|
|
if (e->pipe != crtc)
|
|
|
|
continue;
|
|
|
|
if ((seq - e->event.sequence) > (1<<23))
|
|
|
|
continue;
|
|
|
|
|
|
|
|
DRM_DEBUG("vblank event on %d, current %d\n",
|
|
|
|
e->event.sequence, seq);
|
|
|
|
|
|
|
|
e->event.sequence = seq;
|
|
|
|
e->event.tv_sec = now.tv_sec;
|
|
|
|
e->event.tv_usec = now.tv_usec;
|
|
|
|
drm_vblank_put(dev, e->pipe);
|
|
|
|
list_move_tail(&e->base.link, &e->base.file_priv->event_list);
|
|
|
|
wake_up_interruptible(&e->base.file_priv->event_wait);
|
2010-07-02 07:48:09 +08:00
|
|
|
trace_drm_vblank_event_delivered(e->base.pid, e->pipe,
|
|
|
|
e->event.sequence);
|
2009-09-12 02:33:34 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&dev->event_lock, flags);
|
2010-07-02 07:47:31 +08:00
|
|
|
|
|
|
|
trace_drm_vblank_event(crtc, seq);
|
2009-09-12 02:33:34 +08:00
|
|
|
}
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
/**
|
|
|
|
* drm_handle_vblank - handle a vblank event
|
|
|
|
* @dev: DRM device
|
|
|
|
* @crtc: where this event occurred
|
|
|
|
*
|
|
|
|
* Drivers should call this routine in their vblank interrupt handlers to
|
|
|
|
* update the vblank counter and send any signals that may be pending.
|
|
|
|
*/
|
drm/i915: Suppress spurious vblank interrupts
Hugh Dickins found that characters in xterm were going missing and oft
delayed. Being the curious type, he managed to associate this with the
new high-precision vblank patches; disabling these he found, restored
the orderliness of his characters.
The oddness begins when one realised that Hugh was not using vblanks at
all on his system (fvwm and some xterms). Instead, all he had to go on
were warning of a pipe underrun, curiously enough at around 60Hz. He
poked and found that in addition to the underrun warning, the hardware
was flagging the start of a new frame, a vblank, which in turn was
kicking off the pending vblank processing code.
There is little we can do for the underruns on Hugh's machine, a
Crestline [965GM], which must have its FIFO watermarks set to 8.
However, we do not need to process the vblank if we know that they are
disabled...
Reported-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
2011-01-31 18:48:04 +08:00
|
|
|
bool drm_handle_vblank(struct drm_device *dev, int crtc)
|
2008-10-01 03:14:26 +08:00
|
|
|
{
|
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, ×tamp).
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
|
|
|
u32 vblcount;
|
|
|
|
s64 diff_ns;
|
|
|
|
struct timeval tvblank;
|
|
|
|
unsigned long irqflags;
|
|
|
|
|
2009-09-12 02:33:34 +08:00
|
|
|
if (!dev->num_crtcs)
|
drm/i915: Suppress spurious vblank interrupts
Hugh Dickins found that characters in xterm were going missing and oft
delayed. Being the curious type, he managed to associate this with the
new high-precision vblank patches; disabling these he found, restored
the orderliness of his characters.
The oddness begins when one realised that Hugh was not using vblanks at
all on his system (fvwm and some xterms). Instead, all he had to go on
were warning of a pipe underrun, curiously enough at around 60Hz. He
poked and found that in addition to the underrun warning, the hardware
was flagging the start of a new frame, a vblank, which in turn was
kicking off the pending vblank processing code.
There is little we can do for the underruns on Hugh's machine, a
Crestline [965GM], which must have its FIFO watermarks set to 8.
However, we do not need to process the vblank if we know that they are
disabled...
Reported-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
2011-01-31 18:48:04 +08:00
|
|
|
return false;
|
2009-09-12 02:33:34 +08:00
|
|
|
|
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, ×tamp).
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
|
|
|
/* Need timestamp lock to prevent concurrent execution with
|
|
|
|
* vblank enable/disable, as this would cause inconsistent
|
|
|
|
* or corrupted timestamps and vblank counts.
|
|
|
|
*/
|
|
|
|
spin_lock_irqsave(&dev->vblank_time_lock, irqflags);
|
|
|
|
|
|
|
|
/* Vblank irq handling disabled. Nothing to do. */
|
|
|
|
if (!dev->vblank_enabled[crtc]) {
|
|
|
|
spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
|
drm/i915: Suppress spurious vblank interrupts
Hugh Dickins found that characters in xterm were going missing and oft
delayed. Being the curious type, he managed to associate this with the
new high-precision vblank patches; disabling these he found, restored
the orderliness of his characters.
The oddness begins when one realised that Hugh was not using vblanks at
all on his system (fvwm and some xterms). Instead, all he had to go on
were warning of a pipe underrun, curiously enough at around 60Hz. He
poked and found that in addition to the underrun warning, the hardware
was flagging the start of a new frame, a vblank, which in turn was
kicking off the pending vblank processing code.
There is little we can do for the underruns on Hugh's machine, a
Crestline [965GM], which must have its FIFO watermarks set to 8.
However, we do not need to process the vblank if we know that they are
disabled...
Reported-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
2011-01-31 18:48:04 +08:00
|
|
|
return false;
|
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, ×tamp).
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
|
|
|
}
|
|
|
|
|
|
|
|
/* Fetch corresponding timestamp for this vblank interval from
|
|
|
|
* driver and store it in proper slot of timestamp ringbuffer.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* Get current timestamp and count. */
|
|
|
|
vblcount = atomic_read(&dev->_vblank_count[crtc]);
|
|
|
|
drm_get_last_vbltimestamp(dev, crtc, &tvblank, DRM_CALLED_FROM_VBLIRQ);
|
|
|
|
|
|
|
|
/* Compute time difference to timestamp of last vblank */
|
|
|
|
diff_ns = timeval_to_ns(&tvblank) -
|
|
|
|
timeval_to_ns(&vblanktimestamp(dev, crtc, vblcount));
|
|
|
|
|
|
|
|
/* Update vblank timestamp and count if at least
|
|
|
|
* DRM_REDUNDANT_VBLIRQ_THRESH_NS nanoseconds
|
|
|
|
* difference between last stored timestamp and current
|
|
|
|
* timestamp. A smaller difference means basically
|
|
|
|
* identical timestamps. Happens if this vblank has
|
|
|
|
* been already processed and this is a redundant call,
|
|
|
|
* e.g., due to spurious vblank interrupts. We need to
|
|
|
|
* ignore those for accounting.
|
|
|
|
*/
|
2011-02-21 12:42:00 +08:00
|
|
|
if (abs64(diff_ns) > DRM_REDUNDANT_VBLIRQ_THRESH_NS) {
|
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, ×tamp).
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
|
|
|
/* Store new timestamp in ringbuffer. */
|
|
|
|
vblanktimestamp(dev, crtc, vblcount + 1) = tvblank;
|
|
|
|
|
|
|
|
/* Increment cooked vblank count. This also atomically commits
|
|
|
|
* the timestamp computed above.
|
|
|
|
*/
|
2011-02-21 12:42:01 +08:00
|
|
|
smp_mb__before_atomic_inc();
|
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, ×tamp).
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
|
|
|
atomic_inc(&dev->_vblank_count[crtc]);
|
2011-02-21 12:42:01 +08:00
|
|
|
smp_mb__after_atomic_inc();
|
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, ×tamp).
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
|
|
|
} else {
|
|
|
|
DRM_DEBUG("crtc %d: Redundant vblirq ignored. diff_ns = %d\n",
|
|
|
|
crtc, (int) diff_ns);
|
|
|
|
}
|
|
|
|
|
2008-10-01 03:14:26 +08:00
|
|
|
DRM_WAKEUP(&dev->vbl_queue[crtc]);
|
2009-09-12 02:33:34 +08:00
|
|
|
drm_handle_vblank_events(dev, crtc);
|
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, ×tamp).
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
|
|
|
|
|
|
|
spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
|
drm/i915: Suppress spurious vblank interrupts
Hugh Dickins found that characters in xterm were going missing and oft
delayed. Being the curious type, he managed to associate this with the
new high-precision vblank patches; disabling these he found, restored
the orderliness of his characters.
The oddness begins when one realised that Hugh was not using vblanks at
all on his system (fvwm and some xterms). Instead, all he had to go on
were warning of a pipe underrun, curiously enough at around 60Hz. He
poked and found that in addition to the underrun warning, the hardware
was flagging the start of a new frame, a vblank, which in turn was
kicking off the pending vblank processing code.
There is little we can do for the underruns on Hugh's machine, a
Crestline [965GM], which must have its FIFO watermarks set to 8.
However, we do not need to process the vblank if we know that they are
disabled...
Reported-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
2011-01-31 18:48:04 +08:00
|
|
|
return true;
|
2008-10-01 03:14:26 +08:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(drm_handle_vblank);
|