Hi,
I have tested Thomas drm-ttm(-next) patch-sets in combination with
Daniel Vetter's patch-sets (series see below).
No bad - no good in OpenArena-benchmark.
( It is definitely pageflip-support patches causing a slow-down of 4-5
fps... 13-14 vs. 17-18 fps on a radeon RV250 is a real difference.
Registering radeon display connectors fails due to missing adapter
name. The problem surfaced after this commit to Linus' kernel tree:
commit 2236baa75f704851d3cd3310569058151acb1f06
Author: Jean Delvare
Date: Mon Nov 15 22:40:38 2010 +0100
i2c: Sanity checks on adapter registration
M
On Wed, Nov 17, 2010 at 6:44 PM, Thomas Hellstrom
wrote:
> On 11/17/2010 05:40 PM, Sedat Dilek wrote:
>>
>> Against which GIT tree is this patchset, against latest drm-next GIT they
>> fail?
>> Missing a patch or patchset?
>>
>
> On top of drm-next you first need to apply
>
> drm/ttm: ?Fix up a t
On 11/17/2010 08:11 PM, Jerome Glisse wrote:
> On Wed, Nov 17, 2010 at 7:38 AM, Thomas Hellstrom
> wrote:
>
>> The patch is compile-tested ony and not very intrusive. It should be applied
>> on top of the latest TTM patches.
>>
>> Besides reduced CPU-usage on SMP kernels, there is the benefit
We were previously dropping alignment requests on the floor
when allocating buffers so we always ended up page aligned.
Certain tiling modes on 6xx+ require larger alignment which
wasn't happening before.
Signed-off-by: Alex Deucher
Cc: Jerome Glisse
---
drivers/gpu/drm/radeon/evergreen_blit_km
On 11/17/2010 05:40 PM, Sedat Dilek wrote:
> Against which GIT tree is this patchset, against latest drm-next GIT they
> fail?
> Missing a patch or patchset?
>
On top of drm-next you first need to apply
drm/ttm: Fix up a theoretical deadlock.
drm/ttm: Add a bo list reserve fastpath (need v2
This causes the connector to not be added since i2c init fails
for the adapter. Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=31688
Noticed by Ari Savolainen.
Signed-off-by: Alex Deucher
Cc: Ari Savolainen
Cc: Jean Delvare
Cc: stable at kernel.org
---
drivers/gpu/drm/radeon/radeon_i2c.
Against which GIT tree is this patchset, against latest drm-next GIT they fail?
Missing a patch or patchset?
- Sedat -
$ cd /mnt/sdb5/linux-kernel/drm-2.6
$ git branch
* drm-next
master
$ git describe
v2.6.34-rc5-32040-g01d5a1f
$ cat patches/series
drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_ba
Makes it possible to reserve a list of buffer objects with a single
spin lock / unlock if there is no contention.
Should improve cpu usage on SMP kernels.
v2: Initialize private list members on reserve and don't call
ttm_bo_list_ref_sub() with zero put_count.
Signed-off-by: Thomas Hellstrom
---
SMP i810 systems were practically nonexistent and the configuration
was not officially supported by Intel at the time when Pentium-III
was common.
With this change, it is still possible to build a distribution kernel
that has support for SMP and includes the i810 driver without the BKL.
As a preca
We were previously dropping alignment requests on the floor
when allocating buffers so we always ended up page aligned.
Certain tiling modes on 6xx+ require larger alignment which
wasn't happening before.
Signed-off-by: Alex Deucher
Cc: Jerome Glisse
---
drivers/gpu/drm/radeon/evergreen_blit_km
This causes the connector to not be added since i2c init fails
for the adapter. Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=31688
Noticed by Ari Savolainen.
Signed-off-by: Alex Deucher
Cc: Ari Savolainen
Cc: Jean Delvare
Cc: sta...@kernel.org
---
drivers/gpu/drm/radeon/radeon_i2c.c |
https://bugs.freedesktop.org/show_bug.cgi?id=27332
--- Comment #2 from Paulo César Pereira de Andrade
2010-11-17 14:55:34 PST ---
Just tested with a build of salome version 5.1.4.
Unlike when running on a computer with an intel card (bug #27333),
it now works without setting LIBGL_ALWAYS_INDIRE
https://bugs.freedesktop.org/show_bug.cgi?id=27332
--- Comment #2 from Paulo C?sar Pereira de Andrade
2010-11-17 14:55:34 PST ---
Just tested with a build of salome version 5.1.4.
Unlike when running on a computer with an intel card (bug #27333),
it now works without setting LIBGL_ALWAYS_INDIRE
Hi,
I have tested Thomas drm-ttm(-next) patch-sets in combination with
Daniel Vetter's patch-sets (series see below).
No bad - no good in OpenArena-benchmark.
( It is definitely pageflip-support patches causing a slow-down of 4-5
fps... 13-14 vs. 17-18 fps on a radeon RV250 is a real difference.
Fix many small bugs in I2C adapter registration:
* Properly reject unsupported GPIO pin.
* Fix improper use of I2C_NAME_SIZE (which is the size of
i2c_client.name, not i2c_adapter.name.)
* Prefix adapter names with "i915" so that the user knows what the
I2C channel is connected to.
* Fix swappe
On Wed, Nov 17, 2010 at 7:38 AM, Thomas Hellstrom
wrote:
> The patch is compile-tested ony and not very intrusive. It should be applied
> on top of the latest TTM patches.
>
> Besides reduced CPU-usage on SMP kernels, there is the benefit of using
> shared code. Will also ease implementation of c
On Wed, Nov 17, 2010 at 7:28 AM, Thomas Hellstrom
wrote:
> This patch series improves on the performance and correctness of the
> reserving / fencing sequence.
>
Reviewed-by: Jerome Glisse
Registering radeon display connectors fails due to missing adapter
name. The problem surfaced after this commit to Linus' kernel tree:
commit 2236baa75f704851d3cd3310569058151acb1f06
Author: Jean Delvare
Date: Mon Nov 15 22:40:38 2010 +0100
i2c: Sanity checks on adapter registration
M
Rather than re-implementing in the Radeon driver,
Use the execbuf / cs / pushbuf utilities that comes with TTM.
This comes with an even greater benefit now that many spinlocks have been
optimized away...
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/radeon/radeon.h|4 +-
driver
The patch is compile-tested ony and not very intrusive. It should be applied
on top of the latest TTM patches.
Besides reduced CPU-usage on SMP kernels, there is the benefit of using
shared code. Will also ease implementation of concurrent CS due to the
deadlock prevention mechanisms.
If time all
Rather than having the driver supply the validation sequence, leave that
responsibility to TTM. This saves some confusion and a function argument.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/ttm/ttm_bo.c|1 +
drivers/gpu/drm/ttm/ttm_execbuf_util.c |5 -
drivers/g
Drastically reduce the number of spin lock / unlock operations by performing
unreserving and fencing under global locks.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/ttm/ttm_bo.c | 11 +++--
drivers/gpu/drm/ttm/ttm_execbuf_util.c | 36 ++-
inc
The bo lock used only to protect the bo sync object members, and since it
is a per bo lock, fencing a buffer list will see a lot of locks and unlocks.
Replace it with a per-device lock that protects the sync object members on
*all* bos. Reading and setting these members will always be very quick, s
Add an aid for the driver to detect deadlocks on multi-bo reservations
Update documentation.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/ttm/ttm_bo.c| 15 ---
include/drm/ttm/ttm_bo_driver.h | 25 ++---
2 files changed, 34 insertions(+), 6 deletion
Avoid the ttm_bo_unreserve() spinlocks by calling
ttm_eu_backoff_reservation_locked under the lru spinlock.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/ttm/ttm_execbuf_util.c | 15 ---
1 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/drivers/gpu/drm/ttm/ttm_ex
This patch series improves on the performance and correctness of the
reserving / fencing sequence.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 14 --
1 files changed, 8 insertions(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index 36e129f..5408b1b 100644
--- a/drivers/gpu/d
On Wed, Nov 17, 2010 at 6:44 PM, Thomas Hellstrom wrote:
> On 11/17/2010 05:40 PM, Sedat Dilek wrote:
>>
>> Against which GIT tree is this patchset, against latest drm-next GIT they
>> fail?
>> Missing a patch or patchset?
>>
>
> On top of drm-next you first need to apply
>
> drm/ttm: Fix up a th
We aren't currently using tiling in userspace on evergreen,
but the info we currently return for the tiling info query
(gb_addr_config) is no adequate for userspace tiling alignment
calculations. It does not contain the bank info. Create a custom
tiling info dword with all the necessary info (num
On 11/17/2010 08:11 PM, Jerome Glisse wrote:
On Wed, Nov 17, 2010 at 7:38 AM, Thomas Hellstrom wrote:
The patch is compile-tested ony and not very intrusive. It should be applied
on top of the latest TTM patches.
Besides reduced CPU-usage on SMP kernels, there is the benefit of using
share
On Wed, Nov 17, 2010 at 7:38 AM, Thomas Hellstrom wrote:
> The patch is compile-tested ony and not very intrusive. It should be applied
> on top of the latest TTM patches.
>
> Besides reduced CPU-usage on SMP kernels, there is the benefit of using
> shared code. Will also ease implementation of co
On Wed, Nov 17, 2010 at 7:28 AM, Thomas Hellstrom wrote:
> This patch series improves on the performance and correctness of the
> reserving / fencing sequence.
>
Reviewed-by: Jerome Glisse
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http
On 11/17/2010 05:40 PM, Sedat Dilek wrote:
Against which GIT tree is this patchset, against latest drm-next GIT they fail?
Missing a patch or patchset?
On top of drm-next you first need to apply
drm/ttm: Fix up a theoretical deadlock.
drm/ttm: Add a bo list reserve fastpath (need v2 which
We aren't currently using tiling in userspace on evergreen,
but the info we currently return for the tiling info query
(gb_addr_config) is no adequate for userspace tiling alignment
calculations. It does not contain the bank info. Create a custom
tiling info dword with all the necessary info (num
On Tue, 2010-11-16 at 17:19 -0500, Andrew Lutomirski wrote:
> On Wed, Nov 10, 2010 at 6:04 PM, Andy Lutomirski wrote:
> > Nouveau takes down my system quite reliably when any hotplug event occurs.
> > The bug happens because the IRQ handler didn't acknowledge the hotplug
> > state until the bottom
Against which GIT tree is this patchset, against latest drm-next GIT they fail?
Missing a patch or patchset?
- Sedat -
$ cd /mnt/sdb5/linux-kernel/drm-2.6
$ git branch
* drm-next
master
$ git describe
v2.6.34-rc5-32040-g01d5a1f
$ cat patches/series
drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_ba
Makes it possible to reserve a list of buffer objects with a single
spin lock / unlock if there is no contention.
Should improve cpu usage on SMP kernels.
v2: Initialize private list members on reserve and don't call
ttm_bo_list_ref_sub() with zero put_count.
Signed-off-by: Thomas Hellstrom
---
SMP i810 systems were practically nonexistent and the configuration
was not officially supported by Intel at the time when Pentium-III
was common.
With this change, it is still possible to build a distribution kernel
that has support for SMP and includes the i810 driver without the BKL.
As a preca
https://bugs.freedesktop.org/show_bug.cgi?id=30152
--- Comment #9 from Clément 2010-11-17 07:21:47 PST ---
Hi,
I did a git bisect (only on src/gallium, I have some problem doing it for the
whole mesa since it doesn't compile :
radeong_dri.so.tmp: undefined reference to `operator delete[](void*
https://bugs.freedesktop.org/show_bug.cgi?id=30152
--- Comment #9 from Cl?ment 2010-11-17 07:21:47 PST ---
Hi,
I did a git bisect (only on src/gallium, I have some problem doing it for the
whole mesa since it doesn't compile :
radeong_dri.so.tmp: undefined reference to `operator delete[](void*
https://bugs.freedesktop.org/show_bug.cgi?id=28995
--- Comment #23 from Álmos 2010-11-17 06:36:47 PST ---
nope, it's i686 on a pentium 4, which is not 64bit capable
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
Yo
https://bugs.freedesktop.org/show_bug.cgi?id=28995
--- Comment #23 from ?lmos 2010-11-17 06:36:47 PST ---
nope, it's i686 on a pentium 4, which is not 64bit capable
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
Yo
Fix many small bugs in I2C adapter registration:
* Properly reject unsupported GPIO pin.
* Fix improper use of I2C_NAME_SIZE (which is the size of
i2c_client.name, not i2c_adapter.name.)
* Prefix adapter names with "i915" so that the user knows what the
I2C channel is connected to.
* Fix swappe
https://bugs.freedesktop.org/show_bug.cgi?id=27755
Fabio Pedretti changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|
https://bugs.freedesktop.org/show_bug.cgi?id=27755
Fabio Pedretti changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|
Rather than re-implementing in the Radeon driver,
Use the execbuf / cs / pushbuf utilities that comes with TTM.
This comes with an even greater benefit now that many spinlocks have been
optimized away...
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/radeon/radeon.h|4 +-
driver
The patch is compile-tested ony and not very intrusive. It should be applied
on top of the latest TTM patches.
Besides reduced CPU-usage on SMP kernels, there is the benefit of using
shared code. Will also ease implementation of concurrent CS due to the
deadlock prevention mechanisms.
If time all
Rather than having the driver supply the validation sequence, leave that
responsibility to TTM. This saves some confusion and a function argument.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/ttm/ttm_bo.c|1 +
drivers/gpu/drm/ttm/ttm_execbuf_util.c |5 -
drivers/g
Drastically reduce the number of spin lock / unlock operations by performing
unreserving and fencing under global locks.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/ttm/ttm_bo.c | 11 +++--
drivers/gpu/drm/ttm/ttm_execbuf_util.c | 36 ++-
inc
The bo lock used only to protect the bo sync object members, and since it
is a per bo lock, fencing a buffer list will see a lot of locks and unlocks.
Replace it with a per-device lock that protects the sync object members on
*all* bos. Reading and setting these members will always be very quick, s
Add an aid for the driver to detect deadlocks on multi-bo reservations
Update documentation.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/ttm/ttm_bo.c| 15 ---
include/drm/ttm/ttm_bo_driver.h | 25 ++---
2 files changed, 34 insertions(+), 6 deletion
Avoid the ttm_bo_unreserve() spinlocks by calling
ttm_eu_backoff_reservation_locked under the lru spinlock.
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/ttm/ttm_execbuf_util.c | 15 ---
1 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/drivers/gpu/drm/ttm/ttm_ex
This patch series improves on the performance and correctness of the
reserving / fencing sequence.
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel
Signed-off-by: Thomas Hellstrom
---
drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 14 --
1 files changed, 8 insertions(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index 36e129f..5408b1b 100644
--- a/drivers/gpu/d
https://bugs.freedesktop.org/show_bug.cgi?id=28995
--- Comment #22 from Marek Olšák 2010-11-17 03:59:52 PST ---
Aren't you running an x86_64 kernel?
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the assign
https://bugs.freedesktop.org/show_bug.cgi?id=28995
--- Comment #22 from Marek Ol??k 2010-11-17 03:59:52 PST
---
Aren't you running an x86_64 kernel?
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the assig
https://bugs.freedesktop.org/show_bug.cgi?id=25093
Andy Furniss changed:
What|Removed |Added
Attachment #31208|0 |1
is obsolete|
https://bugs.freedesktop.org/show_bug.cgi?id=25093
Andy Furniss changed:
What|Removed |Added
Attachment #31208|0 |1
is obsolete|
The eDP panel must be powered up for aux transactions, so power it
up for detect and mode probe functions, otherwise power it up or
down based on dpms.
v2:
- only mess with eDP panel on DCE4+
- only mess with eDP panel on eDP connectors, not all DP connectors
v3:
- be extra careful to only mess wi
If there was no connector mapped to the encoder, atombios_get_encoder_mode()
returned 0 which is the id for DP. Return something sane instead based on
the encoder id. This avoids hitting the DP paths on non-DP encoders.
Signed-off-by: Alex Deucher
Cc: stable at kernel.org
---
drivers/gpu/drm/r
The eDP panel must be powered up for aux transactions, so power it
up for detect and mode probe functions, otherwise power it up or
down based on dpms.
v2:
- only mess with eDP panel on DCE4+
- only mess with eDP panel on eDP connectors, not all DP connectors
v3:
- be extra careful to only mess wi
62 matches
Mail list logo
