---
libdrm/xf86drm.h |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/libdrm/xf86drm.h b/libdrm/xf86drm.h
index c1d173c..a7f9763 100644
--- a/libdrm/xf86drm.h
+++ b/libdrm/xf86drm.h
@@ -37,6 +37,7 @@
#include stdarg.h
#include sys/types.h
#include stdint.h
+#include
---
configure.ac | 11 ++-
1 files changed, 10 insertions(+), 1 deletions(-)
diff --git a/configure.ac b/configure.ac
index be2fbbf..763464b 100644
--- a/configure.ac
+++ b/configure.ac
@@ -52,6 +52,11 @@ AC_ARG_ENABLE(radeon-experimental-api,
[Enable support for radeon's
---
libdrm/xf86drm.c |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/libdrm/xf86drm.c b/libdrm/xf86drm.c
index 1e6eb7a..7b05386 100644
--- a/libdrm/xf86drm.c
+++ b/libdrm/xf86drm.c
@@ -34,6 +34,7 @@
#ifdef HAVE_CONFIG_H
# include config.h
#endif
+#define _XOPEN_SOURCE
---
shared-core/drm.h |1 -
1 files changed, 0 insertions(+), 1 deletions(-)
diff --git a/shared-core/drm.h b/shared-core/drm.h
index 7758af4..42a6c23 100644
--- a/shared-core/drm.h
+++ b/shared-core/drm.h
@@ -855,7 +855,6 @@ typedef struct drm_set_version drm_set_version_t;
typedef struct
I did try to compile libdrm with -pedantic -Wall -Werror but there is a lot
of warnings generated.
This patch set fixes compilation of libdrm/xf86drm.c. I also made patch to add
configuration option to make it easy for developers to use strict compilation
with all warnings enabled as errors.
---
libdrm/xf86drm.c | 12 ++--
1 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/libdrm/xf86drm.c b/libdrm/xf86drm.c
index 55df19a..1e6eb7a 100644
--- a/libdrm/xf86drm.c
+++ b/libdrm/xf86drm.c
@@ -240,22 +240,22 @@ static int drmMatchBusID(const char *id1, const char *id2)
---
libdrm/xf86drm.c | 49 ++---
1 files changed, 46 insertions(+), 3 deletions(-)
diff --git a/libdrm/xf86drm.c b/libdrm/xf86drm.c
index 7b05386..bf80fd4 100644
--- a/libdrm/xf86drm.c
+++ b/libdrm/xf86drm.c
@@ -269,6 +269,49 @@ static int
This might also require -std=c99 to set c standard correctly. I did
use that in my CFLAGS when I tried to compile with -Werror.
--
--
___
Dri-devel mailing list
strings.h (strcasecmp is posix function)
On Sat, Jul 4, 2009 at 10:32 PM, Julien Cristaujcris...@debian.org wrote:
On Sat, Jul 4, 2009 at 02:18:50 +0300, Pauli Nieminen wrote:
---
libdrm/xf86drm.h | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/libdrm/xf86drm.h b
strcasecmp is declared in strings.h
---
libdrm/xf86drm.h |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/libdrm/xf86drm.h b/libdrm/xf86drm.h
index c1d173c..a7f9763 100644
--- a/libdrm/xf86drm.h
+++ b/libdrm/xf86drm.h
@@ -37,6 +37,7 @@
#include stdarg.h
#include
Attached improved patch.
On Sat, Jul 4, 2009 at 11:14 AM, Pekka Paalanenp...@iki.fi wrote:
On Sat, 4 Jul 2009 02:18:53 +0300
Pauli Nieminen suok...@gmail.com wrote:
---
libdrm/xf86drm.c | 49 ++---
1 files changed, 46 insertions(+), 3
fc7f62d955f2d7b66cd5179eace2f628a7c66277 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Sun, 5 Jul 2009 18:34:29 +0300
Subject: [PATCH 6/6] libdrm: Add configuration option to enable strict compilation
Source is writen for c99 standard so we have to tell compiler about it with -std=c99
attached to add the -std=c99 if using strict compilation
I should learn to type and test everything well before submitting.
From fc7f62d955f2d7b66cd5179eace2f628a7c66277 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Sun, 5 Jul 2009 18:34:29 +0300
Subject: [PATCH 6/6] libdrm
From 3eb4a3493f13c48bee1b2453f5ecead0ba498c51 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 15:45:41 +0300
Subject: [PATCH 05/15] libdrm: Fix drmOpenDevice to take dev_t type as
parameter.
This fixes signed/unsigned comparission between st.st_dev and dev
From 78bfacedfb3b81968b2171a942101d323b1782cb Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 15:48:48 +0300
Subject: [PATCH 06/15] libdrm: Fix loop index variable type to match
the type of count
---
libdrm/xf86drm.c |4 ++--
1 files changed, 2
From 10a9bf298be4695242bcdb63bcfad720e03bf040 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 16:31:54 +0300
Subject: [PATCH 08/15] libdrm: Make gcc to understand that pointers
are realy pointers.
---
libdrm/xf86drmSL.c |6 +++---
1 files changed, 3
From 7fcfbf715644d2268236b8a19d1f2f83c952d72b Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 16:25:59 +0300
Subject: [PATCH 07/15] libdrm: Fix random number generator to use unsigned seed.
This fixes wanring about unsigned/signed comparision.
Also make
From 69cad004e5355e4adcee753de4da91c98949d3df Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 16:50:20 +0300
Subject: [PATCH 09/15] libdrm: Make drmAllocCpy static.
---
libdrm/xf86drmMode.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff
From 6ad9d2f1495df9bbdecbe4e7676da75d48ba3719 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 16:55:21 +0300
Subject: [PATCH 10/15] libdrm: Make drmAllocCpy use only single memcpy
call instead of calling memcpy in loop.
---
libdrm/xf86drmMode.c |4
From 9043299720ef5ecac1fe575b0bbda19b81720dd8 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 18:00:43 +0300
Subject: [PATCH 13/15] libdrm/radeon: Remove unussed variables and
make printf happy with explicity void* cast from struct radeon_bo*.
---
libdrm
From dea4cebdb49556ab06930bd5a4d303ce1c0b85c7 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 17:19:55 +0300
Subject: [PATCH 12/15] intel: Move debug output after setting the value.
---
libdrm/intel/intel_bufmgr_fake.c |4 ++--
1 files changed, 2
From 0c8b7939fc95084d77063eb8de78b0724d3d4058 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 17:00:45 +0300
Subject: [PATCH 11/15] libdrm: Make pci id numbers unsigned for sscanf.
---
libdrm/xf86drmMode.c |3 ++-
1 files changed, 2 insertions(+), 1
From 9c0cd0ba5bb5966498cad2c5a8c69ec24c4c8e34 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 18:40:49 +0300
Subject: [PATCH 15/15] Fix signed/unsigned comparison warning
---
libdrm/radeon/radeon_cs_gem.c | 13 +++--
1 files changed, 7 insertions
From 36aef4bf07154ef6d206f15ff1c036a7f8e368a9 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 18:33:03 +0300
Subject: [PATCH 14/15] libdrm/radeon: Add config.h include to enable
POSIX extensions for strdup call.
---
libdrm/radeon/radeon_track.c |3
And then follows fixes for all tests.
From 1cd7a914be3fc5ee8aabdd2297710f5c781b555f Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 21:36:57 +0300
Subject: [PATCH 1/4] libdrm/tests: Fix compiler warnings in dristat.c.
---
tests/dristat.c | 12
On Tue, Jul 21, 2009 at 12:21 PM, Julien Cristau jcris...@debian.orgwrote:
On Mon, Jul 20, 2009 at 19:17:01 +0300, Pauli Nieminen wrote:
From b7e77b71d1f2d8ff6741e534911e09f10e3f3d4e Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Mon, 20 Jul 2009 14:39:57 +0300
---
shared-core/drm.h |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/shared-core/drm.h b/shared-core/drm.h
index 42a6c23..155b0f6 100644
--- a/shared-core/drm.h
+++ b/shared-core/drm.h
@@ -243,7 +243,7 @@ enum drm_map_type {
_DRM_SCATTER_GATHER = 4, /**
This will bo tracking also enabled in mesa if it is enabled in libdrm tough
pkg-config.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
configure.ac | 14 ++
libdrm/radeon/Makefile.am |1 +
libdrm/radeon/libdrm_radeon.pc.in |3 ++-
3 files
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
libdrm/radeon/radeon_track.c |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/libdrm/radeon/radeon_track.c b/libdrm/radeon/radeon_track.c
index 09f8725..0068aa9 100644
--- a/libdrm/radeon/radeon_track.c
+++ b/libdrm
s/TRACK_BO/BO_TRACK/
From 5f4ad665d42791416973391a43236d64427f6304 Mon Sep 17 00:00:00 2001
From: Pauli Nieminen suok...@gmail.com
Date: Fri, 7 Aug 2009 19:24:30 +0300
Subject: [PATCH] libdrm/radeon: Add configure option to enable bo tracking.
This will bo tracking also enabled in mesa
Using this call in OUT_BATCH_TABLE reduces radeonEmitState cpu usage from
9% to 5% and emit_vpu goes from 7% to 1.5%. I did use calgrind to profile
gears for cpu hotspots with r500 card.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
libdrm/radeon/radeon_cs.h |9 +
1 files
GCC did war about optimization not possible because possible forever loop.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
libdrm/radeon/radeon_cs_gem.c | 12 ++--
1 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/libdrm/radeon/radeon_cs_gem.c b/libdrm/radeon
Hi!
Idea is to not to block client application when it request buffer swap. This
should give small performance gains when swap buffer is synchronised to
vblank.
I can think of 3 different implementations to archive this continues
execution of client.
First one is to just allocate 2 back buffers
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
libdrm/radeon/radeon_bo.h | 13 +
libdrm/radeon/radeon_bo_gem.c | 18 ++
shared-core/radeon_drm.h |4 +++-
3 files changed, 34 insertions(+), 1 deletions(-)
diff --git a/libdrm/radeon/radeon_bo.h b
documentations.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/r100.c |5 +-
drivers/gpu/drm/radeon/r300.c | 245 ++-
drivers/gpu/drm/radeon/r300_reg.h | 84 -
drivers/gpu/drm/radeon/r420.c |4
be better only to flush when necessary it is very hard to know
when it is required. So we do flush every time because it is better to be safe
than sorry.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/r100.c | 11 ++-
1 files changed, 10 insertions(+), 1
2009/9/9 Jerome Glisse gli...@freedesktop.org
On Wed, 2009-09-09 at 10:50 +0200, Rafał Miłecki wrote:
I'm thinking of organizing power management in KMS. There are my
ideas, would like to get your opinions.
First of all we need a few different power modes. I think radeonhd has
nice
2009/9/17 Rafał Miłecki zaj...@gmail.com
W dniu 18 września 2009 00:55 użytkownik Rafał Miłecki
zaj...@gmail.com napisał:
My new proposition is rather create some user_power_mode and let
user specify all 3 params. For example:
$ echo 20 40 1050 /sys/class/gpu/user_power_mode
$
Hi!
I have been thinking GPU reset as possible DoS attack from
user-space.Problem here is that display doesn't work anymore at all if
attacker chooses to run a application that constantly causes GPU hang. It
would be of course ideal to have CS checker not to let in any problematic
combinations of
On Tue, Sep 22, 2009 at 11:51 PM, Corbin Simpson
mostawesomed...@gmail.comwrote:
On 09/22/2009 01:19 PM, Nicolai Hähnle wrote:
I'm pretty confident that you can write a perfectly legal OpenGL
application
that creates commands that take *minutes* to run on decent graphics
cards.
Just
2009/9/23 Dave Airlie airl...@linux.ie
From: Dave Airlie airl...@redhat.com
This avoids needing to do a kmalloc PAGE_SIZE for the main
indirect buffer chunk, it adds an accessor for all reads from
the chunk and caches a single page at a time for subsequent
reads.
FWIW,
On Wed, Sep 23, 2009 at 10:52 PM, Nicolai Hähnle nhaeh...@gmail.com wrote:
Am Tuesday 22 September 2009 23:25:09 schrieb Pauli Nieminen:
Too bad GPU reset is already now stopping this use case while it doesn't
protect user from possible attack causing multiple GPU reset in row. So
Hi Marc!
Is it an agp card?
If yes you could try to boot with agpmode=-1 parameter for radeon module.
On Tue, Jan 26, 2010 at 7:32 PM, Marc R. O'Connor mrocon...@oel.state.nj.us
wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
I have been having issues with KMS on 2.6.33-rcX kernels.
Fixing kernel memory allocation bugs in radeon command stream parser requires
splitting the buffer to multiple independ pages.
To make splitting one large array to multiple pages first patch adds drm_buffer
structure and functions to reference data in the buffer.
Second patch converts existing
for
data if it is split over boundary in the buffer.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/Makefile |2 +-
drivers/gpu/drm/drm_buffer.c | 184 ++
include/drm/drm_buffer.h | 148 +
3 files
to go tough drm_buffer
functions to get correct pointer.
Outputting the command stream to ring buffer needs to be awear of
the split nature of drm_buffer. The output operation requires the
new OUT_RING_DRM_BUFFER.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon
If there is section size mismatch reusing the section object
makes section start fail.
Reseting the object before doing error checking prevents the
possible flood of errors.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
radeon/radeon_cs_gem.c |2 +-
1 files changed, 1 insertions(+), 1
the section even when there is error. */
+ cs-section_ndw = 0;
+ return -EPIPE;
}
cs-section_ndw = 0;
return 0;
On Mon, Feb 1, 2010 at 10:12 PM, Dave Airlie airl...@gmail.com wrote:
On Tue, Feb 2, 2010 at 4:24 AM, Pauli Nieminen suok...@gmail.com wrote
On Tue, Feb 2, 2010 at 10:50 PM, Dave Airlie airl...@gmail.com wrote:
On Tue, Feb 2, 2010 at 3:11 AM, Pauli Nieminen suok...@gmail.com wrote:
Allocating multiple pages of memory for data that is coming
from user space may fail. To fix memory allocation failures
the buffer object should
Cache flush is required in case CPU is accessing rendered data.
This fixes glean/readPixSanity test case and random rendering
errors in sauerbraten.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/r100.c |6 ++
drivers/gpu/drm/radeon/radeon_reg.h |2
Cache flush is required in case CPU is accessing rendered data.
This fixes glean/readPixSanity test case and random rendering
errors in sauerbraten and warzone2100.
v2 Fix comment ordering in r100_fence_ring_emit and remove extra
defines added in first version.
Signed-off-by: Pauli Nieminen
On Sat, Feb 6, 2010 at 1:54 AM, Paul Menzel
paulepan...@users.sourceforge.net wrote:
Am Donnerstag, den 04.02.2010, 19:49 +0200 schrieb Pauli Nieminen:
[…]
+/* Who ever call radeon_fence_emit should call ring_lock and ask
s/ever call/ever calls/
Too bad you are asking too much from me
radeon_copy_dma is only available for r200 or newer cards.
Call to radeon_copy_dma would result to NULL pointer
dereference if benchmarking asic without dma engine.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/radeon_benchmark.c | 55
that supports
starts from r200 cards.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/r200.c | 46
drivers/gpu/drm/radeon/r300.c | 44 --
drivers/gpu/drm/radeon/radeon_asic.h | 18
that supports
starts from r200 cards.
v2: Created a new asic object for r200 cards.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/r200.c | 46 +
drivers/gpu/drm/radeon/r300.c | 44
drivers/gpu/drm/radeon
Setting global asic structure to point to different function
would cause problem in system where is multiple r300 cards
with different bus type.
r300_asic_pcie is just copy from r300_asic with gart tlb
functions replaced with pcie versions.
Signed-off-by: Pauli Nieminen suok...@gmail.com
This is in the shared code. I see this bug with r200 too. Models have
worked before rewrite so this is some regression in rewrite.
--- Comment #10 from Alex Deucher ag...@yahoo.com 2010-02-12 11:27:49 PST
---
(In reply to comment #9)
The last commit for mesa i have is,
to
load the module.
drmCheckModesettingSupported is now deprecated.
Version bump is also required for the interface change.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
configure.ac |2 +-
xf86drm.c |3 ++-
xf86drmMode.c | 29 -
xf86drmMode.h |6
This prevents ddx from selecting UMS if KMS is enabled but
kernel module is not loaded before X.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
configure.ac |3 +++
src/radeon_probe.c |6 ++
2 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/configure.ac b
for multiple threads.
Based on Jerome Glisse's and Dave Airlie's pool allocator.
Signed-off-by: Jerome Glisse jgli...@redhat.com
Signed-off-by: Dave Airlie airl...@redhat.com
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm/Makefile |2 +-
drivers/gpu/drm/ttm/ttm_memory.c
Type of iterator was promoted to unsigned long in 64bit systems.
*header is small structure so it is alwas safe to cast return value
of sizeof operator to int.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/r300_cmdbuf.c |2 +-
1 files changed, 1 insertions(+), 1
Bit has table will be first checked from BO if we can quarentee this BO is not
in this cs already. When bo is emited the reference to cs is removed with bit
operations.
To quarentee that there is no other cs with same id number of CS that can have
id is limited to number of bits in unsigned long.
Bit has table will be first checked from BO if we can quarentee this BO is not
in this cs already. When bo is emited the reference to cs is removed with bit
operations.
To quarentee that all cs ids are unique number of ids is limited to 32. If
application uses more than 32 cs objects extra cs
2010/3/2 Thomas Hellström tho...@shipmail.org:
Michel Dänzer wrote:
On Tue, 2010-03-02 at 00:32 +0200, Pauli Nieminen wrote:
bo allocation is still too expensive operation for dma buffers in
classic mesa. It takes quite a lot cpu time in bind memory (agp
system) and ttm_mem_global functions
On Wed, Mar 3, 2010 at 4:50 PM, Jerome Glisse gli...@freedesktop.org wrote:
On Wed, Mar 03, 2010 at 04:23:08PM +0200, Pauli Nieminen wrote:
On Wed, Mar 3, 2010 at 3:33 PM, Jerome Glisse gli...@freedesktop.org wrote:
On Tue, Mar 02, 2010 at 12:32:54AM +0200, Pauli Nieminen wrote:
On Sun, Feb
On Tue, Mar 9, 2010 at 4:45 PM, Jerome Glisse jgli...@redhat.com wrote:
This patch cleanup the fence code, it drops the timeout field of
fence as the time to complete each IB is unpredictable and shouldn't
be bound.
The fence cleanup lead to GPU lockup detection improvement, this
patch
On Wed, Mar 10, 2010 at 7:50 AM, Paul Mundt let...@linux-sh.org wrote:
On Tue, Mar 09, 2010 at 10:08:28PM +0100, Rafael J. Wysocki wrote:
On Tuesday 09 March 2010, James Simmons wrote:
Second, in the KMS case, we'd be able to skip the kernel VT switch,
because
the KMS driver
to a bo from multiple contexts.
This optimization decreases cs_write_reloc share of torcs profiling from 4.3%
to 2.6%.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
radeon/radeon_bo_gem.c |1 +
radeon/radeon_cs.c |6 ++
radeon/radeon_cs.h |2 +-
radeon/radeon_cs_gem.c | 133
intel_atomic.h includes very usefull atomic operations for
lock free parrallel access of variables. Moving these to
core libdrm for code sharing with radeon.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
Makefile.am |2 +-
configure.ac |2 +-
intel/intel_atomic.h
2010/3/10 Michel Dänzer mic...@daenzer.net:
On Wed, 2010-03-10 at 18:20 +0200, Pauli Nieminen wrote:
Bit has table will be first checked from BO if we can quarentee this BO is
not
in this cs already.
To quarentee that there is no other cs with same id number of CS that can
have
id
objects rest will get id 0.
This optimization decreases cs_write_reloc share of torcs profiling from 4.3%
to 2.6%.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
radeon/radeon_bo_gem.c |1 +
radeon/radeon_bo_int.h |4 +-
radeon/radeon_cs.c |6 ++
radeon/radeon_cs.h |2
intel_atomic.h includes very usefull atomic operations for
lock free parrallel access of variables. Moving these to
core libdrm for code sharing with radeon.
V2: Fix remaining references to intel_atomic.h and libdrm-intel.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
Makefile.am
2010/3/11 Michel Dänzer mic...@daenzer.net:
On Wed, 2010-03-10 at 13:10 -0500, Alex Deucher wrote:
On Wed, Mar 10, 2010 at 1:05 PM, Alex Deucher alexdeuc...@gmail.com wrote:
On Wed, Mar 10, 2010 at 12:42 PM, James Simmons jsimm...@infradead.org
wrote:
See my other post about what fbdev
cs_write_reloc share of torcs profiling from 4.3%
to 2.6%.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
configure.ac | 27 ---
radeon/radeon_bo_gem.c |9
radeon/radeon_bo_gem.h |1 +
radeon/radeon_cs.c |6 ++
radeon/radeon_cs.h |2 +-
radeon
intel_atomic.h includes very usefull atomic operations for
lock free parrallel access of variables. Moving these to
core libdrm for code sharing with radeon.
V2: Fix remaining references to intel_atomic.h and libdrm-intel.
V3: Remove useless wrapper header intel_atomic.h
Signed-off-by: Pauli
Cleanup make system so that all noninstalled headers are put
to noinst_HEADERS. This quarentees that header will be present
in tar ball but not installed with make install.
CC: nouv...@lists.freedesktop.org
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
Makefile.am |6
This looks like screen is destroyed before context. Can someone
reproduce this in valgrind and attach the log?
It may not crash in valgrind but there should be double free report
with same backtrace.
--
Download
When allocating wc/uc pages cache state transition requires cache flush which
is expensive operation. To avoid cache flushes allocation of wc/uc pages should
be done in large groups when only single cache flush is required for whole group
of pages.
In some cases drivers need t oallocate and
Using single call to set multiple pages to wc reduces number of expensive cache
flushes.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm/ttm_page_alloc.c | 12 +---
1 files changed, 1 insertions(+), 11 deletions(-)
diff --git a/drivers/gpu/drm/ttm
-by: Pauli Nieminen suok...@gmail.com
---
arch/x86/include/asm/cacheflush.h |2 +
arch/x86/mm/pageattr.c| 53 +++-
2 files changed, 47 insertions(+), 8 deletions(-)
diff --git a/arch/x86/include/asm/cacheflush.h
b/arch/x86/include/asm/cacheflush.h
ttm_page_pool file is hooked ttm_page_alloc_debugfs for pool
allocator state.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/radeon_ttm.c | 13 ++---
1 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c
b
ttm_page_alloc_debugfs can be registered to output the state
of pools.
Debugfs file will output number of pages freed from the pool,
number of pages in pool now and the lowes number of pages in
pool since previous shrink.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm
ttm_page_pool file is hooked ttm_page_alloc_debugfs for pool
allocator state.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/nouveau/nouveau_debugfs.c |3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/drivers/gpu/drm/nouveau/nouveau_debugfs.c
b
.
The allocation sizes of pools is doubled everytime refill is done more than
once between shrink calls. While the size is halfed happens if there is no
refills for manager-shrink_alloc_interval seconds.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm/ttm_page_alloc.c | 105
...@redhat.com
Signed-off-by: Dave Airlie airl...@redhat.com
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm/Makefile |2 +-
drivers/gpu/drm/ttm/ttm_memory.c |7 +-
drivers/gpu/drm/ttm/ttm_page_alloc.c | 775 ++
drivers/gpu/drm/ttm
On Thu, Mar 18, 2010 at 1:52 AM, Dave Airlie airl...@gmail.com wrote:
On Thu, Mar 18, 2010 at 6:50 AM, Pauli Nieminen suok...@gmail.com wrote:
Setting single memory pages at a time to wc takes a lot time in cache flush.
To
reduce number of cache flush set_pages_array_wc
On Thu, Mar 18, 2010 at 9:29 PM, Suresh Siddha
suresh.b.sid...@intel.com wrote:
On Thu, 2010-03-18 at 02:41 -0700, Pauli Nieminen wrote:
On Thu, Mar 18, 2010 at 1:52 AM, Dave Airlie airl...@gmail.com wrote:
On Thu, Mar 18, 2010 at 6:50 AM, Pauli Nieminen suok...@gmail.com wrote:
Setting
On Fri, Mar 19, 2010 at 2:35 AM, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
We can get this if the user moves the mouse when we are waiting to move
some stuff around in the validate. Don't fail.
Signed-off-by: Dave Airlie airl...@redhat.com
---
On Fri, Mar 19, 2010 at 8:43 AM, Pauli Nieminen suok...@gmail.com wrote:
On Fri, Mar 19, 2010 at 2:35 AM, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
We can get this if the user moves the mouse when we are waiting to move
some stuff around in the validate. Don't
code.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/radeon_cs.c |8 +---
1 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c
b/drivers/gpu/drm/radeon/radeon_cs.c
index 70ba02e..dffb356 100644
--- a/drivers/gpu/drm
Pauli Nieminen wrote:
When allocating wc/uc pages cache state transition requires cache flush
which
is expensive operation. To avoid cache flushes allocation of wc/uc pages
should
be done in large groups when only single cache flush is required for whole
group
of pages.
In some cases
This patch series simplifies pool management code and adds sysfd interface
to control pool functionality from sysfs.
Pool size reduction only happens when mm shrinker request for free
pages. To prevent pool from wasting too much free memory there is
user configureable maximum size. If pool tries
ttm_page_pool file is hooked ttm_page_alloc_debugfs for pool
allocator state.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/nouveau/nouveau_debugfs.c |3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/drivers/gpu/drm/nouveau/nouveau_debugfs.c
b
ttm_page_pool file is hooked ttm_page_alloc_debugfs for pool
allocator state.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/radeon/radeon_ttm.c | 13 ++---
1 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c
b
Using single call to set multiple pages to wc reduces number of expensive cache
flushes.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm/ttm_page_alloc.c | 12 +---
1 files changed, 1 insertions(+), 11 deletions(-)
diff --git a/drivers/gpu/drm/ttm
ttm_page_alloc_debugfs can be registered to output the state
of pools.
Debugfs file will output number of pages freed from the pool,
number of pages in pool now and the lowes number of pages in
pool since previous shrink.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm
pool allocator.
Signed-off-by: Jerome Glisse jgli...@redhat.com
Signed-off-by: Dave Airlie airl...@redhat.com
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm/Makefile |2 +-
drivers/gpu/drm/ttm/ttm_memory.c |7 +-
drivers/gpu/drm/ttm/ttm_page_alloc.c
-by: Pauli Nieminen suok...@gmail.com
---
arch/x86/include/asm/cacheflush.h |2 +
arch/x86/mm/pageattr.c| 53 +++-
2 files changed, 47 insertions(+), 8 deletions(-)
diff --git a/arch/x86/include/asm/cacheflush.h
b/arch/x86/include/asm/cacheflush.h
Sysfs interface allows user to configure pool allocator functionality and
change limits for the size of pool.
Signed-off-by: Pauli Nieminen suok...@gmail.com
---
drivers/gpu/drm/ttm/ttm_memory.c |2 +-
drivers/gpu/drm/ttm/ttm_page_alloc.c | 113
On Fri, Mar 26, 2010 at 3:59 PM, Jerome Glisse gli...@freedesktop.org wrote:
On Thu, Mar 25, 2010 at 12:36:43AM +0200, Pauli Nieminen wrote:
On AGP system we might allocate/free routinely uncached or wc memory,
changing page from cached (wb) to uc or wc is very expensive and involves
a lot
1 - 100 of 116 matches
Mail list logo