On Fri, 2016-01-15 at 12:55 -0200, Gustavo Padovan wrote:
> From: Gustavo Padovan
>
> This patch series de-stage the sync framework, and in order to accomplish that
> a bunch of cleanups/improvements on the sync and fence were made.
Perhaps add/update a MAINTAINERS entry too?
__
On Fri, Jan 15, 2016 at 03:03:42PM -0800, Laura Abbott wrote:
> (adding linux-arm and a few people)
>
> On 01/14/2016 06:42 PM, Chen Feng wrote:
> >The page is already alloc at ion_alloc function,
> >ion_mmap map the alloced pages to user-space.
> >
> >The default prot can be PTE_RDONLY. Take a lo
(adding linux-arm and a few people)
On 01/14/2016 06:42 PM, Chen Feng wrote:
The page is already alloc at ion_alloc function,
ion_mmap map the alloced pages to user-space.
The default prot can be PTE_RDONLY. Take a look at
here:
set_pte_at()
arch/arm64/include/asm:
if (pte_dirty
On 01/15/2016 08:42 PM, Bhumika Goyal wrote:
> This patch adds apace around '-' operator.Found using checkpatch.pl
>
> Signed-off-by: Bhumika Goyal
> ---
> drivers/staging/iio/adc/ad7280a.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/staging/iio/adc/ad72
This patch adds space around '-' operator.Found using checkpatch.pl
Signed-off-by: Bhumika Goyal
---
drivers/staging/iio/adc/ad7192.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/staging/iio/adc/ad7192.c b/drivers/staging/iio/adc/ad7192.c
index 9221103..54db0f0 100
This patch adds apace around '-' operator.Found using checkpatch.pl
Signed-off-by: Bhumika Goyal
---
drivers/staging/iio/adc/ad7280a.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/staging/iio/adc/ad7280a.c
b/drivers/staging/iio/adc/ad7280a.c
index f45ebed..0c7
These patches adds space around '-' operator.Found using checkpatch.pl.
Bhumika Goyal (2):
staging:iio:adc:added space around '-'
staging:iio:adc:add space around '-'
drivers/staging/iio/adc/ad7192.c | 2 +-
drivers/staging/iio/adc/ad7280a.c | 4 ++--
2 files changed, 3 insertions(+), 3 de
2016-01-15 John Harrison :
> On 15/01/2016 14:55, Gustavo Padovan wrote:
> >From: Gustavo Padovan
> >
> >All changes to timeline value come through the user via
> >fence_timeline_signal() calls. When fence_timeline_destroy() is called no
> >changes on timeline->value happens hence call fence_time
On 15/01/2016 14:55, Gustavo Padovan wrote:
From: Gustavo Padovan
All changes to timeline value come through the user via
fence_timeline_signal() calls. When fence_timeline_destroy() is called no
changes on timeline->value happens hence call fence_timeline_signal() with
no increment is pointles
From: David Kershner
>From napi documentation you should only process the amount your
budget allows, if you go over it just wait for the next napi poll
to continue.
Signed-off-by: David Kershner
Signed-off-by: Benjamin Romer
---
drivers/staging/unisys/visornic/visornic_main.c | 19 +++
From: Gustavo Padovan
As we moved value storage to fence_timeline and fence those two structs
became useless and can be removed now.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sw_sync.c| 15 ++-
drivers/staging/android/sw_sync.h| 24 ++--
From: Gustavo Padovan
Trivial indentation fix.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sync.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/staging/android/sync.h b/drivers/staging/android/sync.h
index 7a4d820..f3d4f75 100644
--- a/drivers/stagi
From: Gustavo Padovan
Updates comments about functions and structures.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sync.h | 45 --
1 file changed, 21 insertions(+), 24 deletions(-)
diff --git a/drivers/staging/android/sync.h b/drivers/sta
From: Gustavo Padovan
signaled_pts is not used in this function.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/dma-buf/fence.c b/drivers/dma-buf/fence.c
index 09faf2e..7a5fc9b 100644
--- a/drivers/dma-buf/fence.c
+++ b/dri
From: Gustavo Padovan
Bring the debug information that belonged to sync_timeline and sync_pt
back.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/Makefile | 2 +-
drivers/dma-buf/fence.c | 4 ++
drivers/dma-buf/fence_debug.c| 128 +++
From: Gustavo Padovan
Move driver name handling to fence and create a default function for it.
Returns the driver which the fence belongs.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c | 19 ++-
drivers/staging/android/sw_sync.c| 3 +--
drivers/st
From: Gustavo Padovan
This functions in intended to replace sync_pt_create() and it does
exactly the same thing sync_pt_create() did.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c| 38 ++
drivers/staging/android/sync.c | 20 ++---
From: Gustavo Padovan
Called when the fence_timeline is destroyed so users can cleanup routines
on fences.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c | 24
include/linux/fence.h | 4
2 files changed, 28 insertions(+)
diff --git a/drivers/dma-bu
From: Gustavo Padovan
fence_timeline_ops was just workarounding over fence_ops and was used
chained in the fence ops. So remove it all to simplify the fence code
flow.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c | 7 ++-
drivers/staging/android/sw_sync.c | 20 +---
From: Gustavo Padovan
get sw_sync_pt_create() to use fence_create_on_timeline() directly and
remove an abstractions layers between fences and its users.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sw_sync.c | 22 ++
drivers/staging/android/sync.c| 20
From: Gustavo Padovan
All changes to timeline value come through the user via
fence_timeline_signal() calls. When fence_timeline_destroy() is called no
changes on timeline->value happens hence call fence_timeline_signal() with
no increment is pointless.
Signed-off-by: Gustavo Padovan
---
drive
From: Gustavo Padovan
When the timeline is being destroyed all the waiters waiting for the
sync_fence to signal need to be woken up and finished as well.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sw_sync.c | 8
drivers/staging/android/sync.c| 9 +
drivers/
From: Gustavo Padovan
The fence timeline struct already receives the name of the timeline so
we can use a default function to return the timeline name.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c| 14 ++
drivers/staging/android/sync.c | 9 +
include/
From: Gustavo Padovan
The value needed by sw_sync is now inside struct fence so we create a
default ops to reply that value for us.
---
drivers/dma-buf/fence.c| 21 +
drivers/staging/android/sync.c | 12 +---
include/linux/fence.h | 1 +
3 files chan
From: Gustavo Padovan
Now fence timeline is aware of the last signaled fence, as it
receives the increment to the current value in fence_timeline_signal().
That allow us to create a fence_default_signaled() using timeline->value
and fence->seqno in the comparison.
Signed-off-by: Gustavo Padovan
From: Gustavo Padovan
The sync framework is ready for mainline inclusion. Here we move it to
drivers/dma-buf and the header files to the appropiated places.
The sync framework contained some abstractions around struct fence and those
were removed in the de-staging process among other changes:
U
From: Gustavo Padovan
Provide simple default functions to help users retrieve the values of the
fence and the timeline.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c | 31 +++
drivers/staging/android/sw_sync.c| 18 --
dr
From: Gustavo Padovan
Add a default .release() op to be used on fence_ops vtable.
It removes the fences from any list it was added, removes a timeline ref
and free the fence.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c| 28
drivers/staging/an
From: Gustavo Padovan
To avoid confusion with struct fence vars that are most of the time called
'fence' as well we should rename all struct sync_fence's to sync_fence.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sync.c | 192 ++-
drivers/st
From: Gustavo Padovan
Add a default .enable_signaling() op to be used on fence_ops vtable.
fence_default_enable_signaling() checks if the was not signaled yet
and adds it to the active_list.
Signed-off-by: Gustavo Padovan
---
drivers/dma-buf/fence.c| 22 ++
drivers/
From: Gustavo Padovan
struct sync_pt was just wrapping around struct fence and creating an
extra abstraction layer. The only two members of struct sync_pt, child_list
and active_list, were moved to struct fence in an earlier commit. After
removing those two members struct sync_pt is nothing more
From: Gustavo Padovan
This remove CONFIG_SW_SYNC_USER and instead compile the sw_sync file into
debugpfs under /sync/sw_sync.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/Kconfig | 9 ---
drivers/staging/android/sw_sync.c| 129 ---
drive
From: Gustavo Padovan
Creates the 'sync' dir on debugfs root dir and move the 'sync' file
to sync/info. This is the preparation to add more debug info and control.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sync_debug.c | 21 +
1 file changed, 17 insertions(
From: Gustavo Padovan
Add the sync timeline from sync framework to fence synchronization system.
This is an attempt to remove some duplication between sync.c and fence.c
The sync_timeline was no more than a wrapper on top of the fence
framework to be used by sw_sync. It simplifies some accesses,
From: Gustavo Padovan
.dup and .compare are not used by the sync framework, so remove them
from sw_sync.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sw_sync.c | 29 +
drivers/staging/android/sync.c| 6 --
drivers/staging/android/sync.h| 1
From: Gustavo Padovan
There isn't any problem on removing the sync_pts from the active_list
when the fence is released. If the user decides to close the fence before
it is signaled we should do it and not warn about anything.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sync.c |
From: Gustavo Padovan
'sync_pt' is actually declared as struct fence so to make the name means
its type we rename it to 'fence'.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sync.c | 20 ++--
drivers/staging/android/sync.h | 2 +-
drivers/staging/andr
From: Gustavo Padovan
As we are de-staging it rename it to a name that is not related to
android.
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sync.c | 40
1 file changed, 20 insertions(+), 20 deletions(-)
diff --git a/drivers/staging/and
From: Gustavo Padovan
To avoid collision with an upcoming rename change the name of
sync_fence_release() to sync_fence_release_file()
Signed-off-by: Gustavo Padovan
---
drivers/staging/android/sync.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/staging/androi
From: Gustavo Padovan
This patch series de-stage the sync framework, and in order to accomplish that
a bunch of cleanups/improvements on the sync and fence were made.
The sync framework contained some abstractions around struct fence and those
were removed in the de-staging process among other c
40 matches
Mail list logo
