Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1407359 > unrolled thread

Re: dma-buf/sync_file: de-stage sync_file

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: dma-buf/sync_file: de-stage sync_file Dave Jones <davej@codemonkey.org.uk> - 2016-05-26 04:50 +0200
    Re: dma-buf/sync_file: de-stage sync_file Gustavo Padovan <gustavo.padovan@collabora.co.uk> - 2016-05-27 23:10 +0200

#1407359 — Re: dma-buf/sync_file: de-stage sync_file

On Sat, May 21, 2016 at 05:31:53AM +0000, Linux Kernel wrote:
 
 >     dma-buf/sync_file: de-stage sync_file
 >     
 >     sync_file is useful to connect one or more fences to the file. The file is
 >     used by userspace to track fences between drivers that share DMA bufs.
 >     
 >     Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
 >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
 >     Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

 ...
 
 > +config SYNC_FILE
 > +	bool "sync_file support for fences"
 > +	default n
 > +	select ANON_INODES
 > +	select DMA_SHARED_BUFFER
 > +	---help---
 > +	  This option enables the fence framework synchronization to export
 > +	  sync_files to userspace that can represent one or more fences.

For such a generic sounding CONFIG_ item, this is one of the more obtuse descriptions
in a kconfig we've had in a while, and the commit message doesn't give any more
clues as to why anyone might want to enable this.

I'm guessing this is some graphics thing given that Daniel reviewed it.
From skimming the other commits, it seems to be some Android thing ?
Are there depends missing perhaps that might make this more obvious ?

Could you elaborate in the help text why someone might want to enable this ?
As is, it's just a bunch of words with no context for anyone who isn't
close to whatever domain this came from.

	Dave

[toc] | [next] | [standalone]

#1408316

2016-05-25 Dave Jones <davej@codemonkey.org.uk>:

> On Sat, May 21, 2016 at 05:31:53AM +0000, Linux Kernel wrote:
>  
>  >     dma-buf/sync_file: de-stage sync_file
>  >     
>  >     sync_file is useful to connect one or more fences to the file. The file is
>  >     used by userspace to track fences between drivers that share DMA bufs.
>  >     
>  >     Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
>  >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
>  >     Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> 
>  ...
>  
>  > +config SYNC_FILE
>  > +	bool "sync_file support for fences"
>  > +	default n
>  > +	select ANON_INODES
>  > +	select DMA_SHARED_BUFFER
>  > +	---help---
>  > +	  This option enables the fence framework synchronization to export
>  > +	  sync_files to userspace that can represent one or more fences.
> 
> For such a generic sounding CONFIG_ item, this is one of the more obtuse descriptions
> in a kconfig we've had in a while, and the commit message doesn't give any more
> clues as to why anyone might want to enable this.
>
> I'm guessing this is some graphics thing given that Daniel reviewed it.
> From skimming the other commits, it seems to be some Android thing ?
> Are there depends missing perhaps that might make this more obvious ?

You are right, only a few people understand this in the way I wrote. I
definitelly reword this. It is android framework to help userspace with
explict synchronization. DRM will require this, but the patches for that
did not land yet.

	Gustavo

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web