[FileAPI] Remove Flush() on write completion on SandboxFileSystems

[FileAPI] Remove Flush() on write completion on SandboxFileSystems

The Flush() call is relatively less important for SandboxFileSystems which
 is stored in a internal storage, than external storage (e.g. USB storage).
And, it hurts performance of truncate()-write() use case on encrypted disk.

This CL removes Flush() on write completion on SandboxFileSystems.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=268914

[tzik, 2014-05-02 04:48:31]

PTL

[kinuko, 2014-05-02 05:26:16]

https://codereview.chromium.org/264993002/diff/40001/webkit/browser/fileapi/f...
File webkit/browser/fileapi/file_system_context.cc (right):

https://codereview.chromium.org/264993002/diff/40001/webkit/browser/fileapi/f...
webkit/browser/fileapi/file_system_context.cc:484: return
!IsSandboxFileSystem(type);
This'd look hacky or cryptic, can you comment some context here?

[tzik, 2014-05-02 06:55:10]

https://codereview.chromium.org/264993002/diff/40001/webkit/browser/fileapi/f...
File webkit/browser/fileapi/file_system_context.cc (right):

https://codereview.chromium.org/264993002/diff/40001/webkit/browser/fileapi/f...
webkit/browser/fileapi/file_system_context.cc:484: return
!IsSandboxFileSystem(type);
On 2014/05/02 05:26:16, kinuko wrote:
> This'd look hacky or cryptic, can you comment some context here?

Added some comment.

[kinuko, 2014-05-02 07:02:48]

lgtm

https://codereview.chromium.org/264993002/diff/60001/webkit/browser/fileapi/f...
File webkit/browser/fileapi/file_system_context.cc (right):

https://codereview.chromium.org/264993002/diff/60001/webkit/browser/fileapi/f...
webkit/browser/fileapi/file_system_context.cc:487: // disk, we don't need to
keep calling fsync() for it.
Maybe... it'd be also better note that some non-sandbox filesystems that could
work on removable devices need flush to make sure its data is written out to the
disk?

[tzik, 2014-05-02 07:37:24]

https://codereview.chromium.org/264993002/diff/60001/webkit/browser/fileapi/f...
File webkit/browser/fileapi/file_system_context.cc (right):

https://codereview.chromium.org/264993002/diff/60001/webkit/browser/fileapi/f...
webkit/browser/fileapi/file_system_context.cc:487: // disk, we don't need to
keep calling fsync() for it.
On 2014/05/02 07:02:48, kinuko wrote:
> Maybe... it'd be also better note that some non-sandbox filesystems that could
> work on removable devices need flush to make sure its data is written out to
the
> disk?

Wrote some more about non-sandboxed filesystem.

[tzik, 2014-05-07 02:33:34]

Michael, please take a look to ctx/b/index_db change as a OWNER.

[michaeln, 2014-05-07 02:45:34]

> Michael, please take a look to ctx/b/index_db change as a OWNER.

lgtm

(i see no behavior change in idb, if thats not right please ask ericu or
cmumford to take a look)

[tzik, 2014-05-07 05:43:10]

On 2014/05/07 02:45:34, michaeln wrote:
> > Michael, please take a look to ctx/b/index_db change as a OWNER.
> 
> lgtm
> 
> (i see no behavior change in idb, if thats not right please ask ericu or
> cmumford to take a look)

Thanks.
Right, it's not affect to IDB behavior.

[tzik, 2014-05-07 05:43:16]

The CQ bit was checked by tzik@chromium.org

[commit-bot: I haz the power, 2014-05-07 05:47:40]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/tzik@chromium.org/264993002/80001

[commit-bot: I haz the power, 2014-05-07 17:56:27]

FYI, CQ is re-trying this CL (attempt #1).
Please consider checking whether the failures are real,
and report flakes to chrome-troopers@google.com.
The failing builders are:
  win_chromium_compile_dbg on tryserver.chromium

[commit-bot: I haz the power, 2014-05-07 20:13:22]

Change committed as 268914