Subject: Re: letting userland issue FUA writes
To: Daniel Carosone <>
From: =?ISO-8859-15?Q?Joachim_K=F6nig-Baltes?= <>
List: tech-kern
Date: 03/16/2006 09:23:50
Daniel Carosone wrote:
> Just for general background, can we please clarify/confirm some of the
> semantics involved?
>  - FUA applies only to the specific write carrying that flag, and
>    says nothing about whether other unordered writes that may be
>    sitting in the disk's write-cache have to hit the platters?  ie, it
>    is distinct from implying a cache flush around the write, and
>    distinct from ordered vs unordered writes. it's up to the host and
>    filesystem and/or application to worry about structural
>    dependencies and transactional boundaries using these tools.
>  - O_*SYNC applies to all IO done with the file descriptor, and
>    basically does imply an fsync() or at least fdatasync(), even for
>    data written by other processes with the same file open.

In addition to fsync(fd) and fdatasync(fd) we also have:

      fsync_range(int fd, int how, off_t start, off_t length)

This would still be two calls, write + fsync_range and a bit of 
bookkeeping for the current position, but at least the semantics
can be expressed without race conditions.