login
Header Space

 
 

Re: NFSv4/pNFS possible POSIX I/O API standards

Score:
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
Cc: Trond Myklebust <trond.myklebust@...>, Christoph Hellwig <hch@...>, Brad Boyer <flar@...>, Anton Altaparmakov <aia21@...>, Gary Grider <ggrider@...>, <linux-fsdevel@...>
Date: Sunday, December 3, 2006 - 12:10 pm

On Sat, 2 Dec 2006, Andreas Dilger wrote:

I'm not suggesting any "synchronization" beyond what opendir()/readdir() 
already require for the directory entries themselves.  If I'm not 
mistaken, readdir() is required only to return directory entries as recent 
as the opendir() (i.e., you shouldn't see entries that were unlink()ed 
before you called opendir(), and intervening changes to the directory may 
or may not be reflected in the result, depending on how your 
implementation is buffering things).  I would think the stat() portion of 
readdirplus() would be similarly (in)consistent (i.e., return a value at 
least as recent as the opendir()) to make life easy for the implementation 
and to align with existing readdir() semantics.  My only concern is the 
"at least as recent as the opendir()" part, in contrast to statlite(), 
which has undefined "recentness" of its result for fields not specified in 
the mask.

Ideally, I'd like to see readdirplus() also take a statlite() style mask, 
so that you can choose between either "vaguely recent" and "at least as 
recent as opendir()".

As you mentioned, by the time you look at the result of any call (in the 
absence of locking) it may be out of date.  But simply establishing an 
ordering is useful, especially in a clustered environment where some nodes 
are waiting for other nodes (via barriers or whatever) and then want to 
see the effects of previously completed fs operations.

Anyway, "synchronization" semantics aside (since I appear to be somewhat 
alone on this :)...

I'm wondering if a corresponding opendirplus() (or similar) would also be 
appropriate to inform the kernel/filesystem that readdirplus() will 
follow, and stat information should be gathered/buffered.  Or do most 
implementations wait for the first readdir() before doing any actual work 
anyway?

sage
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
NFSv4/pNFS possible POSIX I/O API standards, Gary Grider, (Tue Nov 28, 12:34 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Matthew Wilcox, (Tue Nov 28, 11:08 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Tue Nov 28, 1:54 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Wed Nov 29, 5:04 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Matthew Wilcox, (Wed Nov 29, 8:23 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Wed Nov 29, 8:39 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Fri Dec 1, 6:29 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Latchesar Ionkov, (Fri Dec 1, 10:35 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Mon Dec 4, 8:37 pm)
Re: Re: NFSv4/pNFS possible POSIX I/O API standards, Latchesar Ionkov, (Tue Dec 5, 12:47 pm)
Re: Re: NFSv4/pNFS possible POSIX I/O API standards, David Chinner, (Wed Dec 6, 5:48 am)
openg and path_to_handle, Rob Ross, (Wed Dec 6, 11:53 am)
Re: openg and path_to_handle, Latchesar Ionkov, (Wed Dec 6, 7:19 pm)
Re: openg and path_to_handle, Rob Ross, (Thu Dec 14, 5:00 pm)
Re: openg and path_to_handle, Matthew Wilcox, (Thu Dec 14, 5:20 pm)
Re: openg and path_to_handle, Rob Ross, (Thu Dec 14, 7:02 pm)
Re: openg and path_to_handle, David Chinner, (Wed Dec 6, 4:40 pm)
Re: openg and path_to_handle, Rob Ross, (Wed Dec 6, 4:50 pm)
Re: openg and path_to_handle, David Chinner, (Wed Dec 6, 5:01 pm)
Re: openg and path_to_handle, Matthew Wilcox, (Wed Dec 6, 4:50 pm)
Re: openg and path_to_handle, Andreas Dilger, (Wed Dec 6, 6:09 pm)
Re: openg and path_to_handle, Christoph Hellwig, (Wed Dec 6, 7:39 pm)
Re: openg and path_to_handle, Rob Ross, (Thu Dec 14, 6:52 pm)
Re: openg and path_to_handle, Matthew Wilcox, (Wed Dec 6, 6:17 pm)
Re: openg and path_to_handle, Andreas Dilger, (Wed Dec 6, 6:41 pm)
Re: openg and path_to_handle, David Chinner, (Wed Dec 6, 5:09 pm)
Re: openg and path_to_handle, Matthew Wilcox, (Wed Dec 6, 12:04 pm)
Re: openg and path_to_handle, Rob Ross, (Wed Dec 6, 12:20 pm)
Re: openg and path_to_handle, David Chinner, (Wed Dec 6, 4:57 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Tue Dec 5, 5:44 pm)
Re: Re: NFSv4/pNFS possible POSIX I/O API standards, Latchesar Ionkov, (Wed Dec 6, 7:25 pm)
openg, Christoph Hellwig, (Wed Dec 6, 7:01 am)
Re: openg, Rob Ross, (Wed Dec 6, 11:42 am)
Re: openg, Christoph Hellwig, (Wed Dec 6, 7:32 pm)
Re: openg, Rob Ross, (Thu Dec 14, 7:36 pm)
Re: openg, Trond Myklebust, (Wed Dec 6, 11:41 am)
Re: Re: NFSv4/pNFS possible POSIX I/O API standards, Matthew Wilcox, (Tue Dec 5, 1:01 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Peter Staubach, (Tue Dec 5, 5:50 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Tue Dec 5, 6:02 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Matthew Wilcox, (Wed Nov 29, 8:35 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Gary Grider, (Wed Nov 29, 12:26 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Wed Nov 29, 1:18 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Andreas Dilger, (Wed Nov 29, 5:48 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ric Wheeler, (Fri Dec 1, 11:52 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Steven Whitehouse, (Wed Nov 29, 6:25 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Thu Nov 30, 8:29 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Anton Altaparmakov, (Wed Nov 29, 6:18 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Brad Boyer, (Wed Nov 29, 4:26 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Thu Nov 30, 5:25 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Sage Weil, (Thu Nov 30, 1:49 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Fri Dec 1, 1:26 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Sage Weil, (Fri Dec 1, 3:08 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Fri Dec 1, 10:41 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Andreas Dilger, (Sat Dec 2, 9:52 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Sage Weil, (Sun Dec 3, 12:10 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Andreas Dilger, (Mon Dec 4, 3:32 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Mon Dec 4, 11:15 am)
Re: readdirplus() as possible POSIX I/O API, Andreas Dilger, (Tue Dec 5, 6:26 am)
Re: Re: readdirplus() as possible POSIX I/O API, Latchesar Ionkov, (Tue Dec 5, 1:06 pm)
Re: readdirplus() as possible POSIX I/O API, Rob Ross, (Tue Dec 5, 6:48 pm)
Re: readdirplus() as possible POSIX I/O API, Trond Myklebust, (Tue Dec 5, 11:23 am)
Re: readdirplus() as possible POSIX I/O API, Andreas Dilger, (Wed Dec 6, 6:28 am)
Re: readdirplus() as possible POSIX I/O API, Trond Myklebust, (Wed Dec 6, 11:10 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Mon Dec 4, 8:59 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Peter Staubach, (Tue Dec 5, 10:37 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Tue Dec 5, 1:56 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Tue Dec 5, 6:07 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Tue Dec 5, 5:50 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Tue Dec 5, 6:05 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Nikita Danilov, (Thu Dec 7, 7:39 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ulrich Drepper, (Tue Dec 5, 7:55 pm)
statlite(), Rob Ross, (Thu Dec 14, 7:58 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Andreas Dilger, (Wed Dec 6, 6:06 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ulrich Drepper, (Wed Dec 6, 1:19 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Wed Dec 6, 1:27 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ulrich Drepper, (Wed Dec 6, 1:42 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Andreas Dilger, (Thu Dec 7, 1:57 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ulrich Drepper, (Fri Dec 15, 6:37 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Andreas Dilger, (Sat Dec 16, 2:13 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ulrich Drepper, (Sat Dec 16, 3:08 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ulrich Drepper, (Wed Dec 6, 2:13 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ragnar , (Sun Dec 17, 10:41 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ulrich Drepper, (Sun Dec 17, 3:07 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Matthew Wilcox, (Sun Dec 17, 3:38 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ulrich Drepper, (Sun Dec 17, 5:51 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Ragnar , (Sun Dec 17, 10:57 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Gary Grider, (Sun Dec 17, 11:54 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Sage Weil, (Tue Dec 5, 7:18 pm)
Re: Re: NFSv4/pNFS possible POSIX I/O API standards, Latchesar Ionkov, (Tue Dec 5, 12:55 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Tue Dec 5, 6:12 pm)
Re: Re: NFSv4/pNFS possible POSIX I/O API standards, Latchesar Ionkov, (Wed Dec 6, 7:12 pm)
Re: Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Wed Dec 6, 7:33 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Tue Dec 5, 10:55 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Wed Dec 6, 11:14 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Tue Dec 5, 6:11 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Tue Dec 5, 7:24 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Wed Dec 6, 12:42 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Matthew Wilcox, (Tue Dec 5, 10:20 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Wed Dec 6, 11:04 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Matthew Wilcox, (Wed Dec 6, 11:44 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Rob Ross, (Wed Dec 6, 12:15 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Gary Grider, (Tue Dec 5, 12:44 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Tue Dec 5, 6:05 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Sage Weil, (Fri Dec 1, 12:47 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Fri Dec 1, 2:07 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Andreas Dilger, (Sat Dec 2, 9:57 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Sage Weil, (Fri Dec 1, 2:42 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Peter Staubach, (Mon Dec 4, 2:02 pm)
Re: readdirplus() as possible POSIX I/O API, Sage Weil, (Tue Dec 5, 7:20 pm)
Re: readdirplus() as possible POSIX I/O API, Peter Staubach, (Wed Dec 6, 11:48 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Trond Myklebust, (Fri Dec 1, 3:13 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Sage Weil, (Fri Dec 1, 4:32 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Wed Nov 29, 5:14 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Andreas Dilger, (Tue Nov 28, 6:54 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Wendy Cheng, (Tue Nov 28, 7:28 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Christoph Hellwig, (Wed Nov 29, 5:12 am)
Re: NFSv4/pNFS possible POSIX I/O API standards, Russell Cattelan, (Tue Nov 28, 4:17 pm)
Re: NFSv4/pNFS possible POSIX I/O API standards, Anton Altaparmakov, (Tue Nov 28, 7:28 am)
speck-geostationary