Hi, On Wed, Jan 16, 2002 at 10:36:25AM +0100, eric chacron wrote: > Even in synchronous mode (O_SYNC ...), it seems the nfs client sends as many > write request to the server as the user data is splitted into cache pages on > client side (ref. nfs_writepage_sync() , nfs_writepage() in fs/nfs/write.c > ,nfs_updatepage() in fs/nfs/write.c , nfs_commit_write() in fs/nfs/file.c , > generic_file_write() in mm/filemap.c , nfs_file_write in fs/nfs/file.c) That's correct. O_SYNC guarantees that all of the data gets sent to the server before the call returns, but does not guarantee order of data within the write. There *are* no guarantees for that in Unix semantics, anywhere. > I propose the following conditions to provide atomicity of write through nsf + > ext3, with the current implementation: It could change at any time. It is really up to the application to make its own provision for atomicity. Cheers, Stephen
