This is worth a few percent on write performance, and about 15% on
O_SYNC writes. Basically get more information about the type of I/O
we are doing down to the block layer code.
Date: Fri May 17 10:20:52 PDT 2002
Workarea: jen.americas.sgi.com:/src/lord/xfs-baseline
The following file(s) were checked into:
bonnie.engr.sgi.com:/isms/slinx/2.4.x-xfs
Modid: 2.4.x-xfs:slinx:119446a
linux/fs/xfs/linux/xfs_lrw.c - 1.136
- Pass info on the expected new file size down to the allocator
linux/fs/xfs/linux/xfs_iops.c - 1.140
- Add getblock case for passing O_SYNC information down to the
allocator,
don't just use 64K as the default preallocation size.
linux/fs/xfs/pagebuf/page_buf_io.c - 1.35
- pass file pointer into prepare and commit write so they can get
O_SYNC information.
|