Hi,
I am trying to trace a file read to understand how the
inode cache comes in to the picture. I know how it is
suppose to operate but I do not know how the inode cache
fits in the implementation.
Kerner version: 2.0.33
FS:ext2
I can trace the call to fs/buffer.c/generic_readpage() routine
where VFS(?) calls bmap->ext2_bmap to read the inode. In ext2_bmap
bread is called to read the inode from the disk. bread calls get_blk
to gets the buffer.
I have the following questions
1. Looks like get_blk checks the buffer cache and not inode cache to see
whether the inode is in memory. Is this correct. If so is the inode and
the buffer cache unified ? (and code in fs/inode.c to manage inode
cache never used ?)
2. For large files with sequential access would it be a performance
improvement to pre fetch cache the inodes in memory for future referances
at file open time. ? For storage I/O intensive servers (Like file servers)
this seems to be a good idea. Is there any related work on this being done
on this ?
Thank you
Ravi Wijayaratne
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Mar 31 2000 - 21:00:29 EST