TAKE - pagebuf massaging

[Nathan Scott <nathans@snort.melbourne.sgi.com>]

No-op change, effectively, but should help Steve and I to keep our
workareas in sync a bit better.


Date:  Sun Mar 24 19:32:38 PST 2002
Workarea:  snort.melbourne.sgi.com:/home/nathans/pagebuf

The following file(s) were checked into:
  bonnie.engr.sgi.com:/isms/slinx/2.4.x-xfs


Modid:  2.4.x-xfs:slinx:114782a
linux/fs/xfs/xfsidbg.c - 1.175
linux/fs/xfs/xfs_buf.h - 1.84
linux/fs/xfs/pagebuf/page_buf_trace.h - 1.2
	- Get external trace points working again, PB_TRACE macro busted
	  for this case.

linux/fs/xfs/xfs_vnodeops.c - 1.519
linux/fs/xfs/linux/xfs_lrw.h - 1.19
linux/fs/xfs/linux/xfs_lrw.c - 1.127
linux/fs/xfs/linux/xfs_iops.c - 1.129
linux/fs/xfs/linux/xfs_vnode.h - 1.28
linux/fs/xfs/pagebuf/page_buf_io.c - 1.18
linux/fs/xfs/pagebuf/page_buf_locking.c - 1.6
linux/fs/xfs/pagebuf/page_buf.c - 1.14
linux/fs/xfs/pagebuf/page_buf.h - 1.9
	- Prepare pagebuf APIs for multiple blocksize support.  In particular,
	  the IO path needs to get access to blocksize information which
	  currently is only held in the xfs_mount structure -- this change
	  reorganises a few things so that the necessary information is
	  available to the code that needs it.  We also no longer need the
	  pbm_dev field (in pb_bmap_t) as this is now available from the
	  pb_target_t always.

linux/fs/xfs/Makefile - 1.136
linux/fs/xfs/Makefile.in - 1.14
linux/fs/xfs/pagebuf/Makefile.in - 1.4
linux/fs/xfs/pagebuf/Makefile - 1.4
	- Propogate the PAGEBUF_TRACE flag if need be.