Diffs for NetBSD-current 3.99.11

[Ty Sarna]

In article <1137837895.056663.14509.nullmailer at yamt.dyndns.org> you write:
> why you think __NetBSD_Version__ check is necessary?
> otherwise, the diff itself seems logical.

My first thouggh was it was safest to preserve the code for earlier
versions as closely as possible, but because it looks like VOP_UPDATE
was causing a bug (by always skipping the code that followed), maybe it
is best to remove the code entirely. See alternate patch below (I'll
call this patch 2b)

> i'm not sure why these VOP_TRUNCATE and VOP_UPDATE are there
> at the first place, tho.

Yes, this is why I would like to hear from Love, who I think is the
original author.

> btw, doesn't this need something like nfs_bio 1.85 ?

About the call to ubc_alloc? I don't understand much about this. Can you
suggest a patch that I can try?

Thank you for your help!



--- arla-0.41/nnpfs/bsd/nnpfs_vnodeops-netbsd.c.orig	2005-11-14 19:15:43.000000000 -0500
+++ arla-0.41/nnpfs/bsd/nnpfs_vnodeops-netbsd.c	2006-01-21 11:00:19.000000000 -0500
@@ -355,12 +355,8 @@
     }
 
     if (error) {
-	(void) VOP_TRUNCATE(vp, osize, ioflag & IO_SYNC, ap->a_cred,
-	    uio->uio_procp);
 	nnpfs_uio_setoffset(uio, nnpfs_uio_offset(uio) - resid - uio->uio_resid);
 	uio->uio_resid = resid;
-    } else if (resid > uio->uio_resid && (ioflag & IO_SYNC) == IO_SYNC) {
-	error = VOP_UPDATE(vp, NULL, NULL, UPDATE_WAIT);
     }
     if (!error) {
 	struct timespec ts;