| Age | Commit message (Collapse) | Author |
|---|
|
ok blambert@
|
|
pointed out by bluhm@, "obvious" thib@
|
|
(that I applied to a clean tree to commit)
|
|
timeouts. Rrids us of the ugly cur_sec variable, and some shadows.
Also helps with granularity.
Diff from blambert@ who asked me to commit this since he's away for
some days and we wanted to put this in with the timespec changes in
the nfsnode.
|
|
mtime, gives us better granularity, helps with cache consistency.
Idea lifted from NetBSD.
OK blambert@
|
|
VA_EXCLUSIVE. Handle this in NFS, also in NFS use arc4random()
for the create verifier instead of an uninitialized long and
the address of the first interface (which is likely to be lo0).
Lifted from NetBSD with small tweaks;
"looks good", miod@
OK blambert@
|
|
prompted by David <dunnoseriously _att_ gmail.com>
|
|
from David <dunnoseriously _att_ gmail.com>
ok blambert@
|
|
the udp receive space got completely filled up. Even if the next
packet from the server was a small reply, it got dropped by
udp_input(). After a second the client resent the NFS request.
Doubling rcvreserve reduces the chance of retransmits by having
enough recv space for multiple NFS replies even if there comes a
big one.
found with pedro@, ok blambert@, thib@, pedro@
|
|
of 5 cases. In the functions nfs_reconnect() and nfs_reply() put
the TAILQ_FOREACH over nfs_reqq also inside splsoftnet() protection.
found with pedro@, ok blambert@, pedro@, thib@
|
|
requires well-formed mbuf chains to work. Existing function name is now
a wrapper to this interface. Much, much more work to follow which will
build on this.
Truly change we can believe in.
ok thib@, who would have preferred this in two commits, but I'm impatient :)
testing on multiple arches (i386, arm, amd64, and sparc64 at least) by many,
thanks to all who did so
|
|
cache instead of setting n_attrstamp to 0 directly.
Lift the macro name from NetBSD.
prompted by and OK blambert@
|
|
ok thib@
|
|
to kthread_create()
ok thib@
|
|
in diagnostic kernels. The typo comes from 4.4BSD-Lite2 via NetBSD.
Found with pedro@, ok pedro@, blambert@, thib@
|
|
|
|
as for open/read/write etc, that is assume that the file
as already been opened proper, this fixes an ancient
ftruncate bug, that is if ftruncate was called on a file
with no write modes, we'd return EACCES, this breaks posixy
behavior. however, for e.g. truncate() we'll return EACCESS
as we should.
Diff from PR5880 by mickey.
I had forgotten about this PR, and the fix but niklas@ ran
into similar issues which prompted another review of this.
OK kettenis@
|
|
and to print out struct nfsreq.
"get it in so people can pound on it" blambert@
OK and information_s_ on the manpage from miod@
|
|
argument since all of the info we need is there, makes it and
the use it a tad bit nicer.
ok blambert@
|
|
|
|
ok millert@ tedu@
|
|
since they have no business being export to userland.
OK blambert@
|
|
its used only once and nuke its definition.
Stolen from blambert's bag'o'diffs.
OK blambert@.
|
|
file shares the same sane style.
|
|
another comment.
|
|
hurt readability and it was just plain annoying seeing them defined
in every other .c file.
OK blambert@
|
|
|
|
|
|
blambert@--
Real solution is to fix NFS mbuf handling.
|
|
ok thib@
|
|
the server cache on shutdown.
ok thib@
|
|
#define's in nfs_var.h
ok thib@
|
|
|
|
shrinks, same thing as UFS does.
ok/pointers pedro@
ok blambert@
|
|
ok blambert@
|
|
ok tedu@, blambert@, art@
|
|
replaced by a function long ago..
ok blambert@
|
|
changes, okay beck@, tedu@ and thib@.
|
|
|
|
From NetBSD;
OK art@, blambert@
|
|
checks if the mount is actually interrutable, and if not sets it 0.
remove this argument from nfs_vinvalbuf and just do the checking inside
the function.
o give nfs_vinvalbuf() a makeover so it looks nice. (spacing, casts, &c);
o Actually pass PCATCH too tsleep() if the mount it interrutable.
ok art@, blambert@
|
|
and the dead code that was supposed todo that has been removed.
o rename the NFS_SMALLFH constant to NFS_MAXFHSIZE, since it better
reflects what it's for.
ok blambert@
|
|
nfs_iodwant array became unused. Garbage collect and free up
a few bytes.
ok thib@
|
|
as it was never used because there were 64-bit buffers already allocated
for filehandles in nfsnode structs.
ok thib@
|
|
wrong. But this was harmless anyway since sys_statfs() sets this field
corrently after calling VFS_STATFS();
ok pedro@, blambert@
|
|
ok thib@
|
|
|
|
them going negative - this consists of identifying a number of cases of
IO not going through the buffer cache and marking those buffers with
B_RAW - as well as fixing nfs_bio to show pending writes and reads through
the buffer cache via NFS
still has a problem with mishandling the counters I believe in the
async/sync fallback case where counters stay positive which will be
addressed seperately.
ok tedu@ deraadt@
|
|
not just a return value, nfs_lookitup treats it specially when *npp
is not NULL.
- nfs_lookitup - When we match the file handle to the directory
filehandle (lookup on '.'), initialize np so that we don't return
stack garbage to the caller. The function shouldn't ever be
called with '.' at this moment, but that's no reason to not do the
right thing in case future callers use it that way.
blambert@ thib@ ok
|
|
|
