linux-stable.git/fs/nfs/objlayout, branch linux-3.0.y

pnfs-obj: Must return layout on IO error

2012-01-26T01:25:03+00:00

commit fe0fe83585f88346557868a803a479dfaaa0688a upstream.

As mandated by the standard. In case of an IO error, a pNFS
objects layout driver must return it's layout. This is because
all device errors are reported to the server as part of the
layout return buffer.

This is implemented the same way PNFS_LAYOUTRET_ON_SETATTR
is done, through a bit flag on the pnfs_layoutdriver_type->flags
member. The flag is set by the layout driver that wants a
layout_return preformed at pnfs_ld_{write,read}_done in case
of an error.
(Though I have not defined a wrapper like pnfs_ld_layoutret_on_setattr
 because this code is never called outside of pnfs.c and pnfs IO
 paths)

Without this patch 3.[0-2] Kernels leak memory and have an annoying
WARN_ON after every IO error utilizing the pnfs-obj driver.

Signed-off-by: Boaz Harrosh 
Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

pnfs-obj: pNFS errors are communicated on iodata->pnfs_error

2012-01-26T01:25:02+00:00

commit 5c0b4129c07b902b27d3f3ebc087757f534a3abd upstream.

Some time along the way pNFS IO errors were switched to
communicate with a special iodata->pnfs_error member instead
of the regular RPC members. But objlayout was not switched
over.

Fix that!
Without this fix any IO error is hanged, because IO is not
switched to MDS and pages are never cleared or read.

Signed-off-by: Boaz Harrosh 
Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

pnfs-obj: Bug when we are running out of bio

2011-08-29T20:29:08+00:00

commit 20618b21da0796115e81906d24ff1601552701b7 upstream.

When we have a situation that the number of pages we want
to encode is bigger then the size of the bio. (Which can
currently happen only when all IO is going to a single device
.e.g group_width==1) then the IO is submitted short and we
report back only the amount of bytes we actually wrote/read
and all is fine. BUT ...

There was a bug that the current length counter was advanced
before the fail to add the extra page, and we come to a situation
that the CDB length was one-page longer then the actual bio size,
which is of course rejected by the osd-target.

While here also fix the bio size calculation, in the case
that we received more then one group of devices.

Signed-off-by: Boaz Harrosh 
Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

pnfs-obj: Fix the comp_index != 0 case

2011-08-29T20:29:07+00:00

commit 9af7db3228acc286c50e3a0f054ec982efdbc6c6 upstream.

There were bugs in the case of partial layout where olo_comp_index
is not zero. This used to work and was tested but one of the later
cleanup SQUASHMEs broke it and was not tested since.

Also add a dprint that specify those received layout parameters.
Everything else was already printed.

Signed-off-by: Boaz Harrosh 
Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

NFSv4.1: Fix some issues with pnfs_generic_pg_test

2011-06-21T15:54:05+00:00

1. If the intention is to coalesce requests 'prev' and 'req' then we
   have to ensure at least that we have a layout starting at
   req_offset(prev).

2. If we're only requesting a minimal layout of length desc->pg_count,
   we need to test the length actually returned by the server before
   we allow the coalescing to occur.

3. We need to deal correctly with (pgio->lseg == NULL)

4. Fixup the test guarding the pnfs_update_layout.

Signed-off-by: Trond Myklebust

pnfs-obj: No longer needed to take an extra ref at add_device

2011-06-19T18:49:51+00:00

Andy's last device_cache patches, already take an extra
reference on the newly inserted device_id. So we can remove it
from obj-io.

Without this patch the device_ids are leaked.

Andy's patches are not in Linus tree yet. So I'm not sure if they are
scheduled for this Kernel or the next. This patch should be added as
part of these.

CC: Andy Adamson 
Signed-off-by: Boaz Harrosh 
Signed-off-by: Trond Myklebust

NFS: (d)printks should use %zd for ssize_t arguments

2011-06-15T15:24:31+00:00

(d)printks should use %zd for ssize_t arguments not %ld, otherwise they might
get a warning.  I see the following with MN10300.

fs/nfs/objlayout/objlayout.c: In function 'objlayout_read_done':
fs/nfs/objlayout/objlayout.c:294: warning: format '%ld' expects type 'long int', but argument 3 has type 'ssize_t'

Signed-off-by: David Howells 
cc: Trond Myklebust 
cc: linux-nfs@vger.kernel.org
Signed-off-by: Trond Myklebust

pnfs-obj: pg_test check for max_io_size

2011-05-29T18:03:08+00:00

Implement pg_test vector to test for max IO sizes. We calculate
a max_io_size member only once, and cache it in lseg so to not
do so on every page insert.

Signed-off-by: Boaz Harrosh 
[simplify logic]
Signed-off-by: Benny Halevy

NFSv4.1: use pnfs_generic_pg_test directly by layout driver

2011-05-29T17:56:55+00:00

Signed-off-by: Benny Halevy

NFSv4.1: change pg_test return type to bool

2011-05-29T17:56:54+00:00

Signed-off-by: Benny Halevy