Brian Behlendorf writes:
The root cause of the NFS hang we were observing appears to be a rare
deadlock between the kernel provided usermodehelper API and the linux NFS
client. The deadlock can arise because both of these services use the
generic linux work queues. The usermodehelper API run the specified user
application in the context of the work queue. And NFS submits both cleanup
and reconnect work to the generic work queue for handling. Normally this
is fine but a deadlock can result in the following situation.
- NFS client is in a disconnected state
- [events/0] runs a usermodehelper app with an NFS dependent operation,
this triggers an NFS reconnect.
- NFS reconnect happens to be submitted to [events/0] work queue.
- Deadlock, the [events/0] work queue will never process the
reconnect because it is blocked on the previous NFS dependent
operation which will not complete.`
The solution is simply to run reconnect requests on rpciod.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Instead of taking the mutex every time we just need to increment/decrement
rpciod_users, we can optmise by using atomic_inc_not_zero and
atomic_dec_and_test.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The kref now does most of what cl_count + cl_user used to do. The only
remaining role for cl_count is to tell us if we are in a 'shutdown'
phase. We can provide that information using a single bit field instead
of a full atomic counter.
Also rename rpc_destroy_client() to rpc_close_client(), which reflects
better what its role is these days.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Replace it with explicit calls to rpc_shutdown_client() or
rpc_destroy_client() (for the case of asynchronous calls).
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Its use is at best racy, and there is only one user (lockd), which has
additional locking that makes the whole thing redundant.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Also ensure that nfs_inode ncommit and npages are large enough to represent
all possible values for the number of pages.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Also get rid of a redundant call to nfs_setattr_update_inode(). The call to
nfs3_proc_setattr() already takes care of that.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The Linux NFS4 client simply skips over the bitmask in an O_EXCL open
call and so it doesn't bother to reset any fields that may be holding
the verifier. This patch has us save the first two words of the bitmask
(which is all the current client has #defines for). The client then
later checks this bitmask and turns on the appropriate flags in the
sattr->ia_verify field for the following SETATTR call.
This patch only currently checks to see if the server used the atime
and mtime slots for the verifier (which is what the Linux server uses
for this). I'm not sure of what other fields the server could
reasonably use, but adding checks for others should be trivial.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
nfs_symlink() allocates a GFP_KERNEL page for the pagecache. Most
pagecache pages are allocated using GFP_HIGHUSER, and there's no reason
not to do that in nfs_symlink() as well.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
We don't need to revalidate the fsid on the root directory. It suffices to
revalidate it on the current directory.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
nfs4_do_close() does not currently have any way to ensure that the user
won't attempt to unmount the partition while the asynchronous RPC call
is completing. This again may cause Oopses in nfs_update_inode().
Add a vfsmount argument to nfs4_close_state to ensure that the partition
remains mounted while we're closing the file.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
A number of race conditions may currently ensue if the user presses ^C
and then unmounts the partition while an asynchronous open() is in
progress.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Since PG_uptodate may now end up getting set during the call to
nfs_wb_page(), we can avoid putting a read request on the wire in those
situations.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The write may fail, so we should not mark the page as uptodate until we are
certain that the data has been accepted and written to disk by the server.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
There is no need to fail the entire O_DIRECT read/write just because
get_user_pages() returned fewer pages than we requested.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Add Port Multiplier related ATA constants and macros. Some of these
will be used by ata_link implementation.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Separate out ata_eh_handle_dev_fail() from ata_eh_recover(). This is
in preparation of ata_link and PMP support.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Add Hitachi HDS7250SASUN500G 0621KTAWSD to list of devices with broken NCQ.
Signed-off-by: Prarit Bhargava <prarit@redhat.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Workaround for errata A308: turn down the UDMA mode and retry
the DMA command when the data lost condition is detected.
Signed-off-by: Kou Ishizaki <kou.ishizaki@toshiba.co.jp>
Signed-off-by: Akira Iguchi <akira2.iguchi@toshiba.co.jp>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Please warmly welcome the first member from FUJITSU to the prestigious
NCQ spurious completion club.
This is reported by Serge Van Thillo in bugzilla bug 8730.
http://bugzilla.kernel.org/show_bug.cgi?id=8730
Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Serge van Thillo <nulleke@hotmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
The HPT343/345 (aka 363) is a bit of a warped device. For many setups you
need to access the other registers via BAR4 offsets. PIO is now rock
solid, DMA isn't. Unfortunately the drivers/ide hpt34x driver is
completely broken so doesn't help further debug.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Horkage handling had the following problems.
* dev->horkage was positioned after ATA_DEVICE_CLEAR_OFFSET, so it was
cleared before the device is configured. This broke
HORKAGE_DIAGNOSTIC.
* Some used dev->horkage while others called ata_device_blacklisted()
directly. This was at best confusing.
This patch moves dev->horkage right after dev->flags and set the field
according to the blacklist during device configuration. All users
test against dev->horkage. ata_device_blacklisted() now has only one
user, make it static. While at it, rename it to ata_dev_blacklisted()
for consistency.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
The Zip 250 which chokes on MWDMA SET_XFERMODE sometimes have "Floppy"
appeneded to its model number. Quirk it too.
http://bugzilla.kernel.org/show_bug.cgi?id=8563
Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Hans de Bruin <bruinjm@xs4all.nl>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
With PCI resource fix up for legacy hosts. We can use the same code
path to allocate IO resources and initialize host for both legacy and
native SFF hosts. Only IRQ requesting needs to be different.
Rename ata_pci_*_native_host() to ata_pci_*_sff_host(), kill all
legacy specific functions and use the renamed functions instead. This
simplifies code a lot.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Implement suspend and resume routines for mpc52xx ata driver.
Tested on Lite5200b with deep-sleep and low-power (not yet in-tree)
modes.
Signed-off-by: Domen Puncer <domen.puncer@telargo.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
This patch enables hotplugging of SATA devices in the
sata_promise driver. It's been tested successfully on
both first- and second-generation Promise SATA chips:
SATA150 TX2plus, SATAII150 TX2plus, SATAII150 TX4,
SATA300 TX2plus, and SATA300 TX4.
The only quirk I've seen is that hotplugging (insertion)
on the first-generation SATA150 TX2plus requires a lengthier
EH sequence than on the second-generation chips.
On the second-generation chips a simple soft reset seems
to suffice, but on the first-generation chip there's a
"port is slow to respond" after the initial soft reset,
after which libata issues a hard reset, and then the
device is recognised.
The hotplug checks are high up in the interrupt handling
path, not deep down in error_intr as in ahci/sata_sil24.
That's because the chip doesn't signal hotplug status changes
in the per-port status register: instead a global register
contains hotplug control and status flags for all ports.
I considered following the ahci/sata_sil24 structure, but
that would have required non-trivial changes to the interrupt
handling path, so I chose to keep the hotplug changes simple
and unobtrusive.
Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>
--
This patch depends on the "sata_promise: cleanups" patch.
Changes since the previous version (posted June 19):
- Correct pdc_interrupt() to increment 'handled' also in
the hotplug case. This prevents IRQ_NONE from being
returned when an interrupt only has hotplug events to
handle, which could confuse the kernel's IRQ machinery.
- Added testing on the SATAII150 TX4.
drivers/ata/sata_promise.c | 41 ++++++++++++++++++++++++++++++++++++-----
1 files changed, 36 insertions(+), 5 deletions(-)
Signed-off-by: Jeff Garzik <jeff@garzik.org>
