[[email protected] wrote:]
From: Randy Dunlap <[email protected]>
Drivers that use lro functions should depend on INET, otherwise they
may not link correctly. Let's not select INET. Select should be used
only for library-like code, not to enable subsystems.
ERROR: "lro_flush_all" [drivers/net/myri10ge/myri10ge.ko] undefined!
ERROR: "lro_receive_frags" [drivers/net/myri10ge/myri10ge.ko] undefined!
Signed-off-by: Randy Dunlap <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
- make the kconfig NAPI option prompt consistent across all net drivers
(other than EXPERIMENTAL; can it now be removed also, or is the new
napi_struct implementation now EXPERIMENTAL ?)
- remove comment about the now-deleted NAPI_HOWTO.txt file
- clean up typos in Tulip NAPI & Interrupt Mitigation
Signed-off-by: Randy Dunlap <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Missing MODULE_LICENSE(), loading this module taints the kernel.
Signed-off-by: Randy Dunlap <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
DM9000 driver returns success even if it is failed to detect the chip.
Below patch fixes it.
Signed-off-by: Mike Rapoport <[email protected]>
drivers/net/dm9000.c | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
Signed-off-by: Jeff Garzik <[email protected]>
Erez Zadok reports that certain configurations fail to build due to
schedule() TASK_[UN]INTERRUPTIBLE not being declared. Add proper
include files to fix.
Cc: Erez Zadok <[email protected]>
Cc: Trond Myklebust <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
sata_sis has the same restrictions as other SFF controllers, and so must
use LIBATA_MAX_PRD to denote that SCSI may only fill ATA_MAX_PRD/2
entries, due to our need to handle IOMMU merging.
Signed-off-by: Jeff Garzik <[email protected]>
SCR read for controllers which uses PCI configuration space for SCR
access got broken while adding @val argument to SCR accessors. Fix
it.
Signed-off-by: Tejun Heo <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Fix libata kernel-doc parameter name.
Warning(linux-2.6.23-git13//drivers/ata/libata-core.c:1415): No description found for parameter 'sgl'
Signed-off-by: Randy Dunlap <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Whilst looking up what profile=sleep did, I noticed that we missed
adding docs for the most recent addition to the profiler.
Signed-off-by: Dave Jones <[email protected]>
Acked-by: Ingo Molnar <[email protected]>
Signed-off-by: Adrian Bunk <[email protected]>
Weird I thought I had written the makefile so this would be handled. Oh
well this should fix it.
Sorry about that.
Signed-off-by: Eric W. Biederman <[email protected]>
Acked-and-tested-by: Randy Dunlap <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Add crypt prefix to dec_pending to avoid confusing it in backtraces with
the dm core function of the same name.
No functional change here.
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
This patch adds calls to dm_path_event for a failed path and a reinstated
path.
Signed-off-by: Mike Anderson <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
This patch adds support for the dm_path_event dm_send_event functions which
create and send udev events.
Signed-off-by: Mike Anderson <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
This patch adds a uevent skeleton to device-mapper.
Signed-off-by: Mike Anderson <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
This patch adds a function to obtain a copy of a mapped device's name and uuid.
Signed-off-by: Mike Anderson <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Store a pointer to the owning mirror_set structure within each mirror
structure for a subsequent patch to use.
Signed-off-by: Jonathan Brassow <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
There are now two phases to a suspend in device-mapper -
presuspend and postsuspend. This patch removes the
single 'suspend' in the logging API and replaces it with
'presuspend' and 'postsuspend' functions to align it
better with core device-mapper.
A subsequent patch will make use of 'presuspend'.
Signed-off-by: Jonathan Brassow <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
This patch adds retries to the hp hardware handler, and utilizes the
MP_RETRY flag of dm-multipath. For now in the hp handler, if we get a
pg_init completed with a check condition we just assume we can retry the
pg_init command. We make this assumption because of incomplete data on
specific check condition code of the HP hardware, and because testing
has shown the HP path initialization command to be idempotent.
The number of times we retry is settable via the "pg_init_retries"
multipath map feature.
Signed-off-by: Dave Wysochanski <[email protected]>
Acked-by: Chandra Seetharaman <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
This patch adds the most basic dm-multipath hardware support for the
HP active/passive arrays.
Signed-off-by: Dave Wysochanski <[email protected]>
Signed-off-by: Mike Christie <[email protected]>
Acked-by: Chandra Seetharaman <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
This patch allows a failed path group initialisation command to be retried.
It adds a generic MP_RETRY flag and a "pg_init_retries" feature to
device-mapper multipath which limits the number of retries.
1. A hw handler sends a path initialization command to the storage and
the command completes with an error code indicating the command
should be retried.
2. The hardware handler calls dm_pg_init_complete() with MP_RETRY
set in err_flags to ask the dm multipath core to retry.
3. If the retry limit has not been exceeded, pg_init() is retried.
Otherwise fail_path() is called.
If you are using the userspace multipath-tools or device-mapper-multipath
package, you can set pg_init_retries in the 'device' section of your
/etc/multipath.conf file. For example:
features "2 pg_init_retries 7"
The number of PG retries attempted is reported in the 'dmsetup status' output.
Signed-off-by: Dave Wysochanski <[email protected]>
Acked-by: Mike Christie <[email protected]>
Acked-by: Chandra Seetharaman <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Replace numbers with names in labels in error paths, to avoid confusion
when new one get added between existing ones.
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Clean up, convert some spaces to tabs.
No functional change here.
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Add post-processing queue (per crypt device) for read operations.
Current implementation uses only one queue for all operations
and this can lead to starvation caused by many requests waiting
for memory allocation. But the needed memory-releasing operation
is queued after these requests (in the same queue).
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Herbert Xu <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Use a separate single-threaded workqueue for each crypt device
instead of one global workqueue.
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Remove BIO_LIST and DEFINE_BIO_LIST macros that gain us nothing
since contents are initialised to NULL.
Cc: Jan Engelhardt <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
In drivers/md/dm-ioctl.c::copy_params() there's a call to vmalloc()
where we currently cast the return value, but that's pretty pointless
given that vmalloc() returns "void *".
Signed-off-by: Jesper Juhl <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Use bio_io_error() in only two places and tidy the code,
preparing for later patches.
There is no functional change in this patch.
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Kcopyd uses a semaphore as mutex. Use the mutex API instead of the (binary)
semaphore,
Signed-off-by: Matthias Kaehlcke <[email protected]>
Cc: Neil Brown <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Replacing n & (n - 1) for power of 2 check by is_power_of_2(n)
Signed-off-by: vignesh babu <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
This patch fixes a bd_mount_sem counter corruption bug in device-mapper.
thaw_bdev() should be called only when freeze_bdev() was called for the
device.
Otherwise, thaw_bdev() will up bd_mount_sem and corrupt the semaphore counter.
struct block_device with the corrupted semaphore may remain in slab cache
and be reused later.
Attached patch will fix it by calling unlock_fs() instead.
unlock_fs() will determine whether it should call thaw_bdev()
by checking the device is frozen or not.
Easy reproducer is:
#!/bin/sh
while [ 1 ]; do
dmsetup --notable create a
dmsetup --nolockfs suspend a
dmsetup remove a
done
It's not easy to see the effect of corrupted semaphore.
So I have tested with putting printk below in bdev_alloc_inode():
if (atomic_read(&ei->bdev.bd_mount_sem.count) != 1)
printk(KERN_DEBUG "Incorrect semaphore count = %d (%p)\n",
atomic_read(&ei->bdev.bd_mount_sem.count),
&ei->bdev);
Without the patch, I saw something like:
Incorrect semaphore count = 17 (f2ab91c0)
With the patch, the message didn't appear.
The bug was introduced in 2.6.16 with this bug fix:
commit d9dde59ba0
Date: Fri Feb 24 13:04:24 2006 -0800
[PATCH] dm: missing bdput/thaw_bdev at removal
Need to unfreeze and release bdev otherwise the bdev inode with
inconsistent state is reused later and cause problem.
and backported to 2.6.15.5.
It occurs only in free_dev(), which is called only when the dm device is
removed. The buggy code is executed only if md->suspended_bdev is
non-NULL and that can happen only when the device was suspended without
noflush.
Signed-off-by: Jun'ichi Nomura <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Cc: [email protected]
Fix missing space in dm-delay target status output
if separate read and write delay are configured.
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Add a missing 'dm_put_device' in an error path in crypt target constructor.
Signed-off-by: Dmitry Monakhov <[email protected]>
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Make size of dm_ioctl struct always 312 bytes on all supported
architectures.
This change retains compatibility with already-compiled code because
it uses an embedded offset to locate the payload that follows the
structure.
On 64-bit architectures there is no change at all; on 32-bit
we are increasing the size of dm-ioctl from 308 to 312 bytes.
Currently with 32-bit userspace / 64-bit kernel on x86_64
some ioctls (including rename, message) are incorrectly rejected
by the comparison against 'param + 1'. This breaks userspace
lvrename and multipath 'fail_if_no_path' changes, for example.
(BTW Device-mapper uses its own versioning and ignores the ioctl
size bits. Only the generic ioctl compat code on mixed arches
checks them, and that will continue to accept both sizes for now,
but we intend to list 308 as deprecated and eventually remove it.)
Signed-off-by: Milan Broz <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
Cc: Guido Guenther <[email protected]>
Cc: Kevin Corry <[email protected]>
Cc: [email protected]
Re-order the initialisation of dm-rdac to avoid registering the hw
handler before the workqueue has been initialised. Closes a race
that would potentially give an oops.
Signed-off-by: Bryn M. Reeves <[email protected]>
Signed-off-by: Alasdair G Kergon <[email protected]>
