This patch (as1083) combines hub_quiesce() and hub_stop() into a
single routine.  There's no point keeping them separate since they are
usually called together.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as1071) combines hub_activate() and hub_restart() into a
single routine.  There's no point keeping them separate, since they
are always called together.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as1082) makes a small optimization to the way the hub
driver carries out port debouncing immediately after a hub is
activated (i.e., initialized, reset, or resumed).  If any port-change
statuses are observed, the code will delay for a minimal debounce
period -- thereby making a good start at debouncing all the ports at
once.

If this wasn't sufficient then khubd will debounce any port that still
requires attention.  But in most cases it should suffice; it's rare
for a device to need more than a minimal debounce delay.  (In the
cases of hub initialization or reset even that is most likely not
needed, since any devices plugged in at such times have probably been
attached for a while.)
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as1073) adds to khubd a way to recover from power-session
interruption caused by transient connect-change or enable-change
events.  After the debouncing period, khubd attempts to do a
USB-Persist-style reset or reset-resume.  If it works, the connection
will remain unscathed.

The upshot is that we will be more immune to noise caused by EMI.  The
grace period is on the order of 100 ms, so this won't permit recovery
from the "accidentally knocked the USB cable out of its socket" type
of event, but it's a start.

As an added bonus, if a device was suspended when the system goes to
sleep then we no longer need to check for power-session interruptions
when the system wakes up.  Khubd will naturally see the status change
while processing the device's parent hub and will do the right thing.

The remote_wakeup() routine is changed; now it expects the caller to
acquire the device lock rather than acquiring the lock itself.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as1081) straightens out the logic of the hub_restart()
routine.  Each port of the hub is scanned and the driver makes sure
that ports which are supposed to be disabled really _are_ disabled.
Any ports with a significant change in status are flagged in
hub->change_bits, so that khubd can focus on them without the need to
scan all the ports a second time -- which means the hub->activating
flag is no longer needed.

Also, it is now recognized explicitly that the only reason for
resuming a port which was not suspended is to carry out a reset-resume
operation, which happens only in a non-CONFIG_USB_SUSPEND setting.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This reverts Linus's previous patch that is in mainline to make it
easier for the USB hub.c patches that follow this to apply cleanly.  The
functionality will be added back in a followon patch in this series.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as1080) makes a significant change to the way khubd
handles port connect-change and enable-change events.  Both types of
event are now debounced, and the debouncing is carried out _before_ an
existing usb_device is unregistered, instead of afterward.

This means that drivers will have to deal with longer runs of errors
when a device is unplugged, but they are supposed to be prepared for
that in any case.

The advantage is that when an enable-change occurs (caused for example
by electromagnetic interference), the debouncing period will provide
time for the cause of the problem to die away.  A simple port reset
(added in a forthcoming patch) will then allow us to recover from the
fault.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as1070) creates a new subroutine to check whether a device
can be resumed.  This code is needed even when CONFIG_USB_SUSPEND
isn't set, because devices do suspend themselves when the root hub
(and hence the entire bus) is suspended, and power sessions can get
lost during a system sleep even without individual port suspends.

The patch also fixes a loose end in USB-Persist reset-resume handling.
When a low- or full-speed device is attached to an EHCI's companion
controller, the port handoff during resume will cause the companion
port's connect-status-change feature to be set.  If that flag isn't
cleared, the port-reset code will think it indicates that the device
has been unplugged and the reset-resume will fail.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Signed-off-by: NHarvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The bus_id field is going away, use the dev_set_name() function
to set it properly.
Signed-off-by: NKay Sievers <kay.sievers@vrfy.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The bus_id field is going away, use the dev_name() function instead.
Signed-off-by: NKay Sievers <kay.sievers@vrfy.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This reverts Alan's previous patch so that the recent Hub changes will
apply cleanly.  The above mentioned patch was needed for 2.6.26 to work
properly.

Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Lukas Hejtmanek <xhejtman@ics.muni.cz>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This changes usb_create_hcd() to be able to handle the fact that
pci_name() has changed to a constant string.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NMatthew Dharm <mdharm-usb@one-eyed-alien.net>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Fix kernel-doc comments so that they don't produce errors.
Also cut some extraneous copy-paste text.

Error(linhead//drivers/pci/pci.c:1133): duplicate section name 'Description'
Error(linhead//drivers/pci/pci.c:1189): duplicate section name 'Description'
Signed-off-by: NRandy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This is a trivial patch against the hdlcdrv module that fixes its CRC
calculation. The finished CRC was overwriting the first two bytes of
each packet rather than being appended to the end.

I've tested this with 2.6.8 and 2.6.10-rc1, but hdlcdrv hasn't changed
much recently so it should work with many other kernel versions.
Signed-off-by: NMicah Dowty <micah@navi.cx>
Acked-by: NThomas Sailer <t.sailer@alumni.ethz.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commits b02f5ad6 ("istallion: use
tty_port") and a352def2 ("tty: Ldisc
revamp") broke the istallion driver.

Fix the compile error and silence a warning.
Signed-off-by: NWang Chen <wangchen@cn.fujitsu.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Based upon a bug report by Alexander Beregalov and commentary
from Ben Hutchings.

These are totally unnecessary, in particular because this
driver's ->hard_start_xmit() handler takes the same driver
spinlock that the set-multicast-list handler uses.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If the net queue has not been started, we'll get this nice oops
and non-working ethernet:

------------[ cut here ]------------
Kernel BUG at c01f4648 [verbose debug info unavailable]
Oops: Exception in kernel mode, sig: 5 [#1]
MPC836x RDK
Modules linked in:
NIP: c01f4648 LR: c01c0a10 CTR: c01c08e4
REGS: cf839e40 TRAP: 0700   Not tainted  (2.6.26-05254-gc7b9969)
MSR: 00021032 <ME,IR,DR>  CR: 22042044  XER: 00000000
TASK = cf828c30[4] 'events/0' THREAD: cf838000
GPR00: c01c0a10 cf839ef0 cf828c30 c035ceb0 cf8469a0 00000064 00000000 00000000
GPR08: c035ceb0 00000001 00000001 cf99c280 22044044 7ca81020 0fffc000 00000000
GPR16: 0fff2544 0fff63c0 00000000 0fff78e0 0ffa5580 00000004 00000000 00000000
GPR24: 02082000 cf9d0000 d1068000 00009032 cf846800 cf846b80 00000001 00000014
NIP [c01f4648] __netif_schedule+0x28/0x8c
LR [c01c0a10] adjust_link+0x12c/0x1e4
Call Trace:
[cf839ef0] [c0380f50] 0xc0380f50 (unreliable)
[cf839f10] [c01c0a10] adjust_link+0x12c/0x1e4
[cf839f40] [c01c2628] phy_state_machine+0x2e0/0x448
[cf839f60] [c00425e8] run_workqueue+0xc8/0x168
[cf839f90] [c0042c6c] worker_thread+0x70/0xd0
[cf839fd0] [c0046954] kthread+0x48/0x84
[cf839ff0] [c0012488] kernel_thread+0x44/0x60
Instruction dump:
7c0803a6 4e800020 3d20c036 9421ffe0 7c0802a6 7c681b78 3929ceb0 7c694a78
7d290034 90010024 bfa10014 5529d97e <0f090000> 39600002 38030024 7d200028
---[ end trace a57d367843bd2904 ]---

Since the driver is using phylib (which is doing netif_carrier_on/off()),
we should simply remove netif_tx_schedule_all() from adjust_link().
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If the net queue has not been started, we'll get this nice oops
and non-working ethernet:

PHY: 0:01 - Link is Up - 1000/Full
------------[ cut here ]------------
kernel BUG at net/core/dev.c:1328!
Oops: Exception in kernel mode, sig: 5 [#1]
MPC837x RDB
Modules linked in:
NIP: c02544a0 LR: c01a17d0 CTR: c01a16ac
REGS: cf837e40 TRAP: 0700   Not tainted  (2.6.26-05253-g14b395e3)
MSR: 00021032 <ME,IR,DR>  CR: 22042044  XER: 00000000
TASK = cf819400[5] 'events/0' THREAD: cf836000
GPR00: c01a17d0 cf837ef0 cf819400 c03d8d08 cf8469a0 00000064 00000000 00000000
GPR08: c03d8d08 00000001 00000001 cf899ba0 22044044 00000000 0fffd000 00000000
GPR16: 0fff3028 0fff6cf0 00000000 0fff8390 0ff494a0 00000004 00000000 00000000
GPR24: c0361a00 00001058 cf9f6600 00009032 cf846800 cf846b80 00000001 00000014
NIP [c02544a0] __netif_schedule+0x28/0x8c
LR [c01a17d0] adjust_link+0x124/0x1cc
Call Trace:
[cf837ef0] [c03fb3a0] 0xc03fb3a0 (unreliable)
[cf837f10] [c01a17d0] adjust_link+0x124/0x1cc
[cf837f40] [c01a8e28] phy_state_machine+0x2e0/0x448
[cf837f60] [c0040254] run_workqueue+0xc8/0x168
[cf837f90] [c00408d8] worker_thread+0x70/0xd0
[cf837fd0] [c0044630] kthread+0x48/0x84
[cf837ff0] [c0012610] kernel_thread+0x44/0x60
Instruction dump:
7c0803a6 4e800020 3d20c03e 9421ffe0 7c0802a6 7c681b78 39298d08 7c694a78
7d290034 90010024 bfa10014 5529d97e <0f090000> 39600002 38030024 7d200028
---[ end trace 13dfd73ee42d0c30 ]---

Since the driver is using phylib (which is doing netif_carrier_on/off()),
we should simply remove netif_tx_schedule_all() from adjust_link().
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Based upon a bug report by Alexey Dobriyan, the patch is
also tested by him and confirmed to fix the problem.

Packet flow during link state events should not be done by
waking and stopping the TX queue anyways, that is handled
transparently by netif_carrier_{on,off}().

So, remove the netif_{wake,stop}_queue() calls in the link
check code, and add the necessary netif_start_queue() call
to atl1_up().
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch implements biovec merge function for crypt target.

If the underlying device has merge function defined, call it.
If not, keep precomputed value.
Signed-off-by: NMilan Broz <mbroz@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Remove max_sector restriction - merge function replaced it.
Signed-off-by: NMilan Broz <mbroz@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

This patch implements biovec merge function for linear target.

If the underlying device has merge function defined, call it.
If not, keep precomputed value.
Signed-off-by: NMilan Broz <mbroz@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Introduce a bvec merge function for device mapper devices
for dynamic size restrictions.

This code ensures the requested biovec lies within a single
target and then calls a target-specific function to check
against any constraints imposed by underlying devices.
Signed-off-by: NMilan Broz <mbroz@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Change snapshot per-module mempool to per-device mempool.

Per-module mempools could cause a deadlock if multiple
snapshot devices are stacked above each other.
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Fix a race condition that returns incorrect data when a write causes an
exception to be allocated whilst a read is still in flight.

The race condition happens as follows:
* A read to non-reallocated sector in the snapshot is submitted so that the
  read is routed to the original device.
* A write to the original device is submitted. The write causes an exception
  that reallocates the block.  The write proceeds.
* The original read is dequeued and reads the wrong data.

This race can be triggered with CFQ scheduler and one thread writing and
multiple threads reading simultaneously.

(This patch relies upon the earlier dm-kcopyd-per-device.patch to avoid a
deadlock.)
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Whenever a snapshot read gets mapped through to the origin, track it in
a per-snapshot hash table indexed by chunk number, using memory allocated
from a new per-snapshot mempool.

We need to track these reads to avoid race conditions which will be fixed
by patches that follow.
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Fix test for reinstate_path method before attempting to use it.
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>
Cc: Julia Lawall <julia@diku.dk>

Return a specific error message if there are an invalid number of multipath
arguments.

This invalid command returns an "Unknown error" because the ti->error field is
not set

dmsetup create --table '0 2 multipath 0 0 1 1 round-robin 0 1 1 /dev/sdh' mpath0
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Rearrange struct dm_io.
Shrinks size from 40 -> 32 allowing more objects/slab.
Signed-off-by: NRichard Kennedy <richard@rsk.demon.co.uk>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

dm_dirty_log_{init,exit}() can now become static.
Signed-off-by: NAdrian Bunk <bunk@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Free path selector if the arguments are invalid.

This command (note that it is invalid) causes reference leak on module
"dm_round_robin" and prevents the module from being removed.

dmsetup create --table '0 2 multipath 0 0 1 1 round-robin /dev/sdh' mpath0
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

All modifications and most access to the mddev->disks list are made
under the reconfig_mutex lock.  However there are three places where
the list is walked without any locking.  If a reconfig happens at this
time, havoc (and oops) can ensue.

So use RCU to protect these accesses:
  - wrap them in rcu_read_{,un}lock()
  - use list_for_each_entry_rcu
  - add to the list with list_add_rcu
  - delete from the list with list_del_rcu
  - delay the 'free' with call_rcu rather than schedule_work

Note that export_rdev did a list_del_init on this list.  In almost all
cases the entry was not in the list anymore so it was a no-op and so
safe.  It is no longer safe as after list_del_rcu we may not touch
the list_head.
An audit shows that export_rdev is called:
  - after unbind_rdev_from_array, in which case the delete has
     already been done,
  - after bind_rdev_to_array fails, in which case the delete isn't needed.
  - before the device has been put on a list at all (e.g. in
      add_new_disk where reading the superblock fails).
  - and in autorun devices after a failure when the device is on a
      different list.

So remove the list_del_init call from export_rdev, and add it back
immediately before the called to export_rdev for that last case.

Note also that ->same_set is sometimes used for lists other than
mddev->list (e.g. candidates).  In these cases rcu is not needed.
Signed-off-by: NNeilBrown <neilb@suse.de>

Open isn't the only thing that increments ->active.  e.g. reading
/proc/mdstat will increment it briefly.  So to avoid false positives
in testing for concurrent access, introduce a new counter that counts
just the number of times the md device it open.
Signed-off-by: NNeilBrown <neilb@suse.de>

Signed-off-by: NAndre Noll <maan@systemlinux.org>
Signed-off-by: NNeilBrown <neilb@suse.de>

This patch renames the array_size field of struct mddev_s to array_sectors
and converts all instances to use units of 512 byte sectors instead of 1k
blocks.
Signed-off-by: NAndre Noll <maan@systemlinux.org>
Signed-off-by: NNeilBrown <neilb@suse.de>

Also, change the type of the size parameter from unsigned long long to
sector_t and rename it to num_sectors.
Signed-off-by: NAndre Noll <maan@systemlinux.org>
Signed-off-by: NNeilBrown <neilb@suse.de>

The checks in overlaps() expect all parameters either in block-based
or sector-based quantities. However, its single caller passes two
rdev->data_offset arguments as well as two rdev->size arguments, the
former being sector counts while the latter are measured in 1K blocks.

This could cause rdev_size_store() to accept an invalid size from user
space. Fix it by passing only sector-based quantities to overlaps().
Signed-off-by: NAndre Noll <maan@systemlinux.org>
Signed-off-by: NNeilBrown <neilb@suse.de>

 - used strict_strtoull in place of simple_strtoull
 - use my_mddev in place of rdev->mddev (they have the same value)
and more significantly,
 - don't adjust mddev->size to fit, rather reject changes which make
   rdev->size smaller than mddev->size

Adjusting mddev->size is a hangover from bind_rdev_to_array which
does a similar thing.  But it really is a better design to insist that
mddev->size is set as required, then the rdev->sizes are set to allow
for that.  The previous way invites confusion.
Signed-off-by: NNeilBrown <neilb@suse.de>