linux-stable.git/drivers/misc/sgi-xp, branch linux-3.19.y

net: set name_assign_type in alloc_netdev()

2014-07-15T23:12:48+00:00

Extend alloc_netdev{,_mq{,s}}() to take name_assign_type as argument, and convert
all users to pass NET_NAME_UNKNOWN.

Coccinelle patch:

@@
expression sizeof_priv, name, setup, txqs, rxqs, count;
@@

(
-alloc_netdev_mqs(sizeof_priv, name, setup, txqs, rxqs)
+alloc_netdev_mqs(sizeof_priv, name, NET_NAME_UNKNOWN, setup, txqs, rxqs)
|
-alloc_netdev_mq(sizeof_priv, name, setup, count)
+alloc_netdev_mq(sizeof_priv, name, NET_NAME_UNKNOWN, setup, count)
|
-alloc_netdev(sizeof_priv, name, setup)
+alloc_netdev(sizeof_priv, name, NET_NAME_UNKNOWN, setup)
)

v9: move comments here from the wrong commit

Signed-off-by: Tom Gundersen 
Reviewed-by: David Herrmann 
Signed-off-by: David S. Miller

mm: fix GFP_THISNODE callers and clarify

2014-03-11T00:26:19+00:00

GFP_THISNODE is for callers that implement their own clever fallback to
remote nodes.  It restricts the allocation to the specified node and
does not invoke reclaim, assuming that the caller will take care of it
when the fallback fails, e.g.  through a subsequent allocation request
without GFP_THISNODE set.

However, many current GFP_THISNODE users only want the node exclusive
aspect of the flag, without actually implementing their own fallback or
triggering reclaim if necessary.  This results in things like page
migration failing prematurely even when there is easily reclaimable
memory available, unless kswapd happens to be running already or a
concurrent allocation attempt triggers the necessary reclaim.

Convert all callsites that don't implement their own fallback strategy
to __GFP_THISNODE.  This restricts the allocation a single node too, but
at the same time allows the allocator to enter the slowpath, wake
kswapd, and invoke direct reclaim if necessary, to make the allocation
happen when memory is full.

Signed-off-by: Johannes Weiner 
Acked-by: Rik van Riel 
Cc: Jan Stancek 
Cc: Mel Gorman 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

sgi-xp: open-code interruptible_sleep_on_timeout

2014-01-08T23:18:02+00:00

interruptible_sleep_on_timeout is deprecated and going away soon.
The use in the sgi-xp driver leaves me puzzled, so I'd prefer not
to touch it. This patch replaces it with an open-coded prepare_to_wait
and finish_wait pair, which should be completely equivalent, so it
doesn't fix an existing race, but lets us get away with removing
the function so we can not get any new users.

In order to remove the typical sleep_on race, one would have to
replace the call with wait_event_interruptible_timeout and add
a condition to wait for. The fact that there is a one-jiffy timeout
suggests that we don't actually expect to get woken up properly
and the caller just uses this as a short sleeping function
if it doesn't wake up properly.

Signed-off-by: Arnd Bergmann 
Cc: Cliff Whickman 
Acked-by: Robin Holt 
Signed-off-by: Greg Kroah-Hartman

sgi: xpc: Convert use of typedef ctl_table to struct ctl_table

2013-06-14T08:19:25+00:00

This typedef is unnecessary and should just be removed.

Signed-off-by: Joe Perches 
Acked-by: Robin Holt 
Signed-off-by: Jiri Kosina

SGI-XP: handle non-fatal traps

2012-12-21T01:40:20+00:00

We found a user code which was raising a divide-by-zero trap.  That trap
would lead to XPC connections between system-partitions being torn down
due to the die_chain notifier callouts it received.

This also revealed a different issue where multiple callers into
xpc_die_deactivate() would all attempt to do the disconnect in parallel
which would sometimes lock up but often overwhelm the console on very
large machines as each would print at least one line of output at the
end of the deactivate.

I reviewed all the users of the die_chain notifier and changed the code
to ignore the notifier callouts for reasons which will not actually lead
to a system to continue on to call die().

[akpm@linux-foundation.org: fix ia64]
Signed-off-by: Robin Holt 
Cc: Thomas Gleixner 
Cc: Ingo Molnar 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

drivers/misc/sgi-xp/xpc_uv.c: SGI XPC fails to load when cpu 0 is out of IRQ resources

2012-08-21T23:45:03+00:00

On many of our larger systems, CPU 0 has had all of its IRQ resources
consumed before XPC loads.  Worst cases on machines with multiple 10
GigE cards and multiple IB cards have depleted the entire first socket
of IRQs.

This patch makes selecting the node upon which IRQs are allocated (as
well as all the other GRU Message Queue structures) specifiable as a
module load param and has a default behavior of searching all nodes/cpus
for an available resources.

[akpm@linux-foundation.org: fix build: include cpu.h and module.h]
Signed-off-by: Robin Holt 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

sgi-xp: nested calls to spin_lock_irqsave()

2012-07-11T23:04:47+00:00

The code here has a nested spin_lock_irqsave().  It's not needed since
IRQs are already disabled and it causes a problem because it means that
IRQs won't be enabled again at the end.  The second call to
spin_lock_irqsave() will overwrite the value of irq_flags and we can't
restore the proper settings.

Signed-off-by: Dan Carpenter 
Signed-off-by: Robin Holt 
Cc: Jack Steiner 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

Remove all #inclusions of asm/system.h

2012-03-28T17:30:03+00:00

Remove all #inclusions of asm/system.h preparatory to splitting and killing
it.  Performed with the following command:

perl -p -i -e 's!^#\s*include\s*.*\n!!' `grep -Irl '^#\s*include\s*' *`

Signed-off-by: David Howells

net: remove NETIF_F_NO_CSUM feature bit

2011-11-16T22:43:12+00:00

Only distinct use is checking if NETIF_F_NOCACHE_COPY should be
enabled by default. The check heuristics is altered a bit here,
so it hits other people than before. The default shouldn't be
trusted for performance-critical cases anyway.

For all other uses NETIF_F_NO_CSUM is equivalent to NETIF_F_HW_CSUM.

Signed-off-by: Michał Mirosław 
Signed-off-by: David S. Miller

sgi-xp: fix a use after free

2011-06-19T23:28:03+00:00

Its illegal to dereference skb after dev_kfree_skb(skb)

Signed-off-by: Eric Dumazet 
CC: Robin Holt 
Signed-off-by: David S. Miller