Thomas Gleixner [Fri, 25 Mar 2011 15:26:01 +0000 (16:26 +0100)]
powerpc: xilinx: Cleanup flow type handling
The core irq_set_type() function updates the flow type when the chip
callback returns 0. So setting the type is bogus. The core also
updates the LEVEL flag.
Thomas Gleixner [Fri, 25 Mar 2011 15:23:57 +0000 (16:23 +0100)]
powerpc: uic: Cleanup flow type handling
The core irq_set_type() function updates the flow type when the chip
callback returns 0. So setting the type is bogus. The core also
updates IRQ_LEVEL.
Use irq_data to get the level type information in the chip functions.
Thomas Gleixner [Fri, 25 Mar 2011 15:20:15 +0000 (16:20 +0100)]
powerpc: mpic: Cleanup flow type handling
The core irq_set_type() function updates the flow type when the chip
callback returns 0. So setting the type is bogus.
The new core code allows to update the type in irq_data and return
IRQ_SET_MASK_OK_NOCOPY, so the core code will not touch it, except for
setting the IRQ_LEVEL flag.
Retrieve the IRQ_LEVEL information from irq_data which avoids a
redundant sparse irq lookup as well.
Thomas Gleixner [Fri, 25 Mar 2011 15:18:32 +0000 (16:18 +0100)]
powerpc: mpc8xx_pic: Cleanup flow type handling
The core irq_set_type() function updates the flow type when the chip
callback returns 0. So setting the type is bogus. The level flag is
updated in the core as well.
Use the proper accessors for setting the irq handlers.
Thomas Gleixner [Fri, 25 Mar 2011 15:16:30 +0000 (16:16 +0100)]
powerpc: ipic: Cleanup flow type handling
The core irq_set_type() function updates the flow type when the chip
callback returns 0. So setting the type is bogus.
The new core code allows to update the type in irq_data and return
IRQ_SET_MASK_OK_NOCOPY, so the core code will not touch it, except for
setting the IRQ_LEVEL flag.
Use the proper accessors for setting the irq handlers.
Thomas Gleixner [Fri, 25 Mar 2011 15:07:51 +0000 (16:07 +0100)]
powerpc: cpm2_pic: Cleanup flow type handling
The core irq_set_type() function updates the flow type when the chip
callback returns 0. So setting the type is bogus.
The new core code allows to update the type in irq_data and return
IRQ_SET_MASK_OK_NOCOPY, so the core code will not touch it, except for
setting the IRQ_LEVEL flag.
Use the proper accessors for setting the irq handlers.
Thomas Gleixner [Fri, 25 Mar 2011 15:04:41 +0000 (16:04 +0100)]
powerpc: cell: Fix the irq_desc access
The core irq_set_type() function updates the flow type when the chip
callback returns 0. It also updates irq_data, so this can be used in
irq_ack() to check for the level bit. That avoids a redundant sparse
irq lookup.
Thomas Gleixner [Fri, 25 Mar 2011 19:25:06 +0000 (20:25 +0100)]
ia64: iosapic: Remove redundant mask from ack()
The core code calls mask_ack() which calls irq_ack() and irq_mask()
for the case where an interrupt is disabled and marked pending. That
seems to be a leftover from the old __do_IRQ() mode.
Thomas Gleixner [Thu, 24 Mar 2011 11:47:04 +0000 (12:47 +0100)]
arm: davinci: Cleanup irq chip code
Make use of the new functionality which ensures that irq_set_type is
called with the chip masked. Unmask is only done when the interrupt is
not disabled.
Thomas Gleixner [Mon, 28 Mar 2011 09:26:09 +0000 (11:26 +0200)]
arm: dove: Use correct GPIO_BASE and remove orion_gpio_init() leftover
commit 9eac6d0 (ARM: Remove dependency of plat-orion GPIO code on mach
directory includes) missed to convert one instance of
DOVE_GPIO_VIRT_BASE and left the orion_gpio_init() in mpp.c
watchdog: softdog.c: enhancement to optionally invoke panic instead of reboot on timer expiry
This is needed for determining the reason for failure when a softdog
timeout occurs.
We use softdog to watch for critical application failures and at the
minimum a snapshot of the system would help to determine the cause. In
such a scenario the application could fail but there isn't a softlockup as
such, hence the detect softlockup feature does not help.
The patch adds a module parameter soft_panic which when set to 1 causes
softdog to invoke panic instead of reboot when the softdog timer expires.
By invoking panic we execute kdump if it is configured and the vmcore
generated by kdump should provide atleast a minimal idea of the reason for
failure.
Randy Dunlap [Thu, 24 Mar 2011 20:32:44 +0000 (13:32 -0700)]
watchdog: fix nv_tco section mismatch
Fix section mismatch warning:
Mark the called nv_tco_getdevice() as __devinit, just like its caller.
WARNING: drivers/watchdog/nv_tco.o(.devinit.text+0x16): Section mismatch in reference from the function nv_tco_init() to the function .init.text:nv_tco_getdevice()
The function __devinit nv_tco_init() references
a function __init nv_tco_getdevice().
If nv_tco_getdevice is only used by nv_tco_init then
annotate nv_tco_getdevice with a matching annotation.
Steffen Klassert [Mon, 28 Mar 2011 19:47:30 +0000 (19:47 +0000)]
xfrm: Check for esn buffer len in xfrm_new_ae
In xfrm_new_ae() we may overwrite the allocated esn replay state
buffer with a wrong size. So check that the new size matches the
original allocated size and return an error if this is not the case.
Steffen Klassert [Mon, 28 Mar 2011 19:46:39 +0000 (19:46 +0000)]
xfrm: Assign esn pointers when cloning a state
When we clone a xfrm state we have to assign the replay_esn
and the preplay_esn pointers to the state if we use the
new replay detection method. To this end, we add a
xfrm_replay_clone() function that allocates memory for
the replay detection and takes over the necessary values
from the original state.
Steffen Klassert [Mon, 28 Mar 2011 19:45:52 +0000 (19:45 +0000)]
xfrm: Move the test on replay window size into the replay check functions
As it is, the replay check is just performed if the replay window of the
legacy implementation is nonzero. So we move the test on a nonzero replay
window inside the replay check functions to be sure we are testing for the
right implementation.
Anton Blanchard [Sun, 27 Mar 2011 16:50:41 +0000 (16:50 +0000)]
cxgb3: Apply interrupt coalescing settings to all queues
While testing the performance of different receive interrupt
coalescing settings on a single stream TCP benchmark, I noticed two
very different results. With rx-usecs=50, most of the time a
connection would hit 8280 Mbps but once in a while it would hit
9330 Mbps.
It turns out we are only applying the interrupt coalescing settings
to the first queue and whenever the rx hash would direct us onto
that queue we ran faster.
With this patch applied and rx-usecs=50, I get 9330 Mbps
consistently.
Anton Blanchard [Sun, 27 Mar 2011 14:57:26 +0000 (14:57 +0000)]
net: Always allocate at least 16 skb frags regardless of page size
When analysing performance of the cxgb3 on a ppc64 box I noticed that
we weren't doing much GRO merging. It turns out we are limited by the
number of SKB frags:
#define MAX_SKB_FRAGS (65536/PAGE_SIZE + 2)
With a 4kB page size we have 18 frags, but with a 64kB page size we
only have 3 frags.
I ran a single stream TCP bandwidth test to compare the performance of
md: Fix integrity registration error when no devices are capable
We incorrectly returned -EINVAL when none of the devices in the array
had an integrity profile. This in turn prevented mdadm from starting
the metadevice. Fix this so we only return errors on mismatched
profiles and memory allocation failures.
Andres Salomon [Tue, 29 Mar 2011 00:22:30 +0000 (17:22 -0700)]
ARM: mx51_efika: fix build error due to new mfd changes
MFD changes in 4ec1b54c ('mfd: mfd_cell is now implicitly available to
mc13xxx drivers') changed the mc13xxx_platform_data struct layout.
At the time all users were changed, but this driver was introduced in
another tree at the same time. This updates the mc13xxx_platform_data
user, fixing a build error.
Linus Torvalds [Tue, 29 Mar 2011 00:39:54 +0000 (17:39 -0700)]
Merge branches 'irq-cleanup-for-linus' and 'irq-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'irq-cleanup-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
vlynq: Convert irq functions
* 'irq-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
genirq; Fix cleanup fallout
genirq: Fix typo and remove unused variable
genirq: Fix new kernel-doc warnings
genirq: Add setter for AFFINITY_SET in irq_data state
genirq: Provide setter inline for IRQD_IRQ_INPROGRESS
genirq: Remove handle_IRQ_event
arm: Ns9xxx: Remove private irq flow handler
powerpc: cell: Use the core flow handler
genirq: Provide edge_eoi flow handler
genirq: Move INPROGRESS, MASKED and DISABLED state flags to irq_data
genirq: Split irq_set_affinity() so it can be called with lock held.
genirq: Add chip flag for restricting cpu_on/offline calls
genirq: Add chip hooks for taking CPUs on/off line.
genirq: Add irq disabled flag to irq_data state
genirq: Reserve the irq when calling irq_set_chip()
projects / linux.git / log