Since there are only two locations where cpu_mask_to_apicid() is
called from, remove the operation and use only
cpu_mask_to_apicid_and() instead.
Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
Suggested-and-acked-by: Suresh Siddha <suresh.b.siddha@intel.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/20120614074935.GE3383@dhcp-26-207.brq.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Since commit 8637e38 ("x86/apic: Avoid useless scanning thru a
cpumask in assign_irq_vector()") vector_allocation_domain()
operation indicates if a cpumask is dynamic or static. This
update fixes the oversight and makes the operation to return a
value.
Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/20120614103933.GJ3383@dhcp-26-207.brq.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Add "read-mostly" qualifier to the following variables in
smp.h:
- cpu_sibling_map
- cpu_core_map
- cpu_llc_shared_map
- cpu_llc_id
- cpu_number
- x86_cpu_to_apicid
- x86_bios_cpu_apicid
- x86_cpu_to_logical_apicid
As long as all the variables above are only written during the
initialization, this change is meant to prevent the false
sharing. More specifically, on vSMP Foundation platform
x86_cpu_to_apicid shared the same internode_cache_line with
frequently written lapic_events.
From the analysis of the first 33 per_cpu variables out of 219
(memories they describe, to be more specific) the 8 have read_mostly
nature (tlb_vector_offset, cpu_loops_per_jiffy, xen_debug_irq, etc.)
and 25 are frequently written (irq_stack_union, gdt_page,
exception_stacks, idt_desc, etc.).
Assuming that the spread of the rest of the per_cpu variables is
similar, identifying the read mostly memories will make more sense
in terms of long-term code maintenance comparing to identifying
frequently written memories.
Signed-off-by: Vlad Zolotarov <vlad@scalemp.com>
Acked-by: Shai Fultheim <shai@scalemp.com>
Cc: Shai Fultheim (Shai@ScaleMP.com) <Shai@scalemp.com>
Cc: ido@wizery.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1719258.EYKzE4Zbq5@vlad
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Some read-mostly per-cpu data may need to be declared or defined
early, so it can be initialized and accessed before per_cpu
areas are allocated.
Only the data that resides in the per_cpu areas should be
read-mostly, as there is little benefit in optimizing cache
lines on initialization.
Signed-off-by: Ido Yariv <ido@wizery.com>
[ Added the missing declarations in !SMP code. ]
Signed-off-by: Vlad Zolotarov <vlad@scalemp.com>
Acked-by: Shai Fultheim <shai@scalemp.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/46188571.ddB8aVQYWo@vlad
Signed-off-by: Ingo Molnar <mingo@kernel.org>
this patch fixes the following warning:
[ 2.825378] genirq: Threaded irq requested \
with handler=NULL and !ONESHOT for irq 363
Signed-off-by: Felipe Balbi <balbi@ti.com>
A bunch of bugzillas have complained how noisy the nmi_watchdog
is during boot-up especially with its expected failure cases
(like virt and bios resource contention).
This is my attempt to quiet them down and keep it less confusing
for the end user. What I did is print the message for cpu0 and
save it for future comparisons. If future cpus have an
identical message as cpu0, then don't print the redundant info.
However, if a future cpu has a different message, happily print
that loudly.
Before the change, you would see something like:
..TIMER: vector=0x30 apic1=0 pin1=2 apic2=-1 pin2=-1
CPU0: Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz stepping 0a
Performance Events: PEBS fmt0+, Core2 events, Intel PMU driver.
... version: 2
... bit width: 40
... generic registers: 2
... value mask: 000000ffffffffff
... max period: 000000007fffffff
... fixed-purpose events: 3
... event mask: 0000000700000003
NMI watchdog enabled, takes one hw-pmu counter.
Booting Node 0, Processors #1
NMI watchdog enabled, takes one hw-pmu counter.
#2
NMI watchdog enabled, takes one hw-pmu counter.
#3 Ok.
NMI watchdog enabled, takes one hw-pmu counter.
Brought up 4 CPUs
Total of 4 processors activated (22607.24 BogoMIPS).
After the change, it is simplified to:
..TIMER: vector=0x30 apic1=0 pin1=2 apic2=-1 pin2=-1
CPU0: Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz stepping 0a
Performance Events: PEBS fmt0+, Core2 events, Intel PMU driver.
... version: 2
... bit width: 40
... generic registers: 2
... value mask: 000000ffffffffff
... max period: 000000007fffffff
... fixed-purpose events: 3
... event mask: 0000000700000003
NMI watchdog: enabled on all CPUs, permanently consumes one hw-PMU counter.
Booting Node 0, Processors #1#2#3 Ok.
Brought up 4 CPUs
V2: little changes based on Joe Perches' feedback
V3: printk cleanup based on Ingo's feedback; checkpatch fix
V4: keep printk as one long line
V5: Ingo fix ups
Reported-and-tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Don Zickus <dzickus@redhat.com>
Cc: nzimmer@sgi.com
Cc: joe@perches.com
Link: http://lkml.kernel.org/r/1339594548-17227-1-git-send-email-dzickus@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This patch fixes a potential Rx timestamp deadlock that causes the Rx
timestamping to stall indefinitely. The issue could occur when a PTP packet is
timestamped by hardware but never reaches the Rx queue. In order to prevent a
permanent loss of timestamping, the RXSTMP(L/H) registers have to be read to
unlock them. (This used to only occur when a packet that was timestamped
reached the software.) However the registers can't be read early otherwise
there is no way to correlate them to the packet.
This patch introduces a filter function which can be used to determine if a
packet should have been timestamped. Supplied with the filter setup by the
hwtstamp ioctl, check to make sure the PTP protocol and message type match the
expected values. If so, then read the timestamp registers (to free them.) At
this point check the descriptor bit, if the bit is set then we know this
packet correlates to the timestamp stored in the RXTSTAMP registers.
Otherwise, assume that packet was dropped by the hardware, and ignore this
timestamp value. However, we have at least unlocked the rxtstamp registers for
future timestamping.
Due to the way the driver handles skb data, it cannot be directly accessed. In
order to work around this, a copy of the skb data into a linear buffer is
made. From this buffer it becomes possible to read the data correctly
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Richard Cochran <richardcochran@gmail.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
When enabling the hwtstamp mode for Rx timestamping the V2 ptp event type
specific modes (Delay Request and Sync) have been rolled into the V2 all event
packet modes, in order to more accurately represent what hardware is doing.
Hardware always timestamps the Path delay packets when a V2 mode is selected,
regardless of what type was selected (in order to always support Path delay
mode). However this means the user selected modes of timestamping only Sync or
Delay Request is not truly supported. This patch correctly sets the mode for
the hwtstamp config and returns to the user that all V2 event packets will be
timestamped.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This patch fixes two minor nits from Richard Cochran. The first is a case of
ambitious line wrapping that wasn't necessary. The second is to re-order the
flag checks for PPS support. Previously, the hardware test was done first, and
the interrupt flag test was done second. Now, test the interrupt flag and use
the unlikely macro.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
ixgbe_sysfs.c is only needed when CONFIG_IXGBE_HWMON is configured in the
kernel.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Acked-by: Don Skidmore <Donald.c.skidmore@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
The flow control DV macros are used to calculate the flow control
high and low thresholds. This patch annotates these macros slightly
better and fixes the issues below.
The macro variables are renamed LINK to _max_frame_link and TC to
_max_frame_tc. This was to avoid confusion and make them more
readable. It was found that people auditing the code read TC to be
'traffic class' in the 802.1Q definition instead of the max frame
size of the tc. Hopefully it is clear now.
This audit also found the following real deviations from the
theoretical values. Fixed in this patch.
* I multiplied the DV calculations by (36/25) which always
evaluates to 1. This does not match the intended theoretical
value of 1.44.
* IXGBE_BT2KB added 1023 to account for rounding however this
really should be 8 * 1023 - 1 to account for division by 8k.
* x2 multiplication of max frame in DV calculations to account
for updated hardware recommendations.
With this patch the DV values are inline with the recommendations
in the 82599 and 82598 data sheets. Its worth noting I did not
see any dropped frames with flow control on in my experiments without
this patch. However aligning with the hardware specs and
recommendations seems like a good idea here to account for worst
case scenarios.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Based on a report from Ethan Zhao, before calling register_netdev() the
driver should be using logging macros that do not display the potentially
confusing "(unregistered net_device)" yet still display the useful driver
name and PCI bus/device/function.
Reported-by: Ethan Zhao <ethan.kernel@gmail.com>
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
The OMAP dmtimer driver allows you to dynamically configure the functional
clock that drives the timer logic. The dmtimer driver uses the device name and
a "con-id" string to search for the appropriate functional clock.
Currently, we define a clock alias for each functional clock source each timer
supports. Some functional clock sources are common to all of the timers on a
device and so for these clock sources we can use a single alias with a unique
con-id string.
The possible functional clock sources for an OMAP device are a 32kHz clock,
a system (MHz range) clock and (for OMAP2 only) an external clock. By defining
a unique con-id name for each of these (timer_32k_ck, timer_sys_ck and
timer_ext_ck) we can eliminate a lot of the clock aliases for timers. This
reduces code, speeds-up searches and clock initialisation time.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Acked-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
OMAP1 uses an architecture specific function for setting the dmtimer clock
source, where as the OMAP2+ devices use the clock framework. Eventually OMAP1
device should also use the clock framework and hence we should not any
architecture specific functions.
For now move the OMAP2+ function for configuring the clock source into the
dmtimer driver. Therefore, we do no longer need to specify an architecture
specific function for setting the clock source for OMAP2+ devices. This will
simplify device tree migration of the dmtimers for OMAP2+ devices.
From now on, only OMAP1 devices should specify an architecture specific
function for setting the clock source via the platform data set_dmtimer_src()
function pointer.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
OMAP1 dmtimer support is currently broken. When a dmtimer is requested by the
omap_dm_timer_request() function fails to allocate a dmtimer because the call
to clk_get() inside omap_dm_timer_prepare fails. The clk_get() fails simply
because the clock data for the OMAP1 dmtimers is not present.
Ideally this should be fixed by moving OMAP1 dmtimers to use the clock
framework. For now simply fix this by using the "TIMER_NEEDS_RESET" flag to
identify an OMAP1 device and avoid calling clk_get(). Although this is not
the ideal fix and should be corrected, this flag has already been use for the
same purpose in omap_dm_timer_stop().
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
For OMAP1 devices, it is necessary to perform a manual reset of the timer.
Currently, this is indicating by setting the "needs_manual_reset" variable in
the platform data. Instead of using an extra variable to indicate this add a new
timer capabilities flag to indicate this and remove the "needs_manual_reset"
member from the platform data.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
For OMAP2+ devices, a function pointer that returns the number of times a timer
power domain has lost context is passed to the dmtimer driver. This function
pointer is only populated for OMAP2+ devices and it is pointing to a platform
function. Given that this is a platform function, we can simplify the code by
removing the function pointer and referencing the function directly. We can use
the OMAP_TIMER_ALWON flag to determine if we need to call this function for
OMAP1 and OMAP2+ devices.
The benefit of this change is the we can remove the function pointer from the
platform data and simplifies the dmtimer migration to device-tree.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
The platform data variable loses_context is used to determine if the timer may
lose its logic state during power transitions and so needs to be restored. This
information is also provided in the HWMOD device attributes for OMAP2+ devices
via the OMAP_TIMER_ALWON flag. When this flag is set the timer will not lose
context. So use the HWMOD device attributes to determine this.
For OMAP1 devices, loses_context is never set and so set the OMAP_TIMER_ALWON
flag for OMAP1 timers to ensure that code is equivalent.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Currently, the dmtimer determines whether an timer can support an external
clock source (sys_altclk) for driving the timer by the IP version. Only
OMAP24xx devices can support an external clock source, but the IP version
between OMAP24xx and OMAP3xxx is common and so this incorrectly indicates
that OMAP3 devices can use an external clock source.
Rather than use the IP version, just let the clock framework handle this.
If the "alt_ck" does not exist for a timer then the clock framework will fail
to find the clock and hence will return an error. By doing this we can eliminate
the "timer_ip_version" variable passed as part of the platform data and simplify
the code.
We can also remove the timer IP version from the HWMOD data because the dmtimer
driver uses the TIDR register to determine the IP version.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Fix the following issues with the timer device attributes for OMAP2+ devices:
1. For OMAP24xx devices, timers 2-8 have the ALWAYS-ON attribute indicating
that these timers are in an ALWAYS-ON power domain. This is not the case
only timer1 is in an ALWAYS-ON power domain.
2. For OMAP3xxx devices, timers 2-7 have the ALWAYS-ON attribute indicating
that these timers are in an ALWAYS-ON power domain. This is not the case
only timer1 and timer12 are in an ALWAYS-ON power domain.
3. For OMAP3xxx devices, timer12 does not have the ALWAYS-ON attribute but
is in an always-on power domain.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Acked-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Although the OMAP timers share a common hardware design, there are some
differences between the timer instances in a given device. For example, a timer
maybe in a power domain that can be powered-of, so can lose its logic state and
need restoring where as another may be in power domain that is always be on.
Another example, is a timer may support different clock sources to drive the
timer. This information is passed to the dmtimer via the following platform data
structure.
struct dmtimer_platform_data {
int (*set_timer_src)(struct platform_device *pdev, int source);
int timer_ip_version;
u32 needs_manual_reset:1;
bool loses_context;
int (*get_context_loss_count)(struct device *dev);
};
The above structure uses multiple variables to represent the timer features.
HWMOD also stores the timer capabilities using a bit-mask that represents the
features supported. By using the same format for representing the timer
features in the platform data as used by HWMOD, we can ...
1. Use the flags defined in the plat/dmtimer.h to represent the features
supported.
2. For devices using HWMOD, we can retrieve the features supported from HWMOD.
3. Eventually, simplify the platform data structure to be ...
struct dmtimer_platform_data {
int (*set_timer_src)(struct platform_device *pdev, int source);
u32 timer_capability;
}
Another benefit from doing this, is that it will simplify the migration of the
dmtimer driver to device-tree. For example, in the current OMAP2+ timer code the
"loses_context" variable is configured at runtime by calling an architecture
specific function. For device tree this creates a problem, because we would need
to call the architecture specific function from within the dmtimer driver.
However, such attributes do not need to be queried at runtime and we can look up
the attributes via HWMOD or device-tree.
This changes a new "capability" variable to the platform data and timer
structure so we can start removing and simplifying the platform data structure.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
During early boot, one or two dmtimers are reserved by the kernel as system
timers (for clocksource and clockevents). These timers are marked as reserved
and the dmtimer driver is notified which timers have been reserved via the
platform data information.
For OMAP2+ devices the timers reserved may vary depending on device and compile
flags. Therefore, it is not easy to assume which timers we be reserved for the
system timers. In order to migrate the dmtimer driver to support device-tree we
need a way to pass the timers reserved for system timers to the dmtimer driver.
Using the platform data structure will not work in the same way as it is
currently used because the platform data structure will be stored statically in
the dmtimer itself and the platform data will be selected via the device-tree
match device function (of_match_device).
There are a couple ways to workaround this. One option is to store the system
timers reserved for the kernel in the device-tree and query them on boot.
The downside of this approach is that it adds some delay to parse the DT blob
to search for the system timers. Secondly, for OMAP3 devices we have a
dependency on compile time flags and the device-tree would not be aware of that
kernel compile flags and so we would need to address that.
The second option is to add a function to the dmtimer code to reserved the
system timers during boot and so the dmtimer knows exactly which timers are
being used for system timers. This also allows us to remove the "reserved"
member from the timer platform data. This seemed like the simpler approach and
so was implemented here.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
The OMAP2+ timer code has a definition for the maximum number of timers that
OMAP2+ devices have. This defintion is not used anywhere in the code and
appears to be left over. Furthermore the definition is not accurate for OMAP4
devices that only have 11 timers available because the 12th timer is reserved
as a secure timer and for OMAP3 devices the 12th timer is not available on
secure devices. Therefore, remove this definition.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
In the plat/dmtimer.h there is a structure named "clk" declared. This structure
is not used and appears to be left over from previous code. Hence, remove this
unused structure.
Verified that both omap1 and omap2plus kernel configurations build with this
change.
Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
This function combines rgrp functions get_local_rgrp and
gfs2_inplace_reserve so that the double retry loop is gone.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Setting the cpu restart parameters is done in three different fashions:
- directly setting the four parameters individually
- copying the four parameters with memcpy (using 4 * sizeof(long))
- copying the four parameters using a private structure
In addition code in entry*.S relies on a certain order of the restart
members of struct _lowcore.
Make all of this more robust to future changes by adding a
mem_absolute_assign(dest, val) define, which assigns val to dest
using absolute addressing mode. Also the load multiple instructions
in entry*.S have been split into separate load instruction so the
order of the struct _lowcore members doesn't matter anymore.
In addition move the prototypes of memcpy_real/absolute from uaccess.h
to processor.h. These memcpy* variants are not related to uaccess at all.
string.h doesn't seem to match as well, so lets use processor.h.
Also replace the eight byte array in struct _lowcore which represents a
misaliged u64 with a u64. The compiler will always create code that
handles the misaligned u64 correctly.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Generate the proactive PREQ element as defined in
Sec. 13.10.9.3 (Case C) of IEEE Std. 802.11-2012
based on the selection of dot11MeshHWMPRootMode as follow:
dot11MeshHWMPRootMode (2) is proactivePREQnoPREP
dot11MeshHWMPRootMode (3) is proactivePREQwithPREP
The proactive PREQ is generated based on the interval
defined by dot11MeshHWMProotInterval.
With this change, proactive RANN element is now generated
if the dot11MeshHWMPRootMode is set to (4) instead of (1).
Signed-off-by: Chun-Yeow Yeoh <yeohchunyeow@gmail.com>
[line-break commit log]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Add the mesh configuration parameters dot11MeshHWMProotInterval
and dot11MeshHWMPactivePathToRootTimeout to be used by
proactive PREQ mechanism.
Signed-off-by: Chun-Yeow Yeoh <yeohchunyeow@gmail.com>
[line-break commit log]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The size of the slab object is frequently needed. Since we now
have a size field directly in the kmem_cache structure there is no
need anymore of the obj_size macro/function.
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Define a struct that describes common fields used in all slab allocators.
A slab allocator either uses the common definition (like SLOB) or is
required to provide members of kmem_cache with the definition given.
After that it will be possible to share code that
only operates on those fields of kmem_cache.
The patch basically takes the slob definition of kmem cache and
uses the field namees for the other allocators.
It also standardizes the names used for basic object lengths in
allocators:
object_size Struct size specified at kmem_cache_create. Basically
the payload expected to be used by the subsystem.
size The size of memory allocator for each object. This size
is larger than object_size and includes padding, alignment
and extra metadata for each object (f.e. for debugging
and rcu).
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Those are rather trivial now and its better to see inline what is
really going on.
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Add fields to the page struct so that it is properly documented that
slab overlays the lru fields.
This cleans up some casts in slab.
Reviewed-by: Glauber Costa <glommer@parallels.com>
Reviewed-by: Joonsoo Kim <js1304@gmail.com>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Those have become so simple that they are no longer needed.
Reviewed-by: Joonsoo Kim <js1304@gmail.com>
Acked-by: David Rientjes <rientjes@google.com>
signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Reviewed-by: Joonsoo Kim <js1304@gmail.com>
Acked-by: David Rientjes <rientjes@google.com>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Define the fields used by slob in mm_types.h and use struct page instead
of struct slob_page in slob. This cleans up numerous of typecasts in slob.c and
makes readers aware of slob's use of page struct fields.
[Also cleans up some bitrot in slob.c. The page struct field layout
in slob.c is an old layout and does not match the one in mm_types.h]
Reviewed-by: Glauber Costa <glommer@parallels.com>
Acked-by: David Rientjes <rientjes@google.com>
Reviewed-by: Joonsoo Kim <js1304@gmail.com>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
WARNING: vmlinux.o(.cpuinit.text+0x280): Section mismatch in reference from the function cpu_probe() to the function .init.text:sh64_tlb_init()
The function __cpuinit cpu_probe() references
a function __init sh64_tlb_init().
If sh64_tlb_init is only used by cpu_probe then
annotate sh64_tlb_init with a matching annotation.
sh64_tlb_init() simply needs to be __cpuinit annotated, so fix that up.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
There's not much here that we can't get at through alternate means (aside
from the TLB contents, but that doesn't belong here anyways). Most of
this information is already provided by the sh32 routines, which we'll
consolidate on next.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
This switches over from the special-casing that sh64 had and to the model
that's being employed for sh32.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Allocate memory, region, remap and irq for device state using
devm_* helpers to simplify memory accounting.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@linux.intel.com>
pl330_update() stores a pointer to the thrd->req that finished, which
contains a pointer to the corresponding pl330_req. This is done with
the pl330_lock held. Then, it iterates through the req_done list,
calling the callback for each of the requests that are done. The
problem is that the driver releases the lock before calling the
callback for each of the callbacks. pl330_submit_req() running in
another processor can then acquire the lock and insert another request
in one of the thrd->req that hasn't been processed yet, replacing the
pointer to pl330_req there. When the callback returns in
pl330_update() and the next rqdone is popped from the list, it
dereferences the pl330_req pointer to the just scheduled pl330_req,
instead of the one that has finished, calling pl330 with the wrong r.
This patch fixes this by storing the pointer to pl330_req directly in
the list.
Signed-off-by: Javi Merino <javi.merino@arm.com>
Cc: Jassi Brar <jaswinder.singh@linaro.org>
Acked-by: Jassi Brar <jaswinder.singh@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@linux.intel.com>
Rename serpent-avx assembler functions so that they do not collide with
serpent-sse2 assembler functions when linking both versions in to same
kernel image.
Reported-by: Randy Dunlap <rdunlap@xenotime.net>
Cc: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de>
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This patch changes u8 in struct arc4_ctx and variables to u32 (as AMD seems
to have problem with u8 array). Below are tcrypt results of old 1-byte block
cipher versus ecb(arc4) with u8 and ecb(arc4) with u32.
tcrypt results, x86-64 (speed ratios: new-u32/old, new-u8/old):
u32 u8
AMD Phenom II : x3.6 x2.7
Intel Core 2 : x2.0 x1.9
tcrypt results, i386 (speed ratios: new-u32/old, new-u8/old):
u32 u8
Intel Atom N260 : x1.5 x1.4
Cc: Jon Oberheide <jon@oberheide.org>
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Currently arc4.c provides simple one-byte blocksize cipher which is wrapped
by ecb() module, giving function call overhead on every encrypted byte. This
patch adds ecb(arc4) directly into arc4.c for higher performance.
tcrypt results (speed ratios: new/old):
AMD Phenom II, x86-64 : x2.7
Intel Core 2, x86-64 : x1.9
Intel Atom N260, i386 : x1.4
Cc: Jon Oberheide <jon@oberheide.org>
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>