When a kernel change causes the interface that the kernel exposes to
userspace to change, it is recommended that you send the information or
a patch to the manual pages explaining the change to the manual pages
-maintainer at mtk.manpages@gmail.com.
+maintainer at mtk.manpages@gmail.com, and CC the list
+linux-api@vger.kernel.org.
Here is a list of files that are in the kernel source tree that are
required reading:
19: All new userspace interfaces are documented in Documentation/ABI/.
See Documentation/ABI/README for more information.
+ Patches that change userspace interfaces should be CCed to
+ linux-api@vger.kernel.org.
20: Check that it all passes `make headers_check'.
mkdir $ARGV[0],0777;
$state = 0;
while (<STDIN>) {
- if (/^\.TH \"[^\"]*\" 4 \"([^\"]*)\"/) {
+ if (/^\.TH \"[^\"]*\" 9 \"([^\"]*)\"/) {
if ($state == 1) { close OUT }
$state = 1;
- $fn = "$ARGV[0]/$1.4";
+ $fn = "$ARGV[0]/$1.9";
print STDERR "Creating $fn\n";
open OUT, ">$fn" or die "can't open $fn: $!\n";
print OUT $_;
+ =============
+ CFS Scheduler
+ =============
-This is the CFS scheduler.
-
-80% of CFS's design can be summed up in a single sentence: CFS basically
-models an "ideal, precise multi-tasking CPU" on real hardware.
-
-"Ideal multi-tasking CPU" is a (non-existent :-)) CPU that has 100%
-physical power and which can run each task at precise equal speed, in
-parallel, each at 1/nr_running speed. For example: if there are 2 tasks
-running then it runs each at 50% physical power - totally in parallel.
-
-On real hardware, we can run only a single task at once, so while that
-one task runs, the other tasks that are waiting for the CPU are at a
-disadvantage - the current task gets an unfair amount of CPU time. In
-CFS this fairness imbalance is expressed and tracked via the per-task
-p->wait_runtime (nanosec-unit) value. "wait_runtime" is the amount of
-time the task should now run on the CPU for it to become completely fair
-and balanced.
-
-( small detail: on 'ideal' hardware, the p->wait_runtime value would
- always be zero - no task would ever get 'out of balance' from the
- 'ideal' share of CPU time. )
-
-CFS's task picking logic is based on this p->wait_runtime value and it
-is thus very simple: it always tries to run the task with the largest
-p->wait_runtime value. In other words, CFS tries to run the task with
-the 'gravest need' for more CPU time. So CFS always tries to split up
-CPU time between runnable tasks as close to 'ideal multitasking
-hardware' as possible.
-
-Most of the rest of CFS's design just falls out of this really simple
-concept, with a few add-on embellishments like nice levels,
-multiprocessing and various algorithm variants to recognize sleepers.
-
-In practice it works like this: the system runs a task a bit, and when
-the task schedules (or a scheduler tick happens) the task's CPU usage is
-'accounted for': the (small) time it just spent using the physical CPU
-is deducted from p->wait_runtime. [minus the 'fair share' it would have
-gotten anyway]. Once p->wait_runtime gets low enough so that another
-task becomes the 'leftmost task' of the time-ordered rbtree it maintains
-(plus a small amount of 'granularity' distance relative to the leftmost
-task so that we do not over-schedule tasks and trash the cache) then the
-new leftmost task is picked and the current task is preempted.
-
-The rq->fair_clock value tracks the 'CPU time a runnable task would have
-fairly gotten, had it been runnable during that time'. So by using
-rq->fair_clock values we can accurately timestamp and measure the
-'expected CPU time' a task should have gotten. All runnable tasks are
-sorted in the rbtree by the "rq->fair_clock - p->wait_runtime" key, and
-CFS picks the 'leftmost' task and sticks to it. As the system progresses
-forwards, newly woken tasks are put into the tree more and more to the
-right - slowly but surely giving a chance for every task to become the
-'leftmost task' and thus get on the CPU within a deterministic amount of
-time.
-
-Some implementation details:
-
- - the introduction of Scheduling Classes: an extensible hierarchy of
- scheduler modules. These modules encapsulate scheduling policy
- details and are handled by the scheduler core without the core
- code assuming about them too much.
-
- - sched_fair.c implements the 'CFS desktop scheduler': it is a
- replacement for the vanilla scheduler's SCHED_OTHER interactivity
- code.
-
- I'd like to give credit to Con Kolivas for the general approach here:
- he has proven via RSDL/SD that 'fair scheduling' is possible and that
- it results in better desktop scheduling. Kudos Con!
-
- The CFS patch uses a completely different approach and implementation
- from RSDL/SD. My goal was to make CFS's interactivity quality exceed
- that of RSDL/SD, which is a high standard to meet :-) Testing
- feedback is welcome to decide this one way or another. [ and, in any
- case, all of SD's logic could be added via a kernel/sched_sd.c module
- as well, if Con is interested in such an approach. ]
-
- CFS's design is quite radical: it does not use runqueues, it uses a
- time-ordered rbtree to build a 'timeline' of future task execution,
- and thus has no 'array switch' artifacts (by which both the vanilla
- scheduler and RSDL/SD are affected).
-
- CFS uses nanosecond granularity accounting and does not rely on any
- jiffies or other HZ detail. Thus the CFS scheduler has no notion of
- 'timeslices' and has no heuristics whatsoever. There is only one
- central tunable (you have to switch on CONFIG_SCHED_DEBUG):
-
- /proc/sys/kernel/sched_granularity_ns
-
- which can be used to tune the scheduler from 'desktop' (low
- latencies) to 'server' (good batching) workloads. It defaults to a
- setting suitable for desktop workloads. SCHED_BATCH is handled by the
- CFS scheduler module too.
-
- Due to its design, the CFS scheduler is not prone to any of the
- 'attacks' that exist today against the heuristics of the stock
- scheduler: fiftyp.c, thud.c, chew.c, ring-test.c, massive_intr.c all
- work fine and do not impact interactivity and produce the expected
- behavior.
-
- the CFS scheduler has a much stronger handling of nice levels and
- SCHED_BATCH: both types of workloads should be isolated much more
- agressively than under the vanilla scheduler.
-
- ( another detail: due to nanosec accounting and timeline sorting,
- sched_yield() support is very simple under CFS, and in fact under
- CFS sched_yield() behaves much better than under any other
- scheduler i have tested so far. )
-
- - sched_rt.c implements SCHED_FIFO and SCHED_RR semantics, in a simpler
- way than the vanilla scheduler does. It uses 100 runqueues (for all
- 100 RT priority levels, instead of 140 in the vanilla scheduler)
- and it needs no expired array.
-
- - reworked/sanitized SMP load-balancing: the runqueue-walking
- assumptions are gone from the load-balancing code now, and
- iterators of the scheduling modules are used. The balancing code got
- quite a bit simpler as a result.
-
-
-Group scheduler extension to CFS
-================================
-
-Normally the scheduler operates on individual tasks and strives to provide
-fair CPU time to each task. Sometimes, it may be desirable to group tasks
-and provide fair CPU time to each such task group. For example, it may
-be desirable to first provide fair CPU time to each user on the system
-and then to each task belonging to a user.
-
-CONFIG_FAIR_GROUP_SCHED strives to achieve exactly that. It lets
-SCHED_NORMAL/BATCH tasks be be grouped and divides CPU time fairly among such
-groups. At present, there are two (mutually exclusive) mechanisms to group
-tasks for CPU bandwidth control purpose:
-
- - Based on user id (CONFIG_FAIR_USER_SCHED)
- In this option, tasks are grouped according to their user id.
- - Based on "cgroup" pseudo filesystem (CONFIG_FAIR_CGROUP_SCHED)
- This options lets the administrator create arbitrary groups
- of tasks, using the "cgroup" pseudo filesystem. See
- Documentation/cgroups.txt for more information about this
- filesystem.
-Only one of these options to group tasks can be chosen and not both.
+1. OVERVIEW
+
+CFS stands for "Completely Fair Scheduler," and is the new "desktop" process
+scheduler implemented by Ingo Molnar and merged in Linux 2.6.23. It is the
+replacement for the previous vanilla scheduler's SCHED_OTHER interactivity
+code.
+
+80% of CFS's design can be summed up in a single sentence: CFS basically models
+an "ideal, precise multi-tasking CPU" on real hardware.
+
+"Ideal multi-tasking CPU" is a (non-existent :-)) CPU that has 100% physical
+power and which can run each task at precise equal speed, in parallel, each at
+1/nr_running speed. For example: if there are 2 tasks running, then it runs
+each at 50% physical power --- i.e., actually in parallel.
+
+On real hardware, we can run only a single task at once, so we have to
+introduce the concept of "virtual runtime." The virtual runtime of a task
+specifies when its next timeslice would start execution on the ideal
+multi-tasking CPU described above. In practice, the virtual runtime of a task
+is its actual runtime normalized to the total number of running tasks.
+
+
+
+2. FEW IMPLEMENTATION DETAILS
+
+In CFS the virtual runtime is expressed and tracked via the per-task
+p->se.vruntime (nanosec-unit) value. This way, it's possible to accurately
+timestamp and measure the "expected CPU time" a task should have gotten.
+
+[ small detail: on "ideal" hardware, at any time all tasks would have the same
+ p->se.vruntime value --- i.e., tasks would execute simultaneously and no task
+ would ever get "out of balance" from the "ideal" share of CPU time. ]
+
+CFS's task picking logic is based on this p->se.vruntime value and it is thus
+very simple: it always tries to run the task with the smallest p->se.vruntime
+value (i.e., the task which executed least so far). CFS always tries to split
+up CPU time between runnable tasks as close to "ideal multitasking hardware" as
+possible.
+
+Most of the rest of CFS's design just falls out of this really simple concept,
+with a few add-on embellishments like nice levels, multiprocessing and various
+algorithm variants to recognize sleepers.
+
+
+
+3. THE RBTREE
+
+CFS's design is quite radical: it does not use the old data structures for the
+runqueues, but it uses a time-ordered rbtree to build a "timeline" of future
+task execution, and thus has no "array switch" artifacts (by which both the
+previous vanilla scheduler and RSDL/SD are affected).
+
+CFS also maintains the rq->cfs.min_vruntime value, which is a monotonic
+increasing value tracking the smallest vruntime among all tasks in the
+runqueue. The total amount of work done by the system is tracked using
+min_vruntime; that value is used to place newly activated entities on the left
+side of the tree as much as possible.
+
+The total number of running tasks in the runqueue is accounted through the
+rq->cfs.load value, which is the sum of the weights of the tasks queued on the
+runqueue.
+
+CFS maintains a time-ordered rbtree, where all runnable tasks are sorted by the
+p->se.vruntime key (there is a subtraction using rq->cfs.min_vruntime to
+account for possible wraparounds). CFS picks the "leftmost" task from this
+tree and sticks to it.
+As the system progresses forwards, the executed tasks are put into the tree
+more and more to the right --- slowly but surely giving a chance for every task
+to become the "leftmost task" and thus get on the CPU within a deterministic
+amount of time.
+
+Summing up, CFS works like this: it runs a task a bit, and when the task
+schedules (or a scheduler tick happens) the task's CPU usage is "accounted
+for": the (small) time it just spent using the physical CPU is added to
+p->se.vruntime. Once p->se.vruntime gets high enough so that another task
+becomes the "leftmost task" of the time-ordered rbtree it maintains (plus a
+small amount of "granularity" distance relative to the leftmost task so that we
+do not over-schedule tasks and trash the cache), then the new leftmost task is
+picked and the current task is preempted.
+
+
+
+4. SOME FEATURES OF CFS
+
+CFS uses nanosecond granularity accounting and does not rely on any jiffies or
+other HZ detail. Thus the CFS scheduler has no notion of "timeslices" in the
+way the previous scheduler had, and has no heuristics whatsoever. There is
+only one central tunable (you have to switch on CONFIG_SCHED_DEBUG):
+
+ /proc/sys/kernel/sched_granularity_ns
+
+which can be used to tune the scheduler from "desktop" (i.e., low latencies) to
+"server" (i.e., good batching) workloads. It defaults to a setting suitable
+for desktop workloads. SCHED_BATCH is handled by the CFS scheduler module too.
+
+Due to its design, the CFS scheduler is not prone to any of the "attacks" that
+exist today against the heuristics of the stock scheduler: fiftyp.c, thud.c,
+chew.c, ring-test.c, massive_intr.c all work fine and do not impact
+interactivity and produce the expected behavior.
+
+The CFS scheduler has a much stronger handling of nice levels and SCHED_BATCH
+than the previous vanilla scheduler: both types of workloads are isolated much
+more aggressively.
+
+SMP load-balancing has been reworked/sanitized: the runqueue-walking
+assumptions are gone from the load-balancing code now, and iterators of the
+scheduling modules are used. The balancing code got quite a bit simpler as a
+result.
+
+
+
+5. Scheduling policies
+
+CFS implements three scheduling policies:
+
+ - SCHED_NORMAL (traditionally called SCHED_OTHER): The scheduling
+ policy that is used for regular tasks.
+
+ - SCHED_BATCH: Does not preempt nearly as often as regular tasks
+ would, thereby allowing tasks to run longer and make better use of
+ caches but at the cost of interactivity. This is well suited for
+ batch jobs.
+
+ - SCHED_IDLE: This is even weaker than nice 19, but its not a true
+ idle timer scheduler in order to avoid to get into priority
+ inversion problems which would deadlock the machine.
+
+SCHED_FIFO/_RR are implemented in sched_rt.c and are as specified by
+POSIX.
+
+The command chrt from util-linux-ng 2.13.1.1 can set all of these except
+SCHED_IDLE.
-Group scheduler tunables:
-When CONFIG_FAIR_USER_SCHED is defined, a directory is created in sysfs for
-each new user and a "cpu_share" file is added in that directory.
+
+6. SCHEDULING CLASSES
+
+The new CFS scheduler has been designed in such a way to introduce "Scheduling
+Classes," an extensible hierarchy of scheduler modules. These modules
+encapsulate scheduling policy details and are handled by the scheduler core
+without the core code assuming too much about them.
+
+sched_fair.c implements the CFS scheduler described above.
+
+sched_rt.c implements SCHED_FIFO and SCHED_RR semantics, in a simpler way than
+the previous vanilla scheduler did. It uses 100 runqueues (for all 100 RT
+priority levels, instead of 140 in the previous scheduler) and it needs no
+expired array.
+
+Scheduling classes are implemented through the sched_class structure, which
+contains hooks to functions that must be called whenever an interesting event
+occurs.
+
+This is the (partial) list of the hooks:
+
+ - enqueue_task(...)
+
+ Called when a task enters a runnable state.
+ It puts the scheduling entity (task) into the red-black tree and
+ increments the nr_running variable.
+
+ - dequeue_tree(...)
+
+ When a task is no longer runnable, this function is called to keep the
+ corresponding scheduling entity out of the red-black tree. It decrements
+ the nr_running variable.
+
+ - yield_task(...)
+
+ This function is basically just a dequeue followed by an enqueue, unless the
+ compat_yield sysctl is turned on; in that case, it places the scheduling
+ entity at the right-most end of the red-black tree.
+
+ - check_preempt_curr(...)
+
+ This function checks if a task that entered the runnable state should
+ preempt the currently running task.
+
+ - pick_next_task(...)
+
+ This function chooses the most appropriate task eligible to run next.
+
+ - set_curr_task(...)
+
+ This function is called when a task changes its scheduling class or changes
+ its task group.
+
+ - task_tick(...)
+
+ This function is mostly called from time tick functions; it might lead to
+ process switch. This drives the running preemption.
+
+ - task_new(...)
+
+ The core scheduler gives the scheduling module an opportunity to manage new
+ task startup. The CFS scheduling module uses it for group scheduling, while
+ the scheduling module for a real-time task does not use it.
+
+
+
+7. GROUP SCHEDULER EXTENSIONS TO CFS
+
+Normally, the scheduler operates on individual tasks and strives to provide
+fair CPU time to each task. Sometimes, it may be desirable to group tasks and
+provide fair CPU time to each such task group. For example, it may be
+desirable to first provide fair CPU time to each user on the system and then to
+each task belonging to a user.
+
+CONFIG_GROUP_SCHED strives to achieve exactly that. It lets tasks to be
+grouped and divides CPU time fairly among such groups.
+
+CONFIG_RT_GROUP_SCHED permits to group real-time (i.e., SCHED_FIFO and
+SCHED_RR) tasks.
+
+CONFIG_FAIR_GROUP_SCHED permits to group CFS (i.e., SCHED_NORMAL and
+SCHED_BATCH) tasks.
+
+At present, there are two (mutually exclusive) mechanisms to group tasks for
+CPU bandwidth control purposes:
+
+ - Based on user id (CONFIG_USER_SCHED)
+
+ With this option, tasks are grouped according to their user id.
+
+ - Based on "cgroup" pseudo filesystem (CONFIG_CGROUP_SCHED)
+
+ This options needs CONFIG_CGROUPS to be defined, and lets the administrator
+ create arbitrary groups of tasks, using the "cgroup" pseudo filesystem. See
+ Documentation/cgroups.txt for more information about this filesystem.
+
+Only one of these options to group tasks can be chosen and not both.
+
+When CONFIG_USER_SCHED is defined, a directory is created in sysfs for each new
+user and a "cpu_share" file is added in that directory.
# cd /sys/kernel/uids
# cat 512/cpu_share # Display user 512's CPU share
2048
#
-CPU bandwidth between two users are divided in the ratio of their CPU shares.
-For ex: if you would like user "root" to get twice the bandwidth of user
-"guest", then set the cpu_share for both the users such that "root"'s
-cpu_share is twice "guest"'s cpu_share
-
+CPU bandwidth between two users is divided in the ratio of their CPU shares.
+For example: if you would like user "root" to get twice the bandwidth of user
+"guest," then set the cpu_share for both the users such that "root"'s cpu_share
+is twice "guest"'s cpu_share.
-When CONFIG_FAIR_CGROUP_SCHED is defined, a "cpu.shares" file is created
-for each group created using the pseudo filesystem. See example steps
-below to create task groups and modify their CPU share using the "cgroups"
-pseudo filesystem
+When CONFIG_CGROUP_SCHED is defined, a "cpu.shares" file is created for each
+group created using the pseudo filesystem. See example steps below to create
+task groups and modify their CPU share using the "cgroups" pseudo filesystem.
# mkdir /dev/cpuctl
# mount -t cgroup -ocpu none /dev/cpuctl
45 -> Pinnacle PCTV DVB-T (em2870)
46 -> Compro, VideoMate U3 (em2870) [185b:2870]
47 -> KWorld DVB-T 305U (em2880) [eb1a:e305]
- 48 -> KWorld DVB-T 310U (em2880)
+ 48 -> KWorld DVB-T 310U (em2880) [eb1a:e310]
49 -> MSI DigiVox A/D (em2880) [eb1a:e310]
50 -> MSI DigiVox A/D II (em2880) [eb1a:e320]
51 -> Terratec Hybrid XS Secam (em2880) [0ccd:004c]
pac7311 093a:2621 PAC731x
pac7311 093a:2624 PAC7302
pac7311 093a:2626 Labtec 2200
+pac7311 093a:262a Webcam 300k
zc3xx 0ac8:0302 Z-star Vimicro zc0302
vc032x 0ac8:0321 Vimicro generic vc0321
vc032x 0ac8:0323 Vimicro Vc0323
S: Maintained
CPUSETS
-P: Paul Jackson
P: Paul Menage
-M: pj@sgi.com
M: menage@google.com
L: linux-kernel@vger.kernel.org
W: http://www.bullopensource.org/cpuset/
P: Michael Kerrisk
M: mtk.manpages@gmail.com
W: http://www.kernel.org/doc/man-pages
+L: linux-man@vger.kernel.org
S: Supported
MARVELL LIBERTAS WIRELESS DRIVER
VERSION = 2
PATCHLEVEL = 6
SUBLEVEL = 27
-EXTRAVERSION = -rc8
+EXTRAVERSION = -rc9
NAME = Rotary Wombat
# *DOCUMENTATION*
atomic_inc(&init_mm.mm_count);
current->active_mm = &init_mm;
+ /* inform the notifiers about the new cpu */
+ notify_cpu_starting(cpuid);
+
/* Must have completely accurate bogos. */
local_irq_enable();
/*
* Enable local interrupts.
*/
+ notify_cpu_starting(cpu);
local_irq_enable();
local_fiq_enable();
unmask_irq(IPI_INTR_VECT);
unmask_irq(TIMER0_INTR_VECT);
preempt_disable();
+ notify_cpu_starting(cpu);
local_irq_enable();
cpu_set(cpu, cpu_online_map);
spin_lock(&vector_lock);
/* Setup the per cpu irq handling data structures */
__setup_vector_irq(cpuid);
+ notify_cpu_starting(cpuid);
cpu_set(cpuid, cpu_online_map);
per_cpu(cpu_state, cpuid) = CPU_ONLINE;
spin_unlock(&vector_lock);
{
int cpu_id = smp_processor_id();
+ notify_cpu_starting(cpu_id);
+
local_irq_enable();
/* Get our bogomips. */
select SYS_SUPPORTS_64BIT_KERNEL
select SYS_SUPPORTS_BIG_ENDIAN
select SYS_SUPPORTS_LITTLE_ENDIAN
+ select SYS_SUPPORTS_MIPS_CMP if BROKEN # because SYNC_R4K is broken
select SYS_SUPPORTS_MULTITHREADING
select SYS_SUPPORTS_SMARTMIPS
help
depends on CPU_MIPS32_R2
#depends on CPU_MIPS64_R2 # once there is hardware ...
depends on SYS_SUPPORTS_MULTITHREADING
- select GENERIC_CLOCKEVENTS_BROADCAST
select CPU_MIPSR2_IRQ_VI
select CPU_MIPSR2_IRQ_EI
select MIPS_MT
Includes a loader for loading an elf relocatable object
onto another VPE and running it.
-config MIPS_MT_SMTC_INSTANT_REPLAY
- bool "Low-latency Dispatch of Deferred SMTC IPIs"
- depends on MIPS_MT_SMTC && !PREEMPT
- default y
- help
- SMTC pseudo-interrupts between TCs are deferred and queued
- if the target TC is interrupt-inhibited (IXMT). In the first
- SMTC prototypes, these queued IPIs were serviced on return
- to user mode, or on entry into the kernel idle loop. The
- INSTANT_REPLAY option dispatches them as part of local_irq_restore()
- processing, which adds runtime overhead (hence the option to turn
- it off), but ensures that IPIs are handled promptly even under
- heavy I/O interrupt load.
-
config MIPS_MT_SMTC_IM_BACKSTOP
bool "Use per-TC register bits as backstop for inhibited IM bits"
depends on MIPS_MT_SMTC
- default y
+ default n
help
To support multiple TC microthreads acting as "CPUs" within
a VPE, VPE-wide interrupt mask bits must be specially manipulated
during interrupt handling. To support legacy drivers and interrupt
controller management code, SMTC has a "backstop" to track and
if necessary restore the interrupt mask. This has some performance
- impact on interrupt service overhead. Disable it only if you know
- what you are doing.
+ impact on interrupt service overhead.
config MIPS_MT_SMTC_IRQAFF
bool "Support IRQ affinity API"
Enables SMP IRQ affinity API (/proc/irq/*/smp_affinity, etc.)
for SMTC Linux kernel. Requires platform support, of which
an example can be found in the MIPS kernel i8259 and Malta
- platform code. It is recommended that MIPS_MT_SMTC_INSTANT_REPLAY
- be enabled if MIPS_MT_SMTC_IRQAFF is used. Adds overhead to
- interrupt dispatch, and should be used only if you know what
- you are doing.
+ platform code. Adds some overhead to interrupt dispatch, and
+ should be used only if you know what you are doing.
config MIPS_VPE_LOADER_TOM
bool "Load VPE program into memory hidden from linux"
"exit" syscall notifying other kernel modules the SP program is
exiting. You probably want to say yes here.
+config MIPS_CMP
+ bool "MIPS CMP framework support"
+ depends on SYS_SUPPORTS_MIPS_CMP
+ select SYNC_R4K if BROKEN
+ select SYS_SUPPORTS_SMP
+ select SYS_SUPPORTS_SCHED_SMT if SMP
+ select WEAK_ORDERING
+ default n
+ help
+ This is a placeholder option for the GCMP work. It will need to
+ be handled differently...
+
config SB1_PASS_1_WORKAROUNDS
bool
depends on CPU_SB1_PASS_1
config SMP_UP
bool
+config SYS_SUPPORTS_MIPS_CMP
+ bool
+
config SYS_SUPPORTS_SMP
bool
performance should round up your number of processors to the next
power of two.
-config MIPS_CMP
- bool "MIPS CMP framework support"
- depends on SMP
- select SYNC_R4K
- select SYS_SUPPORTS_SCHED_SMT
- select WEAK_ORDERING
- default n
- help
- This is a placeholder option for the GCMP work. It will need to
- be handled differently...
-
source "kernel/time/Kconfig"
#
obj-$(CONFIG_CEVT_BCM1480) += cevt-bcm1480.o
obj-$(CONFIG_CEVT_R4K) += cevt-r4k.o
+obj-$(CONFIG_MIPS_MT_SMTC) += cevt-smtc.o
obj-$(CONFIG_CEVT_DS1287) += cevt-ds1287.o
obj-$(CONFIG_CEVT_GT641XX) += cevt-gt641xx.o
obj-$(CONFIG_CEVT_SB1250) += cevt-sb1250.o
#include <asm/smtc_ipi.h>
#include <asm/time.h>
+#include <asm/cevt-r4k.h>
+
+/*
+ * The SMTC Kernel for the 34K, 1004K, et. al. replaces several
+ * of these routines with SMTC-specific variants.
+ */
+
+#ifndef CONFIG_MIPS_MT_SMTC
static int mips_next_event(unsigned long delta,
struct clock_event_device *evt)
unsigned int cnt;
int res;
-#ifdef CONFIG_MIPS_MT_SMTC
- {
- unsigned long flags, vpflags;
- local_irq_save(flags);
- vpflags = dvpe();
-#endif
cnt = read_c0_count();
cnt += delta;
write_c0_compare(cnt);
res = ((int)(read_c0_count() - cnt) > 0) ? -ETIME : 0;
-#ifdef CONFIG_MIPS_MT_SMTC
- evpe(vpflags);
- local_irq_restore(flags);
- }
-#endif
return res;
}
-static void mips_set_mode(enum clock_event_mode mode,
- struct clock_event_device *evt)
+#endif /* CONFIG_MIPS_MT_SMTC */
+
+void mips_set_clock_mode(enum clock_event_mode mode,
+ struct clock_event_device *evt)
{
/* Nothing to do ... */
}
-static DEFINE_PER_CPU(struct clock_event_device, mips_clockevent_device);
-static int cp0_timer_irq_installed;
+DEFINE_PER_CPU(struct clock_event_device, mips_clockevent_device);
+int cp0_timer_irq_installed;
-/*
- * Timer ack for an R4k-compatible timer of a known frequency.
- */
-static void c0_timer_ack(void)
-{
- write_c0_compare(read_c0_compare());
-}
+#ifndef CONFIG_MIPS_MT_SMTC
-/*
- * Possibly handle a performance counter interrupt.
- * Return true if the timer interrupt should not be checked
- */
-static inline int handle_perf_irq(int r2)
-{
- /*
- * The performance counter overflow interrupt may be shared with the
- * timer interrupt (cp0_perfcount_irq < 0). If it is and a
- * performance counter has overflowed (perf_irq() == IRQ_HANDLED)
- * and we can't reliably determine if a counter interrupt has also
- * happened (!r2) then don't check for a timer interrupt.
- */
- return (cp0_perfcount_irq < 0) &&
- perf_irq() == IRQ_HANDLED &&
- !r2;
-}
-
-static irqreturn_t c0_compare_interrupt(int irq, void *dev_id)
+irqreturn_t c0_compare_interrupt(int irq, void *dev_id)
{
const int r2 = cpu_has_mips_r2;
struct clock_event_device *cd;
* interrupt. Being the paranoiacs we are we check anyway.
*/
if (!r2 || (read_c0_cause() & (1 << 30))) {
- c0_timer_ack();
-#ifdef CONFIG_MIPS_MT_SMTC
- if (cpu_data[cpu].vpe_id)
- goto out;
- cpu = 0;
-#endif
+ /* Clear Count/Compare Interrupt */
+ write_c0_compare(read_c0_compare());
cd = &per_cpu(mips_clockevent_device, cpu);
cd->event_handler(cd);
}
return IRQ_HANDLED;
}
-static struct irqaction c0_compare_irqaction = {
+#endif /* Not CONFIG_MIPS_MT_SMTC */
+
+struct irqaction c0_compare_irqaction = {
.handler = c0_compare_interrupt,
-#ifdef CONFIG_MIPS_MT_SMTC
- .flags = IRQF_DISABLED,
-#else
.flags = IRQF_DISABLED | IRQF_PERCPU,
-#endif
.name = "timer",
};
-#ifdef CONFIG_MIPS_MT_SMTC
-DEFINE_PER_CPU(struct clock_event_device, smtc_dummy_clockevent_device);
-
-static void smtc_set_mode(enum clock_event_mode mode,
- struct clock_event_device *evt)
-{
-}
-
-static void mips_broadcast(cpumask_t mask)
-{
- unsigned int cpu;
-
- for_each_cpu_mask(cpu, mask)
- smtc_send_ipi(cpu, SMTC_CLOCK_TICK, 0);
-}
-
-static void setup_smtc_dummy_clockevent_device(void)
-{
- //uint64_t mips_freq = mips_hpt_^frequency;
- unsigned int cpu = smp_processor_id();
- struct clock_event_device *cd;
- cd = &per_cpu(smtc_dummy_clockevent_device, cpu);
-
- cd->name = "SMTC";
- cd->features = CLOCK_EVT_FEAT_DUMMY;
-
- /* Calculate the min / max delta */
- cd->mult = 0; //div_sc((unsigned long) mips_freq, NSEC_PER_SEC, 32);
- cd->shift = 0; //32;
- cd->max_delta_ns = 0; //clockevent_delta2ns(0x7fffffff, cd);
- cd->min_delta_ns = 0; //clockevent_delta2ns(0x30, cd);
-
- cd->rating = 200;
- cd->irq = 17; //-1;
-// if (cpu)
-// cd->cpumask = CPU_MASK_ALL; // cpumask_of_cpu(cpu);
-// else
- cd->cpumask = cpumask_of_cpu(cpu);
-
- cd->set_mode = smtc_set_mode;
-
- cd->broadcast = mips_broadcast;
-
- clockevents_register_device(cd);
-}
-#endif
-
-static void mips_event_handler(struct clock_event_device *dev)
+void mips_event_handler(struct clock_event_device *dev)
{
}
return (read_c0_cause() >> cp0_compare_irq) & 0x100;
}
-static int c0_compare_int_usable(void)
+/*
+ * Compare interrupt can be routed and latched outside the core,
+ * so a single execution hazard barrier may not be enough to give
+ * it time to clear as seen in the Cause register. 4 time the
+ * pipeline depth seems reasonably conservative, and empirically
+ * works better in configurations with high CPU/bus clock ratios.
+ */
+
+#define compare_change_hazard() \
+ do { \
+ irq_disable_hazard(); \
+ irq_disable_hazard(); \
+ irq_disable_hazard(); \
+ irq_disable_hazard(); \
+ } while (0)
+
+int c0_compare_int_usable(void)
{
unsigned int delta;
unsigned int cnt;
*/
if (c0_compare_int_pending()) {
write_c0_compare(read_c0_count());
- irq_disable_hazard();
+ compare_change_hazard();
if (c0_compare_int_pending())
return 0;
}
cnt = read_c0_count();
cnt += delta;
write_c0_compare(cnt);
- irq_disable_hazard();
+ compare_change_hazard();
if ((int)(read_c0_count() - cnt) < 0)
break;
/* increase delta if the timer was already expired */
while ((int)(read_c0_count() - cnt) <= 0)
; /* Wait for expiry */
+ compare_change_hazard();
if (!c0_compare_int_pending())
return 0;
write_c0_compare(read_c0_count());
- irq_disable_hazard();
+ compare_change_hazard();
if (c0_compare_int_pending())
return 0;
return 1;
}
+#ifndef CONFIG_MIPS_MT_SMTC
+
int __cpuinit mips_clockevent_init(void)
{
uint64_t mips_freq = mips_hpt_frequency;
if (!cpu_has_counter || !mips_hpt_frequency)
return -ENXIO;
-#ifdef CONFIG_MIPS_MT_SMTC
- setup_smtc_dummy_clockevent_device();
-
- /*
- * On SMTC we only register VPE0's compare interrupt as clockevent
- * device.
- */
- if (cpu)
- return 0;
-#endif
-
if (!c0_compare_int_usable())
return -ENXIO;
cd->rating = 300;
cd->irq = irq;
-#ifdef CONFIG_MIPS_MT_SMTC
- cd->cpumask = CPU_MASK_ALL;
-#else
cd->cpumask = cpumask_of_cpu(cpu);
-#endif
cd->set_next_event = mips_next_event;
- cd->set_mode = mips_set_mode;
+ cd->set_mode = mips_set_clock_mode;
cd->event_handler = mips_event_handler;
clockevents_register_device(cd);
cp0_timer_irq_installed = 1;
-#ifdef CONFIG_MIPS_MT_SMTC
-#define CPUCTR_IMASKBIT (0x100 << cp0_compare_irq)
- setup_irq_smtc(irq, &c0_compare_irqaction, CPUCTR_IMASKBIT);
-#else
setup_irq(irq, &c0_compare_irqaction);
-#endif
return 0;
}
+
+#endif /* Not CONFIG_MIPS_MT_SMTC */
--- /dev/null
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2007 MIPS Technologies, Inc.
+ * Copyright (C) 2007 Ralf Baechle <ralf@linux-mips.org>
+ * Copyright (C) 2008 Kevin D. Kissell, Paralogos sarl
+ */
+#include <linux/clockchips.h>
+#include <linux/interrupt.h>
+#include <linux/percpu.h>
+
+#include <asm/smtc_ipi.h>
+#include <asm/time.h>
+#include <asm/cevt-r4k.h>
+
+/*
+ * Variant clock event timer support for SMTC on MIPS 34K, 1004K
+ * or other MIPS MT cores.
+ *
+ * Notes on SMTC Support:
+ *
+ * SMTC has multiple microthread TCs pretending to be Linux CPUs.
+ * But there's only one Count/Compare pair per VPE, and Compare
+ * interrupts are taken opportunisitically by available TCs
+ * bound to the VPE with the Count register. The new timer
+ * framework provides for global broadcasts, but we really
+ * want VPE-level multicasts for best behavior. So instead
+ * of invoking the high-level clock-event broadcast code,
+ * this version of SMTC support uses the historical SMTC
+ * multicast mechanisms "under the hood", appearing to the
+ * generic clock layer as if the interrupts are per-CPU.
+ *
+ * The approach taken here is to maintain a set of NR_CPUS
+ * virtual timers, and track which "CPU" needs to be alerted
+ * at each event.
+ *
+ * It's unlikely that we'll see a MIPS MT core with more than
+ * 2 VPEs, but we *know* that we won't need to handle more
+ * VPEs than we have "CPUs". So NCPUs arrays of NCPUs elements
+ * is always going to be overkill, but always going to be enough.
+ */
+
+unsigned long smtc_nexttime[NR_CPUS][NR_CPUS];
+static int smtc_nextinvpe[NR_CPUS];
+
+/*
+ * Timestamps stored are absolute values to be programmed
+ * into Count register. Valid timestamps will never be zero.
+ * If a Zero Count value is actually calculated, it is converted
+ * to be a 1, which will introduce 1 or two CPU cycles of error
+ * roughly once every four billion events, which at 1000 HZ means
+ * about once every 50 days. If that's actually a problem, one
+ * could alternate squashing 0 to 1 and to -1.
+ */
+
+#define MAKEVALID(x) (((x) == 0L) ? 1L : (x))
+#define ISVALID(x) ((x) != 0L)
+
+/*
+ * Time comparison is subtle, as it's really truncated
+ * modular arithmetic.
+ */
+
+#define IS_SOONER(a, b, reference) \
+ (((a) - (unsigned long)(reference)) < ((b) - (unsigned long)(reference)))
+
+/*
+ * CATCHUP_INCREMENT, used when the function falls behind the counter.
+ * Could be an increasing function instead of a constant;
+ */
+
+#define CATCHUP_INCREMENT 64
+
+static int mips_next_event(unsigned long delta,
+ struct clock_event_device *evt)
+{
+ unsigned long flags;
+ unsigned int mtflags;
+ unsigned long timestamp, reference, previous;
+ unsigned long nextcomp = 0L;
+ int vpe = current_cpu_data.vpe_id;
+ int cpu = smp_processor_id();
+ local_irq_save(flags);
+ mtflags = dmt();
+
+ /*
+ * Maintain the per-TC virtual timer
+ * and program the per-VPE shared Count register
+ * as appropriate here...
+ */
+ reference = (unsigned long)read_c0_count();
+ timestamp = MAKEVALID(reference + delta);
+ /*
+ * To really model the clock, we have to catch the case
+ * where the current next-in-VPE timestamp is the old
+ * timestamp for the calling CPE, but the new value is
+ * in fact later. In that case, we have to do a full
+ * scan and discover the new next-in-VPE CPU id and
+ * timestamp.
+ */
+ previous = smtc_nexttime[vpe][cpu];
+ if (cpu == smtc_nextinvpe[vpe] && ISVALID(previous)
+ && IS_SOONER(previous, timestamp, reference)) {
+ int i;
+ int soonest = cpu;
+
+ /*
+ * Update timestamp array here, so that new
+ * value gets considered along with those of
+ * other virtual CPUs on the VPE.
+ */
+ smtc_nexttime[vpe][cpu] = timestamp;
+ for_each_online_cpu(i) {
+ if (ISVALID(smtc_nexttime[vpe][i])
+ && IS_SOONER(smtc_nexttime[vpe][i],
+ smtc_nexttime[vpe][soonest], reference)) {
+ soonest = i;
+ }
+ }
+ smtc_nextinvpe[vpe] = soonest;
+ nextcomp = smtc_nexttime[vpe][soonest];
+ /*
+ * Otherwise, we don't have to process the whole array rank,
+ * we just have to see if the event horizon has gotten closer.
+ */
+ } else {
+ if (!ISVALID(smtc_nexttime[vpe][smtc_nextinvpe[vpe]]) ||
+ IS_SOONER(timestamp,
+ smtc_nexttime[vpe][smtc_nextinvpe[vpe]], reference)) {
+ smtc_nextinvpe[vpe] = cpu;
+ nextcomp = timestamp;
+ }
+ /*
+ * Since next-in-VPE may me the same as the executing
+ * virtual CPU, we update the array *after* checking
+ * its value.
+ */
+ smtc_nexttime[vpe][cpu] = timestamp;
+ }
+
+ /*
+ * It may be that, in fact, we don't need to update Compare,
+ * but if we do, we want to make sure we didn't fall into
+ * a crack just behind Count.
+ */
+ if (ISVALID(nextcomp)) {
+ write_c0_compare(nextcomp);
+ ehb();
+ /*
+ * We never return an error, we just make sure
+ * that we trigger the handlers as quickly as
+ * we can if we fell behind.
+ */
+ while ((nextcomp - (unsigned long)read_c0_count())
+ > (unsigned long)LONG_MAX) {
+ nextcomp += CATCHUP_INCREMENT;
+ write_c0_compare(nextcomp);
+ ehb();
+ }
+ }
+ emt(mtflags);
+ local_irq_restore(flags);
+ return 0;
+}
+
+
+void smtc_distribute_timer(int vpe)
+{
+ unsigned long flags;
+ unsigned int mtflags;
+ int cpu;
+ struct clock_event_device *cd;
+ unsigned long nextstamp = 0L;
+ unsigned long reference;
+
+
+repeat:
+ for_each_online_cpu(cpu) {
+ /*
+ * Find virtual CPUs within the current VPE who have
+ * unserviced timer requests whose time is now past.
+ */
+ local_irq_save(flags);
+ mtflags = dmt();
+ if (cpu_data[cpu].vpe_id == vpe &&
+ ISVALID(smtc_nexttime[vpe][cpu])) {
+ reference = (unsigned long)read_c0_count();
+ if ((smtc_nexttime[vpe][cpu] - reference)
+ > (unsigned long)LONG_MAX) {
+ smtc_nexttime[vpe][cpu] = 0L;
+ emt(mtflags);
+ local_irq_restore(flags);
+ /*
+ * We don't send IPIs to ourself.
+ */
+ if (cpu != smp_processor_id()) {
+ smtc_send_ipi(cpu, SMTC_CLOCK_TICK, 0);
+ } else {
+ cd = &per_cpu(mips_clockevent_device, cpu);
+ cd->event_handler(cd);
+ }
+ } else {
+ /* Local to VPE but Valid Time not yet reached. */
+ if (!ISVALID(nextstamp) ||
+ IS_SOONER(smtc_nexttime[vpe][cpu], nextstamp,
+ reference)) {
+ smtc_nextinvpe[vpe] = cpu;
+ nextstamp = smtc_nexttime[vpe][cpu];
+ }
+ emt(mtflags);
+ local_irq_restore(flags);
+ }
+ } else {
+ emt(mtflags);
+ local_irq_restore(flags);
+
+ }
+ }
+ /* Reprogram for interrupt at next soonest timestamp for VPE */
+ if (ISVALID(nextstamp)) {
+ write_c0_compare(nextstamp);
+ ehb();
+ if ((nextstamp - (unsigned long)read_c0_count())
+ > (unsigned long)LONG_MAX)
+ goto repeat;
+ }
+}
+
+
+irqreturn_t c0_compare_interrupt(int irq, void *dev_id)
+{
+ int cpu = smp_processor_id();
+
+ /* If we're running SMTC, we've got MIPS MT and therefore MIPS32R2 */
+ handle_perf_irq(1);
+
+ if (read_c0_cause() & (1 << 30)) {
+ /* Clear Count/Compare Interrupt */
+ write_c0_compare(read_c0_compare());
+ smtc_distribute_timer(cpu_data[cpu].vpe_id);
+ }
+ return IRQ_HANDLED;
+}
+
+
+int __cpuinit mips_clockevent_init(void)
+{
+ uint64_t mips_freq = mips_hpt_frequency;
+ unsigned int cpu = smp_processor_id();
+ struct clock_event_device *cd;
+ unsigned int irq;
+ int i;
+ int j;
+
+ if (!cpu_has_counter || !mips_hpt_frequency)
+ return -ENXIO;
+ if (cpu == 0) {
+ for (i = 0; i < num_possible_cpus(); i++) {
+ smtc_nextinvpe[i] = 0;
+ for (j = 0; j < num_possible_cpus(); j++)
+ smtc_nexttime[i][j] = 0L;
+ }
+ /*
+ * SMTC also can't have the usablility test
+ * run by secondary TCs once Compare is in use.
+ */
+ if (!c0_compare_int_usable())
+ return -ENXIO;
+ }
+
+ /*
+ * With vectored interrupts things are getting platform specific.
+ * get_c0_compare_int is a hook to allow a platform to return the
+ * interrupt number of it's liking.
+ */
+ irq = MIPS_CPU_IRQ_BASE + cp0_compare_irq;
+ if (get_c0_compare_int)
+ irq = get_c0_compare_int();
+
+ cd = &per_cpu(mips_clockevent_device, cpu);
+
+ cd->name = "MIPS";
+ cd->features = CLOCK_EVT_FEAT_ONESHOT;
+
+ /* Calculate the min / max delta */
+ cd->mult = div_sc((unsigned long) mips_freq, NSEC_PER_SEC, 32);
+ cd->shift = 32;
+ cd->max_delta_ns = clockevent_delta2ns(0x7fffffff, cd);
+ cd->min_delta_ns = clockevent_delta2ns(0x300, cd);
+
+ cd->rating = 300;
+ cd->irq = irq;
+ cd->cpumask = cpumask_of_cpu(cpu);
+ cd->set_next_event = mips_next_event;
+ cd->set_mode = mips_set_clock_mode;
+ cd->event_handler = mips_event_handler;
+
+ clockevents_register_device(cd);
+
+ /*
+ * On SMTC we only want to do the data structure
+ * initialization and IRQ setup once.
+ */
+ if (cpu)
+ return 0;
+ /*
+ * And we need the hwmask associated with the c0_compare
+ * vector to be initialized.
+ */
+ irq_hwmask[irq] = (0x100 << cp0_compare_irq);
+ if (cp0_timer_irq_installed)
+ return 0;
+
+ cp0_timer_irq_installed = 1;
+
+ setup_irq(irq, &c0_compare_irqaction);
+
+ return 0;
+}
* interrupt is requested" restriction in the MIPS32/MIPS64 architecture makes
* using this version a gamble.
*/
-static void r4k_wait_irqoff(void)
+void r4k_wait_irqoff(void)
{
local_irq_disable();
if (!need_resched())
- __asm__(" .set mips3 \n"
+ __asm__(" .set push \n"
+ " .set mips3 \n"
" wait \n"
- " .set mips0 \n");
+ " .set pop \n");
local_irq_enable();
+ __asm__(" .globl __pastwait \n"
+ "__pastwait: \n");
+ return;
}
/*
FEXPORT(restore_all) # restore full frame
#ifdef CONFIG_MIPS_MT_SMTC
-/* Detect and execute deferred IPI "interrupts" */
- LONG_L s0, TI_REGS($28)
- LONG_S sp, TI_REGS($28)
- jal deferred_smtc_ipi
- LONG_S s0, TI_REGS($28)
#ifdef CONFIG_MIPS_MT_SMTC_IM_BACKSTOP
/* Re-arm any temporarily masked interrupts not explicitly "acked" */
mfc0 v0, CP0_TCSTATUS
xor t0, t0, t3
mtc0 t0, CP0_TCCONTEXT
#endif /* CONFIG_MIPS_MT_SMTC_IM_BACKSTOP */
+/* Detect and execute deferred IPI "interrupts" */
+ LONG_L s0, TI_REGS($28)
+ LONG_S sp, TI_REGS($28)
+ jal deferred_smtc_ipi
+ LONG_S s0, TI_REGS($28)
#endif /* CONFIG_MIPS_MT_SMTC */
.set noat
RESTORE_TEMP
and t0, a0, t1
#ifdef CONFIG_MIPS_MT_SMTC_IM_BACKSTOP
mfc0 t2, CP0_TCCONTEXT
- or t0, t0, t2
- mtc0 t0, CP0_TCCONTEXT
+ or t2, t0, t2
+ mtc0 t2, CP0_TCCONTEXT
#endif /* CONFIG_MIPS_MT_SMTC_IM_BACKSTOP */
xor t1, t1, t0
mtc0 t1, CP0_STATUS
#include <asm/irqflags.h>
#include <asm/regdef.h>
#include <asm/page.h>
+#include <asm/pgtable-bits.h>
#include <asm/mipsregs.h>
#include <asm/stackframe.h>
/*
* FPU Use Factor empirically derived from experiments on 34K
*/
-#define FPUSEFACTOR 333
+#define FPUSEFACTOR 2000
static __init int mt_fp_affinity_init(void)
{
while (1) {
tick_nohz_stop_sched_tick(1);
while (!need_resched()) {
-#ifdef CONFIG_SMTC_IDLE_HOOK_DEBUG
+#ifdef CONFIG_MIPS_MT_SMTC
extern void smtc_idle_loop_hook(void);
smtc_idle_loop_hook();
*/
p->thread.cp0_status = read_c0_status() & ~(ST0_CU2|ST0_CU1);
childregs->cp0_status &= ~(ST0_CU2|ST0_CU1);
+
+#ifdef CONFIG_MIPS_MT_SMTC
+ /*
+ * SMTC restores TCStatus after Status, and the CU bits
+ * are aliased there.
+ */
+ childregs->cp0_tcstatus &= ~(ST0_CU2|ST0_CU1);
+#endif
clear_tsk_thread_flag(p, TIF_USEDFPU);
#ifdef CONFIG_MIPS_MT_FPAFF
clear_tsk_thread_flag(p, TIF_FPUBOUND);
-
- /*
- * FPU affinity support is cleaner if we track the
- * user-visible CPU affinity from the very beginning.
- * The generic cpus_allowed mask will already have
- * been copied from the parent before copy_thread
- * is invoked.
- */
- p->thread.user_cpus_allowed = p->cpus_allowed;
#endif /* CONFIG_MIPS_MT_FPAFF */
if (clone_flags & CLONE_SETTLS)
case FPC_EIR: { /* implementation / version register */
unsigned int flags;
#ifdef CONFIG_MIPS_MT_SMTC
- unsigned int irqflags;
+ unsigned long irqflags;
unsigned int mtflags;
#endif /* CONFIG_MIPS_MT_SMTC */
cpu = smp_processor_id();
cpu_data[cpu].udelay_val = loops_per_jiffy;
+ notify_cpu_starting(cpu);
+
mp_ops->smp_finish();
set_cpu_sibling_map(cpu);
-/* Copyright (C) 2004 Mips Technologies, Inc */
+/*
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright (C) 2004 Mips Technologies, Inc
+ * Copyright (C) 2008 Kevin D. Kissell
+ */
#include <linux/clockchips.h>
#include <linux/kernel.h>
#include <asm/time.h>
#include <asm/addrspace.h>
#include <asm/smtc.h>
-#include <asm/smtc_ipi.h>
#include <asm/smtc_proc.h>
/*
asiduse smtc_live_asid[MAX_SMTC_TLBS][MAX_SMTC_ASIDS];
-/*
- * Clock interrupt "latch" buffers, per "CPU"
- */
-
-static atomic_t ipi_timer_latch[NR_CPUS];
/*
* Number of InterProcessor Interrupt (IPI) message buffers to allocate
#define IPIBUF_PER_CPU 4
-static struct smtc_ipi_q IPIQ[NR_CPUS];
+struct smtc_ipi_q IPIQ[NR_CPUS];
static struct smtc_ipi_q freeIPIq;
* phys_cpu_present_map and the logical/physical mappings.
*/
-int __init mipsmt_build_cpu_map(int start_cpu_slot)
+int __init smtc_build_cpu_map(int start_cpu_slot)
{
int i, ntcs;
write_tc_c0_tcstatus((read_tc_c0_tcstatus()
& ~(TCSTATUS_TKSU | TCSTATUS_DA | TCSTATUS_IXMT))
| TCSTATUS_A);
- write_tc_c0_tccontext(0);
+ /*
+ * TCContext gets an offset from the base of the IPIQ array
+ * to be used in low-level code to detect the presence of
+ * an active IPI queue
+ */
+ write_tc_c0_tccontext((sizeof(struct smtc_ipi_q) * cpu) << 16);
/* Bind tc to vpe */
write_tc_c0_tcbind(vpe);
/* In general, all TCs should have the same cpu_data indications */
cpu_data[cpu].options &= ~MIPS_CPU_FPU;
cpu_data[cpu].vpe_id = vpe;
cpu_data[cpu].tc_id = tc;
+ /* Multi-core SMTC hasn't been tested, but be prepared */
+ cpu_data[cpu].core = (read_vpe_c0_ebase() >> 1) & 0xff;
}
+/*
+ * Tweak to get Count registes in as close a sync as possible.
+ * Value seems good for 34K-class cores.
+ */
+
+#define CP0_SKEW 8
-void mipsmt_prepare_cpus(void)
+void smtc_prepare_cpus(int cpus)
{
int i, vpe, tc, ntc, nvpe, tcpervpe[NR_CPUS], slop, cpu;
unsigned long flags;
IPIQ[i].head = IPIQ[i].tail = NULL;
spin_lock_init(&IPIQ[i].lock);
IPIQ[i].depth = 0;
- atomic_set(&ipi_timer_latch[i], 0);
}
/* cpu_data index starts at zero */
cpu = 0;
cpu_data[cpu].vpe_id = 0;
cpu_data[cpu].tc_id = 0;
+ cpu_data[cpu].core = (read_c0_ebase() >> 1) & 0xff;
cpu++;
/* Report on boot-time options */
write_vpe_c0_compare(0);
/* Propagate Config7 */
write_vpe_c0_config7(read_c0_config7());
- write_vpe_c0_count(read_c0_count());
+ write_vpe_c0_count(read_c0_count() + CP0_SKEW);
+ ehb();
}
/* enable multi-threading within VPE */
write_vpe_c0_vpecontrol(read_vpe_c0_vpecontrol() | VPECONTROL_TE);
void __cpuinit smtc_boot_secondary(int cpu, struct task_struct *idle)
{
extern u32 kernelsp[NR_CPUS];
- long flags;
+ unsigned long flags;
int mtflags;
LOCK_MT_PRA();
void smtc_init_secondary(void)
{
- /*
- * Start timer on secondary VPEs if necessary.
- * plat_timer_setup has already have been invoked by init/main
- * on "boot" TC. Like per_cpu_trap_init() hack, this assumes that
- * SMTC init code assigns TCs consdecutively and in ascending order
- * to across available VPEs.
- */
- if (((read_c0_tcbind() & TCBIND_CURTC) != 0) &&
- ((read_c0_tcbind() & TCBIND_CURVPE)
- != cpu_data[smp_processor_id() - 1].vpe_id)){
- write_c0_compare(read_c0_count() + mips_hpt_frequency/HZ);
- }
-
local_irq_enable();
}
void smtc_smp_finish(void)
{
+ int cpu = smp_processor_id();
+
+ /*
+ * Lowest-numbered CPU per VPE starts a clock tick.
+ * Like per_cpu_trap_init() hack, this assumes that
+ * SMTC init code assigns TCs consdecutively and
+ * in ascending order across available VPEs.
+ */
+ if (cpu > 0 && (cpu_data[cpu].vpe_id != cpu_data[cpu - 1].vpe_id))
+ write_c0_compare(read_c0_count() + mips_hpt_frequency/HZ);
+
printk("TC %d going on-line as CPU %d\n",
cpu_data[smp_processor_id()].tc_id, smp_processor_id());
}
{
int tcstatus;
struct smtc_ipi *pipi;
- long flags;
+ unsigned long flags;
int mtflags;
+ unsigned long tcrestart;
+ extern void r4k_wait_irqoff(void), __pastwait(void);
if (cpu == smp_processor_id()) {
printk("Cannot Send IPI to self!\n");
pipi->arg = (void *)action;
pipi->dest = cpu;
if (cpu_data[cpu].vpe_id != cpu_data[smp_processor_id()].vpe_id) {
- if (type == SMTC_CLOCK_TICK)
- atomic_inc(&ipi_timer_latch[cpu]);
/* If not on same VPE, enqueue and send cross-VPE interrupt */
smtc_ipi_nq(&IPIQ[cpu], pipi);
LOCK_CORE_PRA();
if ((tcstatus & TCSTATUS_IXMT) != 0) {
/*
- * Spin-waiting here can deadlock,
- * so we queue the message for the target TC.
+ * If we're in the the irq-off version of the wait
+ * loop, we need to force exit from the wait and
+ * do a direct post of the IPI.
+ */
+ if (cpu_wait == r4k_wait_irqoff) {
+ tcrestart = read_tc_c0_tcrestart();
+ if (tcrestart >= (unsigned long)r4k_wait_irqoff
+ && tcrestart < (unsigned long)__pastwait) {
+ write_tc_c0_tcrestart(__pastwait);
+ tcstatus &= ~TCSTATUS_IXMT;
+ write_tc_c0_tcstatus(tcstatus);
+ goto postdirect;
+ }
+ }
+ /*
+ * Otherwise we queue the message for the target TC
+ * to pick up when he does a local_irq_restore()
*/
write_tc_c0_tchalt(0);
UNLOCK_CORE_PRA();
- /* Try to reduce redundant timer interrupt messages */
- if (type == SMTC_CLOCK_TICK) {
- if (atomic_postincrement(&ipi_timer_latch[cpu])!=0){
- smtc_ipi_nq(&freeIPIq, pipi);
- return;
- }
- }
smtc_ipi_nq(&IPIQ[cpu], pipi);
} else {
- if (type == SMTC_CLOCK_TICK)
- atomic_inc(&ipi_timer_latch[cpu]);
+postdirect:
post_direct_ipi(cpu, pipi);
write_tc_c0_tchalt(0);
UNLOCK_CORE_PRA();
smp_call_function_interrupt();
}
-DECLARE_PER_CPU(struct clock_event_device, smtc_dummy_clockevent_device);
+DECLARE_PER_CPU(struct clock_event_device, mips_clockevent_device);
void ipi_decode(struct smtc_ipi *pipi)
{
struct clock_event_device *cd;
void *arg_copy = pipi->arg;
int type_copy = pipi->type;
- int ticks;
-
smtc_ipi_nq(&freeIPIq, pipi);
switch (type_copy) {
case SMTC_CLOCK_TICK:
irq_enter();
kstat_this_cpu.irqs[MIPS_CPU_IRQ_BASE + 1]++;
- cd = &per_cpu(smtc_dummy_clockevent_device, cpu);
- ticks = atomic_read(&ipi_timer_latch[cpu]);
- atomic_sub(ticks, &ipi_timer_latch[cpu]);
- while (ticks) {
- cd->event_handler(cd);
- ticks--;
- }
+ cd = &per_cpu(mips_clockevent_device, cpu);
+ cd->event_handler(cd);
irq_exit();
break;
}
}
+/*
+ * Similar to smtc_ipi_replay(), but invoked from context restore,
+ * so it reuses the current exception frame rather than set up a
+ * new one with self_ipi.
+ */
+
void deferred_smtc_ipi(void)
{
- struct smtc_ipi *pipi;
- unsigned long flags;
-/* DEBUG */
- int q = smp_processor_id();
+ int cpu = smp_processor_id();
/*
* Test is not atomic, but much faster than a dequeue,
* and the vast majority of invocations will have a null queue.
+ * If irq_disabled when this was called, then any IPIs queued
+ * after we test last will be taken on the next irq_enable/restore.
+ * If interrupts were enabled, then any IPIs added after the
+ * last test will be taken directly.
*/
- if (IPIQ[q].head != NULL) {
- while((pipi = smtc_ipi_dq(&IPIQ[q])) != NULL) {
- /* ipi_decode() should be called with interrupts off */
- local_irq_save(flags);
+
+ while (IPIQ[cpu].head != NULL) {
+ struct smtc_ipi_q *q = &IPIQ[cpu];
+ struct smtc_ipi *pipi;
+ unsigned long flags;
+
+ /*
+ * It may be possible we'll come in with interrupts
+ * already enabled.
+ */
+ local_irq_save(flags);
+
+ spin_lock(&q->lock);
+ pipi = __smtc_ipi_dq(q);
+ spin_unlock(&q->lock);
+ if (pipi != NULL)
ipi_decode(pipi);
- local_irq_restore(flags);
- }
+ /*
+ * The use of the __raw_local restore isn't
+ * as obviously necessary here as in smtc_ipi_replay(),
+ * but it's more efficient, given that we're already
+ * running down the IPI queue.
+ */
+ __raw_local_irq_restore(flags);
}
}
struct smtc_ipi *pipi;
unsigned long tcstatus;
int sent;
- long flags;
+ unsigned long flags;
unsigned int mtflags;
unsigned int vpflags;
/*
* SMTC-specific hacks invoked from elsewhere in the kernel.
- *
- * smtc_ipi_replay is called from raw_local_irq_restore which is only ever
- * called with interrupts disabled. We do rely on interrupts being disabled
- * here because using spin_lock_irqsave()/spin_unlock_irqrestore() would
- * result in a recursive call to raw_local_irq_restore().
*/
-static void __smtc_ipi_replay(void)
+ /*
+ * smtc_ipi_replay is called from raw_local_irq_restore
+ */
+
+void smtc_ipi_replay(void)
{
unsigned int cpu = smp_processor_id();
/*
* To the extent that we've ever turned interrupts off,
* we may have accumulated deferred IPIs. This is subtle.
- * If we use the smtc_ipi_qdepth() macro, we'll get an
- * exact number - but we'll also disable interrupts
- * and create a window of failure where a new IPI gets
- * queued after we test the depth but before we re-enable
- * interrupts. So long as IXMT never gets set, however,
* we should be OK: If we pick up something and dispatch
* it here, that's great. If we see nothing, but concurrent
* with this operation, another TC sends us an IPI, IXMT
* is clear, and we'll handle it as a real pseudo-interrupt
- * and not a pseudo-pseudo interrupt.
+ * and not a pseudo-pseudo interrupt. The important thing
+ * is to do the last check for queued message *after* the
+ * re-enabling of interrupts.
*/
- if (IPIQ[cpu].depth > 0) {
- while (1) {
- struct smtc_ipi_q *q = &IPIQ[cpu];
- struct smtc_ipi *pipi;
- extern void self_ipi(struct smtc_ipi *);
-
- spin_lock(&q->lock);
- pipi = __smtc_ipi_dq(q);
- spin_unlock(&q->lock);
- if (!pipi)
- break;
+ while (IPIQ[cpu].head != NULL) {
+ struct smtc_ipi_q *q = &IPIQ[cpu];
+ struct smtc_ipi *pipi;
+ unsigned long flags;
+
+ /*
+ * It's just possible we'll come in with interrupts
+ * already enabled.
+ */
+ local_irq_save(flags);
+
+ spin_lock(&q->lock);
+ pipi = __smtc_ipi_dq(q);
+ spin_unlock(&q->lock);
+ /*
+ ** But use a raw restore here to avoid recursion.
+ */
+ __raw_local_irq_restore(flags);
+ if (pipi) {
self_ipi(pipi);
smtc_cpu_stats[cpu].selfipis++;
}
}
}
-void smtc_ipi_replay(void)
-{
- raw_local_irq_disable();
- __smtc_ipi_replay();
-}
-
EXPORT_SYMBOL(smtc_ipi_replay);
void smtc_idle_loop_hook(void)
}
}
- /*
- * Now that we limit outstanding timer IPIs, check for hung TC
- */
- for (tc = 0; tc < NR_CPUS; tc++) {
- /* Don't check ourself - we'll dequeue IPIs just below */
- if ((tc != smp_processor_id()) &&
- atomic_read(&ipi_timer_latch[tc]) > timerq_limit) {
- if (clock_hang_reported[tc] == 0) {
- pdb_msg += sprintf(pdb_msg,
- "TC %d looks hung with timer latch at %d\n",
- tc, atomic_read(&ipi_timer_latch[tc]));
- clock_hang_reported[tc]++;
- }
- }
- }
emt(mtflags);
local_irq_restore(flags);
if (pdb_msg != &id_ho_db_msg[0])
printk("CPU%d: %s", smp_processor_id(), id_ho_db_msg);
#endif /* CONFIG_SMTC_IDLE_HOOK_DEBUG */
- /*
- * Replay any accumulated deferred IPIs. If "Instant Replay"
- * is in use, there should never be any.
- */
-#ifndef CONFIG_MIPS_MT_SMTC_INSTANT_REPLAY
- {
- unsigned long flags;
-
- local_irq_save(flags);
- __smtc_ipi_replay();
- local_irq_restore(flags);
- }
-#endif /* CONFIG_MIPS_MT_SMTC_INSTANT_REPLAY */
+ smtc_ipi_replay();
}
void smtc_soft_dump(void)
printk("%d: %ld\n", i, smtc_cpu_stats[i].selfipis);
}
smtc_ipi_qdump();
- printk("Timer IPI Backlogs:\n");
- for (i=0; i < NR_CPUS; i++) {
- printk("%d: %d\n", i, atomic_read(&ipi_timer_latch[i]));
- }
printk("%d Recoveries of \"stolen\" FPU\n",
atomic_read(&smtc_fpu_recoveries));
}
if (cpus_intersects(current->cpus_allowed, mt_fpu_cpumask)) {
cpumask_t tmask;
- cpus_and(tmask, current->thread.user_cpus_allowed,
- mt_fpu_cpumask);
+ current->thread.user_cpus_allowed
+ = current->cpus_allowed;
+ cpus_and(tmask, current->cpus_allowed,
+ mt_fpu_cpumask);
set_cpus_allowed(current, tmask);
set_thread_flag(TIF_FPUBOUND);
}
obj-$(CONFIG_PCI) += malta-pci.o
# FIXME FIXME FIXME
-obj-$(CONFIG_MIPS_MT_SMTC) += malta_smtc.o
+obj-$(CONFIG_MIPS_MT_SMTC) += malta-smtc.o
EXTRA_CFLAGS += -Werror
static void __init msmtc_smp_setup(void)
{
- mipsmt_build_cpu_map(0);
+ /*
+ * we won't get the definitive value until
+ * we've run smtc_prepare_cpus later, but
+ * we would appear to need an upper bound now.
+ */
+ smp_num_siblings = smtc_build_cpu_map(0);
}
static void __init msmtc_prepare_cpus(unsigned int max_cpus)
{
- mipsmt_prepare_cpus();
+ smtc_prepare_cpus(max_cpus);
}
struct plat_smp_ops msmtc_smp_ops = {
-obj-y := setup.o rtc_xicor1241.o rtc_m41t81.o
+obj-y := platform.o setup.o rtc_xicor1241.o \
+ rtc_m41t81.o
obj-$(CONFIG_I2C_BOARDINFO) += swarm-i2c.o
--- /dev/null
+#include <linux/err.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/platform_device.h>
+#include <linux/ata_platform.h>
+
+#include <asm/sibyte/board.h>
+#include <asm/sibyte/sb1250_genbus.h>
+#include <asm/sibyte/sb1250_regs.h>
+
+#define DRV_NAME "pata-swarm"
+
+#define SWARM_IDE_SHIFT 5
+#define SWARM_IDE_BASE 0x1f0
+#define SWARM_IDE_CTRL 0x3f6
+
+static struct resource swarm_pata_resource[] = {
+ {
+ .name = "Swarm GenBus IDE",
+ .flags = IORESOURCE_MEM,
+ }, {
+ .name = "Swarm GenBus IDE",
+ .flags = IORESOURCE_MEM,
+ }, {
+ .name = "Swarm GenBus IDE",
+ .flags = IORESOURCE_IRQ,
+ .start = K_INT_GB_IDE,
+ .end = K_INT_GB_IDE,
+ },
+};
+
+static struct pata_platform_info pata_platform_data = {
+ .ioport_shift = SWARM_IDE_SHIFT,
+};
+
+static struct platform_device swarm_pata_device = {
+ .name = "pata_platform",
+ .id = -1,
+ .resource = swarm_pata_resource,
+ .num_resources = ARRAY_SIZE(swarm_pata_resource),
+ .dev = {
+ .platform_data = &pata_platform_data,
+ .coherent_dma_mask = ~0, /* grumble */
+ },
+};
+
+static int __init swarm_pata_init(void)
+{
+ u8 __iomem *base;
+ phys_t offset, size;
+ struct resource *r;
+
+ if (!SIBYTE_HAVE_IDE)
+ return -ENODEV;
+
+ base = ioremap(A_IO_EXT_BASE, 0x800);
+ offset = __raw_readq(base + R_IO_EXT_REG(R_IO_EXT_START_ADDR, IDE_CS));
+ size = __raw_readq(base + R_IO_EXT_REG(R_IO_EXT_MULT_SIZE, IDE_CS));
+ iounmap(base);
+
+ offset = G_IO_START_ADDR(offset) << S_IO_ADDRBASE;
+ size = (G_IO_MULT_SIZE(size) + 1) << S_IO_REGSIZE;
+ if (offset < A_PHYS_GENBUS || offset >= A_PHYS_GENBUS_END) {
+ pr_info(DRV_NAME ": PATA interface at GenBus disabled\n");
+
+ return -EBUSY;
+ }
+
+ pr_info(DRV_NAME ": PATA interface at GenBus slot %i\n", IDE_CS);
+
+ r = swarm_pata_resource;
+ r[0].start = offset + (SWARM_IDE_BASE << SWARM_IDE_SHIFT);
+ r[0].end = offset + ((SWARM_IDE_BASE + 8) << SWARM_IDE_SHIFT) - 1;
+ r[1].start = offset + (SWARM_IDE_CTRL << SWARM_IDE_SHIFT);
+ r[1].end = offset + ((SWARM_IDE_CTRL + 1) << SWARM_IDE_SHIFT) - 1;
+
+ return platform_device_register(&swarm_pata_device);
+}
+
+device_initcall(swarm_pata_init);
secondary_cpu_time_init();
ipi_call_lock();
+ notify_cpu_starting(cpu);
cpu_set(cpu, cpu_online_map);
/* Update sibling maps */
base = cpu_first_thread_in_core(cpu);
/* Enable pfault pseudo page faults on this cpu. */
pfault_init();
+ /* call cpu notifiers */
+ notify_cpu_starting(smp_processor_id());
/* Mark this cpu as online */
spin_lock(&call_lock);
cpu_set(smp_processor_id(), cpu_online_map);
static void clock_comparator_interrupt(__u16 code)
{
+ if (S390_lowcore.clock_comparator == -1ULL)
+ set_clock_comparator(S390_lowcore.clock_comparator);
}
static void etr_timing_alert(struct etr_irq_parm *);
/*
- * arch/s390/lib/delay.c
* Precise Delay Loops for S390
*
- * S390 version
- * Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation
- * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
- *
- * Derived from "arch/i386/lib/delay.c"
- * Copyright (C) 1993 Linus Torvalds
- * Copyright (C) 1997 Martin Mares <mj@atrey.karlin.mff.cuni.cz>
+ * Copyright IBM Corp. 1999,2008
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>,
+ * Heiko Carstens <heiko.carstens@de.ibm.com>,
*/
#include <linux/sched.h>
asm volatile("0: brct %0,0b" : : "d" ((loops/2) + 1));
}
-/*
- * Waits for 'usecs' microseconds using the TOD clock comparator.
- */
-void __udelay(unsigned long usecs)
+static void __udelay_disabled(unsigned long usecs)
{
- u64 end, time, old_cc = 0;
- unsigned long flags, cr0, mask, dummy;
- int irq_context;
+ unsigned long mask, cr0, cr0_saved;
+ u64 clock_saved;
- irq_context = in_interrupt();
- if (!irq_context)
- local_bh_disable();
- local_irq_save(flags);
- if (raw_irqs_disabled_flags(flags)) {
- old_cc = local_tick_disable();
- S390_lowcore.clock_comparator = -1ULL;
- __ctl_store(cr0, 0, 0);
- dummy = (cr0 & 0xffff00e0) | 0x00000800;
- __ctl_load(dummy , 0, 0);
- mask = psw_kernel_bits | PSW_MASK_WAIT | PSW_MASK_EXT;
- } else
- mask = psw_kernel_bits | PSW_MASK_WAIT |
- PSW_MASK_EXT | PSW_MASK_IO;
+ clock_saved = local_tick_disable();
+ set_clock_comparator(get_clock() + ((u64) usecs << 12));
+ __ctl_store(cr0_saved, 0, 0);
+ cr0 = (cr0_saved & 0xffff00e0) | 0x00000800;
+ __ctl_load(cr0 , 0, 0);
+ mask = psw_kernel_bits | PSW_MASK_WAIT | PSW_MASK_EXT;
+ trace_hardirqs_on();
+ __load_psw_mask(mask);
+ local_irq_disable();
+ __ctl_load(cr0_saved, 0, 0);
+ local_tick_enable(clock_saved);
+ set_clock_comparator(S390_lowcore.clock_comparator);
+}
+static void __udelay_enabled(unsigned long usecs)
+{
+ unsigned long mask;
+ u64 end, time;
+
+ mask = psw_kernel_bits | PSW_MASK_WAIT | PSW_MASK_EXT | PSW_MASK_IO;
end = get_clock() + ((u64) usecs << 12);
do {
time = end < S390_lowcore.clock_comparator ?
__load_psw_mask(mask);
local_irq_disable();
} while (get_clock() < end);
+ set_clock_comparator(S390_lowcore.clock_comparator);
+}
- if (raw_irqs_disabled_flags(flags)) {
- __ctl_load(cr0, 0, 0);
- local_tick_enable(old_cc);
+/*
+ * Waits for 'usecs' microseconds using the TOD clock comparator.
+ */
+void __udelay(unsigned long usecs)
+{
+ unsigned long flags;
+
+ preempt_disable();
+ local_irq_save(flags);
+ if (in_irq()) {
+ __udelay_disabled(usecs);
+ goto out;
+ }
+ if (in_softirq()) {
+ if (raw_irqs_disabled_flags(flags))
+ __udelay_disabled(usecs);
+ else
+ __udelay_enabled(usecs);
+ goto out;
}
- if (!irq_context)
+ if (raw_irqs_disabled_flags(flags)) {
+ local_bh_disable();
+ __udelay_disabled(usecs);
_local_bh_enable();
- set_clock_comparator(S390_lowcore.clock_comparator);
+ goto out;
+ }
+ __udelay_enabled(usecs);
+out:
local_irq_restore(flags);
+ preempt_enable();
}
preempt_disable();
+ notify_cpu_starting(smp_processor_id());
+
local_irq_enable();
calibrate_delay();
local_flush_cache_all();
local_flush_tlb_all();
+ notify_cpu_starting(cpuid);
/*
* Unblock the master CPU _only_ when the scheduler state
* of all secondary CPUs will be up-to-date, so after
local_flush_cache_all();
local_flush_tlb_all();
+ notify_cpu_starting(cpuid);
+
/* Get our local ticker going. */
smp_setup_percpu_timer();
while (!cpu_isset(cpu, smp_commenced_mask))
cpu_relax();
+ notify_cpu_starting(cpu);
cpu_set(cpu, cpu_online_map);
default_idle();
return 0;
continue;
}
sh_symtab = sec_symtab->symtab;
- sym_strtab = sec->link->strtab;
+ sym_strtab = sec_symtab->link->strtab;
for (j = 0; j < sec->shdr.sh_size/sizeof(Elf32_Rel); j++) {
Elf32_Rel *rel;
Elf32_Sym *sym;
* is not connected at all. Force ignoring BIOS IRQ0 pin2
* override in that cases.
*/
+ {
+ .callback = dmi_ignore_irq0_timer_override,
+ .ident = "HP nx6115 laptop",
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq nx6115"),
+ },
+ },
{
.callback = dmi_ignore_irq0_timer_override,
.ident = "HP NX6125 laptop",
DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq nx6325"),
},
},
+ {
+ .callback = dmi_ignore_irq0_timer_override,
+ .ident = "HP 6715b laptop",
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq 6715b"),
+ },
+ },
{}
};
return NOTIFY_DONE;
case DIE_NMI_IPI:
- if (atomic_read(&kgdb_active) != -1) {
- /* KGDB CPU roundup */
- kgdb_nmicallback(raw_smp_processor_id(), regs);
- was_in_debug_nmi[raw_smp_processor_id()] = 1;
- touch_nmi_watchdog();
- }
+ /* Just ignore, we will handle the roundup on DIE_NMI. */
return NOTIFY_DONE;
case DIE_NMIUNKNOWN:
struct pci_dev *dev;
void *gatt;
int i, error;
- unsigned long start_pfn, end_pfn;
printk(KERN_INFO "PCI-DMA: Disabling AGP.\n");
aper_size = aper_base = info->aper_size = 0;
printk(KERN_INFO "PCI-DMA: aperture base @ %x size %u KB\n",
aper_base, aper_size>>10);
- /* need to map that range */
- end_pfn = (aper_base>>PAGE_SHIFT) + (aper_size>>PAGE_SHIFT);
- if (end_pfn > max_low_pfn_mapped) {
- start_pfn = (aper_base>>PAGE_SHIFT);
- init_memory_mapping(start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT);
- }
return 0;
nommu:
{
struct agp_kern_info info;
unsigned long iommu_start;
- unsigned long aper_size;
+ unsigned long aper_base, aper_size;
+ unsigned long start_pfn, end_pfn;
unsigned long scratch;
long i;
return;
}
+ /* need to map that range */
+ aper_size = info.aper_size << 20;
+ aper_base = info.aper_base;
+ end_pfn = (aper_base>>PAGE_SHIFT) + (aper_size>>PAGE_SHIFT);
+ if (end_pfn > max_low_pfn_mapped) {
+ start_pfn = (aper_base>>PAGE_SHIFT);
+ init_memory_mapping(start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT);
+ }
+
printk(KERN_INFO "PCI-DMA: using GART IOMMU.\n");
- aper_size = info.aper_size * 1024 * 1024;
iommu_size = check_iommu_size(info.aper_base, aper_size);
iommu_pages = iommu_size >> PAGE_SHIFT;
end_local_APIC_setup();
map_cpu_to_logical_apicid();
+ notify_cpu_starting(cpuid);
/*
* Get our bogomips.
*
VDEBUG(("VOYAGER SMP: CPU%d, stack at about %p\n", cpuid, &cpuid));
+ notify_cpu_starting(cpuid);
+
/* enable interrupts */
local_irq_enable();
"firmware_node");
ret = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
"physical_node");
- if (acpi_dev->wakeup.flags.valid)
+ if (acpi_dev->wakeup.flags.valid) {
device_set_wakeup_capable(dev, true);
+ device_set_wakeup_enable(dev,
+ acpi_dev->wakeup.state.enabled);
+ }
}
return 0;
return 0;
}
+static void physical_device_enable_wakeup(struct acpi_device *adev)
+{
+ struct device *dev = acpi_get_physical_device(adev->handle);
+
+ if (dev && device_can_wakeup(dev))
+ device_set_wakeup_enable(dev, adev->wakeup.state.enabled);
+}
+
static ssize_t
acpi_system_write_wakeup_device(struct file *file,
const char __user * buffer,
}
}
if (found_dev) {
+ physical_device_enable_wakeup(found_dev);
list_for_each_safe(node, next, &acpi_wakeup_device_list) {
struct acpi_device *dev = container_of(node,
struct
dev->pnp.bus_id, found_dev->pnp.bus_id);
dev->wakeup.state.enabled =
found_dev->wakeup.state.enabled;
+ physical_device_enable_wakeup(dev);
}
}
}
int i;
status_block = dma_readl(dw, RAW.BLOCK);
- status_xfer = dma_readl(dw, RAW.BLOCK);
+ status_xfer = dma_readl(dw, RAW.XFER);
status_err = dma_readl(dw, RAW.ERROR);
dev_vdbg(dw->dma.dev, "tasklet: status_block=%x status_err=%x\n",
to transfer data to and from memory. Saying Y is safe and improves
performance.
-config BLK_DEV_IDE_SWARM
- tristate "IDE for Sibyte evaluation boards"
- depends on SIBYTE_SB1xxx_SOC
-
config BLK_DEV_IDE_AU1XXX
bool "IDE for AMD Alchemy Au1200"
depends on SOC_AU1200
cdi->mask &= ~CDC_PLAY_AUDIO;
mechtype = buf[8 + 6] >> 5;
- if (mechtype == mechtype_caddy || mechtype == mechtype_popup)
+ if (mechtype == mechtype_caddy ||
+ mechtype == mechtype_popup ||
+ (drive->atapi_flags & IDE_AFLAG_NO_AUTOCLOSE))
cdi->mask |= CDC_CLOSE_TRAY;
if (cdi->sanyo_slot > 0) {
{ "MATSHITADVD-ROM SR-8176", NULL, IDE_AFLAG_PLAY_AUDIO_OK },
{ "MATSHITADVD-ROM SR-8174", NULL, IDE_AFLAG_PLAY_AUDIO_OK },
{ "Optiarc DVD RW AD-5200A", NULL, IDE_AFLAG_PLAY_AUDIO_OK },
+ { "Optiarc DVD RW AD-7200A", NULL, IDE_AFLAG_PLAY_AUDIO_OK },
+ { "Optiarc DVD RW AD-7543A", NULL, IDE_AFLAG_NO_AUTOCLOSE },
{ NULL, NULL, 0 }
};
xcount = bcount & 0xffff;
if (is_trm290)
xcount = ((xcount >> 2) - 1) << 16;
- if (xcount == 0x0000) {
+ else if (xcount == 0x0000) {
/*
* Most chipsets correctly interpret a length of 0x0000 as 64KB,
* but at least one (e.g. CS5530) misinterprets it as zero (!).
static int ide_sysfs_register_port(ide_hwif_t *hwif)
{
- int i, rc;
+ int i, uninitialized_var(rc);
for (i = 0; ide_port_attrs[i]; i++) {
rc = device_create_file(hwif->portdev, ide_port_attrs[i]);
-obj-$(CONFIG_BLK_DEV_IDE_SWARM) += swarm.o
obj-$(CONFIG_BLK_DEV_IDE_AU1XXX) += au1xxx-ide.o
EXTRA_CFLAGS := -Idrivers/ide
+++ /dev/null
-/*
- * Copyright (C) 2001, 2002, 2003 Broadcom Corporation
- * Copyright (C) 2004 MontaVista Software Inc.
- * Author: Manish Lachwani, mlachwani@mvista.com
- * Copyright (C) 2004 MIPS Technologies, Inc. All rights reserved.
- * Author: Maciej W. Rozycki <macro@mips.com>
- * Copyright (c) 2006, 2008 Maciej W. Rozycki
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version 2
- * of the License, or (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
- */
-
-/*
- * Derived loosely from ide-pmac.c, so:
- * Copyright (C) 1998 Paul Mackerras.
- * Copyright (C) 1995-1998 Mark Lord
- */
-
-/*
- * Boards with SiByte processors so far have supported IDE devices via
- * the Generic Bus, PCI bus, and built-in PCMCIA interface. In all
- * cases, byte-swapping must be avoided for these devices (whereas
- * other PCI devices, for example, will require swapping). Any
- * SiByte-targetted kernel including IDE support will include this
- * file. Probing of a Generic Bus for an IDE device is controlled by
- * the definition of "SIBYTE_HAVE_IDE", which is provided by
- * <asm/sibyte/board.h> for Broadcom boards.
- */
-
-#include <linux/ide.h>
-#include <linux/ioport.h>
-#include <linux/kernel.h>
-#include <linux/types.h>
-#include <linux/platform_device.h>
-
-#include <asm/io.h>
-
-#include <asm/sibyte/board.h>
-#include <asm/sibyte/sb1250_genbus.h>
-#include <asm/sibyte/sb1250_regs.h>
-
-#define DRV_NAME "ide-swarm"
-
-static char swarm_ide_string[] = DRV_NAME;
-
-static struct resource swarm_ide_resource = {
- .name = "SWARM GenBus IDE",
- .flags = IORESOURCE_MEM,
-};
-
-static struct platform_device *swarm_ide_dev;
-
-static const struct ide_port_info swarm_port_info = {
- .name = DRV_NAME,
- .host_flags = IDE_HFLAG_MMIO | IDE_HFLAG_NO_DMA,
-};
-
-/*
- * swarm_ide_probe - if the board header indicates the existence of
- * Generic Bus IDE, allocate a HWIF for it.
- */
-static int __devinit swarm_ide_probe(struct device *dev)
-{
- u8 __iomem *base;
- struct ide_host *host;
- phys_t offset, size;
- int i, rc;
- hw_regs_t hw, *hws[] = { &hw, NULL, NULL, NULL };
-
- if (!SIBYTE_HAVE_IDE)
- return -ENODEV;
-
- base = ioremap(A_IO_EXT_BASE, 0x800);
- offset = __raw_readq(base + R_IO_EXT_REG(R_IO_EXT_START_ADDR, IDE_CS));
- size = __raw_readq(base + R_IO_EXT_REG(R_IO_EXT_MULT_SIZE, IDE_CS));
- iounmap(base);
-
- offset = G_IO_START_ADDR(offset) << S_IO_ADDRBASE;
- size = (G_IO_MULT_SIZE(size) + 1) << S_IO_REGSIZE;
- if (offset < A_PHYS_GENBUS || offset >= A_PHYS_GENBUS_END) {
- printk(KERN_INFO DRV_NAME
- ": IDE interface at GenBus disabled\n");
- return -EBUSY;
- }
-
- printk(KERN_INFO DRV_NAME ": IDE interface at GenBus slot %i\n",
- IDE_CS);
-
- swarm_ide_resource.start = offset;
- swarm_ide_resource.end = offset + size - 1;
- if (request_resource(&iomem_resource, &swarm_ide_resource)) {
- printk(KERN_ERR DRV_NAME
- ": can't request I/O memory resource\n");
- return -EBUSY;
- }
-
- base = ioremap(offset, size);
-
- memset(&hw, 0, sizeof(hw));
- for (i = 0; i <= 7; i++)
- hw.io_ports_array[i] =
- (unsigned long)(base + ((0x1f0 + i) << 5));
- hw.io_ports.ctl_addr =
- (unsigned long)(base + (0x3f6 << 5));
- hw.irq = K_INT_GB_IDE;
- hw.chipset = ide_generic;
-
- rc = ide_host_add(&swarm_port_info, hws, &host);
- if (rc)
- goto err;
-
- dev_set_drvdata(dev, host);
-
- return 0;
-err:
- release_resource(&swarm_ide_resource);
- iounmap(base);
- return rc;
-}
-
-static struct device_driver swarm_ide_driver = {
- .name = swarm_ide_string,
- .bus = &platform_bus_type,
- .probe = swarm_ide_probe,
-};
-
-static void swarm_ide_platform_release(struct device *device)
-{
- struct platform_device *pldev;
-
- /* free device */
- pldev = to_platform_device(device);
- kfree(pldev);
-}
-
-static int __devinit swarm_ide_init_module(void)
-{
- struct platform_device *pldev;
- int err;
-
- printk(KERN_INFO "SWARM IDE driver\n");
-
- if (driver_register(&swarm_ide_driver)) {
- printk(KERN_ERR "Driver registration failed\n");
- err = -ENODEV;
- goto out;
- }
-
- if (!(pldev = kzalloc(sizeof (*pldev), GFP_KERNEL))) {
- err = -ENOMEM;
- goto out_unregister_driver;
- }
-
- pldev->name = swarm_ide_string;
- pldev->id = 0;
- pldev->dev.release = swarm_ide_platform_release;
-
- if (platform_device_register(pldev)) {
- err = -ENODEV;
- goto out_free_pldev;
- }
-
- if (!pldev->dev.driver) {
- /*
- * The driver was not bound to this device, there was
- * no hardware at this address. Unregister it, as the
- * release fuction will take care of freeing the
- * allocated structure
- */
- platform_device_unregister (pldev);
- }
-
- swarm_ide_dev = pldev;
-
- return 0;
-
-out_free_pldev:
- kfree(pldev);
-
-out_unregister_driver:
- driver_unregister(&swarm_ide_driver);
-out:
- return err;
-}
-
-module_init(swarm_ide_init_module);
{
int ret;
+ /* Map the LED chip select address space */
+ latch_address = (unsigned short *) ioremap(IXP4XX_EXP_BUS_BASE(2), 512);
+ if (!latch_address) {
+ ret = -ENOMEM;
+ goto failremap;
+ }
+
+ latch_value = 0xffff;
+ *latch_address = latch_value;
+
ret = led_classdev_register(&pdev->dev, &fsg_wlan_led);
if (ret < 0)
goto failwlan;
if (ret < 0)
goto failring;
- /* Map the LED chip select address space */
- latch_address = (unsigned short *) ioremap(IXP4XX_EXP_BUS_BASE(2), 512);
- if (!latch_address) {
- ret = -ENOMEM;
- goto failremap;
- }
-
- latch_value = 0xffff;
- *latch_address = latch_value;
-
return ret;
- failremap:
- led_classdev_unregister(&fsg_ring_led);
failring:
led_classdev_unregister(&fsg_sync_led);
failsync:
failwan:
led_classdev_unregister(&fsg_wlan_led);
failwlan:
+ iounmap(latch_address);
+ failremap:
return ret;
}
static int fsg_led_remove(struct platform_device *pdev)
{
- iounmap(latch_address);
-
led_classdev_unregister(&fsg_wlan_led);
led_classdev_unregister(&fsg_wan_led);
led_classdev_unregister(&fsg_sata_led);
led_classdev_unregister(&fsg_sync_led);
led_classdev_unregister(&fsg_ring_led);
+ iounmap(latch_address);
+
return 0;
}
const struct i2c_device_id *id)
{
struct pca955x_led *pca955x;
- int i;
- int err = -ENODEV;
struct pca955x_chipdef *chip;
struct i2c_adapter *adapter;
struct led_platform_data *pdata;
+ int i, err;
chip = &pca955x_chipdefs[id->driver_data];
adapter = to_i2c_adapter(client->dev.parent);
}
}
+ pca955x = kzalloc(sizeof(*pca955x) * chip->bits, GFP_KERNEL);
+ if (!pca955x)
+ return -ENOMEM;
+
+ i2c_set_clientdata(client, pca955x);
+
for (i = 0; i < chip->bits; i++) {
- pca955x = kzalloc(sizeof(struct pca955x_led), GFP_KERNEL);
- if (!pca955x) {
- err = -ENOMEM;
- goto exit;
- }
+ pca955x[i].chipdef = chip;
+ pca955x[i].client = client;
+ pca955x[i].led_num = i;
- pca955x->chipdef = chip;
- pca955x->client = client;
- pca955x->led_num = i;
/* Platform data can specify LED names and default triggers */
if (pdata) {
if (pdata->leds[i].name)
- snprintf(pca955x->name, 32, "pca955x:%s",
- pdata->leds[i].name);
+ snprintf(pca955x[i].name,
+ sizeof(pca955x[i].name), "pca955x:%s",
+ pdata->leds[i].name);
if (pdata->leds[i].default_trigger)
- pca955x->led_cdev.default_trigger =
+ pca955x[i].led_cdev.default_trigger =
pdata->leds[i].default_trigger;
} else {
- snprintf(pca955x->name, 32, "pca955x:%d", i);
+ snprintf(pca955x[i].name, sizeof(pca955x[i].name),
+ "pca955x:%d", i);
}
- spin_lock_init(&pca955x->lock);
- pca955x->led_cdev.name = pca955x->name;
- pca955x->led_cdev.brightness_set =
- pca955x_led_set;
+ spin_lock_init(&pca955x[i].lock);
- /*
- * Client data is a pointer to the _first_ pca955x_led
- * struct
- */
- if (i == 0)
- i2c_set_clientdata(client, pca955x);
+ pca955x[i].led_cdev.name = pca955x[i].name;
+ pca955x[i].led_cdev.brightness_set = pca955x_led_set;
- INIT_WORK(&(pca955x->work), pca955x_led_work);
+ INIT_WORK(&pca955x[i].work, pca955x_led_work);
- led_classdev_register(&client->dev, &(pca955x->led_cdev));
+ err = led_classdev_register(&client->dev, &pca955x[i].led_cdev);
+ if (err < 0)
+ goto exit;
}
/* Turn off LEDs */
pca955x_write_psc(client, 1, 0);
return 0;
+
exit:
+ while (i--) {
+ led_classdev_unregister(&pca955x[i].led_cdev);
+ cancel_work_sync(&pca955x[i].work);
+ }
+
+ kfree(pca955x);
+ i2c_set_clientdata(client, NULL);
+
return err;
}
static int __devexit pca955x_remove(struct i2c_client *client)
{
struct pca955x_led *pca955x = i2c_get_clientdata(client);
- int leds = pca955x->chipdef->bits;
int i;
- for (i = 0; i < leds; i++) {
- led_classdev_unregister(&(pca955x->led_cdev));
- cancel_work_sync(&(pca955x->work));
- kfree(pca955x);
- pca955x = pca955x + 1;
+ for (i = 0; i < pca955x->chipdef->bits; i++) {
+ led_classdev_unregister(&pca955x[i].led_cdev);
+ cancel_work_sync(&pca955x[i].work);
}
+ kfree(pca955x);
+ i2c_set_clientdata(client, NULL);
+
return 0;
}
#include "dvb_frontend.h"
#define XC2028_DEFAULT_FIRMWARE "xc3028-v27.fw"
+#define XC3028L_DEFAULT_FIRMWARE "xc3028L-v36.fw"
/* Dmoduler IF (kHz) */
#define XC3028_FE_DEFAULT 0 /* Don't load SCODE */
.demod_address = 0x53,
.invert = 1,
.repeated_start_workaround = 1,
+ .serial_mpeg = 1,
};
static struct itd1000_config skystar2_rev2_7_itd1000_config = {
enum dmx_success success)
{
struct dmxdev_filter *dmxdevfilter = filter->priv;
+ unsigned long flags;
int ret;
if (dmxdevfilter->buffer.error) {
wake_up(&dmxdevfilter->buffer.queue);
return 0;
}
- spin_lock(&dmxdevfilter->dev->lock);
+ spin_lock_irqsave(&dmxdevfilter->dev->lock, flags);
if (dmxdevfilter->state != DMXDEV_STATE_GO) {
- spin_unlock(&dmxdevfilter->dev->lock);
+ spin_unlock_irqrestore(&dmxdevfilter->dev->lock, flags);
return 0;
}
del_timer(&dmxdevfilter->timer);
}
if (dmxdevfilter->params.sec.flags & DMX_ONESHOT)
dmxdevfilter->state = DMXDEV_STATE_DONE;
- spin_unlock(&dmxdevfilter->dev->lock);
+ spin_unlock_irqrestore(&dmxdevfilter->dev->lock, flags);
wake_up(&dmxdevfilter->buffer.queue);
return 0;
}
{
struct dmxdev_filter *dmxdevfilter = feed->priv;
struct dvb_ringbuffer *buffer;
+ unsigned long flags;
int ret;
- spin_lock(&dmxdevfilter->dev->lock);
+ spin_lock_irqsave(&dmxdevfilter->dev->lock, flags);
if (dmxdevfilter->params.pes.output == DMX_OUT_DECODER) {
- spin_unlock(&dmxdevfilter->dev->lock);
+ spin_unlock_irqrestore(&dmxdevfilter->dev->lock, flags);
return 0;
}
else
buffer = &dmxdevfilter->dev->dvr_buffer;
if (buffer->error) {
- spin_unlock(&dmxdevfilter->dev->lock);
+ spin_unlock_irqrestore(&dmxdevfilter->dev->lock, flags);
wake_up(&buffer->queue);
return 0;
}
dvb_ringbuffer_flush(buffer);
buffer->error = ret;
}
- spin_unlock(&dmxdevfilter->dev->lock);
+ spin_unlock_irqrestore(&dmxdevfilter->dev->lock, flags);
wake_up(&buffer->queue);
return 0;
}
void dvb_dmx_swfilter_packets(struct dvb_demux *demux, const u8 *buf,
size_t count)
{
- spin_lock(&demux->lock);
+ unsigned long flags;
+
+ spin_lock_irqsave(&demux->lock, flags);
while (count--) {
if (buf[0] == 0x47)
buf += 188;
}
- spin_unlock(&demux->lock);
+ spin_unlock_irqrestore(&demux->lock, flags);
}
EXPORT_SYMBOL(dvb_dmx_swfilter_packets);
void dvb_dmx_swfilter(struct dvb_demux *demux, const u8 *buf, size_t count)
{
+ unsigned long flags;
int p = 0, i, j;
- spin_lock(&demux->lock);
+ spin_lock_irqsave(&demux->lock, flags);
if (demux->tsbufp) {
i = demux->tsbufp;
}
bailout:
- spin_unlock(&demux->lock);
+ spin_unlock_irqrestore(&demux->lock, flags);
}
EXPORT_SYMBOL(dvb_dmx_swfilter);
void dvb_dmx_swfilter_204(struct dvb_demux *demux, const u8 *buf, size_t count)
{
+ unsigned long flags;
int p = 0, i, j;
u8 tmppack[188];
- spin_lock(&demux->lock);
+ spin_lock_irqsave(&demux->lock, flags);
if (demux->tsbufp) {
i = demux->tsbufp;
}
bailout:
- spin_unlock(&demux->lock);
+ spin_unlock_irqrestore(&demux->lock, flags);
}
EXPORT_SYMBOL(dvb_dmx_swfilter_204);
* it does not support repeated-start, workaround: write addr-1
* and then read
*/
- u8 shadow[255];
+ u8 shadow[256];
};
static u32 s5h1420_getsymbolrate(struct s5h1420_state* state);
if (ret != 3)
return ret;
} else {
- ret = i2c_transfer(state->i2c, &msg[1], 2);
- if (ret != 2)
+ ret = i2c_transfer(state->i2c, &msg[1], 1);
+ if (ret != 1)
+ return ret;
+ ret = i2c_transfer(state->i2c, &msg[2], 1);
+ if (ret != 1)
return ret;
}
struct s5h1420_state* state = fe->demodulator_priv;
/* disable power down and do reset */
- state->CON_1_val = 0x10;
+ state->CON_1_val = state->config->serial_mpeg << 4;
s5h1420_writereg(state, 0x02, state->CON_1_val);
msleep(10);
s5h1420_reset(state);
u8 demod_address;
/* does the inversion require inversion? */
- u8 invert : 1;
+ u8 invert:1;
- u8 repeated_start_workaround : 1;
- u8 cdclk_polarity : 1; /* 1 == falling edge, 0 == raising edge */
+ u8 repeated_start_workaround:1;
+ u8 cdclk_polarity:1; /* 1 == falling edge, 0 == raising edge */
+
+ u8 serial_mpeg:1;
};
#if defined(CONFIG_DVB_S5H1420) || (defined(CONFIG_DVB_S5H1420_MODULE) && defined(MODULE))
.driver_info = SMS1XXX_BOARD_HAUPPAUGE_OKEMO_B },
{ USB_DEVICE(0x2040, 0x5500),
.driver_info = SMS1XXX_BOARD_HAUPPAUGE_WINDHAM },
+ { USB_DEVICE(0x2040, 0x5510),
+ .driver_info = SMS1XXX_BOARD_HAUPPAUGE_WINDHAM },
{ USB_DEVICE(0x2040, 0x5580),
.driver_info = SMS1XXX_BOARD_HAUPPAUGE_WINDHAM },
{ USB_DEVICE(0x2040, 0x5590),
.fw[DEVICE_MODE_DVBT_BDA] = "sms1xxx-nova-b-dvbt-01.fw",
},
[SMS1XXX_BOARD_HAUPPAUGE_WINDHAM] = {
- .name = "Hauppauge WinTV-Nova-T-MiniStick",
+ .name = "Hauppauge WinTV MiniStick",
.type = SMS_NOVA_B0,
.fw[DEVICE_MODE_DVBT_BDA] = "sms1xxx-hcw-55xxx-dvbt-01.fw",
},
dprintk("bttv: open minor=%d\n",minor);
for (i = 0; i < bttv_num; i++) {
- if (bttvs[i].radio_dev->minor == minor) {
+ if (bttvs[i].radio_dev && bttvs[i].radio_dev->minor == minor) {
btv = &bttvs[i];
break;
}
#include <linux/module.h>
#include <linux/init.h>
#include <linux/fs.h>
+#include <linux/mm.h>
#include <linux/pci.h>
#include <linux/i2c.h>
#include <linux/interrupt.h>
static int submit_urbs(struct camera_data *cam)
{
struct urb *urb;
- int fx, err, i;
+ int fx, err, i, j;
for(i=0; i<NUM_SBUF; ++i) {
if (cam->sbuf[i].data)
}
urb = usb_alloc_urb(FRAMES_PER_DESC, GFP_KERNEL);
if (!urb) {
+ ERR("%s: usb_alloc_urb error!\n", __func__);
+ for (j = 0; j < i; j++)
+ usb_free_urb(cam->sbuf[j].urb);
return -ENOMEM;
}
},
.audio_inputs = {
{ CX18_CARD_INPUT_AUD_TUNER,
- CX18_AV_AUDIO8, 0 },
+ CX18_AV_AUDIO5, 0 },
{ CX18_CARD_INPUT_LINE_IN1,
CX18_AV_AUDIO_SERIAL1, 0 },
},
if (oldptr + length >= runtime->buffer_size) {
unsigned int cnt =
- runtime->buffer_size - oldptr - 1;
+ runtime->buffer_size - oldptr;
memcpy(runtime->dma_area + oldptr * stride, cp,
cnt * stride);
- memcpy(runtime->dma_area, cp + cnt,
+ memcpy(runtime->dma_area, cp + cnt * stride,
length * stride - cnt * stride);
} else {
memcpy(runtime->dma_area + oldptr * stride, cp,
memset(dev->adev->transfer_buffer[i], 0x80, sb_size);
urb = usb_alloc_urb(EM28XX_NUM_AUDIO_PACKETS, GFP_ATOMIC);
- if (!urb)
+ if (!urb) {
+ em28xx_errdev("usb_alloc_urb failed!\n");
+ for (j = 0; j < i; j++) {
+ usb_free_urb(dev->adev->urb[j]);
+ kfree(dev->adev->transfer_buffer[j]);
+ }
return -ENOMEM;
+ }
urb->dev = dev->udev;
urb->context = dev;
.amux = 0,
} },
},
- [EM2800_BOARD_KWORLD_USB2800] = {
- .name = "Kworld USB2800",
- .valid = EM28XX_BOARD_NOT_VALIDATED,
- .is_em2800 = 1,
- .vchannels = 3,
- .tuner_type = TUNER_PHILIPS_FCV1236D,
- .tda9887_conf = TDA9887_PRESENT,
- .decoder = EM28XX_SAA7113,
- .input = { {
- .type = EM28XX_VMUX_TELEVISION,
- .vmux = SAA7115_COMPOSITE2,
- .amux = 0,
- }, {
- .type = EM28XX_VMUX_COMPOSITE1,
- .vmux = SAA7115_COMPOSITE0,
- .amux = 1,
- }, {
- .type = EM28XX_VMUX_SVIDEO,
- .vmux = SAA7115_SVIDEO3,
- .amux = 1,
- } },
- },
[EM2820_BOARD_KWORLD_PVRTV2800RF] = {
.name = "Kworld PVR TV 2800 RF",
.is_em2800 = 0,
}, {
.type = EM28XX_VMUX_COMPOSITE1,
.vmux = TVP5150_COMPOSITE1,
- .amux = 1,
+ .amux = 3,
}, {
.type = EM28XX_VMUX_SVIDEO,
.vmux = TVP5150_SVIDEO,
},
[EM2880_BOARD_KWORLD_DVB_310U] = {
.name = "KWorld DVB-T 310U",
- .valid = EM28XX_BOARD_NOT_VALIDATED,
.vchannels = 3,
.tuner_type = TUNER_XC2028,
+ .has_dvb = 1,
+ .mts_firmware = 1,
.decoder = EM28XX_TVP5150,
.input = { {
.type = EM28XX_VMUX_TELEVISION,
.vmux = TVP5150_COMPOSITE0,
- .amux = 0,
+ .amux = EM28XX_AMUX_VIDEO,
}, {
.type = EM28XX_VMUX_COMPOSITE1,
.vmux = TVP5150_COMPOSITE1,
- .amux = 1,
- }, {
+ .amux = EM28XX_AMUX_AC97_LINE_IN,
+ }, { /* S-video has not been tested yet */
.type = EM28XX_VMUX_SVIDEO,
.vmux = TVP5150_SVIDEO,
- .amux = 1,
+ .amux = EM28XX_AMUX_AC97_LINE_IN,
} },
},
[EM2881_BOARD_DNT_DA2_HYBRID] = {
static struct em28xx_hash_table em28xx_eeprom_hash [] = {
/* P/N: SA 60002070465 Tuner: TVF7533-MF */
{0x6ce05a8f, EM2820_BOARD_PROLINK_PLAYTV_USB2, TUNER_YMEC_TVF_5533MF},
+ {0x966a0441, EM2880_BOARD_KWORLD_DVB_310U, TUNER_XC2028},
};
/* I2C devicelist hash table for devices with generic USB IDs */
/* djh - Not sure which demod we need here */
ctl->demod = XC3028_FE_DEFAULT;
break;
+ case EM2880_BOARD_AMD_ATI_TV_WONDER_HD_600:
+ ctl->demod = XC3028_FE_DEFAULT;
+ ctl->fname = XC3028L_DEFAULT_FIRMWARE;
+ break;
case EM2883_BOARD_HAUPPAUGE_WINTV_HVR_950:
case EM2880_BOARD_PINNACLE_PCTV_HD_PRO:
- case EM2880_BOARD_AMD_ATI_TV_WONDER_HD_600:
/* FIXME: Better to specify the needed IF */
ctl->demod = XC3028_FE_DEFAULT;
break;
break;
case EM2820_BOARD_UNKNOWN:
case EM2800_BOARD_UNKNOWN:
+ /*
+ * The K-WORLD DVB-T 310U is detected as an MSI Digivox AD.
+ *
+ * This occurs because they share identical USB vendor and
+ * product IDs.
+ *
+ * What we do here is look up the EEPROM hash of the K-WORLD
+ * and if it is found then we decide that we do not have
+ * a DIGIVOX and reset the device to the K-WORLD instead.
+ *
+ * This solution is only valid if they do not share eeprom
+ * hash identities which has not been determined as yet.
+ */
+ case EM2880_BOARD_MSI_DIGIVOX_AD:
if (!em28xx_hint_board(dev))
em28xx_set_model(dev);
break;
goto out_free;
}
break;
+ case EM2880_BOARD_KWORLD_DVB_310U:
+ dvb->frontend = dvb_attach(zl10353_attach,
+ &em28xx_zl10353_with_xc3028,
+ &dev->i2c_adap);
+ if (attach_xc3028(0x61, dev) < 0) {
+ result = -EINVAL;
+ goto out_free;
+ }
+ break;
default:
printk(KERN_ERR "%s/2: The frontend of your DVB/ATSC card"
" isn't supported yet\n",
urb = usb_alloc_urb(npkt, GFP_KERNEL);
if (!urb) {
err("usb_alloc_urb failed");
+ destroy_urbs(gspca_dev);
return -ENOMEM;
}
urb->transfer_buffer = usb_buffer_alloc(gspca_dev->dev,
if (urb->transfer_buffer == NULL) {
usb_free_urb(urb);
- destroy_urbs(gspca_dev);
err("usb_buffer_urb failed");
+ destroy_urbs(gspca_dev);
return -ENOMEM;
}
gspca_dev->urb[n] = urb;
{USB_DEVICE(0x093a, 0x2621), .driver_info = SENSOR_PAC7302},
{USB_DEVICE(0x093a, 0x2624), .driver_info = SENSOR_PAC7302},
{USB_DEVICE(0x093a, 0x2626), .driver_info = SENSOR_PAC7302},
+ {USB_DEVICE(0x093a, 0x262a), .driver_info = SENSOR_PAC7302},
{}
};
MODULE_DEVICE_TABLE(usb, device_table);
static struct v4l2_pix_format vga_mode[] = {
{160, 120, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE,
.bytesperline = 160,
- .sizeimage = 160 * 120 * 5 / 4,
+ .sizeimage = 160 * 120,
.colorspace = V4L2_COLORSPACE_SRGB,
.priv = 2 | MODE_RAW},
{160, 120, V4L2_PIX_FMT_SN9C10X, V4L2_FIELD_NONE,
.priv = 1 | MODE_REDUCED_SIF},
{176, 144, V4L2_PIX_FMT_SBGGR8, V4L2_FIELD_NONE,
.bytesperline = 176,
- .sizeimage = 176 * 144 * 5 / 4,
+ .sizeimage = 176 * 144,
.colorspace = V4L2_COLORSPACE_SRGB,
.priv = 1 | MODE_RAW},
{176, 144, V4L2_PIX_FMT_SN9C10X, V4L2_FIELD_NONE,
0x08, 0, /* value, index */
gspca_dev->usb_buf, 8,
500);
+ msleep(2);
}
/* read 5 bytes in gspca_dev->usb_buf */
case BRIDGE_SN9C105:
if (regF1 != 0x11)
return -ENODEV;
- reg_w(gspca_dev, 0x02, regGpio, 2);
+ reg_w(gspca_dev, 0x01, regGpio, 2);
break;
case BRIDGE_SN9C120:
if (regF1 != 0x12)
return -ENODEV;
regGpio[1] = 0x70;
- reg_w(gspca_dev, 0x02, regGpio, 2);
+ reg_w(gspca_dev, 0x01, regGpio, 2);
break;
default:
/* case BRIDGE_SN9C110: */
static const __u8 CA[] = { 0x28, 0xd8, 0x14, 0xec };
static const __u8 CE[] = { 0x32, 0xdd, 0x2d, 0xdd }; /* MI0360 */
static const __u8 CE_ov76xx[] =
- { 0x32, 0xdd, 0x32, 0xdd }; /* OV7630/48 */
+ { 0x32, 0xdd, 0x32, 0xdd };
sn9c1xx = sn_tb[(int) sd->sensor];
configure_gpio(gspca_dev, sn9c1xx);
reg_w(gspca_dev, 0x20, gamma_def, sizeof gamma_def);
for (i = 0; i < 8; i++)
reg_w(gspca_dev, 0x84, reg84, sizeof reg84);
+ switch (sd->sensor) {
+ case SENSOR_OV7660:
+ reg_w1(gspca_dev, 0x9a, 0x05);
+ break;
+ default:
reg_w1(gspca_dev, 0x9a, 0x08);
reg_w1(gspca_dev, 0x99, 0x59);
+ break;
+ }
mode = gspca_dev->cam.cam_mode[(int) gspca_dev->curr_mode].priv;
if (mode)
/* reg1 = 0x44; */
/* reg1 = 0x46; (done) */
} else {
- reg17 = 0x22; /* 640 MCKSIZE */
- reg1 = 0x06;
+ reg17 = 0xa2; /* 640 */
+ reg1 = 0x44;
}
break;
}
switch (sd->sensor) {
case SENSOR_OV7630:
case SENSOR_OV7648:
+ case SENSOR_OV7660:
reg_w(gspca_dev, 0xce, CE_ov76xx, 4);
break;
default:
reg_w_val(gspca_dev->dev, 0x8802, (mode | 0x01));
do {
reg_r(gspca_dev, 0x8803, 1);
- if (!gspca_dev->usb_buf)
+ if (!gspca_dev->usb_buf[0])
break;
} while (--retry);
if (retry == 0)
cs2102_60HZ, cs2102_60HZScale},
/* SENSOR_CS2102K 1 */
{cs2102_NoFliker, cs2102_NoFlikerScale,
- cs2102_50HZ, cs2102_50HZScale,
- cs2102_60HZ, cs2102_60HZScale},
+ NULL, NULL, /* currently disabled */
+ NULL, NULL},
/* SENSOR_GC0305 2 */
{gc0305_NoFliker, gc0305_NoFliker,
gc0305_50HZ, gc0305_50HZ,
ov51x_init_isoc(struct usb_ov511 *ov)
{
struct urb *urb;
- int fx, err, n, size;
+ int fx, err, n, i, size;
PDEBUG(3, "*** Initializing capture ***");
urb = usb_alloc_urb(FRAMES_PER_DESC, GFP_KERNEL);
if (!urb) {
err("init isoc: usb_alloc_urb ret. NULL");
+ for (i = 0; i < n; i++)
+ usb_free_urb(ov->sbuf[i].urb);
return -ENOMEM;
}
ov->sbuf[n].urb = urb;
if (!ov->dev)
return -ENODEV;
sensor_get_exposure(ov, &exp);
- return sprintf(buf, "%d\n", exp >> 8);
+ return sprintf(buf, "%d\n", exp);
}
static DEVICE_ATTR(exposure, S_IRUGO, show_exposure, NULL);
struct usb_device_id pvr2_device_table[] = {
{ USB_DEVICE(0x2040, 0x2900),
.driver_info = (kernel_ulong_t)&pvr2_device_29xxx},
+ { USB_DEVICE(0x2040, 0x2950), /* Logically identical to 2900 */
+ .driver_info = (kernel_ulong_t)&pvr2_device_29xxx},
{ USB_DEVICE(0x2040, 0x2400),
.driver_info = (kernel_ulong_t)&pvr2_device_24xxx},
{ USB_DEVICE(0x1164, 0x0622),
(unsigned long)vbuf, pos);
/* tell v4l buffer was filled */
- buf->vb.field_count++;
+ buf->vb.field_count = dev->frame_count[chn] * 2;
do_gettimeofday(&ts);
buf->vb.ts = ts;
buf->vb.state = VIDEOBUF_DONE;
dev->last_frame[chn] = -1;
dev->bad_payload[chn] = 0;
dev->cur_frame[chn] = 0;
+ dev->frame_count[chn] = 0;
for (j = 0; j < SYS_FRAMES; j++) {
dev->buffer[chn].frame[j].ulState = 0;
dev->buffer[chn].frame[j].cur_size = 0;
if (ctrl == NULL)
return -EINVAL;
- data = kmalloc(8, GFP_KERNEL);
+ data = kmalloc(ctrl->info->size, GFP_KERNEL);
if (data == NULL)
return -ENOMEM;
for (i = 0; i < W9968CF_URBS; i++) {
urb = usb_alloc_urb(W9968CF_ISO_PACKETS, GFP_KERNEL);
- cam->urb[i] = urb;
if (!urb) {
for (j = 0; j < i; j++)
usb_free_urb(cam->urb[j]);
return -ENOMEM;
}
+ cam->urb[i] = urb;
urb->dev = udev;
urb->context = (void*)cam;
urb->pipe = usb_rcvisocpipe(udev, 1);
client->addr << 1, client->adapter->name);
state = kmalloc(sizeof(struct wm8739_state), GFP_KERNEL);
- if (state == NULL) {
- kfree(client);
+ if (state == NULL)
return -ENOMEM;
- }
state->vol_l = 0x17; /* 0dB */
state->vol_r = 0x17; /* 0dB */
state->muted = 0;
zr->v4l_grab_seq = 0;
zr->v4l_settings.width = 192;
zr->v4l_settings.height = 144;
- zr->v4l_settings.format = &zoran_formats[4]; /* YUY2 - YUV-4:2:2 packed */
+ zr->v4l_settings.format = &zoran_formats[7]; /* YUY2 - YUV-4:2:2 packed */
zr->v4l_settings.bytesperline =
zr->v4l_settings.width *
((zr->v4l_settings.format->depth + 7) / 8);
}, {
.name = "16-bit RGB BE",
ZFMT(-1,
- V4L2_PIX_FMT_RGB565, V4L2_COLORSPACE_SRGB),
+ V4L2_PIX_FMT_RGB565X, V4L2_COLORSPACE_SRGB),
.depth = 16,
.flags = ZORAN_FORMAT_CAPTURE |
ZORAN_FORMAT_OVERLAY,
fh->v4l_settings.format->fourcc;
fmt->fmt.pix.colorspace =
fh->v4l_settings.format->colorspace;
- fmt->fmt.pix.bytesperline = 0;
+ fmt->fmt.pix.bytesperline =
+ fh->v4l_settings.bytesperline;
if (BUZ_MAX_HEIGHT <
(fh->v4l_settings.height * 2))
fmt->fmt.pix.field =
fmt->fmt.pix.pixelformat,
(char *) &printformat);
- if (fmt->fmt.pix.bytesperline > 0) {
- dprintk(5,
- KERN_ERR "%s: bpl not supported\n",
- ZR_DEVNAME(zr));
- return -EINVAL;
- }
-
/* we can be requested to do JPEG/raw playback/capture */
if (!
(fmt->type == V4L2_BUF_TYPE_VIDEO_CAPTURE ||
fh->jpg_buffers.buffer_size =
zoran_v4l2_calc_bufsize(&fh->
jpg_settings);
+ fmt->fmt.pix.bytesperline = 0;
fmt->fmt.pix.sizeimage =
fh->jpg_buffers.buffer_size;
/* tell the user the
* results/missing stuff */
+ fmt->fmt.pix.bytesperline =
+ fh->v4l_settings.bytesperline;
fmt->fmt.pix.sizeimage =
fh->v4l_settings.height *
fh->v4l_settings.bytesperline;
host->sg = NULL;
host->data = data;
- mci_writel(host, BLKR, MCI_BCNT(data->blocks)
- | MCI_BLKLEN(data->blksz));
dev_vdbg(&mmc->class_dev, "BLKR=0x%08x\n",
MCI_BCNT(data->blocks) | MCI_BLKLEN(data->blksz));
if (data->blocks > 1 && data->blksz & 3)
goto fail;
atmci_set_timeout(host, data);
+
+ /* Must set block count/size before sending command */
+ mci_writel(host, BLKR, MCI_BCNT(data->blocks)
+ | MCI_BLKLEN(data->blksz));
}
iflags = MCI_CMDRDY;
u32 extcnf_ctrl;
u32 timeout = PHY_CFG_TIMEOUT;
- WARN_ON(preempt_count());
+ might_sleep();
if (!mutex_trylock(&nvm_mutex)) {
WARN(1, KERN_ERR "e1000e mutex contention. Owned by pid %d\n",
return err;
}
+static int rtc_dev_fasync(int fd, struct file *file, int on)
+{
+ struct rtc_device *rtc = file->private_data;
+ return fasync_helper(fd, file, on, &rtc->async_queue);
+}
+
static int rtc_dev_release(struct inode *inode, struct file *file)
{
struct rtc_device *rtc = file->private_data;
if (rtc->ops->release)
rtc->ops->release(rtc->dev.parent);
+ if (file->f_flags & FASYNC)
+ rtc_dev_fasync(-1, file, 0);
+
clear_bit_unlock(RTC_DEV_BUSY, &rtc->flags);
return 0;
}
-static int rtc_dev_fasync(int fd, struct file *file, int on)
-{
- struct rtc_device *rtc = file->private_data;
- return fasync_helper(fd, file, on, &rtc->async_queue);
-}
-
static const struct file_operations rtc_dev_fops = {
.owner = THIS_MODULE,
.llseek = no_llseek,
{
char s[80];
- sprintf(s, "%s sc:%x ", cdev->dev.bus_id, irq_ptr->schid.sch_no);
-
+ sprintf(s, "qdio: %s ", dev_name(&cdev->dev));
switch (irq_ptr->qib.qfmt) {
case QDIO_QETH_QFMT:
- sprintf(s + strlen(s), "OSADE ");
+ sprintf(s + strlen(s), "OSA ");
break;
case QDIO_ZFCP_QFMT:
sprintf(s + strlen(s), "ZFCP ");
break;
case QDIO_IQDIO_QFMT:
- sprintf(s + strlen(s), "HiperSockets ");
+ sprintf(s + strlen(s), "HS ");
break;
}
- sprintf(s + strlen(s), "using: ");
-
- if (!is_thinint_irq(irq_ptr))
- sprintf(s + strlen(s), "no");
- sprintf(s + strlen(s), "AdapterInterrupts ");
- if (!(irq_ptr->sch_token != 0))
- sprintf(s + strlen(s), "no");
- sprintf(s + strlen(s), "QEBSM ");
- if (!(irq_ptr->qib.ac & QIB_AC_OUTBOUND_PCI_SUPPORTED))
- sprintf(s + strlen(s), "no");
- sprintf(s + strlen(s), "OutboundPCI ");
- if (!css_general_characteristics.aif_tdd)
- sprintf(s + strlen(s), "no");
- sprintf(s + strlen(s), "TDD\n");
- printk(KERN_INFO "qdio: %s", s);
-
- memset(s, 0, sizeof(s));
- sprintf(s, "%s SIGA required: ", cdev->dev.bus_id);
- if (irq_ptr->siga_flag.input)
- sprintf(s + strlen(s), "Read ");
- if (irq_ptr->siga_flag.output)
- sprintf(s + strlen(s), "Write ");
- if (irq_ptr->siga_flag.sync)
- sprintf(s + strlen(s), "Sync ");
- if (!irq_ptr->siga_flag.no_sync_ti)
- sprintf(s + strlen(s), "SyncAI ");
- if (!irq_ptr->siga_flag.no_sync_out_ti)
- sprintf(s + strlen(s), "SyncOutAI ");
- if (!irq_ptr->siga_flag.no_sync_out_pci)
- sprintf(s + strlen(s), "SyncOutPCI");
+ sprintf(s + strlen(s), "on SC %x using ", irq_ptr->schid.sch_no);
+ sprintf(s + strlen(s), "AI:%d ", is_thinint_irq(irq_ptr));
+ sprintf(s + strlen(s), "QEBSM:%d ", (irq_ptr->sch_token) ? 1 : 0);
+ sprintf(s + strlen(s), "PCI:%d ",
+ (irq_ptr->qib.ac & QIB_AC_OUTBOUND_PCI_SUPPORTED) ? 1 : 0);
+ sprintf(s + strlen(s), "TDD:%d ", css_general_characteristics.aif_tdd);
+ sprintf(s + strlen(s), "SIGA:");
+ sprintf(s + strlen(s), "%s", (irq_ptr->siga_flag.input) ? "R" : " ");
+ sprintf(s + strlen(s), "%s", (irq_ptr->siga_flag.output) ? "W" : " ");
+ sprintf(s + strlen(s), "%s", (irq_ptr->siga_flag.sync) ? "S" : " ");
+ sprintf(s + strlen(s), "%s",
+ (!irq_ptr->siga_flag.no_sync_ti) ? "A" : " ");
+ sprintf(s + strlen(s), "%s",
+ (!irq_ptr->siga_flag.no_sync_out_ti) ? "O" : " ");
+ sprintf(s + strlen(s), "%s",
+ (!irq_ptr->siga_flag.no_sync_out_pci) ? "P" : " ");
sprintf(s + strlen(s), "\n");
- printk(KERN_INFO "qdio: %s", s);
+ printk(KERN_INFO "%s", s);
}
int __init qdio_setup_init(void)
goto msg_rejected;
}
- if (t->speed_hz < orion_spi->min_speed) {
+ if (t->speed_hz && t->speed_hz < orion_spi->min_speed) {
dev_err(&spi->dev,
"message rejected : "
"device min speed (%d Hz) exceeds "
if (!fbcon_is_inactive(vc, info)) {
if (ops->blank_state != blank) {
+ int ret = 1;
+
ops->blank_state = blank;
fbcon_cursor(vc, blank ? CM_ERASE : CM_DRAW);
ops->cursor_flash = (!blank);
- if (fb_blank(info, blank))
+ if (info->fbops->fb_blank)
+ ret = info->fbops->fb_blank(blank, info);
+ if (ret)
fbcon_generic_blank(vc, info, blank);
}
--- /dev/null
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2008 Kevin D. Kissell
+ */
+
+/*
+ * Definitions used for common event timer implementation
+ * for MIPS 4K-type processors and their MIPS MT variants.
+ * Avoids unsightly extern declarations in C files.
+ */
+#ifndef __ASM_CEVT_R4K_H
+#define __ASM_CEVT_R4K_H
+
+DECLARE_PER_CPU(struct clock_event_device, mips_clockevent_device);
+
+void mips_event_handler(struct clock_event_device *dev);
+int c0_compare_int_usable(void);
+void mips_set_clock_mode(enum clock_event_mode, struct clock_event_device *);
+irqreturn_t c0_compare_interrupt(int, void *);
+
+extern struct irqaction c0_compare_irqaction;
+extern int cp0_timer_irq_installed;
+
+/*
+ * Possibly handle a performance counter interrupt.
+ * Return true if the timer interrupt should not be checked
+ */
+
+static inline int handle_perf_irq(int r2)
+{
+ /*
+ * The performance counter overflow interrupt may be shared with the
+ * timer interrupt (cp0_perfcount_irq < 0). If it is and a
+ * performance counter has overflowed (perf_irq() == IRQ_HANDLED)
+ * and we can't reliably determine if a counter interrupt has also
+ * happened (!r2) then don't check for a timer interrupt.
+ */
+ return (cp0_perfcount_irq < 0) &&
+ perf_irq() == IRQ_HANDLED &&
+ !r2;
+}
+
+#endif /* __ASM_CEVT_R4K_H */
" .set pop \n"
" .endm");
+extern void smtc_ipi_replay(void);
+
static inline void raw_local_irq_enable(void)
{
+#ifdef CONFIG_MIPS_MT_SMTC
+ /*
+ * SMTC kernel needs to do a software replay of queued
+ * IPIs, at the cost of call overhead on each local_irq_enable()
+ */
+ smtc_ipi_replay();
+#endif
__asm__ __volatile__(
"raw_local_irq_enable"
: /* no outputs */
: "memory");
}
+
/*
* For cli() we have to insert nops to make sure that the new value
* has actually arrived in the status register before the end of this
" .set pop \n"
" .endm \n");
-extern void smtc_ipi_replay(void);
static inline void raw_local_irq_restore(unsigned long flags)
{
unsigned long __tmp1;
-#ifdef CONFIG_MIPS_MT_SMTC_INSTANT_REPLAY
+#ifdef CONFIG_MIPS_MT_SMTC
/*
- * CONFIG_MIPS_MT_SMTC_INSTANT_REPLAY does prompt replay of deferred
+ * SMTC kernel needs to do a software replay of queued
* IPIs, at the cost of branch and call overhead on each
* local_irq_restore()
*/
: "memory");
}
+static inline void __raw_local_irq_restore(unsigned long flags)
+{
+ unsigned long __tmp1;
+
+ __asm__ __volatile__(
+ "raw_local_irq_restore\t%0"
+ : "=r" (__tmp1)
+ : "0" (flags)
+ : "memory");
+}
+
static inline int raw_irqs_disabled_flags(unsigned long flags)
{
#ifdef CONFIG_MIPS_MT_SMTC
{ \
unsigned int res; \
unsigned int omt; \
- unsigned int flags; \
+ unsigned long flags; \
\
local_irq_save(flags); \
omt = __dmt(); \
{ \
unsigned int res; \
unsigned int omt; \
- unsigned int flags; \
+ unsigned long flags; \
\
local_irq_save(flags); \
omt = __dmt(); \
{ \
unsigned int res; \
unsigned int omt; \
- unsigned int flags; \
+ unsigned long flags; \
\
local_irq_save(flags); \
\
*/
#include <asm/mips_mt.h>
+#include <asm/smtc_ipi.h>
/*
* System-wide SMTC status information
struct task_struct;
void smtc_get_new_mmu_context(struct mm_struct *mm, unsigned long cpu);
-
+void self_ipi(struct smtc_ipi *);
void smtc_flush_tlb_asid(unsigned long asid);
-extern int mipsmt_build_cpu_map(int startslot);
-extern void mipsmt_prepare_cpus(void);
+extern int smtc_build_cpu_map(int startslot);
+extern void smtc_prepare_cpus(int cpus);
extern void smtc_smp_finish(void);
extern void smtc_boot_secondary(int cpu, struct task_struct *t);
extern void smtc_cpus_done(void);
+
/*
* Sharing the TLB between multiple VPEs means that the
* "random" index selection function is not allowed to
#ifndef __ASM_SN_MAPPED_KERNEL_H
#define __ASM_SN_MAPPED_KERNEL_H
+#include <linux/mmzone.h>
+
/*
* Note on how mapped kernels work: the text and data section is
* compiled at cksseg segment (LOADADDR = 0xc001c000), and the
#define MAPPED_ADDR_RO_TO_PHYS(x) (x - REP_BASE)
#define MAPPED_ADDR_RW_TO_PHYS(x) (x - REP_BASE - 16777216)
-#define MAPPED_KERN_RO_PHYSBASE(n) \
- (PLAT_NODE_DATA(n)->kern_vars.kv_ro_baseaddr)
-#define MAPPED_KERN_RW_PHYSBASE(n) \
- (PLAT_NODE_DATA(n)->kern_vars.kv_rw_baseaddr)
+#define MAPPED_KERN_RO_PHYSBASE(n) (hub_data(n)->kern_vars.kv_ro_baseaddr)
+#define MAPPED_KERN_RW_PHYSBASE(n) (hub_data(n)->kern_vars.kv_rw_baseaddr)
#define MAPPED_KERN_RO_TO_PHYS(x) \
((unsigned long)MAPPED_ADDR_RO_TO_PHYS(x) | \
#ifdef CONFIG_MIPS_MT_SMTC
.set mips32r2
/*
- * This may not really be necessary if ints are already
- * inhibited here.
+ * We need to make sure the read-modify-write
+ * of Status below isn't perturbed by an interrupt
+ * or cross-TC access, so we need to do at least a DMT,
+ * protected by an interrupt-inhibit. But setting IXMT
+ * also creates a few-cycle window where an IPI could
+ * be queued and not be detected before potentially
+ * returning to a WAIT or user-mode loop. It must be
+ * replayed.
+ *
+ * We're in the middle of a context switch, and
+ * we can't dispatch it directly without trashing
+ * some registers, so we'll try to detect this unlikely
+ * case and program a software interrupt in the VPE,
+ * as would be done for a cross-VPE IPI. To accomodate
+ * the handling of that case, we're doing a DVPE instead
+ * of just a DMT here to protect against other threads.
+ * This is a lot of cruft to cover a tiny window.
+ * If you can find a better design, implement it!
+ *
*/
mfc0 v0, CP0_TCSTATUS
ori v0, TCSTATUS_IXMT
mtc0 v0, CP0_TCSTATUS
_ehb
- DMT 5 # dmt a1
+ DVPE 5 # dvpe a1
jal mips_ihb
#endif /* CONFIG_MIPS_MT_SMTC */
mfc0 a0, CP0_STATUS
*/
LONG_L v1, PT_TCSTATUS(sp)
_ehb
- mfc0 v0, CP0_TCSTATUS
+ mfc0 a0, CP0_TCSTATUS
andi v1, TCSTATUS_IXMT
- /* We know that TCStatua.IXMT should be set from above */
- xori v0, v0, TCSTATUS_IXMT
- or v0, v0, v1
- mtc0 v0, CP0_TCSTATUS
- _ehb
- andi a1, a1, VPECONTROL_TE
+ bnez v1, 0f
+
+/*
+ * We'd like to detect any IPIs queued in the tiny window
+ * above and request an software interrupt to service them
+ * when we ERET.
+ *
+ * Computing the offset into the IPIQ array of the executing
+ * TC's IPI queue in-line would be tedious. We use part of
+ * the TCContext register to hold 16 bits of offset that we
+ * can add in-line to find the queue head.
+ */
+ mfc0 v0, CP0_TCCONTEXT
+ la a2, IPIQ
+ srl v0, v0, 16
+ addu a2, a2, v0
+ LONG_L v0, 0(a2)
+ beqz v0, 0f
+/*
+ * If we have a queue, provoke dispatch within the VPE by setting C_SW1
+ */
+ mfc0 v0, CP0_CAUSE
+ ori v0, v0, C_SW1
+ mtc0 v0, CP0_CAUSE
+0:
+ /*
+ * This test should really never branch but
+ * let's be prudent here. Having atomized
+ * the shared register modifications, we can
+ * now EVPE, and must do so before interrupts
+ * are potentially re-enabled.
+ */
+ andi a1, a1, MVPCONTROL_EVP
beqz a1, 1f
- emt
+ evpe
1:
+ /* We know that TCStatua.IXMT should be set from above */
+ xori a0, a0, TCSTATUS_IXMT
+ or a0, a0, v1
+ mtc0 a0, CP0_TCSTATUS
+ _ehb
+
.set mips0
#endif /* CONFIG_MIPS_MT_SMTC */
LONG_L v1, PT_EPC(sp)
#include <linux/wait.h>
+/**
+ * struct completion - structure used to maintain state for a "completion"
+ *
+ * This is the opaque structure used to maintain the state for a "completion".
+ * Completions currently use a FIFO to queue threads that have to wait for
+ * the "completion" event.
+ *
+ * See also: complete(), wait_for_completion() (and friends _timeout,
+ * _interruptible, _interruptible_timeout, and _killable), init_completion(),
+ * and macros DECLARE_COMPLETION(), DECLARE_COMPLETION_ONSTACK(), and
+ * INIT_COMPLETION().
+ */
struct completion {
unsigned int done;
wait_queue_head_t wait;
#define COMPLETION_INITIALIZER_ONSTACK(work) \
({ init_completion(&work); work; })
+/**
+ * DECLARE_COMPLETION: - declare and initialize a completion structure
+ * @work: identifier for the completion structure
+ *
+ * This macro declares and initializes a completion structure. Generally used
+ * for static declarations. You should use the _ONSTACK variant for automatic
+ * variables.
+ */
#define DECLARE_COMPLETION(work) \
struct completion work = COMPLETION_INITIALIZER(work)
* completions - so we use the _ONSTACK() variant for those that
* are on the kernel stack:
*/
+/**
+ * DECLARE_COMPLETION_ONSTACK: - declare and initialize a completion structure
+ * @work: identifier for the completion structure
+ *
+ * This macro declares and initializes a completion structure on the kernel
+ * stack.
+ */
#ifdef CONFIG_LOCKDEP
# define DECLARE_COMPLETION_ONSTACK(work) \
struct completion work = COMPLETION_INITIALIZER_ONSTACK(work)
# define DECLARE_COMPLETION_ONSTACK(work) DECLARE_COMPLETION(work)
#endif
+/**
+ * init_completion: - Initialize a dynamically allocated completion
+ * @x: completion structure that is to be initialized
+ *
+ * This inline function will initialize a dynamically created completion
+ * structure.
+ */
static inline void init_completion(struct completion *x)
{
x->done = 0;
extern void complete(struct completion *);
extern void complete_all(struct completion *);
+/**
+ * INIT_COMPLETION: - reinitialize a completion structure
+ * @x: completion structure to be reinitialized
+ *
+ * This macro should be used to reinitialize a completion structure so it can
+ * be reused. This is especially important after complete_all() is used.
+ */
#define INIT_COMPLETION(x) ((x).done = 0)
#endif
int cpu_up(unsigned int cpu);
+void notify_cpu_starting(unsigned int cpu);
extern void cpu_hotplug_init(void);
extern void cpu_maps_update_begin(void);
extern void cpu_maps_update_done(void);
/* Currently on a filemark */
IDE_AFLAG_FILEMARK = (1 << 25),
/* 0 = no tape is loaded, so we don't rewind after ejecting */
- IDE_AFLAG_MEDIUM_PRESENT = (1 << 26)
+ IDE_AFLAG_MEDIUM_PRESENT = (1 << 26),
+
+ IDE_AFLAG_NO_AUTOCLOSE = (1 << 27),
};
struct ide_drive_s {
#define CPU_DOWN_FAILED 0x0006 /* CPU (unsigned)v NOT going down */
#define CPU_DEAD 0x0007 /* CPU (unsigned)v dead */
#define CPU_DYING 0x0008 /* CPU (unsigned)v not running any task,
- * not handling interrupts, soon dead */
+ * not handling interrupts, soon dead.
+ * Called on the dying cpu, interrupts
+ * are already disabled. Must not
+ * sleep, must not fail */
#define CPU_POST_DEAD 0x0009 /* CPU (unsigned)v dead, cpu_hotplug
* lock is dropped */
+#define CPU_STARTING 0x000A /* CPU (unsigned)v soon running.
+ * Called on the new cpu, just before
+ * enabling interrupts. Must not sleep,
+ * must not fail */
/* Used for CPU hotplug events occuring while tasks are frozen due to a suspend
* operation in progress
#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
#define CPU_DEAD_FROZEN (CPU_DEAD | CPU_TASKS_FROZEN)
#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
+#define CPU_STARTING_FROZEN (CPU_STARTING | CPU_TASKS_FROZEN)
/* Hibernation and suspend events */
#define PM_HIBERNATION_PREPARE 0x0001 /* Going to hibernate */
* snapshot of the last seen global state
* and a lock protecting this state
*/
- int shift;
unsigned long period;
+ int shift;
spinlock_t lock; /* protect the snapshot state */
};
* - everyone except group_exit_task is stopped during signal delivery
* of fatal signals, group_exit_task processes the signal.
*/
- struct task_struct *group_exit_task;
int notify_count;
+ struct task_struct *group_exit_task;
/* thread group stop support, overloads group_exit_code too */
int group_stop_count;
void (*yield_task) (struct rq *rq);
int (*select_task_rq)(struct task_struct *p, int sync);
- void (*check_preempt_curr) (struct rq *rq, struct task_struct *p);
+ void (*check_preempt_curr) (struct rq *rq, struct task_struct *p, int sync);
struct task_struct * (*pick_next_task) (struct rq *rq);
void (*put_prev_task) (struct rq *rq, struct task_struct *p);
struct sched_rt_entity {
struct list_head run_list;
- unsigned int time_slice;
unsigned long timeout;
+ unsigned int time_slice;
int nr_cpus_allowed;
struct sched_rt_entity *back;
#ifndef __LINUX_STACKTRACE_H
#define __LINUX_STACKTRACE_H
+struct task_struct;
+
#ifdef CONFIG_STACKTRACE
struct stack_trace {
unsigned int nr_entries, max_entries;
int result;
if (initcall_debug) {
- print_fn_descriptor_symbol("calling %s\n", fn);
+ printk("calling %pF\n", fn);
t0 = ktime_get();
}
t1 = ktime_get();
delta = ktime_sub(t1, t0);
- print_fn_descriptor_symbol("initcall %s", fn);
- printk(" returned %d after %Ld msecs\n", result,
+ printk("initcall %pF returned %d after %Ld msecs\n",
+ fn, result,
(unsigned long long) delta.tv64 >> 20);
}
local_irq_enable();
}
if (msgbuf[0]) {
- print_fn_descriptor_symbol(KERN_WARNING "initcall %s", fn);
- printk(" returned with %s\n", msgbuf);
+ printk("initcall %pF returned with %s\n", fn, msgbuf);
}
return result;
struct take_cpu_down_param *param = _param;
int err;
- raw_notifier_call_chain(&cpu_chain, CPU_DYING | param->mod,
- param->hcpu);
/* Ensure this CPU doesn't handle any more interrupts. */
err = __cpu_disable();
if (err < 0)
return err;
+ raw_notifier_call_chain(&cpu_chain, CPU_DYING | param->mod,
+ param->hcpu);
+
/* Force idle task to run as soon as we yield: it should
immediately notice cpu is offline and die quickly. */
sched_idle_next();
}
#endif /* CONFIG_PM_SLEEP_SMP */
+/**
+ * notify_cpu_starting(cpu) - call the CPU_STARTING notifiers
+ * @cpu: cpu that just started
+ *
+ * This function calls the cpu_chain notifiers with CPU_STARTING.
+ * It must be called by the arch code on the new cpu, before the new cpu
+ * enables interrupts and before the "boot" cpu returns from __cpu_up().
+ */
+void notify_cpu_starting(unsigned int cpu)
+{
+ unsigned long val = CPU_STARTING;
+
+#ifdef CONFIG_PM_SLEEP_SMP
+ if (cpu_isset(cpu, frozen_cpus))
+ val = CPU_STARTING_FROZEN;
+#endif /* CONFIG_PM_SLEEP_SMP */
+ raw_notifier_call_chain(&cpu_chain, val, (void *)(long)cpu);
+}
+
#endif /* CONFIG_SMP */
/*
/* Signal the primary CPU that we are done: */
atomic_set(&cpu_in_kgdb[cpu], 0);
+ touch_softlockup_watchdog();
clocksource_touch_watchdog();
local_irq_restore(flags);
}
atomic_read(&kgdb_cpu_doing_single_step) != cpu) {
atomic_set(&kgdb_active, -1);
+ touch_softlockup_watchdog();
clocksource_touch_watchdog();
local_irq_restore(flags);
kgdb_restore:
/* Free kgdb_active */
atomic_set(&kgdb_active, -1);
+ touch_softlockup_watchdog();
clocksource_touch_watchdog();
local_irq_restore(flags);
rt_b->rt_period_timer.cb_mode = HRTIMER_CB_IRQSAFE_UNLOCKED;
}
+static inline int rt_bandwidth_enabled(void)
+{
+ return sysctl_sched_rt_runtime >= 0;
+}
+
static void start_rt_bandwidth(struct rt_bandwidth *rt_b)
{
ktime_t now;
- if (rt_b->rt_runtime == RUNTIME_INF)
+ if (rt_bandwidth_enabled() && rt_b->rt_runtime == RUNTIME_INF)
return;
if (hrtimer_active(&rt_b->rt_period_timer))
static DEFINE_PER_CPU(struct sched_rt_entity, init_sched_rt_entity);
static DEFINE_PER_CPU(struct rt_rq, init_rt_rq) ____cacheline_aligned_in_smp;
#endif /* CONFIG_RT_GROUP_SCHED */
-#else /* !CONFIG_FAIR_GROUP_SCHED */
+#else /* !CONFIG_USER_SCHED */
#define root_task_group init_task_group
-#endif /* CONFIG_FAIR_GROUP_SCHED */
+#endif /* CONFIG_USER_SCHED */
/* task_group_lock serializes add/remove of task groups and also changes to
* a task group's cpu shares.
static DEFINE_PER_CPU_SHARED_ALIGNED(struct rq, runqueues);
-static inline void check_preempt_curr(struct rq *rq, struct task_struct *p)
+static inline void check_preempt_curr(struct rq *rq, struct task_struct *p, int sync)
{
- rq->curr->sched_class->check_preempt_curr(rq, p);
+ rq->curr->sched_class->check_preempt_curr(rq, p, sync);
}
static inline int cpu_of(struct rq *rq)
hrtimer_start(&rq->hrtick_timer, ns_to_ktime(delay), HRTIMER_MODE_REL);
}
-static void init_hrtick(void)
+static inline void init_hrtick(void)
{
}
#endif /* CONFIG_SMP */
rq->hrtick_timer.function = hrtick;
rq->hrtick_timer.cb_mode = HRTIMER_CB_IRQSAFE_PERCPU;
}
-#else
+#else /* CONFIG_SCHED_HRTICK */
static inline void hrtick_clear(struct rq *rq)
{
}
static inline void init_hrtick(void)
{
}
-#endif
+#endif /* CONFIG_SCHED_HRTICK */
/*
* resched_task - mark a task 'to be rescheduled now'.
update_load_sub(&rq->load, load);
}
-#ifdef CONFIG_SMP
-static unsigned long source_load(int cpu, int type);
-static unsigned long target_load(int cpu, int type);
-static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd);
-
-static unsigned long cpu_avg_load_per_task(int cpu)
-{
- struct rq *rq = cpu_rq(cpu);
-
- if (rq->nr_running)
- rq->avg_load_per_task = rq->load.weight / rq->nr_running;
-
- return rq->avg_load_per_task;
-}
-
-#ifdef CONFIG_FAIR_GROUP_SCHED
-
-typedef void (*tg_visitor)(struct task_group *, int, struct sched_domain *);
+#if (defined(CONFIG_SMP) && defined(CONFIG_FAIR_GROUP_SCHED)) || defined(CONFIG_RT_GROUP_SCHED)
+typedef int (*tg_visitor)(struct task_group *, void *);
/*
* Iterate the full tree, calling @down when first entering a node and @up when
* leaving it for the final time.
*/
-static void
-walk_tg_tree(tg_visitor down, tg_visitor up, int cpu, struct sched_domain *sd)
+static int walk_tg_tree(tg_visitor down, tg_visitor up, void *data)
{
struct task_group *parent, *child;
+ int ret;
rcu_read_lock();
parent = &root_task_group;
down:
- (*down)(parent, cpu, sd);
+ ret = (*down)(parent, data);
+ if (ret)
+ goto out_unlock;
list_for_each_entry_rcu(child, &parent->children, siblings) {
parent = child;
goto down;
up:
continue;
}
- (*up)(parent, cpu, sd);
+ ret = (*up)(parent, data);
+ if (ret)
+ goto out_unlock;
child = parent;
parent = parent->parent;
if (parent)
goto up;
+out_unlock:
rcu_read_unlock();
+
+ return ret;
}
+static int tg_nop(struct task_group *tg, void *data)
+{
+ return 0;
+}
+#endif
+
+#ifdef CONFIG_SMP
+static unsigned long source_load(int cpu, int type);
+static unsigned long target_load(int cpu, int type);
+static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd);
+
+static unsigned long cpu_avg_load_per_task(int cpu)
+{
+ struct rq *rq = cpu_rq(cpu);
+
+ if (rq->nr_running)
+ rq->avg_load_per_task = rq->load.weight / rq->nr_running;
+
+ return rq->avg_load_per_task;
+}
+
+#ifdef CONFIG_FAIR_GROUP_SCHED
+
static void __set_se_shares(struct sched_entity *se, unsigned long shares);
/*
* This needs to be done in a bottom-up fashion because the rq weight of a
* parent group depends on the shares of its child groups.
*/
-static void
-tg_shares_up(struct task_group *tg, int cpu, struct sched_domain *sd)
+static int tg_shares_up(struct task_group *tg, void *data)
{
unsigned long rq_weight = 0;
unsigned long shares = 0;
+ struct sched_domain *sd = data;
int i;
for_each_cpu_mask(i, sd->span) {
__update_group_shares_cpu(tg, i, shares, rq_weight);
spin_unlock_irqrestore(&rq->lock, flags);
}
+
+ return 0;
}
/*
* This needs to be done in a top-down fashion because the load of a child
* group is a fraction of its parents load.
*/
-static void
-tg_load_down(struct task_group *tg, int cpu, struct sched_domain *sd)
+static int tg_load_down(struct task_group *tg, void *data)
{
unsigned long load;
+ long cpu = (long)data;
if (!tg->parent) {
load = cpu_rq(cpu)->load.weight;
}
tg->cfs_rq[cpu]->h_load = load;
-}
-static void
-tg_nop(struct task_group *tg, int cpu, struct sched_domain *sd)
-{
+ return 0;
}
static void update_shares(struct sched_domain *sd)
if (elapsed >= (s64)(u64)sysctl_sched_shares_ratelimit) {
sd->last_update = now;
- walk_tg_tree(tg_nop, tg_shares_up, 0, sd);
+ walk_tg_tree(tg_nop, tg_shares_up, sd);
}
}
spin_lock(&rq->lock);
}
-static void update_h_load(int cpu)
+static void update_h_load(long cpu)
{
- walk_tg_tree(tg_load_down, tg_nop, cpu, NULL);
+ walk_tg_tree(tg_load_down, tg_nop, (void *)cpu);
}
#else
running = task_running(rq, p);
on_rq = p->se.on_rq;
ncsw = 0;
- if (!match_state || p->state == match_state) {
- ncsw = p->nivcsw + p->nvcsw;
- if (unlikely(!ncsw))
- ncsw = 1;
- }
+ if (!match_state || p->state == match_state)
+ ncsw = p->nvcsw | LONG_MIN; /* sets MSB */
task_rq_unlock(rq, &flags);
/*
trace_mark(kernel_sched_wakeup,
"pid %d state %ld ## rq %p task %p rq->curr %p",
p->pid, p->state, rq, p, rq->curr);
- check_preempt_curr(rq, p);
+ check_preempt_curr(rq, p, sync);
p->state = TASK_RUNNING;
#ifdef CONFIG_SMP
trace_mark(kernel_sched_wakeup_new,
"pid %d state %ld ## rq %p task %p rq->curr %p",
p->pid, p->state, rq, p, rq->curr);
- check_preempt_curr(rq, p);
+ check_preempt_curr(rq, p, 0);
#ifdef CONFIG_SMP
if (p->sched_class->task_wake_up)
p->sched_class->task_wake_up(rq, p);
* Note that idle threads have a prio of MAX_PRIO, for this test
* to be always true for them.
*/
- check_preempt_curr(this_rq, p);
+ check_preempt_curr(this_rq, p, 0);
}
/*
}
EXPORT_SYMBOL_GPL(__wake_up_sync); /* For internal use only */
+/**
+ * complete: - signals a single thread waiting on this completion
+ * @x: holds the state of this particular completion
+ *
+ * This will wake up a single thread waiting on this completion. Threads will be
+ * awakened in the same order in which they were queued.
+ *
+ * See also complete_all(), wait_for_completion() and related routines.
+ */
void complete(struct completion *x)
{
unsigned long flags;
}
EXPORT_SYMBOL(complete);
+/**
+ * complete_all: - signals all threads waiting on this completion
+ * @x: holds the state of this particular completion
+ *
+ * This will wake up all threads waiting on this particular completion event.
+ */
void complete_all(struct completion *x)
{
unsigned long flags;
wait.flags |= WQ_FLAG_EXCLUSIVE;
__add_wait_queue_tail(&x->wait, &wait);
do {
- if ((state == TASK_INTERRUPTIBLE &&
- signal_pending(current)) ||
- (state == TASK_KILLABLE &&
- fatal_signal_pending(current))) {
+ if (signal_pending_state(state, current)) {
timeout = -ERESTARTSYS;
break;
}
return timeout;
}
+/**
+ * wait_for_completion: - waits for completion of a task
+ * @x: holds the state of this particular completion
+ *
+ * This waits to be signaled for completion of a specific task. It is NOT
+ * interruptible and there is no timeout.
+ *
+ * See also similar routines (i.e. wait_for_completion_timeout()) with timeout
+ * and interrupt capability. Also see complete().
+ */
void __sched wait_for_completion(struct completion *x)
{
wait_for_common(x, MAX_SCHEDULE_TIMEOUT, TASK_UNINTERRUPTIBLE);
}
EXPORT_SYMBOL(wait_for_completion);
+/**
+ * wait_for_completion_timeout: - waits for completion of a task (w/timeout)
+ * @x: holds the state of this particular completion
+ * @timeout: timeout value in jiffies
+ *
+ * This waits for either a completion of a specific task to be signaled or for a
+ * specified timeout to expire. The timeout is in jiffies. It is not
+ * interruptible.
+ */
unsigned long __sched
wait_for_completion_timeout(struct completion *x, unsigned long timeout)
{
}
EXPORT_SYMBOL(wait_for_completion_timeout);
+/**
+ * wait_for_completion_interruptible: - waits for completion of a task (w/intr)
+ * @x: holds the state of this particular completion
+ *
+ * This waits for completion of a specific task to be signaled. It is
+ * interruptible.
+ */
int __sched wait_for_completion_interruptible(struct completion *x)
{
long t = wait_for_common(x, MAX_SCHEDULE_TIMEOUT, TASK_INTERRUPTIBLE);
}
EXPORT_SYMBOL(wait_for_completion_interruptible);
+/**
+ * wait_for_completion_interruptible_timeout: - waits for completion (w/(to,intr))
+ * @x: holds the state of this particular completion
+ * @timeout: timeout value in jiffies
+ *
+ * This waits for either a completion of a specific task to be signaled or for a
+ * specified timeout to expire. It is interruptible. The timeout is in jiffies.
+ */
unsigned long __sched
wait_for_completion_interruptible_timeout(struct completion *x,
unsigned long timeout)
}
EXPORT_SYMBOL(wait_for_completion_interruptible_timeout);
+/**
+ * wait_for_completion_killable: - waits for completion of a task (killable)
+ * @x: holds the state of this particular completion
+ *
+ * This waits to be signaled for completion of a specific task. It can be
+ * interrupted by a kill signal.
+ */
int __sched wait_for_completion_killable(struct completion *x)
{
long t = wait_for_common(x, MAX_SCHEDULE_TIMEOUT, TASK_KILLABLE);
* Do not allow realtime tasks into groups that have no runtime
* assigned.
*/
- if (rt_policy(policy) && task_group(p)->rt_bandwidth.rt_runtime == 0)
+ if (rt_bandwidth_enabled() && rt_policy(policy) &&
+ task_group(p)->rt_bandwidth.rt_runtime == 0)
return -EPERM;
#endif
set_task_cpu(p, dest_cpu);
if (on_rq) {
activate_task(rq_dest, p, 0);
- check_preempt_curr(rq_dest, p);
+ check_preempt_curr(rq_dest, p, 0);
}
done:
ret = 1;
#ifdef in_atomic
static unsigned long prev_jiffy; /* ratelimiting */
- if ((in_atomic() || irqs_disabled()) &&
- system_state == SYSTEM_RUNNING && !oops_in_progress) {
- if (time_before(jiffies, prev_jiffy + HZ) && prev_jiffy)
- return;
- prev_jiffy = jiffies;
- printk(KERN_ERR "BUG: sleeping function called from invalid"
- " context at %s:%d\n", file, line);
- printk("in_atomic():%d, irqs_disabled():%d\n",
- in_atomic(), irqs_disabled());
- debug_show_held_locks(current);
- if (irqs_disabled())
- print_irqtrace_events(current);
- dump_stack();
- }
+ if ((!in_atomic() && !irqs_disabled()) ||
+ system_state != SYSTEM_RUNNING || oops_in_progress)
+ return;
+ if (time_before(jiffies, prev_jiffy + HZ) && prev_jiffy)
+ return;
+ prev_jiffy = jiffies;
+
+ printk(KERN_ERR
+ "BUG: sleeping function called from invalid context at %s:%d\n",
+ file, line);
+ printk(KERN_ERR
+ "in_atomic(): %d, irqs_disabled(): %d, pid: %d, name: %s\n",
+ in_atomic(), irqs_disabled(),
+ current->pid, current->comm);
+
+ debug_show_held_locks(current);
+ if (irqs_disabled())
+ print_irqtrace_events(current);
+ dump_stack();
#endif
}
EXPORT_SYMBOL(__might_sleep);
static unsigned long to_ratio(u64 period, u64 runtime)
{
if (runtime == RUNTIME_INF)
- return 1ULL << 16;
+ return 1ULL << 20;
- return div64_u64(runtime << 16, period);
+ return div64_u64(runtime << 20, period);
}
-#ifdef CONFIG_CGROUP_SCHED
-static int __rt_schedulable(struct task_group *tg, u64 period, u64 runtime)
+/* Must be called with tasklist_lock held */
+static inline int tg_has_rt_tasks(struct task_group *tg)
{
- struct task_group *tgi, *parent = tg->parent;
- unsigned long total = 0;
+ struct task_struct *g, *p;
- if (!parent) {
- if (global_rt_period() < period)
- return 0;
+ do_each_thread(g, p) {
+ if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
+ return 1;
+ } while_each_thread(g, p);
- return to_ratio(period, runtime) <
- to_ratio(global_rt_period(), global_rt_runtime());
- }
+ return 0;
+}
- if (ktime_to_ns(parent->rt_bandwidth.rt_period) < period)
- return 0;
+struct rt_schedulable_data {
+ struct task_group *tg;
+ u64 rt_period;
+ u64 rt_runtime;
+};
- rcu_read_lock();
- list_for_each_entry_rcu(tgi, &parent->children, siblings) {
- if (tgi == tg)
- continue;
+static int tg_schedulable(struct task_group *tg, void *data)
+{
+ struct rt_schedulable_data *d = data;
+ struct task_group *child;
+ unsigned long total, sum = 0;
+ u64 period, runtime;
- total += to_ratio(ktime_to_ns(tgi->rt_bandwidth.rt_period),
- tgi->rt_bandwidth.rt_runtime);
+ period = ktime_to_ns(tg->rt_bandwidth.rt_period);
+ runtime = tg->rt_bandwidth.rt_runtime;
+
+ if (tg == d->tg) {
+ period = d->rt_period;
+ runtime = d->rt_runtime;
}
- rcu_read_unlock();
- return total + to_ratio(period, runtime) <=
- to_ratio(ktime_to_ns(parent->rt_bandwidth.rt_period),
- parent->rt_bandwidth.rt_runtime);
-}
-#elif defined CONFIG_USER_SCHED
-static int __rt_schedulable(struct task_group *tg, u64 period, u64 runtime)
-{
- struct task_group *tgi;
- unsigned long total = 0;
- unsigned long global_ratio =
- to_ratio(global_rt_period(), global_rt_runtime());
+ /*
+ * Cannot have more runtime than the period.
+ */
+ if (runtime > period && runtime != RUNTIME_INF)
+ return -EINVAL;
- rcu_read_lock();
- list_for_each_entry_rcu(tgi, &task_groups, list) {
- if (tgi == tg)
- continue;
+ /*
+ * Ensure we don't starve existing RT tasks.
+ */
+ if (rt_bandwidth_enabled() && !runtime && tg_has_rt_tasks(tg))
+ return -EBUSY;
+
+ total = to_ratio(period, runtime);
+
+ /*
+ * Nobody can have more than the global setting allows.
+ */
+ if (total > to_ratio(global_rt_period(), global_rt_runtime()))
+ return -EINVAL;
+
+ /*
+ * The sum of our children's runtime should not exceed our own.
+ */
+ list_for_each_entry_rcu(child, &tg->children, siblings) {
+ period = ktime_to_ns(child->rt_bandwidth.rt_period);
+ runtime = child->rt_bandwidth.rt_runtime;
+
+ if (child == d->tg) {
+ period = d->rt_period;
+ runtime = d->rt_runtime;
+ }
- total += to_ratio(ktime_to_ns(tgi->rt_bandwidth.rt_period),
- tgi->rt_bandwidth.rt_runtime);
+ sum += to_ratio(period, runtime);
}
- rcu_read_unlock();
- return total + to_ratio(period, runtime) < global_ratio;
+ if (sum > total)
+ return -EINVAL;
+
+ return 0;
}
-#endif
-/* Must be called with tasklist_lock held */
-static inline int tg_has_rt_tasks(struct task_group *tg)
+static int __rt_schedulable(struct task_group *tg, u64 period, u64 runtime)
{
- struct task_struct *g, *p;
- do_each_thread(g, p) {
- if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
- return 1;
- } while_each_thread(g, p);
- return 0;
+ struct rt_schedulable_data data = {
+ .tg = tg,
+ .rt_period = period,
+ .rt_runtime = runtime,
+ };
+
+ return walk_tg_tree(tg_schedulable, tg_nop, &data);
}
static int tg_set_bandwidth(struct task_group *tg,
mutex_lock(&rt_constraints_mutex);
read_lock(&tasklist_lock);
- if (rt_runtime == 0 && tg_has_rt_tasks(tg)) {
- err = -EBUSY;
+ err = __rt_schedulable(tg, rt_period, rt_runtime);
+ if (err)
goto unlock;
- }
- if (!__rt_schedulable(tg, rt_period, rt_runtime)) {
- err = -EINVAL;
- goto unlock;
- }
spin_lock_irq(&tg->rt_bandwidth.rt_runtime_lock);
tg->rt_bandwidth.rt_period = ns_to_ktime(rt_period);
static int sched_rt_global_constraints(void)
{
- struct task_group *tg = &root_task_group;
- u64 rt_runtime, rt_period;
+ u64 runtime, period;
int ret = 0;
if (sysctl_sched_rt_period <= 0)
return -EINVAL;
- rt_period = ktime_to_ns(tg->rt_bandwidth.rt_period);
- rt_runtime = tg->rt_bandwidth.rt_runtime;
+ runtime = global_rt_runtime();
+ period = global_rt_period();
+
+ /*
+ * Sanity check on the sysctl variables.
+ */
+ if (runtime > period && runtime != RUNTIME_INF)
+ return -EINVAL;
mutex_lock(&rt_constraints_mutex);
- if (!__rt_schedulable(tg, rt_period, rt_runtime))
- ret = -EINVAL;
+ read_lock(&tasklist_lock);
+ ret = __rt_schedulable(NULL, 0, 0);
+ read_unlock(&tasklist_lock);
mutex_unlock(&rt_constraints_mutex);
return ret;
if (!cgrp->parent) {
/* This is early initialization for the top cgroup */
- init_task_group.css.cgroup = cgrp;
return &init_task_group.css;
}
if (IS_ERR(tg))
return ERR_PTR(-ENOMEM);
- /* Bind the cgroup to task_group object we just created */
- tg->css.cgroup = cgrp;
-
return &tg->css;
}
return __sched_period(nr_running);
}
-/*
- * The goal of calc_delta_asym() is to be asymmetrically around NICE_0_LOAD, in
- * that it favours >=0 over <0.
- *
- * -20 |
- * |
- * 0 --------+-------
- * .'
- * 19 .'
- *
- */
-static unsigned long
-calc_delta_asym(unsigned long delta, struct sched_entity *se)
-{
- struct load_weight lw = {
- .weight = NICE_0_LOAD,
- .inv_weight = 1UL << (WMULT_SHIFT-NICE_0_SHIFT)
- };
-
- for_each_sched_entity(se) {
- struct load_weight *se_lw = &se->load;
- unsigned long rw = cfs_rq_of(se)->load.weight;
-
-#ifdef CONFIG_FAIR_SCHED_GROUP
- struct cfs_rq *cfs_rq = se->my_q;
- struct task_group *tg = NULL
-
- if (cfs_rq)
- tg = cfs_rq->tg;
-
- if (tg && tg->shares < NICE_0_LOAD) {
- /*
- * scale shares to what it would have been had
- * tg->weight been NICE_0_LOAD:
- *
- * weight = 1024 * shares / tg->weight
- */
- lw.weight *= se->load.weight;
- lw.weight /= tg->shares;
-
- lw.inv_weight = 0;
-
- se_lw = &lw;
- rw += lw.weight - se->load.weight;
- } else
-#endif
-
- if (se->load.weight < NICE_0_LOAD) {
- se_lw = &lw;
- rw += NICE_0_LOAD - se->load.weight;
- }
-
- delta = calc_delta_mine(delta, rw, se_lw);
- }
-
- return delta;
-}
-
/*
* Update the current task's runtime statistics. Skip current tasks that
* are not in our scheduling class.
update_load_add(&cfs_rq->load, se->load.weight);
if (!parent_entity(se))
inc_cpu_load(rq_of(cfs_rq), se->load.weight);
- if (entity_is_task(se))
+ if (entity_is_task(se)) {
add_cfs_task_weight(cfs_rq, se->load.weight);
+ list_add(&se->group_node, &cfs_rq->tasks);
+ }
cfs_rq->nr_running++;
se->on_rq = 1;
- list_add(&se->group_node, &cfs_rq->tasks);
}
static void
update_load_sub(&cfs_rq->load, se->load.weight);
if (!parent_entity(se))
dec_cpu_load(rq_of(cfs_rq), se->load.weight);
- if (entity_is_task(se))
+ if (entity_is_task(se)) {
add_cfs_task_weight(cfs_rq, -se->load.weight);
+ list_del_init(&se->group_node);
+ }
cfs_rq->nr_running--;
se->on_rq = 0;
- list_del_init(&se->group_node);
}
static void enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se)
long wl, long wg)
{
struct sched_entity *se = tg->se[cpu];
- long more_w;
if (!tg->parent)
return wl;
if (!wl && sched_feat(ASYM_EFF_LOAD))
return wl;
- /*
- * Instead of using this increment, also add the difference
- * between when the shares were last updated and now.
- */
- more_w = se->my_q->load.weight - se->my_q->rq_weight;
- wl += more_w;
- wg += more_w;
-
for_each_sched_entity(se) {
-#define D(n) (likely(n) ? (n) : 1)
-
long S, rw, s, a, b;
+ long more_w;
+
+ /*
+ * Instead of using this increment, also add the difference
+ * between when the shares were last updated and now.
+ */
+ more_w = se->my_q->load.weight - se->my_q->rq_weight;
+ wl += more_w;
+ wg += more_w;
S = se->my_q->tg->shares;
s = se->my_q->shares;
a = S*(rw + wl);
b = S*rw + s*wg;
- wl = s*(a-b)/D(b);
+ wl = s*(a-b);
+
+ if (likely(b))
+ wl /= b;
+
/*
* Assume the group is already running and will
* thus already be accounted for in the weight.
* alter the group weight.
*/
wg = 0;
-#undef D
}
return wl;
#endif
static int
-wake_affine(struct rq *rq, struct sched_domain *this_sd, struct rq *this_rq,
+wake_affine(struct sched_domain *this_sd, struct rq *this_rq,
struct task_struct *p, int prev_cpu, int this_cpu, int sync,
int idx, unsigned long load, unsigned long this_load,
unsigned int imbalance)
schedstat_inc(p, se.nr_wakeups_affine_attempts);
tl_per_task = cpu_avg_load_per_task(this_cpu);
- if ((tl <= load && tl + target_load(prev_cpu, idx) <= tl_per_task) ||
- balanced) {
+ if (balanced || (tl <= load && tl + target_load(prev_cpu, idx) <=
+ tl_per_task)) {
/*
* This domain has SD_WAKE_AFFINE and
* p is cache cold in this domain, and
struct sched_domain *sd, *this_sd = NULL;
int prev_cpu, this_cpu, new_cpu;
unsigned long load, this_load;
- struct rq *rq, *this_rq;
+ struct rq *this_rq;
unsigned int imbalance;
int idx;
prev_cpu = task_cpu(p);
- rq = task_rq(p);
this_cpu = smp_processor_id();
this_rq = cpu_rq(this_cpu);
new_cpu = prev_cpu;
+ if (prev_cpu == this_cpu)
+ goto out;
/*
* 'this_sd' is the first domain that both
* this_cpu and prev_cpu are present in:
load = source_load(prev_cpu, idx);
this_load = target_load(this_cpu, idx);
- if (wake_affine(rq, this_sd, this_rq, p, prev_cpu, this_cpu, sync, idx,
+ if (wake_affine(this_sd, this_rq, p, prev_cpu, this_cpu, sync, idx,
load, this_load, imbalance))
return this_cpu;
- if (prev_cpu == this_cpu)
- goto out;
-
/*
* Start passive balancing when half the imbalance_pct
* limit is reached.
* + nice tasks.
*/
if (sched_feat(ASYM_GRAN))
- gran = calc_delta_asym(sysctl_sched_wakeup_granularity, se);
- else
- gran = calc_delta_fair(sysctl_sched_wakeup_granularity, se);
+ gran = calc_delta_mine(gran, NICE_0_LOAD, &se->load);
return gran;
}
-/*
- * Should 'se' preempt 'curr'.
- *
- * |s1
- * |s2
- * |s3
- * g
- * |<--->|c
- *
- * w(c, s1) = -1
- * w(c, s2) = 0
- * w(c, s3) = 1
- *
- */
-static int
-wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se)
-{
- s64 gran, vdiff = curr->vruntime - se->vruntime;
-
- if (vdiff < 0)
- return -1;
-
- gran = wakeup_gran(curr);
- if (vdiff > gran)
- return 1;
-
- return 0;
-}
-
-/* return depth at which a sched entity is present in the hierarchy */
-static inline int depth_se(struct sched_entity *se)
-{
- int depth = 0;
-
- for_each_sched_entity(se)
- depth++;
-
- return depth;
-}
-
/*
* Preempt the current task with a newly woken task if needed:
*/
-static void check_preempt_wakeup(struct rq *rq, struct task_struct *p)
+static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int sync)
{
struct task_struct *curr = rq->curr;
struct cfs_rq *cfs_rq = task_cfs_rq(curr);
struct sched_entity *se = &curr->se, *pse = &p->se;
- int se_depth, pse_depth;
+ s64 delta_exec;
if (unlikely(rt_prio(p->prio))) {
update_rq_clock(rq);
cfs_rq_of(pse)->next = pse;
+ /*
+ * We can come here with TIF_NEED_RESCHED already set from new task
+ * wake up path.
+ */
+ if (test_tsk_need_resched(curr))
+ return;
+
/*
* Batch tasks do not preempt (their preemption is driven by
* the tick):
if (!sched_feat(WAKEUP_PREEMPT))
return;
- /*
- * preemption test can be made between sibling entities who are in the
- * same cfs_rq i.e who have a common parent. Walk up the hierarchy of
- * both tasks until we find their ancestors who are siblings of common
- * parent.
- */
-
- /* First walk up until both entities are at same depth */
- se_depth = depth_se(se);
- pse_depth = depth_se(pse);
-
- while (se_depth > pse_depth) {
- se_depth--;
- se = parent_entity(se);
- }
-
- while (pse_depth > se_depth) {
- pse_depth--;
- pse = parent_entity(pse);
- }
-
- while (!is_same_group(se, pse)) {
- se = parent_entity(se);
- pse = parent_entity(pse);
+ if (sched_feat(WAKEUP_OVERLAP) && sync &&
+ se->avg_overlap < sysctl_sched_migration_cost &&
+ pse->avg_overlap < sysctl_sched_migration_cost) {
+ resched_task(curr);
+ return;
}
- if (wakeup_preempt_entity(se, pse) == 1)
+ delta_exec = se->sum_exec_runtime - se->prev_sum_exec_runtime;
+ if (delta_exec > wakeup_gran(pse))
resched_task(curr);
}
if (next == &cfs_rq->tasks)
return NULL;
- /* Skip over entities that are not tasks */
- do {
- se = list_entry(next, struct sched_entity, group_node);
- next = next->next;
- } while (next != &cfs_rq->tasks && !entity_is_task(se));
-
- if (next == &cfs_rq->tasks)
- return NULL;
-
- cfs_rq->balance_iterator = next;
-
- if (entity_is_task(se))
- p = task_of(se);
+ se = list_entry(next, struct sched_entity, group_node);
+ p = task_of(se);
+ cfs_rq->balance_iterator = next->next;
return p;
}
rcu_read_lock();
update_h_load(busiest_cpu);
- list_for_each_entry(tg, &task_groups, list) {
+ list_for_each_entry_rcu(tg, &task_groups, list) {
struct cfs_rq *busiest_cfs_rq = tg->cfs_rq[busiest_cpu];
unsigned long busiest_h_load = busiest_cfs_rq->h_load;
unsigned long busiest_weight = busiest_cfs_rq->load.weight;
* 'current' within the tree based on its new key value.
*/
swap(curr->vruntime, se->vruntime);
+ resched_task(rq->curr);
}
enqueue_task_fair(rq, p, 0);
- resched_task(rq->curr);
}
/*
if (p->prio > oldprio)
resched_task(rq->curr);
} else
- check_preempt_curr(rq, p);
+ check_preempt_curr(rq, p, 0);
}
/*
if (running)
resched_task(rq->curr);
else
- check_preempt_curr(rq, p);
+ check_preempt_curr(rq, p, 0);
}
/* Account for a task changing its policy or group.
SCHED_FEAT(LB_BIAS, 1)
SCHED_FEAT(LB_WAKEUP_UPDATE, 1)
SCHED_FEAT(ASYM_EFF_LOAD, 1)
+SCHED_FEAT(WAKEUP_OVERLAP, 0)
/*
* Idle tasks are unconditionally rescheduled:
*/
-static void check_preempt_curr_idle(struct rq *rq, struct task_struct *p)
+static void check_preempt_curr_idle(struct rq *rq, struct task_struct *p, int sync)
{
resched_task(rq->idle);
}
if (running)
resched_task(rq->curr);
else
- check_preempt_curr(rq, p);
+ check_preempt_curr(rq, p, 0);
}
static void prio_changed_idle(struct rq *rq, struct task_struct *p,
if (p->prio > oldprio)
resched_task(rq->curr);
} else
- check_preempt_curr(rq, p);
+ check_preempt_curr(rq, p, 0);
}
/*
static void sched_rt_rq_enqueue(struct rt_rq *rt_rq)
{
+ struct task_struct *curr = rq_of_rt_rq(rt_rq)->curr;
struct sched_rt_entity *rt_se = rt_rq->rt_se;
- if (rt_se && !on_rt_rq(rt_se) && rt_rq->rt_nr_running) {
- struct task_struct *curr = rq_of_rt_rq(rt_rq)->curr;
-
- enqueue_rt_entity(rt_se);
+ if (rt_rq->rt_nr_running) {
+ if (rt_se && !on_rt_rq(rt_se))
+ enqueue_rt_entity(rt_se);
if (rt_rq->highest_prio < curr->prio)
resched_task(curr);
}
#endif /* CONFIG_RT_GROUP_SCHED */
#ifdef CONFIG_SMP
+/*
+ * We ran out of runtime, see if we can borrow some from our neighbours.
+ */
static int do_balance_runtime(struct rt_rq *rt_rq)
{
struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq);
continue;
spin_lock(&iter->rt_runtime_lock);
+ /*
+ * Either all rqs have inf runtime and there's nothing to steal
+ * or __disable_runtime() below sets a specific rq to inf to
+ * indicate its been disabled and disalow stealing.
+ */
if (iter->rt_runtime == RUNTIME_INF)
goto next;
+ /*
+ * From runqueues with spare time, take 1/n part of their
+ * spare time, but no more than our period.
+ */
diff = iter->rt_runtime - iter->rt_time;
if (diff > 0) {
diff = div_u64((u64)diff, weight);
return more;
}
+/*
+ * Ensure this RQ takes back all the runtime it lend to its neighbours.
+ */
static void __disable_runtime(struct rq *rq)
{
struct root_domain *rd = rq->rd;
spin_lock(&rt_b->rt_runtime_lock);
spin_lock(&rt_rq->rt_runtime_lock);
+ /*
+ * Either we're all inf and nobody needs to borrow, or we're
+ * already disabled and thus have nothing to do, or we have
+ * exactly the right amount of runtime to take out.
+ */
if (rt_rq->rt_runtime == RUNTIME_INF ||
rt_rq->rt_runtime == rt_b->rt_runtime)
goto balanced;
spin_unlock(&rt_rq->rt_runtime_lock);
+ /*
+ * Calculate the difference between what we started out with
+ * and what we current have, that's the amount of runtime
+ * we lend and now have to reclaim.
+ */
want = rt_b->rt_runtime - rt_rq->rt_runtime;
+ /*
+ * Greedy reclaim, take back as much as we can.
+ */
for_each_cpu_mask(i, rd->span) {
struct rt_rq *iter = sched_rt_period_rt_rq(rt_b, i);
s64 diff;
+ /*
+ * Can't reclaim from ourselves or disabled runqueues.
+ */
if (iter == rt_rq || iter->rt_runtime == RUNTIME_INF)
continue;
}
spin_lock(&rt_rq->rt_runtime_lock);
+ /*
+ * We cannot be left wanting - that would mean some runtime
+ * leaked out of the system.
+ */
BUG_ON(want);
balanced:
+ /*
+ * Disable all the borrow logic by pretending we have inf
+ * runtime - in which case borrowing doesn't make sense.
+ */
rt_rq->rt_runtime = RUNTIME_INF;
spin_unlock(&rt_rq->rt_runtime_lock);
spin_unlock(&rt_b->rt_runtime_lock);
if (unlikely(!scheduler_running))
return;
+ /*
+ * Reset each runqueue's bandwidth settings
+ */
for_each_leaf_rt_rq(rt_rq, rq) {
struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq);
int i, idle = 1;
cpumask_t span;
- if (rt_b->rt_runtime == RUNTIME_INF)
+ if (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF)
return 1;
span = sched_rt_period_mask();
curr->se.exec_start = rq->clock;
cpuacct_charge(curr, delta_exec);
+ if (!rt_bandwidth_enabled())
+ return;
+
for_each_sched_rt_entity(rt_se) {
rt_rq = rt_rq_of_se(rt_se);
/*
* Preempt the current task with a newly woken task if needed:
*/
-static void check_preempt_curr_rt(struct rq *rq, struct task_struct *p)
+static void check_preempt_curr_rt(struct rq *rq, struct task_struct *p, int sync)
{
if (p->prio < rq->curr->prio) {
resched_task(rq->curr);
case CLOCK_EVT_NOTIFY_BROADCAST_FORCE:
if (!cpu_isset(cpu, tick_broadcast_mask)) {
cpu_set(cpu, tick_broadcast_mask);
- if (bc->mode == TICKDEV_MODE_PERIODIC)
+ if (tick_broadcast_device.mode ==
+ TICKDEV_MODE_PERIODIC)
clockevents_shutdown(dev);
}
if (*reason == CLOCK_EVT_NOTIFY_BROADCAST_FORCE)
if (!tick_broadcast_force &&
cpu_isset(cpu, tick_broadcast_mask)) {
cpu_clear(cpu, tick_broadcast_mask);
- if (bc->mode == TICKDEV_MODE_PERIODIC)
+ if (tick_broadcast_device.mode ==
+ TICKDEV_MODE_PERIODIC)
tick_setup_periodic(dev, 0);
}
break;
{
struct user_struct *up = container_of(kobj, struct user_struct, kobj);
- return sprintf(buf, "%lu\n", sched_group_rt_runtime(up->tg));
+ return sprintf(buf, "%ld\n", sched_group_rt_runtime(up->tg));
}
static ssize_t cpu_rt_runtime_store(struct kobject *kobj,
unsigned long rt_runtime;
int rc;
- sscanf(buf, "%lu", &rt_runtime);
+ sscanf(buf, "%ld", &rt_runtime);
rc = sched_group_set_rt_runtime(up->tg, rt_runtime);
sp = (struct slob_page *)virt_to_page(block);
if (slob_page(sp))
- return ((slob_t *)block - 1)->units + SLOB_UNIT;
+ return (((slob_t *)block - 1)->units - 1) * SLOB_UNIT;
else
return sp->page.private;
}
mod = find_module(modname);
if (!mod) {
- if (is_vmlinux(modname))
- have_vmlinux = 1;
mod = new_module(NOFAIL(strdup(modname)));
mod->skip = 1;
}
+ if (is_vmlinux(modname)) {
+ have_vmlinux = 1;
+ mod->skip = 0;
+ }
if (!mod->skip)
add_marker(mod, marker, fmt);
if (ctx == NULL)
goto netlbl_secattr_to_sid_return;
+ context_init(&ctx_new);
ctx_new.user = ctx->user;
ctx_new.role = ctx->role;
ctx_new.type = ctx->type;
if (ebitmap_netlbl_import(&ctx_new.range.level[0].cat,
secattr->attr.mls.cat) != 0)
goto netlbl_secattr_to_sid_return;
- ctx_new.range.level[1].cat.highbit =
- ctx_new.range.level[0].cat.highbit;
- ctx_new.range.level[1].cat.node =
- ctx_new.range.level[0].cat.node;
- } else {
- ebitmap_init(&ctx_new.range.level[0].cat);
- ebitmap_init(&ctx_new.range.level[1].cat);
+ memcpy(&ctx_new.range.level[1].cat,
+ &ctx_new.range.level[0].cat,
+ sizeof(ctx_new.range.level[0].cat));
}
if (mls_context_isvalid(&policydb, &ctx_new) != 1)
goto netlbl_secattr_to_sid_return_cleanup;
projects / linux-2.6-omap-h63xx.git / commitdiff