Commit graph

321,296 commits

Author SHA1 Message Date
Laurent Pinchart
217e9c4353 fbdev: sh_mobile_lcdc: Mark init-only symbols with __devinit(const)
default_720p and sh_mobile_lcdc_check_interface are used at device
initialization time only. Mark them as __devinitconst and __devinit
respectively.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2012-03-12 22:40:41 +01:00
Laurent Pinchart
f1f60b5f55 fbdev: sh_mobile_lcdc: Reorder code into sections
Make the driver more readable by reordering code and splitting it into
logical sections. Reorder the headers alphabetically.

No modification to the code have been performed.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2012-03-12 22:40:41 +01:00
Magnus Damm
2854903ad1 ARM: mach-shmobile: default to no earlytimer
Now when all SoCs and boards are converted to use
shmobile_earlytimer_init(), change the default behavior
of shmobile_timer.init() from using early timer to
do nothing which is suitable for upcoming DT support.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:46 +01:00
Magnus Damm
df27a2d8f1 ARM: mach-shmobile: r8a7779 and Marzen timer rework
Move the SoC specific timer code from Marzen board code
to r8a7779 setup code. This makes is possible to share
the SoC specific timer code across boards and it also
removes the need for a board specific timer structure.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:46 +01:00
Magnus Damm
23e5bc03e3 ARM: mach-shmobile: r8a7740 and Bonito timer rework
Copy the SoC specific timer code from Bonito board code
to r8a7740 setup code. This makes is possible to share
the SoC specific timer code across boards. The Bonito
specific timer setup code tied to the FPGA is kept as-is.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:45 +01:00
Magnus Damm
3be26fdba8 ARM: mach-shmobile: sh73a0, AG5EVM and Kota2 timer rework
Move the SoC specific timer code from AG5EVM and Kota2
to sh73a0 setup code. This makes is possible to share
the SoC specific timer code across boards and it also
removes the need for a board specific timer structure.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:45 +01:00
Magnus Damm
17254bffd6 ARM: mach-shmobile: sh7372, AP4EVB and Mackerel timer rework
Move the SoC specific timer code from AP4EVB and Mackerel
to sh7372 setup code. This makes is possible to share
the SoC specific timer code across boards and it also
removes the need for a board specific timer structure.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:45 +01:00
Magnus Damm
03f7beeff4 ARM: mach-shmobile: sh7377 and G4EVM timer rework
Move the SoC specific timer code from G4EVM board code
to sh7377 setup code. This makes is possible to share
the SoC specific timer code across boards and it also
removes the need for a board specific timer structure.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:44 +01:00
Magnus Damm
9e8de440da ARM: mach-shmobile: sh7367 and G3EVM timer rework
Move the SoC specific timer code from G3EVM board code
to sh7367 setup code. This makes is possible to share
the SoC specific timer code across boards and it also
removes the need for a board specific timer structure.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:44 +01:00
Magnus Damm
08ad42fb77 ARM: mach-shmobile: add shmobile_earlytimer_init()
Add shmobile_earlytimer_init() that can be used to
enable the earlytimer probing from the SoC code.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:44 +01:00
Magnus Damm
e3b0161b3c ARM: mach-shmobile: Move sh7372 AP4EVB external clk setup
Move the board specific external clock setting for sh7372 AP4EVB
from machine_desc->sys_timer->init() to machine_desc->init_machine().

This is ground work for shmobile timer code consolidation.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:43 +01:00
Magnus Damm
12bb16d230 ARM: mach-shmobile: Move sh7372 Mackerel external clk setup
Move the board specific external clock setting for sh7372 Mackerel
from machine_desc->sys_timer->init() to machine_desc->init_machine().

This is ground work for shmobile timer code consolidation.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:43 +01:00
Magnus Damm
6b6a4c067c ARM: mach-shmobile: rename clk_init() to shmobile_clk_init()
Rename clk_init() to shmobile_clk_init() to avoid a potential
future name space collision with the common clock framework.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:43 +01:00
Magnus Damm
8bac13f591 ARM: mach-shmobile: r8a7779 L2 cache support
L2 Cache support for r8a7779. Settings taken from
out-of-tree kernel patches by Kouei Abe.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:42 +01:00
Magnus Damm
3e353b875b ARM: mach-shmobile: r8a7779 map_io and init_early update
Update the r8a7779 SoC and the Marzen board to make use of
the functions r8a7779_map_io() and r8a7779_add_early_devices().

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:42 +01:00
Magnus Damm
d3ab722148 ARM: mach-shmobile: r8a7740 map_io and init_early update
Update the r8a7740 SoC and the Bonito board to make use of
the functions r8a7740_map_io() and r8a7740_add_early_devices().

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:42 +01:00
Magnus Damm
50e15c34f0 ARM: mach-shmobile: sh73a0 map_io and init_early update
Update the sh73a0 SoC and the AG5EVM and Kota2 boards to make use
of the functions sh73a0_map_io() and sh73a0_add_early_devices().

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:41 +01:00
Magnus Damm
5d7220ec00 ARM: mach-shmobile: sh7372 map_io and init_early update
Update the sh7372 SoC and the AP4EVB and Mackerel boards to make use
of the functions sh7372_map_io() and sh7372_add_early_devices().

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:41 +01:00
Magnus Damm
bfc46f3ef8 ARM: mach-shmobile: sh7377 map_io and init_early update
Update the sh7377 SoC and the G4EVM board to make use of the
functions sh7377_map_io() and sh7377_add_early_devices().

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:41 +01:00
Magnus Damm
237caf9e6b ARM: mach-shmobile: sh7367 map_io and init_early update
Update the sh7367 SoC and the G3EVM board to make use of the
functions sh7367_map_io() and sh7367_add_early_devices().

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:40 +01:00
Magnus Damm
84c36ffd7c sh: remove clk_ops
Now when all clk_ops have been renamed it is
safe to rename clk_ops to sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:13 +01:00
Magnus Damm
c953efdbb1 sh: board sh_clk_ops rename
Convert remaining sh board code to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:12 +01:00
Magnus Damm
2e679b0b7a sh: sh5 sh_clk_ops rename
Convert sh5 to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:12 +01:00
Magnus Damm
33cb61a400 sh: sh4a sh_clk_ops rename
Convert sh4a SoCs to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:12 +01:00
Magnus Damm
3b8744156d sh: sh4 sh_clk_ops rename
Convert sh4 SoCs to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:11 +01:00
Magnus Damm
43ebacd00a sh: sh3 sh_clk_ops rename
Convert sh3 SoCs to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:11 +01:00
Magnus Damm
4ad2c06155 sh: sh2a sh_clk_ops rename
Convert sh2a SoCs to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:11 +01:00
Magnus Damm
71984236d6 sh: sh2 sh_clk_ops rename
Convert sh2 to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:10 +01:00
Magnus Damm
d6ef333b62 sh: sh header sh_clk_ops rename
Convert the sh include asm/clock.h to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:10 +01:00
Magnus Damm
ae8d194902 ARM: mach-shmobile: r8a7779 sh_clk_ops rename
Convert r8a7779 to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:10 +01:00
Magnus Damm
d9f8670df8 ARM: mach-shmobile: r8a7740 sh_clk_ops rename
Convert r8a7740 to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:09 +01:00
Magnus Damm
7bcda508bf ARM: mach-shmobile: sh73a0 sh_clk_ops rename
Convert sh73a0 to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:09 +01:00
Magnus Damm
628f4561ce ARM: mach-shmobile: sh7372 sh_clk_ops rename
Convert sh7372 to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:09 +01:00
Magnus Damm
9c8c7abc25 ARM: mach-shmobile: sh7377 sh_clk_ops rename
Convert sh7377 to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:08 +01:00
Magnus Damm
79ff312103 ARM: mach-shmobile: sh7367 sh_clk_ops rename
Convert sh7367 to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:07 +01:00
Magnus Damm
a0ec360f6b sh: convert cpg code to sh_clk_ops
Convert the CPG code to use sh_clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:07 +01:00
Magnus Damm
e34828298e sh: introduce sh_clk_ops in parallel with clk_ops
Introduce sh_clk_ops in parallel with clk_ops.

Signed-off-by: Magnus Damm <damm@opensource.se>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2012-03-12 22:19:07 +01:00
Uwe Kleine-König
30c766bdec ARM: davinci: DA850: move da850_register_pm to .init.text
All callers (board-mityomapl138.c and board-da850-evm.c) use it in
__init functions only.

This fixes:

	WARNING: vmlinux.o(.text+0xd664): Section mismatch in reference from the function da850_register_pm() to the function .init.text:da8xx_get_mem_ctlr()
	The function da850_register_pm() references
	the function __init da8xx_get_mem_ctlr().
	This is often because da850_register_pm lacks a __init
	annotation or the annotation of da8xx_get_mem_ctlr is wrong.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sekhar Nori <nsekhar@ti.com>
2012-03-13 02:11:33 +05:30
Uwe Kleine-König
d870df6827 ARM: davinci: cpufreq: fix compiler warning
This fixes:
	arch/arm/mach-davinci/cpufreq.c: In function davinci_target:
	arch/arm/mach-davinci/cpufreq.c:98: warning: passing argument 2 of dev_printk from incompatible pointer type

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sekhar Nori <nsekhar@ti.com>
2012-03-13 01:24:14 +05:30
Ingo Molnar
bea95c152d Merge branch 'perf/hw-branch-sampling' into perf/core
Merge reason: The 'perf record -b' hardware branch sampling feature is ready for upstream.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:47:05 +01:00
Stephane Eranian
24bff2dc0f perf report: Fix annotate double quit issue in branch view mode
This patch fixes perf report to not go back two levels when
pressing the 'q' key while annotating in branch view mode.

When pressing 'q' in annotate mode and if the branch source
and target belong to different functions, perf now brings
up the annotation popup menu again to offer the option to
annotate the other branch source or target.

As part of the code restructuring in perf_evsel__hists_browse()
we also fix a memory leak on options[] in case of error.

Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: peterz@infradead.org
Cc: acme@redhat.com
Cc: asharma@fb.com
Cc: ravitillo@lbl.gov
Cc: vweaver1@eecs.utk.edu
Cc: khandual@linux.vnet.ibm.com
Cc: dsahern@gmail.com
Link: http://lkml.kernel.org/r/1331565210-10865-3-git-send-email-eranian@google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:46:16 +01:00
Stephane Eranian
8bcd65fd29 perf report: Remove duplicate annotate choice in branch view mode
This patch removes the duplicated annotate selection when
browsing in branch view mode. If the sym and dso oof the branch
source and target are the same, then only one annotate choice is
proposed.

Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: peterz@infradead.org
Cc: acme@redhat.com
Cc: asharma@fb.com
Cc: ravitillo@lbl.gov
Cc: vweaver1@eecs.utk.edu
Cc: khandual@linux.vnet.ibm.com
Cc: dsahern@gmail.com
Link: http://lkml.kernel.org/r/1331565210-10865-2-git-send-email-eranian@google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:46:16 +01:00
Peter Zijlstra
f9b4eeb809 perf/x86: Prettify pmu config literals
I got somewhat tired of having to decode hex numbers..

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Stephane Eranian <eranian@google.com>
Cc: Robert Richter <robert.richter@amd.com>
Link: http://lkml.kernel.org/n/tip-0vsy1sgywc4uar3mu1szm0rg@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:44:54 +01:00
Ingo Molnar
35239e23c6 Merge branch 'perf/urgent' into perf/core
Merge reason: We are going to queue up a dependent patch.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:44:11 +01:00
Peter Zijlstra
87e24f4b67 perf/x86: Fix local vs remote memory events for NHM/WSM
Verified using the below proglet.. before:

[root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 0
remote write

 Performance counter stats for './numa 0':

         2,101,554 node-stores
         2,096,931 node-store-misses

       5.021546079 seconds time elapsed

[root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 1
local write

 Performance counter stats for './numa 1':

           501,137 node-stores
               199 node-store-misses

       5.124451068 seconds time elapsed

After:

[root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 0
remote write

 Performance counter stats for './numa 0':

         2,107,516 node-stores
         2,097,187 node-store-misses

       5.012755149 seconds time elapsed

[root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 1
local write

 Performance counter stats for './numa 1':

         2,063,355 node-stores
               165 node-store-misses

       5.082091494 seconds time elapsed

#define _GNU_SOURCE

#include <sched.h>
#include <stdio.h>
#include <errno.h>
#include <sys/mman.h>
#include <sys/types.h>
#include <dirent.h>
#include <signal.h>
#include <unistd.h>
#include <numaif.h>
#include <stdlib.h>

#define SIZE (32*1024*1024)

volatile int done;

void sig_done(int sig)
{
	done = 1;
}

int main(int argc, char **argv)
{
	cpu_set_t *mask, *mask2;
	size_t size;
	int i, err, t;
	int nrcpus = 1024;
	char *mem;
	unsigned long nodemask = 0x01; /* node 0 */
	DIR *node;
	struct dirent *de;
	int read = 0;
	int local = 0;

	if (argc < 2) {
		printf("usage: %s [0-3]\n", argv[0]);
		printf("  bit0 - local/remote\n");
		printf("  bit1 - read/write\n");
		exit(0);
	}

	switch (atoi(argv[1])) {
	case 0:
		printf("remote write\n");
		break;
	case 1:
		printf("local write\n");
		local = 1;
		break;
	case 2:
		printf("remote read\n");
		read = 1;
		break;
	case 3:
		printf("local read\n");
		local = 1;
		read = 1;
		break;
	}

	mask = CPU_ALLOC(nrcpus);
	size = CPU_ALLOC_SIZE(nrcpus);
	CPU_ZERO_S(size, mask);

	node = opendir("/sys/devices/system/node/node0/");
	if (!node)
		perror("opendir");
	while ((de = readdir(node))) {
		int cpu;

		if (sscanf(de->d_name, "cpu%d", &cpu) == 1)
			CPU_SET_S(cpu, size, mask);
	}
	closedir(node);

	mask2 = CPU_ALLOC(nrcpus);
	CPU_ZERO_S(size, mask2);
	for (i = 0; i < size; i++)
		CPU_SET_S(i, size, mask2);
	CPU_XOR_S(size, mask2, mask2, mask); // invert

	if (!local)
		mask = mask2;

	err = sched_setaffinity(0, size, mask);
	if (err)
		perror("sched_setaffinity");

	mem = mmap(0, SIZE, PROT_READ|PROT_WRITE,
			MAP_PRIVATE|MAP_ANONYMOUS, -1, 0);
	err = mbind(mem, SIZE, MPOL_BIND, &nodemask, 8*sizeof(nodemask), MPOL_MF_MOVE);
	if (err)
		perror("mbind");

	signal(SIGALRM, sig_done);
	alarm(5);

	if (!read) {
		while (!done) {
			for (i = 0; i < SIZE; i++)
				mem[i] = 0x01;
		}
	} else {
		while (!done) {
			for (i = 0; i < SIZE; i++)
				t += *(volatile char *)(mem + i);
		}
	}

	return 0;
}

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Cc: <stable@kernel.org>
Link: http://lkml.kernel.org/n/tip-tq73sxus35xmqpojf7ootxgs@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:43:41 +01:00
Peter Zijlstra
c308b56b53 sched: Fix nohz load accounting -- again!
Various people reported nohz load tracking still being wrecked, but Doug
spotted the actual problem. We fold the nohz remainder in too soon,
causing us to loose samples and under-account.

So instead of playing catch-up up-front, always do a single load-fold
with whatever state we encounter and only then fold the nohz remainder
and play catch-up.

Reported-by: Doug Smythies <dsmythies@telus.net>
Reported-by: LesÅ=82aw Kope=C4=87 <leslaw.kopec@nasza-klasa.pl>
Reported-by: Aman Gupta <aman@tmm1.net>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/n/tip-4v31etnhgg9kwd6ocgx3rxl8@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:43:17 +01:00
Peter Zijlstra
8e3fabfde4 sched: Update yield() docs
Suggested-by: Joe Perches <joe@perches.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1331056466.11248.327.camel@twins
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:43:17 +01:00
Peter Zijlstra
3ccf3e8306 printk/sched: Introduce special printk_sched() for those awkward moments
There's a few awkward printk()s inside of scheduler guts that people
prefer to keep but really are rather deadlock prone. Fudge around it
by storing the text in a per-cpu buffer and poll it using the existing
printk_tick() handler.

This will drop output when its more frequent than once a tick, however
only the affinity thing could possible go that fast and for that just
one should suffice to notify the admin he's done something silly..

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/n/tip-wua3lmkt3dg8nfts66o6brne@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:43:16 +01:00
Diwakar Tundlam
554cecaf73 sched/nohz: Correctly initialize 'next_balance' in 'nohz' idle balancer
The 'next_balance' field of 'nohz' idle balancer must be initialized
to jiffies. Since jiffies is initialized to negative 300 seconds the
'nohz' idle balancer does not run for the first 300s (5mins) after
bootup. If no new processes are spawed or no idle cycles happen, the
load on the cpus will remain unbalanced for that duration.

Signed-off-by: Diwakar Tundlam <dtundlam@nvidia.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1DD7BFEDD3147247B1355BEFEFE4665237994F30EF@HQMAIL04.nvidia.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:43:16 +01:00
Peter Zijlstra
5fbd036b55 sched: Cleanup cpu_active madness
Stepan found:

CPU0		CPUn

_cpu_up()
  __cpu_up()

		boostrap()
		  notify_cpu_starting()
		  set_cpu_online()
		  while (!cpu_active())
		    cpu_relax()

<PREEMPT-out>

smp_call_function(.wait=1)
  /* we find cpu_online() is true */
  arch_send_call_function_ipi_mask()

  /* wait-forever-more */

<PREEMPT-in>
		  local_irq_enable()

  cpu_notify(CPU_ONLINE)
    sched_cpu_active()
      set_cpu_active()

Now the purpose of cpu_active is mostly with bringing down a cpu, where
we mark it !active to avoid the load-balancer from moving tasks to it
while we tear down the cpu. This is required because we only update the
sched_domain tree after we brought the cpu-down. And this is needed so
that some tasks can still run while we bring it down, we just don't want
new tasks to appear.

On cpu-up however the sched_domain tree doesn't yet include the new cpu,
so its invisible to the load-balancer, regardless of the active state.
So instead of setting the active state after we boot the new cpu (and
consequently having to wait for it before enabling interrupts) set the
cpu active before we set it online and avoid the whole mess.

Reported-by: Stepan Moskovchenko <stepanm@codeaurora.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1323965362.18942.71.camel@twins
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-12 20:43:15 +01:00