23 Jan, 2011
15 commits
-
Close to perf_mmap__read_head() and the perf_mmap struct definition.
This is useful for any recorder, and we will need it in 'perf test'.Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
Paving the way to using perf_evsel->mmap, do this to reduce the patch
noise in the next ones.Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
Out of the code in 'perf top'. Record is next in line.
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
Now its time to factor out the mmap handling bits into the perf_evsel
class.Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
Now that it handles group_fd and inherit we can use it, sharing it with
stat.Next step: 'perf record' should use, then move the mmap_array out of
->priv and into perf_evsel, with top and record sharing this, and at the
same time, write a 'perf test' stress test.Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
As this is a per-cpu attribute, we can't set it up in advance and use it
for all the calls.Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
The perf_evsel__open now have an extra boolean argument specifying if
event grouping is desired.The first file descriptor created on a CPU becomes the group leader.
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
Allocating just the space needed for nr_cpus * nr_threads * nr_evsels,
not the MAX_NR_CPUS and counters.LKML-Reference:
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
Signed-off-by: Arnaldo Carvalho de Melo -
Killing two more perf wide global variables: nr_counters and evsel_list
as a list_head.There are more operations that will need more fields in perf_evlist,
like the pollfd for polling all the fds in a list of evsel instances.Use option->value to pass the evsel_list to parse_{events,filters}.
LKML-Reference:
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
Signed-off-by: Arnaldo Carvalho de Melo -
It's enough to include the local "debug.h" file to trigger it.
man time reveals this is already declared in glibc:
time - get time in seconds
-> rename the variable.Cc: Ingo Molnar
Cc: arjan@infradead.org
LPU-Reference:
Signed-off-by: Thomas Renninger
Signed-off-by: Arnaldo Carvalho de Melo -
The -Wstack-protector and -Wvolatile-register-var warnings, for
instance, are not supported by gcc 3.4.6.So fix by doing the same check we already do for -fstack-protector-all.
With this and the other patches in this series, perf builds unmodified
on, for instance, RHEL4.Cc: Eric Dumazet
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
[acme@localhost linux]$ make O=~acme/git/build/perf -C tools/perf
make: Entering directory `/home/acme/git/linux/tools/perf'
Makefile:526: No libdw.h found or old libdw.h found or elfutils is older than 0.138, disables dwarf support. Please install new elfutils-devel/libdw-dev
Makefile:582: newt not found, disables TUI support. Please install newt-devel or libnewt-dev
CC /home/acme/git/build/perf/builtin-annotate.o
In file included from builtin-annotate.c:23:
util/parse-events.h:26: warning: declaration of 'evsel_list' shadows a global declaration
util/parse-events.h:12: warning: shadowed declaration is here
make: *** [/home/acme/git/build/perf/builtin-annotate.o] Error 1
make: Leaving directory `/home/acme/git/linux/tools/perf'
[acme@localhost linux]$ gcc --version | head -1
gcc (GCC) 3.4.6 20060404 (Red Hat 3.4.6-11)
[acme@localhost linux]$Fix it by renaming the parameter to evlist.
Cc: Eric Dumazet
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
We need the definiton for __always_inline in bitops.h to fix the build
on distros where it isn't available or compiler.h doesn't get included
indirectly.One of the fixes needed to build perf on RHEL4 systems, for instance.
Cc: Eric Dumazet
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Stephane Eranian
Cc: Tom Zanussi
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo -
Using %L[uxd] has issues in some architectures, like on ppc64. Fix it
by making our 64 bit integers typedefs of stdint.h types and using
PRI[ux]64 like, for instance, git does.Reported by Denis Kirjanov that provided a patch for one case, I went
and changed all cases.Reported-by: Denis Kirjanov
Tested-by: Denis Kirjanov
LKML-Reference:
Cc: Denis Kirjanov
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Mike Galbraith
Cc: Paul Mackerras
Cc: Peter Zijlstra
Cc: Pingtian Han
Cc: Stephane Eranian
Cc: Tom Zanussi
Signed-off-by: Arnaldo Carvalho de Melo -
Where we don't have CPU_ALLOC & friends. As the tools are being used in older
distros where the only allowed change are to replace the kernel, like RHEL4 and
5.Reported-by: Eric Dumazet
Cc: Eric Dumazet
Cc: Frederic Weisbecker
Cc: Mike Galbraith
Cc: Peter Zijlstra
Cc: Paul Mackerras
Cc: Stephane Eranian
LKML-Reference:
Signed-off-by: Arnaldo Carvalho de Melo
22 Jan, 2011
4 commits
-
…/linux-2.6 into perf/urgent
-
In theory, almost every user of task->child->perf_event_ctxp[]
is wrong. find_get_context() can install the new context at any
moment, we need read_barrier_depends().dbe08d82ce3967ccdf459f7951d02589cf967300 "perf: Fix
find_get_context() vs perf_event_exit_task() race" added
rcu_dereference() into perf_event_exit_task_context() to make
the precedent, but this makes __rcu_dereference_check() unhappy.
Use rcu_dereference_raw() to shut up the warning.Reported-by: Ingo Molnar
Signed-off-by: Oleg Nesterov
Cc: acme@redhat.com
Cc: paulus@samba.org
Cc: stern@rowland.harvard.edu
Cc: a.p.zijlstra@chello.nl
Cc: fweisbec@gmail.com
Cc: roland@redhat.com
Cc: prasad@linux.vnet.ibm.com
Cc: Paul E. McKenney
LKML-Reference:
Signed-off-by: Ingo Molnar -
When some of CPUs are offline:
# cat /sys/devices/system/cpu/online
0,6-31perf test will fail on #3 testcase:
3: detect open syscall event on all cpus:
--- start ---
perf_evsel__read_on_cpu: expected to intercept 111 calls on cpu 0, got 681
perf_evsel__read_on_cpu: expected to intercept 112 calls on cpu 1, got 117
perf_evsel__read_on_cpu: expected to intercept 113 calls on cpu 2, got 118
perf_evsel__read_on_cpu: expected to intercept 114 calls on cpu 3, got 119
perf_evsel__read_on_cpu: expected to intercept 115 calls on cpu 4, got 120
perf_evsel__read_on_cpu: expected to intercept 116 calls on cpu 5, got 121
perf_evsel__read_on_cpu: expected to intercept 117 calls on cpu 6, got 122
perf_evsel__read_on_cpu: expected to intercept 118 calls on cpu 7, got 123
perf_evsel__read_on_cpu: expected to intercept 119 calls on cpu 8, got 124
perf_evsel__read_on_cpu: expected to intercept 120 calls on cpu 9, got 125
perf_evsel__read_on_cpu: expected to intercept 121 calls on cpu 10, got 126
....This patch try to use 'cpus->map[cpu]' when setting cpu affinity, and
will check the return code of sched_setaffinity()LKML-Reference:
Signed-off-by: Han Pingtian
Signed-off-by: Arnaldo Carvalho de Melo -
In ARM's Thumb mode the bottom bit of the symbol address is set to mark
the function as Thumb; the instructions are in reality 2 or 4 byte on 2
byte alignments, and when the +1 address is used in annotate it causes
objdump to disassemble invalid instructions.The patch removes that bottom bit during symbol loading.
Many thinks to Dave Martin for comments on an initial version of the
patch.(For reference this corresponds to this bug
https://bugs.launchpad.net/linux-linaro/+bug/677547 )Cc: Ingo Molnar
Cc: Dave Martin
LKML-Reference:
Signed-off-by: Dr. David Alan Gilbert
Signed-off-by: Arnaldo Carvalho de Melo
21 Jan, 2011
1 commit
-
Lockdep spotted:
loop_1b_instruc/1899 is trying to acquire lock:
(event_mutex){+.+.+.}, at: [] perf_trace_init+0x3b/0x2f7but task is already holding lock:
(&ctx->mutex){+.+.+.}, at: [] perf_event_init_context+0xc0/0x218which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #3 (&ctx->mutex){+.+.+.}:
-> #2 (cpu_hotplug.lock){+.+.+.}:
-> #1 (module_mutex){+.+...}:
-> #0 (event_mutex){+.+.+.}:But because the deadlock would be cpuhotplug (cpu-event) vs fork
(task-event) it cannot, in fact, happen. We can annotate this by giving the
perf_event_context used for the cpuctx a different lock class from those
used by tasks.Signed-off-by: Peter Zijlstra
LKML-Reference:
Signed-off-by: Ingo Molnar
20 Jan, 2011
3 commits
-
When fixing the frequency calculations for perf on powerpc I
forgot to fix the FSL version.If we dont set event->hw.last_period the frequency to period
calculations in perf go haywire and we continually
throttle/unthrottle the PMU.Signed-off-by: Anton Blanchard
Acked-by: Benjamin Herrenschmidt
Cc: Scott Wood
Cc: Paul Mackerras
Cc: Arnaldo Carvalho de Melo
Cc: linuxppc-dev@lists.ozlabs.org
Cc: Peter Zijlstra
LKML-Reference:
Signed-off-by: Ingo Molnar -
perf_event_init_task() should clear child->perf_event_ctxp[]
before anything else. Otherwise, if
perf_event_init_context(perf_hw_context) fails,
perf_event_free_task() can free perf_event_ctxp[perf_sw_context]
copied from parent->perf_event_ctxp[] by dup_task_struct().Also move the initialization of perf_event_mutex and
perf_event_list from perf_event_init_context() to
perf_event_init_context().Signed-off-by: Oleg Nesterov
Acked-by: Peter Zijlstra
Cc: Alan Stern
Cc: Arnaldo Carvalho de Melo
Cc: Frederic Weisbecker
Cc: Paul Mackerras
Cc: Prasad
Cc: Roland McGrath
LKML-Reference:
Signed-off-by: Ingo Molnar -
find_get_context() must not install the new perf_event_context
if the task has already passed perf_event_exit_task().If nothing else, this means the memory leak. Initially
ctx->refcount == 2, it is supposed that
perf_event_exit_task_context() should participate and do the
necessary put_ctx().find_lively_task_by_vpid() checks PF_EXITING but this buys
nothing, by the time we call find_get_context() this task can be
already dead. To the point, cmpxchg() can succeed when the task
has already done the last schedule().Change find_get_context() to populate task->perf_event_ctxp[]
under task->perf_event_mutex, this way we can trust PF_EXITING
because perf_event_exit_task() takes the same mutex.Also, change perf_event_exit_task_context() to use
rcu_dereference(). Probably this is not strictly needed, but
with or without this change find_get_context() can race with
setup_new_exec()->perf_event_exit_task(), rcu_dereference()
looks better.Signed-off-by: Oleg Nesterov
Acked-by: Peter Zijlstra
Cc: Alan Stern
Cc: Arnaldo Carvalho de Melo
Cc: Frederic Weisbecker
Cc: Paul Mackerras
Cc: Prasad
Cc: Roland McGrath
LKML-Reference:
Signed-off-by: Ingo Molnar
19 Jan, 2011
17 commits
-
* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/staging:
hwmon: (lm93) Add support for LM94 -
…/git/tip/linux-2.6-tip
* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
perf: Validate cpu early in perf_event_alloc()
perf: Find_get_context: fix the per-cpu-counter check
perf: Fix contexted inheritance -
…git/tip/linux-2.6-tip
* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86: Clear irqstack thread_info
x86: Make relocatable kernel work with new binutils -
* 'upstream' of git://git.linux-mips.org/pub/scm/upstream-linus: (26 commits)
MIPS: Malta: enable Cirrus FB console
MIPS: add CONFIG_VIRTUALIZATION for virtio support
MIPS: Implement __read_mostly
MIPS: ath79: add common WMAC device for AR913X based boards
MIPS: ath79: Add initial support for the Atheros AP81 reference board
MIPS: ath79: add common SPI controller device
SPI: Add SPI controller driver for the Atheros AR71XX/AR724X/AR913X SoCs
MIPS: ath79: add common GPIO buttons device
MIPS: ath79: add common watchdog device
MIPS: ath79: add common GPIO LEDs device
MIPS: ath79: add initial support for the Atheros PB44 reference board
MIPS: ath79: utilize the MIPS multi-machine support
MIPS: ath79: add GPIOLIB support
MIPS: Add initial support for the Atheros AR71XX/AR724X/AR931X SoCs
MIPS: jump label: Add MIPS support.
MIPS: Use WARN() in uasm for better diagnostics.
MIPS: Optimize TLB handlers for Octeon CPUs
MIPS: Add LDX and LWX instructions to uasm.
MIPS: Use BBIT instructions in TLB handlers
MIPS: Declare uasm bbit0 and bbit1 functions.
... -
This patch adds basic support for LM94 to the LM93 driver. LM94 specific
sensors and features are not supported.Signed-off-by: Guenter Roeck
Acked-by: Jean Delvare -
Starting from perf_event_alloc()->perf_init_event(), the kernel
assumes that event->cpu is either -1 or the valid CPU number.Change perf_event_alloc() to validate this argument early. This
also means we can remove the similar check in
find_get_context().Signed-off-by: Oleg Nesterov
Acked-by: Peter Zijlstra
Cc: Alan Stern
Cc: Arnaldo Carvalho de Melo
Cc: Frederic Weisbecker
Cc: Paul Mackerras
Cc: Prasad
Cc: Roland McGrath
Cc: gregkh@suse.de
Cc: stable@kernel.org
LKML-Reference:
Signed-off-by: Ingo Molnar -
If task == NULL, find_get_context() should always check that cpu
is correct.Afaics, the bug was introduced by 38a81da2 "perf events: Clean
up pid passing", but even before that commit "&& cpu != -1" was
not exactly right, -ESRCH from find_task_by_vpid() is not
accurate.Signed-off-by: Oleg Nesterov
Acked-by: Peter Zijlstra
Cc: Alan Stern
Cc: Arnaldo Carvalho de Melo
Cc: Frederic Weisbecker
Cc: Paul Mackerras
Cc: Prasad
Cc: Roland McGrath
Cc: gregkh@suse.de
Cc: stable@kernel.org
LKML-Reference:
Signed-off-by: Ingo Molnar -
While most users of a physical Malta board are using the serial port
as the console, a lot of QEMU users would prefer to interact with a
graphical console. Enable the Cirrus FB support in the Malta default
configuration to make that possible. Note that the default console will
still be the serial port, users have to pass "console=tty0" to the
kernel to use the Cirrus FB.Signed-off-by: Aurelien Jarno
To: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/2001/
Signed-off-by: Ralf Baechle -
Add CONFIG_VIRTUALIZATION to the MIPS architecture and include the
the virtio code there. Used to enable the virtio drivers under QEMU.Signed-off-by: Aurelien Jarno
To: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/2002/
Signed-off-by: Ralf Baechle -
Just do what everyone else is doing by placing __read_mostly things in
the .data.read_mostly section.mips_io_port_base can not be read-only (const) and writable
(__read_mostly) at the same time. One of them has to go, so I chose
to eliminate the __read_mostly. It will still get stuck in a portion
of memory that is not adjacent to things that are written, and thus
not be on a dirty cache line, for whatever that is worth.Signed-off-by: David Daney
To: linux-mips@linux-mips.org
Patchwork: http://patchwork.linux-mips.org/patch/1702/
Signed-off-by: Ralf Baechle -
Add common platform_device and helper code to make the registration
of the built-in wireless MAC easier on the Atheros AR9130/AR9132
based boards. Also register the WMAC device on the AR81 board.Signed-off-by: Gabor Juhos
Cc: linux-mips@linux-mips.org
Cc: Imre Kaloz ,
Cc: Luis R. Rodriguez
Cc: Cliff Holden
Cc: Kathy Giori
Patchwork: https://patchwork.linux-mips.org/patch/1962/
Signed-off-by: Ralf Baechle -
Signed-off-by: Gabor Juhos
Signed-off-by: Imre Kaloz
Cc: linux-mips@linux-mips.org
Cc: Luis R. Rodriguez
Cc: Cliff Holden
Cc: Kathy Giori
Patchwork: https://patchwork.linux-mips.org/patch/1952/
Signed-off-by: Ralf Baechle -
Several boards are using the built-in SPI controller of the
AR71XX/AR724X/AR913X SoCs. This patch adds common platform_device
and helper code to register it. Additionally, the patch registers
the SPI bus on the PB44 board.Signed-off-by: Gabor Juhos
Cc: linux-mips@linux-mips.org
Cc: Imre Kaloz
Cc: Luis R. Rodriguez
Cc: Cliff Holden
Cc: Kathy Giori
Patchwork: https://patchwork.linux-mips.org/patch/1956/
Signed-off-by: Ralf Baechle -
The Atheros AR71XX/AR724X/AR913X SoCs have a built-in SPI controller. This
patch implements a driver for that.Signed-off-by: Gabor Juhos
Cc: David Brownell
Cc: spi-devel-general@lists.sourceforge.net
Acked-by: Grant Likely
Cc: linux-mips@linux-mips.org
Cc: Imre Kaloz
Cc: Luis R. Rodriguez
Cc: Cliff Holden
Cc: Kathy Giori
Patchwork: https://patchwork.linux-mips.org/patch/1960/
Signed-off-by: Ralf Baechle -
Almost all boards have one or more push buttons connected to GPIO lines.
This patch adds common code to register a platform_device for them.The patch also adds support for the buttons on the PB44 board.
Signed-off-by: Gabor Juhos
Signed-off-by: Imre Kaloz
Cc: linux-mips@linux-mips.org
Cc: Luis R. Rodriguez
Cc: Cliff Holden
Cc: Kathy Giori
Patchwork: https://patchwork.linux-mips.org/patch/1954/
Signed-off-by: Ralf Baechle -
All supported SoCs have a built-in hardware watchdog driver. This patch
registers a platform_device for that to make it usable.Signed-off-by: Gabor Juhos
Signed-off-by: Imre Kaloz
Cc: linux-mips@linux-mips.org
Cc: Luis R. Rodriguez
Cc: Cliff Holden
Cc: Kathy Giori
Patchwork: https://patchwork.linux-mips.org/patch/1955/
Signed-off-by: Ralf Baechle
