ATM pthread_cond_signal is unnecessarily slow, because it wakes one waiter
(which at least on UP usually means an immediate context switch to one of
the waiter threads). This waiter wakes up and after a few instructions it
attempts to acquire the cv internal lock, but that lock is still held by
the thread calling pthread_cond_signal. So it goes to sleep and eventually
the signalling thread is scheduled in, unlocks the internal lock and wakes
the waiter again.
Now, before 2003-09-21 NPTL was using FUTEX_REQUEUE in pthread_cond_signal
to avoid this performance issue, but it was removed when locks were
redesigned to the 3 state scheme (unlocked, locked uncontended, locked
contended).
Following scenario shows why simply using FUTEX_REQUEUE in
pthread_cond_signal together with using lll_mutex_unlock_force in place of
lll_mutex_unlock is not enough and probably why it has been disabled at
that time:
The number is value in cv->__data.__lock.
thr1 thr2 thr3
0 pthread_cond_wait
1 lll_mutex_lock (cv->__data.__lock)
0 lll_mutex_unlock (cv->__data.__lock)
0 lll_futex_wait (&cv->__data.__futex, futexval)
0 pthread_cond_signal
1 lll_mutex_lock (cv->__data.__lock)
1 pthread_cond_signal
2 lll_mutex_lock (cv->__data.__lock)
2 lll_futex_wait (&cv->__data.__lock, 2)
2 lll_futex_requeue (&cv->__data.__futex, 0, 1, &cv->__data.__lock)
# FUTEX_REQUEUE, not FUTEX_CMP_REQUEUE
2 lll_mutex_unlock_force (cv->__data.__lock)
0 cv->__data.__lock = 0
0 lll_futex_wake (&cv->__data.__lock, 1)
1 lll_mutex_lock (cv->__data.__lock)
0 lll_mutex_unlock (cv->__data.__lock)
# Here, lll_mutex_unlock doesn't know there are threads waiting
# on the internal cv's lock
Now, I believe it is possible to use FUTEX_REQUEUE in pthread_cond_signal,
but it will cost us not one, but 2 extra syscalls and, what's worse, one of
these extra syscalls will be done for every single waiting loop in
pthread_cond_*wait.
We would need to use lll_mutex_unlock_force in pthread_cond_signal after
requeue and lll_mutex_cond_lock in pthread_cond_*wait after lll_futex_wait.
Another alternative is to do the unlocking pthread_cond_signal needs to do
(the lock can't be unlocked before lll_futex_wake, as that is racy) in the
kernel.
I have implemented both variants, futex-requeue-glibc.patch is the first
one and futex-wake_op{,-glibc}.patch is the unlocking inside of the kernel.
The kernel interface allows userland to specify how exactly an unlocking
operation should look like (some atomic arithmetic operation with optional
constant argument and comparison of the previous futex value with another
constant).
It has been implemented just for ppc*, x86_64 and i?86, for other
architectures I'm including just a stub header which can be used as a
starting point by maintainers to write support for their arches and ATM
will just return -ENOSYS for FUTEX_WAKE_OP. The requeue patch has been
(lightly) tested just on x86_64, the wake_op patch on ppc64 kernel running
32-bit and 64-bit NPTL and x86_64 kernel running 32-bit and 64-bit NPTL.
With the following benchmark on UP x86-64 I get:
for i in nptl-orig nptl-requeue nptl-wake_op; do echo time elf/ld.so --library-path .:$i /tmp/bench; \
for j in 1 2; do echo ( time elf/ld.so --library-path .:$i /tmp/bench ) 2>&1; done; done
time elf/ld.so --library-path .:nptl-orig /tmp/bench
real 0m0.655s user 0m0.253s sys 0m0.403s
real 0m0.657s user 0m0.269s sys 0m0.388s
time elf/ld.so --library-path .:nptl-requeue /tmp/bench
real 0m0.496s user 0m0.225s sys 0m0.271s
real 0m0.531s user 0m0.242s sys 0m0.288s
time elf/ld.so --library-path .:nptl-wake_op /tmp/bench
real 0m0.380s user 0m0.176s sys 0m0.204s
real 0m0.382s user 0m0.175s sys 0m0.207s
The benchmark is at:
http://sourceware.org/ml/libc-alpha/2005-03/txt00001.txt
Older futex-requeue-glibc.patch version is at:
http://sourceware.org/ml/libc-alpha/2005-03/txt00002.txt
Older futex-wake_op-glibc.patch version is at:
http://sourceware.org/ml/libc-alpha/2005-03/txt00003.txt
Will post a new version (just x86-64 fixes so that the patch
applies against pthread_cond_signal.S) to libc-hacker ml soon.
Attached is the kernel FUTEX_WAKE_OP patch as well as a simple-minded
testcase that will not test the atomicity of the operation, but at least
check if the threads that should have been woken up are woken up and
whether the arithmetic operation in the kernel gave the expected results.
Acked-by: Ingo Molnar <mingo@redhat.com>
Cc: Ulrich Drepper <drepper@redhat.com>
Cc: Jamie Lokier <jamie@shareable.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Yoichi Yuasa <yuasa@hh.iij4u.or.jp>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Patch from Richard Purdie
This patch updates the PCMCIA pxa2xx_sharpsl driver to support multiple scoop
devices by adding a scoop to pcmcia slot mapping structure. It adds platform
support for poodle, is known to work on spitz (which is dual slot) and
should also support collie with a minor amount of further work.
Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This is used only in slab.c and each architecture gets to define whcih
underlying type is to be used.
Seems a bit silly - move it to slab.c and use the same type for all
architectures: unsigned int.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Someone mentioned that almost all the architectures used basically the same
implementation of get_order. This patch consolidates them into
asm-generic/page.h and includes that in the appropriate places. The
exceptions are ia64 and ppc which have their own (presumably optimised)
versions.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This is part of Thomas Gleixner's generic IRQ patch, which converts
ARM to use the generic IRQ subsystem. Here, we wrap calls to
desc->handler() in an inline function, desc_handle_irq(). This
reduces the size of Thomas' patch since the changes become more
localised.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This is part of Thomas Gleixner's generic IRQ patch, which converts
ARM to use the generic IRQ subsystem. Here, we rename two of the
irq_chip methods - wake becomes set_wake, and type becomes set_type.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Nicolas Pitre
The prototype for sys_fadvise64_64() is:
long sys_fadvise64_64(int fd, loff_t offset, loff_t len, int advice)
The argument list is therefore as follows on legacy ABI:
fd: type int (r0)
offset: type long long (r1-r2)
len: type long long (r3-sp[0])
advice: type int (sp[4])
With EABI this becomes:
fd: type int (r0)
offset: type long long (r2-r3)
len: type long long (sp[0]-sp[4])
advice: type int (sp[8])
Not only do we have ABI differences here, but the EABI version requires
one additional word on the syscall stack.
To avoid the ABI mismatch and the extra stack space required with EABI
this syscall is now defined with a different argument ordering
on ARM as follows:
long sys_arm_fadvise64_64(int fd, int advice, loff_t offset, loff_t len)
This gives us the following ABI independent argument distribution:
fd: type int (r0)
advice: type int (r1)
offset: type long long (r2-r3)
len: type long long (sp[0]-sp[4])
Now, since the syscall entry code takes care of 5 registers only by
default including the store of r4 to the stack, we need a wrapper to
store r5 to the stack as well. Because that wrapper was missing and was
always required this means that sys_fadvise64_64 never worked on ARM and
therefore we can safely reuse its syscall number for our new
sys_arm_fadvise64_64 interface.
Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from David Vrabel
Correct the ioread* and iowrite* functions. In particular, add an offset to the cookie in ioport_map so we can map I/O port ranges starting from 0 (0 is for reporting errors).
Signed-off-by: David Vrabel <dvrabel@arcom.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Richard Purdie
Add some extra pxa27x register definitions needed for the Sharp
SL-C3000 (Spitz).
Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Nicolas Pitre
There is no need to define __ARCH_WANT_SYS_FADVISE64 on ARM since it
only serves to compile in a compatibility wrapper for sys_fadvise64
which never was tied to any syscall number.
Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Ben Dooks
Add the definitions for the S3C2410_CLKSLOW registers to
the header files, and show the values when the system
starts up
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Deepak Saxena
This patch implements the set_irq_type() hooks for configuring GPIO
IRQ type and updates all the platforms to use it instead of the
gpio_line_config() function which is now used to configure input
vs. output on the pins.
Signed-off-by: Deepak Saxena <dsaxena@plexity.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
As pointed out in the following thread, the CLOCK_TICK_RATE setting for
IXP4xx is incorrect b/c the HW ignores the lowest 2 bits of the LATCH
value.
http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2005-August/030950.html
Tnx to George Anziger and Egil Hjelmeland for finding the issue.
Signed-off-by: Deepak Saxena <dsaxena@plexity.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
a bunch of functions switched from volatile to __attribute__((noreturn)) and
from const to __attribute_pure__
Signed-off-by: Al Viro <viro@parcelfarce.linux.theplanet.co.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Add platform device data for the SA11x0 MCP device. This allows
platforms to customise the configuration of the SA11x0 MCP device
according to their needs.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Ben Dooks
Rename the s3c2410_report_oc() to s3c2410_usb_report_oc()
as this is an usb specific function.
Change port power on the usb-simtec implementation to only
power up the output if both are set, as per the usb 1.1
specification
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Unfortunately, we can't use the "user" bit in the page tables to
control whether a page table entry is "global" or "asid" specific,
since the vector page is mapped as "user" accessible but is not
process specific.
Therefore, give direct control of the ARMv6 "nG" (not global)
bit to the mm layers.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
In yenta_socket, we default to using the resource setting of the CardBus
bridge. However, this is a PCI-bus-centric view of resources and thus needs
to be converted to generic resources first. Therefore, add a call to
pcibios_bus_to_resource() call in between. This function is a mere wrapper on
x86 and friends, however on some others it already exists, is added in this
patch (alpha, arm, ppc, ppc64) or still needs to be provided (parisc -- where
is its pcibios_resource_to_bus() ?).
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
We sometimes forgot to check whether the exclusive store succeeded.
Ensure that we always check. Also ensure that we always use the
out of line versions, since the inline versions are not SMP safe.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
When the kernel is working well and we want to restart cleanly
kernel_restart is the function to use. But in many instances
the kernel wants to reboot when thing are expected to be working
very badly such as from panic or a software watchdog handler.
This patch adds the function emergency_restart() so that
callers can be clear what semantics they expect when calling
restart. emergency_restart() is expected to be callable
from interrupt context and possibly reliable in even more
trying circumstances.
This is an initial generic implementation for all architectures.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Convert explicit gcc asm-based memory barriers into smp_mb() calls.
These change between barrier() and the ARMv6 data memory barrier
instruction depending on whether ARMv6 SMP is enabled.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Dimitry Andric
- Change S3C2440_IISCON_MPLL to S3C2440_IISMOD_MPLL:
The S3C2440 IISCON register doesn\'t control the master clock selection, this is done with the IISMOD register.
- Correct S3C2410_IISMOD_256FS and S3C2410_IISMOD_384FS:
This is set via bit 2 of IISMOD, not bit 1.
- Add S3C2410_IISCON_PSCEN (prescaler enable), for completeness\' sake.
Signed-off-by: Dimitry Andric <dimitry.andric@tomtom.com>
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Sascha Hauer
The dmacr needs different settings on some boards. This patch makes the
register configurable by the platform part.
Also we have imxfb_disable_controller(), so lets use it.
Signed-off-by: Steven Scholz
Signed-off-by: Sascha Hauer
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Alexander Schulz
This patch brings a new default config file for the shark and
fixes a compilation issue with io addressing and a runtime
problem with the serial ports, where I corrected a wrong
regshift value.
These are all shark specific files so I hope it is ok to
put them in one patch.
Signed-off-by: Alexander Schulz <alex@shark-linux.de>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
http://bugme.osdl.org/show_bug.cgi?id=4016
Written-by: David Shaohua Li <shaohua.li@intel.com>
Acked-by: Adam Belay <abelay@novell.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Patch from Tony Lindgren
This patch by various OMAP developers syncs the OMAP
specific arch files with the linux-omap tree.
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Tony Lindgren
This patch by various OMAP developers syncs the OMAP
specific arch files with the linux-omap tree.
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Tony Lindgren
This patch by various OMAP developers syncs the OMAP
specific include files with the linux-omap tree.
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Lennert Buytenhek
This patch converts the ixp2000 serial port over to a platform
serial device.
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Deepak Saxena <dsaxena@plexity.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Patch from Deepak Saxena
This patch implements the iomap API for Intel IXP4xx NPU systems.
We need to implement our own version of the API functions b/c of the
PCI hostbridge does not provide the capability to map PCI I/O space
into the CPU's physical memory space. In addition, if a system has
more than 64M of PCI memory mapped BARs, PCI memory must also be
accessed indirectly. This patch changes the assignment of PCI I/O
resources to fall into to 0x0000:0xffff range so that we can trap
I/O areas in our ioread/iowrite macros.
Signed-off-by: Deepak Saxena
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Jakub Jelinek