3 Commits (4d7d8c82c181711d28c8336108330a9121f5ef07)

Author	SHA1	Message	Date
Anton Blanchard	144b9c135b	[PATCH] powerpc: use lwsync in atomics, bitops, lock functions ... eieio is only a store - store ordering. When used to order an unlock operation loads may leak out of the critical region. This is potentially buggy, one example is if a user wants to atomically read a couple of values. We can solve this with an lwsync which orders everything except store - load. I removed the (now unused) EIEIO_ON_SMP macros and the c versions isync_on_smp and eieio_on_smp now we dont use them. I also removed some old comments that were used to identify inline spinlocks in assembly, they dont make sense now our locks are out of line. Another interesting thing was that read_unlock was using an eieio even though the rest of the spinlock code had already been converted to use lwsync. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Paul Mackerras <paulus@samba.org>	19 years ago
David Gibson	3ddfbcf19b	[PATCH] powerpc: Consolidate asm compatibility macros ... This patch consolidates macros used to generate assembly for compatibility across different CPUs or configs. A new header, asm-powerpc/asm-compat.h contains the main compatibility macros. It uses some preprocessor magic to make the macros suitable both for use in .S files, and in inline asm in .c files. Headers (bitops.h, uaccess.h, atomic.h, bug.h) which had their own such compatibility macros are changed to use asm-compat.h. ppc_asm.h is now for use in .S files *only*, and a #error enforces that. As such, we're a lot more careless about namespace pollution here than in asm-compat.h. While we're at it, this patch adds a call to the PPC405_ERR77 macro in futex.h which should have had it already, but didn't. Built and booted on pSeries, Maple and iSeries (ARCH=powerpc). Built for 32-bit powermac (ARCH=powerpc) and Walnut (ARCH=ppc). Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Paul Mackerras <paulus@samba.org>	19 years ago
David Gibson	2ff2ae7a0d	[PATCH] powerpc: Merge futex.h ... This patch merges the ppc32 and ppc64 versions of futex.h, essentially by taking the ppc64 version as the powerpc version. The old ppc32 version did not implement the futex_atomic_op_inuser() callback (it always returned -ENOSYS), so FUTEX_WAKE_OP would not work on ppc32. In fact the ppc64 version of this function is almost suitable for ppc32 as well - the only change needed is to extend ppc_asm.h with a macro expanding to to the right pseudo-op to store a pointer (either ".long" or ".llong"). Built and booted on pSeries. Built for 32-bit powermac. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Paul Mackerras <paulus@samba.org>	19 years ago
Becky Bruce	feaf7cf153	[PATCH] powerpc: merge atomic.h, memory.h ... powerpc: Merge atomic.h and memory.h into powerpc Merged atomic.h into include/powerpc. Moved asm-style HMT_ defines from memory.h into ppc_asm.h, where there were already HMT_defines; moved c-style HMT_ defines to processor.h. Renamed memory.h to synch.h to better reflect its contents. Signed-off-by: Kumar Gala <kumar.gala@freescale.com> Signed-off-by: Becky Bruce <becky.bruce@freescale.com> Signed-off-by: Jon Loeliger <linuxppc@jdl.com> Signed-off-by: Paul Mackerras <paulus@samba.org>	20 years ago
Jakub Jelinek	4732efbeb9	[PATCH] FUTEX_WAKE_OP: pthread_cond_signal() speedup ... 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>	20 years ago