Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759900AbbBIUeH (ORCPT ); Mon, 9 Feb 2015 15:34:07 -0500 Received: from mail.linuxfoundation.org ([140.211.169.12]:54597 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753469AbbBIUeE (ORCPT ); Mon, 9 Feb 2015 15:34:04 -0500 Date: Mon, 9 Feb 2015 12:34:02 -0800 From: Andrew Morton To: "Wang, Yalin" Cc: "'Kirill A. Shutemov'" , "'arnd@arndb.de'" , "'linux-arch@vger.kernel.org'" , "'linux-kernel@vger.kernel.org'" , "'linux@arm.linux.org.uk'" , "'linux-arm-kernel@lists.infradead.org'" Subject: Re: [RFC] change non-atomic bitops method Message-Id: <20150209123402.1759cb791113cc64483579c5@linux-foundation.org> In-Reply-To: <35FD53F367049845BC99AC72306C23D1044A02027E12@CNBJMBX05.corpusers.net> References: <35FD53F367049845BC99AC72306C23D1044A02027E0A@CNBJMBX05.corpusers.net> <20150202152909.13bfd11f192fb0268b2ab4bf@linux-foundation.org> <20150203011730.GA15653@node.dhcp.inet.fi> <35FD53F367049845BC99AC72306C23D1044A02027E0B@CNBJMBX05.corpusers.net> <35FD53F367049845BC99AC72306C23D1044A02027E0C@CNBJMBX05.corpusers.net> <20150202223851.f30768d0.akpm@linux-foundation.org> <35FD53F367049845BC99AC72306C23D1044A02027E0D@CNBJMBX05.corpusers.net> <35FD53F367049845BC99AC72306C23D1044A02027E0E@CNBJMBX05.corpusers.net> <20150203025925.d1c95fb8.akpm@linux-foundation.org> <35FD53F367049845BC99AC72306C23D1044A02027E12@CNBJMBX05.corpusers.net> X-Mailer: Sylpheed 3.4.1 (GTK+ 2.24.23; x86_64-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3405 Lines: 94 On Mon, 9 Feb 2015 16:18:10 +0800 "Wang, Yalin" wrote: > > That we're running clear_bit against a cleared bit 10% of the time is a > > bit alarming. I wonder where that's coming from. > > > > The enormous miss count in test_and_clear_bit() might indicate an > > inefficiency somewhere. > I te-test the patch on 3.10 kernel. > The result like this: > > VmallocChunk: 251498164 kB > __set_bit_miss_count:11730 __set_bit_success_count:1036316 > __clear_bit_miss_count:209640 __clear_bit_success_count:4806556 > __test_and_set_bit_miss_count:0 __test_and_set_bit_success_count:121 > __test_and_clear_bit_miss_count:0 __test_and_clear_bit_success_count:445 > > __clear_bit miss rate is a little high, > I check the log, and most miss coming from this code: > > <6>[ 442.701798] [] warn_slowpath_fmt+0x4c/0x58 > <6>[ 442.701805] [] __clear_bit+0x98/0xa4 > <6>[ 442.701813] [] __alloc_fd+0xc8/0x124 > <6>[ 442.701821] [] get_unused_fd_flags+0x28/0x34 > <6>[ 442.701828] [] do_sys_open+0x10c/0x1c0 > <6>[ 442.701835] [] SyS_openat+0xc/0x18 > In __clear_close_on_exec(fd, fdt); > > > > <6>[ 442.695354] [] warn_slowpath_fmt+0x4c/0x58 > <6>[ 442.695359] [] __clear_bit+0x98/0xa4 > <6>[ 442.695367] [] dup_fd+0x1d4/0x280 > <6>[ 442.695375] [] copy_process.part.56+0x42c/0xe38 > <6>[ 442.695382] [] do_fork+0xe0/0x360 > <6>[ 442.695389] [] SyS_clone+0x10/0x1c > In __clear_open_fd(open_files - i, new_fdt); > > Do we need test_bit() before clear_bit()at these 2 place? I don't know. I was happily typing in this: diff -puN include/linux/bitops.h~a include/linux/bitops.h --- a/include/linux/bitops.h~a +++ a/include/linux/bitops.h @@ -226,5 +226,37 @@ extern unsigned long find_last_bit(const unsigned long size); #endif +/** + * __set_clear_bit - non-atomically set a bit if it is presently clear + * @nr: The bit number + * @addr: The base address of the operation + * + * __set_clear_bit() and similar functions avoid unnecessarily dirtying a + * cacheline when the operation will have no effect. + */ +static inline void __set_clear_bit(unsigned nr, volatile unsigned long *addr) +{ + if (!test_bit(nr, addr)) + __set_bit(nr, addr); +} + +static inline void __clear_set_bit(unsigned nr, volatile unsigned long *addr) +{ + if (test_bit(nr, addr)) + __clear_bit(nr, addr); +} + +static inline void set_clear_bit(unsigned nr, volatile unsigned long *addr) +{ + if (!test_bit(nr, addr)) + set_bit(nr, addr); +} + +static inline void clear_set_bit(unsigned nr, volatile unsigned long *addr) +{ + if (test_bit(nr, addr)) + clear_bit(nr, addr); +} + #endif /* __KERNEL__ */ #endif (maybe __set_bit_if_clear would be a better name) But I don't know if it will do anything useful. The CPU *should* be able to avoid dirtying the cacheline on its own: it has all the info it needs to know that no writeback will be needed. But I don't know which (if any) CPUs perform this optimisation. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/