Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1764591AbYARRnJ (ORCPT ); Fri, 18 Jan 2008 12:43:09 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1761952AbYARRm5 (ORCPT ); Fri, 18 Jan 2008 12:42:57 -0500 Received: from pentafluge.infradead.org ([213.146.154.40]:48011 "EHLO pentafluge.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1759846AbYARRm5 (ORCPT ); Fri, 18 Jan 2008 12:42:57 -0500 Date: Fri, 18 Jan 2008 09:42:16 -0800 From: Arjan van de Ven To: Arjan van de Ven Cc: Linux Kernel Mailing List , Linus Torvalds , Ingo Molnar , Andrew Morton Subject: [patch 3/3] LatencyTOP instrumentations part 2 Message-ID: <20080118094216.1bf78ff3@laptopd505.fenrus.org> In-Reply-To: <4790E3A6.7060807@linux.intel.com> References: <4790E3A6.7060807@linux.intel.com> Organization: Intel X-Mailer: Claws Mail 3.2.0 (GTK+ 2.12.3; i386-redhat-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Bad-Reply: References and In-Reply-To but no 'Re:' in Subject. Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3615 Lines: 121 A few more high level annotations for LatencyTOP; split into a separate patch since they're at a higher level than the first part (and thus hide some lower level details potentially, but could also obsolete several other low level instrumentations) --- fs/read_write.c | 7 ++++++- fs/readdir.c | 4 +++- lib/kernel_lock.c | 3 +++ 3 files changed, 12 insertions(+), 2 deletions(-) Index: linux-2.6.24-rc7/fs/read_write.c =================================================================== --- linux-2.6.24-rc7.orig/fs/read_write.c +++ linux-2.6.24-rc7/fs/read_write.c @@ -255,6 +255,7 @@ EXPORT_SYMBOL(do_sync_read); ssize_t vfs_read(struct file *file, char __user *buf, size_t count, loff_t *pos) { + struct latency_entry reason; ssize_t ret; if (!(file->f_mode & FMODE_READ)) @@ -264,6 +265,7 @@ ssize_t vfs_read(struct file *file, char if (unlikely(!access_ok(VERIFY_WRITE, buf, count))) return -EFAULT; + set_latency_reason("Reading from file", &reason); ret = rw_verify_area(READ, file, pos, count); if (ret >= 0) { count = ret; @@ -281,6 +283,7 @@ ssize_t vfs_read(struct file *file, char } } + restore_latency_reason(&reason); return ret; } @@ -314,6 +317,7 @@ EXPORT_SYMBOL(do_sync_write); ssize_t vfs_write(struct file *file, const char __user *buf, size_t count, loff_t *pos) { ssize_t ret; + struct latency_entry reason; if (!(file->f_mode & FMODE_WRITE)) return -EBADF; @@ -322,6 +326,7 @@ ssize_t vfs_write(struct file *file, con if (unlikely(!access_ok(VERIFY_READ, buf, count))) return -EFAULT; + set_latency_reason("Writing to file", &reason); ret = rw_verify_area(WRITE, file, pos, count); if (ret >= 0) { count = ret; @@ -338,7 +343,7 @@ ssize_t vfs_write(struct file *file, con inc_syscw(current); } } - + restore_latency_reason(&reason); return ret; } Index: linux-2.6.24-rc7/fs/readdir.c =================================================================== --- linux-2.6.24-rc7.orig/fs/readdir.c +++ linux-2.6.24-rc7/fs/readdir.c @@ -21,6 +21,7 @@ int vfs_readdir(struct file *file, filldir_t filler, void *buf) { + struct latency_entry reason; struct inode *inode = file->f_path.dentry->d_inode; int res = -ENOTDIR; if (!file->f_op || !file->f_op->readdir) @@ -29,7 +30,7 @@ int vfs_readdir(struct file *file, filld res = security_file_permission(file, MAY_READ); if (res) goto out; - + set_latency_reason("Unlinking file", &reason); mutex_lock(&inode->i_mutex); res = -ENOENT; if (!IS_DEADDIR(inode)) { @@ -37,6 +38,7 @@ int vfs_readdir(struct file *file, filld file_accessed(file); } mutex_unlock(&inode->i_mutex); + restore_latency_reason(&reason); out: return res; } Index: linux-2.6.24-rc7/lib/kernel_lock.c =================================================================== --- linux-2.6.24-rc7.orig/lib/kernel_lock.c +++ linux-2.6.24-rc7/lib/kernel_lock.c @@ -64,15 +64,18 @@ void __lockfunc __release_kernel_lock(vo */ void __lockfunc lock_kernel(void) { + struct latency_entry reason; struct task_struct *task = current; int depth = task->lock_depth + 1; + set_latency_reason("Big kernel lock contention", &reason); if (likely(!depth)) /* * No recursion worries - we set up lock_depth _after_ */ down(&kernel_sem); + restore_latency_reason(&reason); task->lock_depth = depth; } -- 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/