From: Andrew Morton Subject: Re: jbd : config_jbd_debug cannot create /proc entry Date: Wed, 26 Sep 2007 14:35:39 -0700 Message-ID: <20070926143539.b01bc0bc.akpm@linux-foundation.org> References: <1190713240.3154.20.camel@castor.rsk.org> <20070925105112.GB8126@atrey.karlin.mff.cuni.cz> <46F8EB26.2090104@linux.vnet.ibm.com> <20070925115046.GA15394@atrey.karlin.mff.cuni.cz> <20070925074938.2321bc15@gara> <20070925084153.36ab65a5@gara> <20070925143608.GB18758@atrey.karlin.mff.cuni.cz> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: jrs@us.ibm.com, aneesh.kumar@linux.vnet.ibm.com, richard@rsk.demon.co.uk, sct@redhat.com, linux-kernel@vger.kernel.org, linux-ext4@vger.kernel.org To: Jan Kara Return-path: Received: from smtp2.linux-foundation.org ([207.189.120.14]:58282 "EHLO smtp2.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754106AbXIZVhE (ORCPT ); Wed, 26 Sep 2007 17:37:04 -0400 In-Reply-To: <20070925143608.GB18758@atrey.karlin.mff.cuni.cz> Sender: linux-ext4-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org On Tue, 25 Sep 2007 16:36:08 +0200 Jan Kara wrote: > > On Tue, 25 Sep 2007 07:49:38 -0500 > > "Jose R. Santos" wrote: > > > > > On Tue, 25 Sep 2007 13:50:46 +0200 > > > Jan Kara wrote: > > > > > Jan Kara wrote: > > > > > >> > > > > > >-#define create_jbd_proc_entry() do {} while (0) > > > > > >-#define remove_jbd_proc_entry() do {} while (0) > > > > > >+static ctl_table fs_table[] = { > > > > > >+ { > > > > > >+ .ctl_name = -1, /* Don't want it */ > > > > > > > > > > > > > > > > > > > > shouldn't this be CTL_UNNUMBERED ? > > > > Oh, it should be. I didn't notice we have this :) Thanks for notifying > > > > me. Attached is a fixed version. > > > > > > This was fixed in JBD2 by moving the jbd-debug file to debugfs: > > > http://lkml.org/lkml/2007/7/11/334 > > > > > > Since this code is already in the kernel, we should keep it consistent. > > > > > > > OK. Here's a quick patch to fix this. Adapted from the JBD2 patch. > > Let me know what you think. > Looks fine - exactly what I've just done here :). hm. I found rather a lot of issues. If this patch is derived from the JBD2 patch then perhaps the JBD2 patch needs some looking at. > > Signed-off-by: Jose R. Santos > You can add Signed-off-by: Jan Kara I suspect you might be getting your signed-off-bys and acked-bys mixed up. (If not this patch, then the previous one). Please see Documentation/SubmittingPatches section 13 for the difference. Jose, please review and if possible runtime test these proposed changes? From: Andrew Morton - use `#ifdef foo' instead of `#if defined(foo)' - CONFIG_JBD_DEBUG depends on CONFIG_DEBUG_FS so we don't need to duplicate that logic in the .c file ifdefs - Make journal_enable_debug __read_mostly just for the heck of it - Make jbd_debugfs_dir and jbd_debug static - debugfs_remove(NULL) is legal: remove unneeded tests - jbd_create_debugfs_entry is a better name than create_jbd_debugfs_entry - ditto remove_jbd_debugfs_entry - C functions are preferred over macros Cc: "Jose R. Santos" Cc: Cc: Jan Kara Cc: Jose R. Santos Signed-off-by: Andrew Morton --- diff -puN fs/jbd/journal.c~jbd-config_jbd_debug-cannot-create-proc-entry-fix fs/jbd/journal.c --- a/fs/jbd/journal.c~jbd-config_jbd_debug-cannot-create-proc-entry-fix +++ a/fs/jbd/journal.c @@ -1853,16 +1853,15 @@ void journal_put_journal_head(struct jou /* * debugfs tunables */ -#if defined(CONFIG_JBD_DEBUG) -u8 journal_enable_debug; -EXPORT_SYMBOL(journal_enable_debug); -#endif +#ifdef CONFIG_JBD_DEBUG -#if defined(CONFIG_JBD_DEBUG) && defined(CONFIG_DEBUG_FS) +u8 journal_enable_debug __read_mostly; +EXPORT_SYMBOL(journal_enable_debug); -struct dentry *jbd_debugfs_dir, *jbd_debug; +static struct dentry *jbd_debugfs_dir; +static struct dentry *jbd_debug; -static void __init create_jbd_debugfs_entry(void) +static void __init jbd_create_debugfs_entry(void) { jbd_debugfs_dir = debugfs_create_dir("jbd", NULL); if (jbd_debugfs_dir) @@ -1871,18 +1870,21 @@ static void __init create_jbd_debugfs_en &journal_enable_debug); } -static void __exit remove_jbd_debugfs_entry(void) +static void __exit jbd_remove_debugfs_entry(void) { - if (jbd_debug) - debugfs_remove(jbd_debug); - if (jbd_debugfs_dir) - debugfs_remove(jbd_debugfs_dir); + debugfs_remove(jbd_debug); + debugfs_remove(jbd_debugfs_dir); } #else -#define create_jbd_debugfs_entry() do {} while (0) -#define remove_jbd_debugfs_entry() do {} while (0) +static inline void jbd_create_debugfs_entry(void) +{ +} + +static inline void jbd_remove_debugfs_entry(void) +{ +} #endif @@ -1940,7 +1942,7 @@ static int __init journal_init(void) ret = journal_init_caches(); if (ret != 0) journal_destroy_caches(); - create_jbd_debugfs_entry(); + jbd_create_debugfs_entry(); return ret; } @@ -1951,7 +1953,7 @@ static void __exit journal_exit(void) if (n) printk(KERN_EMERG "JBD: leaked %d journal_heads!\n", n); #endif - remove_jbd_debugfs_entry(); + jbd_remove_debugfs_entry(); journal_destroy_caches(); } _