From mboxrd@z Thu Jan 1 00:00:00 1970 From: Russell King - ARM Linux Subject: Re: [PATCH] arm: mach-tegra: check wether we have debugfs enabled or not Date: Wed, 17 Oct 2012 23:14:12 +0100 Message-ID: <20121017221412.GO21164@n2100.arm.linux.org.uk> References: <1350501196-5632-1-git-send-email-devendra.aaru@gmail.com> <20121017201244.GN21164@n2100.arm.linux.org.uk> <507F2185.3070408@wwwdotorg.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <507F2185.3070408-3lzwWm7+Weoh9ZMKESR00Q@public.gmane.org> Sender: linux-tegra-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Stephen Warren Cc: Devendra Naga , linux-tegra-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-Id: linux-tegra@vger.kernel.org On Wed, Oct 17, 2012 at 03:22:13PM -0600, Stephen Warren wrote: > On 10/17/2012 02:12 PM, Russell King - ARM Linux wrote: > > On Wed, Oct 17, 2012 at 03:13:16PM -0400, Devendra Naga wrote: > >> if debugfs is not enabled in kernel it returns -ENODEV, check it > >> with IS_ERR_OR_NULL along with the null return if it fails to > >> create a file in debugfs. > >> > >> Signed-off-by: Devendra Naga > >> --- > >> arch/arm/mach-tegra/powergate.c | 2 +- > >> 1 file changed, 1 insertion(+), 1 deletion(-) > >> > >> diff --git a/arch/arm/mach-tegra/powergate.c b/arch/arm/mach-tegra/powergate.c > >> index de0662d..1809b91d 100644 > >> --- a/arch/arm/mach-tegra/powergate.c > >> +++ b/arch/arm/mach-tegra/powergate.c > >> @@ -269,7 +269,7 @@ int __init tegra_powergate_debugfs_init(void) > >> if (powergate_name) { > >> d = debugfs_create_file("powergate", S_IRUGO, NULL, NULL, > >> &powergate_fops); > >> - if (!d) > >> + if (IS_ERR_OR_NULL(d)) > >> return -ENOMEM; > > > > So, you want to make this function hard-fail if a debugfs error occurs > > _or_ debugfs is not configured? That doesn't seem to be a clever thing > > to do. > > In practice, that quoted chunk of code is inside an #ifdef > CONFIG_DEBUG_FS, so according to the documentation for > debugfs_create_file(), the IS_ERR case of IS_ERR_OR_NULL cannot happen > here. So, the above change is a no-op, so there's not much point > applying it. Ok. Another reason not to apply it is that should the ifdefs get removed, it'll start failing when debugfs is disabled with -ENOMEM, which isn't bright. A further reason is it's another place that'd need reviewing for bad usage of the IS_ERR_OR_NULL() problem I've pointed out this evening.