[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.DEB.2.20.1802011158110.3135@hadrien>
Date: Thu, 1 Feb 2018 12:06:35 +0100 (CET)
From: Julia Lawall <julia.lawall@...6.fr>
To: Dan Carpenter <dan.carpenter@...cle.com>
cc: Masahiro Yamada <yamada.masahiro@...ionext.com>,
kernel-janitors@...r.kernel.org,
Gilles Muller <Gilles.Muller@...6.fr>,
Nicolas Palix <nicolas.palix@...g.fr>,
Michal Marek <michal.lkml@...kovi.net>, cocci@...teme.lip6.fr,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH] coccinelle: reduce false positives
Here are the results that are eliminated by my change:
drivers/clk/axs10x/pll_clock.c:323:1-6 kfree(pll_clk);
drivers/clk/clk-gpio.c:131:2-7 kfree(clk_gpio);
drivers/clk/clk-hsdk-pll.c:410:1-6 kfree(pll_clk);
drivers/clk/hisilicon/clk.c:97:1-6 kfree(clk_data);
drivers/mfd/syscon.c:130:1-8 iounmap(base);
drivers/mfd/syscon.c:132:1-6 kfree(syscon);
drivers/pinctrl/freescale/pinctrl-mxs.c:139:2-7 kfree(group);
drivers/pinctrl/samsung/pinctrl-exynos5440.c:264:1-6 kfree(gname);
drivers/platform/chrome/cros_ec_debugfs.c:248:1-6 kfree(msg);
drivers/pwm/pwm-lp3943.c:56:3-8 kfree(pwm_map);
The semantic patch is pretty naive in that it assumes that all uses of the
same name point to the same thing.
Looking through the code, it looks like sometimes both an __init and a
probe function are provided, and the __init function doesn't have access to
a device object.
julia
Powered by blists - more mailing lists