[PATCH 0/6] treewide: test clock rate to avoid division by 0

Here is the outcome of researching if the result of clk_get_rate() was directly
used as a divisor without checking if it is 0. Inspired by a Coverity report.

Wolfram Sang (6):
  ide: palm_bk3710: test clock rate to avoid division by 0
  net: ethernet: renesas: ravb_main: test clock rate to avoid division
    by 0
  pwm: pwm-img: test clock rate to avoid division by 0
  pwm: pwm-lpc18xx-sct: test clock rate to avoid division by 0
  watchdog: atlas7_wdt: test clock rate to avoid division by 0
  watchdog: tangox_wdt: test clock rate to avoid division by 0

 drivers/ide/palm_bk3710.c                | 2 ++
 drivers/net/ethernet/renesas/ravb_main.c | 3 +++
 drivers/pwm/pwm-img.c                    | 5 +++++
 drivers/pwm/pwm-lpc18xx-sct.c            | 2 ++
 drivers/watchdog/atlas7_wdt.c            | 5 +++++
 drivers/watchdog/tangox_wdt.c            | 2 ++
 6 files changed, 19 insertions(+)

-- 
2.7.0

From: Wolfram Sang <wsa+renesas@sang-engineering.com>

The clk API may return 0 on clk_get_rate, so we should check the result before
using it as a divisor.

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
---

Should go individually via subsystem tree.

 drivers/pwm/pwm-lpc18xx-sct.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/pwm/pwm-lpc18xx-sct.c b/drivers/pwm/pwm-lpc18xx-sct.c
index 9163085101bc94..6487962c355c03 100644
--- a/drivers/pwm/pwm-lpc18xx-sct.c
+++ b/drivers/pwm/pwm-lpc18xx-sct.c

@@ -360,6 +360,8 @@ static int lpc18xx_pwm_probe(struct platform_device *pdev)
 	}
 
 	lpc18xx_pwm->clk_rate = clk_get_rate(lpc18xx_pwm->pwm_clk);
+	if (!lpc18xx_pwm->clk_rate)
+		return -EINVAL;
 
 	mutex_init(&lpc18xx_pwm->res_lock);
 	mutex_init(&lpc18xx_pwm->period_lock);

-- 
2.7.0

Hi Wolfram,

On 2 March 2016 at 23:33, Wolfram Sang [off-list ref] wrote:

This needs to be:
if (!lpc18xx_pwm->clk_rate) {
    ret = -EINVAL;
    goto disable_pwmclk;
}

I would also prefer an explicit check against 0 here. ie.:
'lpc18xx_pwm->clk_rate == 0'
A dev_err() message would also be nice to have.


regards,
Joachim Eastwood

On Wed, Mar 02, 2016 at 11:44:02PM +0100, Joachim Eastwood wrote:

Hi Wolfram,

On 2 March 2016 at 23:33, Wolfram Sang [off-list ref] wrote:

quoted

From: Wolfram Sang <wsa+renesas@sang-engineering.com>

The clk API may return 0 on clk_get_rate, so we should check the result before
using it as a divisor.

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
---

Should go individually via subsystem tree.

 drivers/pwm/pwm-lpc18xx-sct.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/pwm/pwm-lpc18xx-sct.c b/drivers/pwm/pwm-lpc18xx-sct.c
index 9163085101bc94..6487962c355c03 100644
--- a/drivers/pwm/pwm-lpc18xx-sct.c
+++ b/drivers/pwm/pwm-lpc18xx-sct.c

@@ -360,6 +360,8 @@ static int lpc18xx_pwm_probe(struct platform_device *pdev)
        }

        lpc18xx_pwm->clk_rate = clk_get_rate(lpc18xx_pwm->pwm_clk);
+       if (!lpc18xx_pwm->clk_rate)
+               return -EINVAL;

This needs to be:
if (!lpc18xx_pwm->clk_rate) {
    ret = -EINVAL;
    goto disable_pwmclk;
}

Yes, that slipped through. Thanks!

I would also prefer an explicit check against 0 here. ie.:

Well, I like the reading "if not rate then error"

Will send V2 now...

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160302/e24b2786/attachment.sig>