2019-02-06 02:25:45

by Yizhuo Zhai

Subject: [PATCH] phy: Variable "caldone" in function rockchip_emmc_phy_power() could be uninitialized

In function rockchip_emmc_phy_power(), local variable "caldone"
could be uninitialized if function regmap_read() returns -EINVAL.
However, it will be used directly in the later context, which
is potentially unsafe.

Signed-off-by: Yizhuo <[email protected]>
---
drivers/phy/rockchip/phy-rockchip-emmc.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/phy/rockchip/phy-rockchip-emmc.c b/drivers/phy/rockchip/phy-rockchip-emmc.c
index f1b24f18e9b2..ca921ae08309 100644
--- a/drivers/phy/rockchip/phy-rockchip-emmc.c
+++ b/drivers/phy/rockchip/phy-rockchip-emmc.c
@@ -85,11 +85,12 @@ struct rockchip_emmc_phy {
static int rockchip_emmc_phy_power(struct phy *phy, bool on_off)
{
struct rockchip_emmc_phy *rk_phy = phy_get_drvdata(phy);
- unsigned int caldone;
+ unsigned int caldone = 0;
unsigned int dllrdy;
unsigned int freqsel = PHYCTRL_FREQSEL_200M;
unsigned long rate;
unsigned long timeout;
+ int err;

/*
* Keep phyctrl_pdb and phyctrl_endll low to allow
@@ -164,9 +165,14 @@ static int rockchip_emmc_phy_power(struct phy *phy, bool on_off)
* wait 5us for calpad busy trimming
*/
udelay(5);
- regmap_read(rk_phy->reg_base,
+ err = regmap_read(rk_phy->reg_base,
rk_phy->reg_offset + GRF_EMMCPHY_STATUS,
&caldone);
+ if (err) {
+ pr_err("Failed to read status.\n");
+ return err;
+ }
+
caldone = (caldone >> PHYCTRL_CALDONE_SHIFT) & PHYCTRL_CALDONE_MASK;
if (caldone != PHYCTRL_CALDONE_DONE) {
pr_err("rockchip_emmc_phy_power: caldone timeout.\n");
--
2.17.1

2019-02-12 11:20:07

by Heiko Stuebner

Subject: Re: [PATCH] phy: Variable "caldone" in function rockchip_emmc_phy_power() could be uninitialized

Am Mittwoch, 6. Februar 2019, 02:47:33 CET schrieb Yizhuo:
> In function rockchip_emmc_phy_power(), local variable "caldone"
> could be uninitialized if function regmap_read() returns -EINVAL.
> However, it will be used directly in the later context, which
> is potentially unsafe.
>
> Signed-off-by: Yizhuo <[email protected]>
> ---
> drivers/phy/rockchip/phy-rockchip-emmc.c | 10 ++++++++--
> 1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/phy/rockchip/phy-rockchip-emmc.c
> b/drivers/phy/rockchip/phy-rockchip-emmc.c index f1b24f18e9b2..ca921ae08309
> 100644
> --- a/drivers/phy/rockchip/phy-rockchip-emmc.c
> +++ b/drivers/phy/rockchip/phy-rockchip-emmc.c
> @@ -85,11 +85,12 @@ struct rockchip_emmc_phy {
> static int rockchip_emmc_phy_power(struct phy *phy, bool on_off)
> {
> struct rockchip_emmc_phy *rk_phy = phy_get_drvdata(phy);
> - unsigned int caldone;
> + unsigned int caldone = 0;

hmm, this may hide actual uninitialized uses in the future?
With the added error handling below, there shouldn't be a case
where caldone might be used uninitialized still?


Heiko

> unsigned int dllrdy;
> unsigned int freqsel = PHYCTRL_FREQSEL_200M;
> unsigned long rate;
> unsigned long timeout;
> + int err;
>
> /*
> * Keep phyctrl_pdb and phyctrl_endll low to allow
> @@ -164,9 +165,14 @@ static int rockchip_emmc_phy_power(struct phy *phy,
> bool on_off) * wait 5us for calpad busy trimming
> */
> udelay(5);
> - regmap_read(rk_phy->reg_base,
> + err = regmap_read(rk_phy->reg_base,
> rk_phy->reg_offset + GRF_EMMCPHY_STATUS,
> &caldone);
> + if (err) {
> + pr_err("Failed to read status.\n");
> + return err;
> + }
> +
> caldone = (caldone >> PHYCTRL_CALDONE_SHIFT) & PHYCTRL_CALDONE_MASK;
> if (caldone != PHYCTRL_CALDONE_DONE) {
> pr_err("rockchip_emmc_phy_power: caldone timeout.\n");