On Sat, Jan 23, 2021 at 01:40:30PM +0800, 慕冬亮 wrote:
> Dear kernel developers,
>
> I found that on the syzbot dashboard, “KMSAN: uninit-value in
> smsc75xx_read_eeprom (2)” [1],
> "KMSAN: uninit-value in smsc95xx_read_eeprom (2)" [2], "KMSAN:
> uninit-value in smsc75xx_bind" [3],
> "KMSAN: uninit-value in smsc95xx_reset" [4], "KMSAN: uninit-value in
> smsc95xx_wait_eeprom (2)" [5]
> should share the same root cause.
>
> ## Root Cause Analysis && Different behaviors
>
> The root cause of these crash reports resides in the
> "__smsc75xx_read_reg/__smsc95xx_read_reg". Take __smsc95xx_read_reg as
> an example,
>
> -----------------------------------------------------------------------------------------------------------------
> static int __must_check __smsc95xx_read_reg(struct usbnet *dev, u32 index,
> u32 *data, int in_pm)
> {
> u32 buf;
> int ret;
> int (*fn)(struct usbnet *, u8, u8, u16, u16, void *, u16);
>
> BUG_ON(!dev);
>
> if (!in_pm)
> fn = usbnet_read_cmd;
> else
> fn = usbnet_read_cmd_nopm;
>
> ret = fn(dev, USB_VENDOR_REQUEST_READ_REGISTER, USB_DIR_IN
> | USB_TYPE_VENDOR | USB_RECIP_DEVICE,
> 0, index, &buf, 4);
> if (unlikely(ret < 0)) {
> netdev_warn(dev->net, "Failed to read reg index 0x%08x: %d\n",
> index, ret);
> return ret;
> }
>
> le32_to_cpus(&buf);
> *data = buf;
>
> return ret;
> }
>
>
> static int __must_check smsc95xx_eeprom_confirm_not_busy(struct usbnet *dev)
> {
> unsigned long start_time = jiffies;
> u32 val;
> int ret;
>
> do {
> ret = smsc95xx_read_reg(dev, E2P_CMD, &val);
> if (ret < 0) {
> netdev_warn(dev->net, "Error reading E2P_CMD\n");
> return ret;
> }
>
> if (!(val & E2P_CMD_BUSY_))
> return 0;
> ......
> }
> -----------------------------------------------------------------------------------------------------------------
>
> In a special situation, local variable "buf" is not initialized with
> "fn" function invocation. And the ret is bigger than zero, and buf is
> assigned to "*data". In its parent function -
> smsc95xx_eeprom_confirm_not_busy, KMSAN reports "uninit-value" when
> accessing variable "val".
> Note, due to the lack of testing environment, I don't know the
> concrete reason for the uninitialization of "buf" local variable.
>
> The reason for such different crash behaviors is that the event -
> "buf" is not initialized is random when
> "__smsc75xx_read_reg/__smsc95xx_read_reg" is invoked.
>
> ## Patch
>
> diff --git a/drivers/net/usb/smsc75xx.c b/drivers/net/usb/smsc75xx.c
> index 4353b370249f..a8e500d92285 100644
> --- a/drivers/net/usb/smsc75xx.c
> +++ b/drivers/net/usb/smsc75xx.c
> @@ -76,7 +76,7 @@ static int smsc75xx_phy_gig_workaround(struct usbnet *dev);
> static int __must_check __smsc75xx_read_reg(struct usbnet *dev, u32 index,
> u32 *data, int in_pm)
> {
> - u32 buf;
> + u32 buf = 0;
> int ret;
> int (*fn)(struct usbnet *, u8, u8, u16, u16, void *, u16);
>
> diff --git a/drivers/net/usb/smsc95xx.c b/drivers/net/usb/smsc95xx.c
> index 4c8ee1cff4d4..dae3be723e0c 100644
> --- a/drivers/net/usb/smsc95xx.c
> +++ b/drivers/net/usb/smsc95xx.c
> @@ -70,7 +70,7 @@ MODULE_PARM_DESC(turbo_mode, "Enable multiple frames
> per Rx transaction");
> static int __must_check __smsc95xx_read_reg(struct usbnet *dev, u32 index,
> u32 *data, int in_pm)
> {
> - u32 buf;
> + u32 buf = 0;
> int ret;
> int (*fn)(struct usbnet *, u8, u8, u16, u16, void *, u16);
>
> If you can have any issues with this statement or our information is
> useful to you, please let us know. Thanks very much.
>
> [1] “KMSAN: uninit-value in smsc75xx_read_eeprom (2)” - url
> [2] “KMSAN: uninit-value in smsc95xx_read_eeprom (2)” - URL
> [3] "KMSAN: uninit-value in smsc75xx_bind" -
> [4] "KMSAN: uninit-value in smsc95xx_reset" -
> [5] "KMSAN: uninit-value in smsc95xx_wait_eeprom (2)" -
As I asked before, please just turn this into a real patch and submit it
to the syzbot to see if it fixes the issue. If so, then submit it
normally, no need to do any huge writeup.
thnaks,
greg k-h
