Re: [PATCH] Fix Atmel TPM crash caused by too frequent queries

[Mimi Zohar <zohar@linux.ibm.com>]

Hi Hao,

When posting to the mailing list, please respond using plain text and
inline/bottom post.

On Wed, 2020-09-30 at 22:26 -0700, Hao Wu wrote:
> 7.1
> 
> - Jarkko’s proposal: Using msleep(TPM_TIMEOUT) for wait_for_tpm_stat(), resolve
>   the Atmel crash as we expect.  
> 
> ---
>  drivers/char/tpm/tpm_tis_core.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/char/tpm/tpm_tis_core.c b/drivers/char/tpm/tpm_tis_core.c
> index 65ab1b027949..7dd9bcff542d 100644
> --- a/drivers/char/tpm/tpm_tis_core.c
> +++ b/drivers/char/tpm/tpm_tis_core.c
> @@ -80,8 +80,9 @@ static int wait_for_tpm_stat(struct tpm_chip *chip, u8 mask,
>  		}
>  	} else {
>  		do {
> -			usleep_range(TPM_TIMEOUT_USECS_MIN,
> -				     TPM_TIMEOUT_USECS_MAX);
> +            msleep(TPM_TIMEOUT);
> +			// usleep_range(TPM_TIMEOUT_USECS_MIN,
> +			//	     TPM_TIMEOUT_USECS_MAX);
>  			status = chip->ops->status(chip);
>  			if ((status & mask) == mask)
>  				return 0;
> --
> 2.17.1
> 
> But I think tpm_msleep(15) is still the right way to go.

Using msleep is definitely not the right way of going.   Please refer
to commit a233a0289cf9 ("tpm: msleep() delays - replace with
usleep_range() in i2c nuvoton driver") for a detailed explanation.

thanks,

Mimi