Re: [PATCH] target/riscv: Smepmp: Skip applying default rules when address matches

[LIU Zhiwei]

On 2023/2/9 13:52, Himanshu Chauhan wrote:
> When MSECCFG.MML is set, after checking the address range in PMP if the
> asked permissions are not same as programmed in PMP, the default
> permissions are applied. This should only be the case when there
> is no matching address is found.
>
> This patch skips applying default rules when matching address range
> is found. It returns the index of the match PMP entry.
>
> fixes: 824cac681c3 (target/riscv: Fix PMP propagation for tlb)
>
> Signed-off-by: Himanshu Chauhan <hchauhan@ventanamicro.com>
> ---
>   target/riscv/pmp.c | 9 ++++++---
>   1 file changed, 6 insertions(+), 3 deletions(-)
>
> diff --git a/target/riscv/pmp.c b/target/riscv/pmp.c
> index d85ad07caa..0dfdb35828 100644
> --- a/target/riscv/pmp.c
> +++ b/target/riscv/pmp.c
> @@ -446,9 +446,12 @@ int pmp_hart_has_privs(CPURISCVState *env, target_ulong 
> addr,
>                   }
>               }
>   
> -            if ((privs & *allowed_privs) == privs) {
> -                ret = i;
> -            }
> +            /*
> +             * If matching address range was found, the protection bits
> +             * defined with PMP must be used. We shouldn't fallback on
> +             * finding default privileges.
> +             */
> +            ret = i;

Notice the return value is the matching rule index, which includes

1) the address range is matching.

2) the permission of the PMP rule and the memory access type are matching.


So we can't simply remove the second check.  I think the right fix is:

           if ((privs & *allowed_privs) == privs) {
                ret = i;
 -         }
 +         } else {
 +              ret = -2;
 +         }

The -2 return value avoids finding the default privileges. And it implies no 
matching rule is found.

Zhiwei

>               break;
>           }
>       }