Re: [PATCH 7/7] target/riscv: Honour -semihosting-config userspace=on and enable=on

[Richard Henderson]

On 8/15/22 14:03, Peter Maydell wrote:
> The riscv target incorrectly enabled semihosting always, whether the
> user asked for it or not.  Call semihosting_enabled() passing the
> correct value to the is_userspace argument, which fixes this and also
> handles the userspace=on argument.
>
> Note that this is a behaviour change: we used to default to
> semihosting being enabled, and now the user must pass
> "-semihosting-config enable=on" if they want it.
>
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
>   target/riscv/cpu_helper.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c
> index 59b3680b1b2..49c4ea98ac9 100644
> --- a/target/riscv/cpu_helper.c
> +++ b/target/riscv/cpu_helper.c
> @@ -24,6 +24,7 @@
>   #include "exec/exec-all.h"
>   #include "tcg/tcg-op.h"
>   #include "trace.h"
> +#include "semihosting/semihost.h"
>   #include "semihosting/common-semi.h"
>   
>   int riscv_cpu_mmu_index(CPURISCVState *env, bool ifetch)
> @@ -1342,7 +1343,7 @@ void riscv_cpu_do_interrupt(CPUState *cs)
>       target_ulong mtval2 = 0;
>   
>       if  (cause == RISCV_EXCP_SEMIHOST) {
> -        if (env->priv >= PRV_S) {
> +        if (semihosting_enabled(env->priv < PRV_S)) {
>               do_common_semihosting(cs);
>               env->pc += 4;
>               return;

I think this should be done in translate.  We should not have the overhead of checking the 
three insn sequence around ebreak unless semihosting is enabled.  Note that ctx->mem_idx 
== env->priv, per cpu_mem_index().


r~