Re: [Qemu-devel] [PATCH-for-4.1] target/arm: Add missing break statement for Hypervisor Trap Exception

[Peter Maydell <peter.maydell@linaro.org>]

On Fri, 19 Jul 2019 at 12:15, Philippe Mathieu-Daudé <philmd@redhat.com> wrote:
>
> Reported by GCC9 when building with  -Wimplicit-fallthrough=2:
>
>   target/arm/helper.c: In function ‘arm_cpu_do_interrupt_aarch32_hyp’:
>   target/arm/helper.c:7958:14: error: this statement may fall through [-Werror=implicit-fallthrough=]
>    7958 |         addr = 0x14;
>         |         ~~~~~^~~~~~
>   target/arm/helper.c:7959:5: note: here
>    7959 |     default:
>         |     ^~~~~~~
>   cc1: all warnings being treated as errors
>
> Fixes: b9bc21ff9f9
> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
>  target/arm/helper.c | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/target/arm/helper.c b/target/arm/helper.c
> index 20f8728be1..b74c23a9bc 100644
> --- a/target/arm/helper.c
> +++ b/target/arm/helper.c
> @@ -7956,6 +7956,7 @@ static void arm_cpu_do_interrupt_aarch32_hyp(CPUState *cs)
>          break;
>      case EXCP_HYP_TRAP:
>          addr = 0x14;
> +        break;
>      default:
>          cpu_abort(cs, "Unhandled exception 0x%x\n", cs->exception_index);
>      }

I think this is right, but EXCP_HYP_TRAP is a bit odd -- we appear
to use this only for the case of "SMC instruction is trapped from
NS EL1 to EL2 by HCR.TSC". I was expecting more traps-to-EL2
to use this EXCP_ variable... Mostly we seem to use EXCP_UDEF,
eg for CP_ACCESS_TRAP_UNCATEGORIZED_EL2 coprocessor/sysreg accesses:
this has the same behaviour as EXCP_HYP_TRAP as long as we know
we are going from an EL below 2 to EL2. Which I think we could
also use in the one place we use EXCP_HYP_TRAP; or we could make
wider use of EXCP_HYP_TRAP, since feeding everything through
EXCP_UDEF is rather confusing.

Anyway, for 4.1 we should do this.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

thanks
-- PMM