Re: [PATCH 0/2] SEV firmware error list touchups

[Connor Kuehl]

On 2/18/21 9:48 AM, Philippe Mathieu-Daudé wrote:
> On 2/18/21 4:16 PM, Connor Kuehl wrote:
> > Connor Kuehl (2):
> >    sev: use explicit indices for mapping firmware error codes to strings
> >    sev: add missing firmware error conditions
> >
> >   target/i386/sev.c | 48 ++++++++++++++++++++++++-----------------------
> >   1 file changed, 25 insertions(+), 23 deletions(-)
>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>

Thank you! :-)

> To avoid this problem in future (new error code added on the Linux
> kernel side) would it be acceptable to add a 3rd patch as:
>
> -- >8 --
> diff --git a/target/i386/sev.c b/target/i386/sev.c
> index 0f414df02f3..e086d3198e8 100644
> --- a/target/i386/sev.c
> +++ b/target/i386/sev.c
> @@ -155,9 +155,12 @@ sev_platform_ioctl(int fd, int cmd, void *data, int
> *error)
>   static const char *
>   fw_error_to_str(int code)
>   {
> +    QEMU_BUILD_BUG_ON(SEV_RET_SECURE_DATA_INVALID + 1 == SEV_RET_MAX);
> +
>       if (code < 0 || code >= SEV_FW_MAX_ERROR) {
>           return "unknown error";
>       }
> +    assert(sev_fw_errlist[code]);
>
>       return sev_fw_errlist[code];
>   }
> ---
>
> which triggers a build error if scripts/update-linux-headers.sh
> added another sev_ret_code entry?

I like this a lot. Should I send a v2 of the series with a third patch 
like this or shall I wait to see if these patches get applied then send 
something like this as a follow up patch?

Thank you,

Connor