On Thu, 21 Jan 2021 07:57:03 +0100
Thomas Huth <thuth@xxxxxxxxxx> wrote:
> The "struct CPUEntry entries[0]" in the ReadInfo structure is misleading
> since the entries could be add a completely different spot. Replace it
s/add/at/
> by a proper comment instead.
>
> Signed-off-by: Thomas Huth <thuth@xxxxxxxxxx>
> ---
> lib/s390x/sclp.h | 9 +++++++--
> 1 file changed, 7 insertions(+), 2 deletions(-)
>
> diff --git a/lib/s390x/sclp.h b/lib/s390x/sclp.h
> index 9f81c0f..8523133 100644
> --- a/lib/s390x/sclp.h
> +++ b/lib/s390x/sclp.h
> @@ -131,10 +131,15 @@ typedef struct ReadInfo {
> uint16_t highest_cpu;
> uint8_t _reserved5[124 - 122]; /* 122-123 */
> uint32_t hmfai;
> - uint8_t reserved7[134 - 128];
> + uint8_t reserved7[134 - 128]; /* 128-133 */
> uint8_t byte_134_diag318 : 1;
> uint8_t : 7;
> - struct CPUEntry entries[0];
> + /*
> + * At the end of the ReadInfo, there are also the CPU entries (see
> + * struct CPUEntry). When the Extended-Length SCCB (ELS) feature is
> + * enabled, the start of the CPU entries array begins at an offset
> + * denoted by the offset_cpu field, otherwise it's at offset 128.
> + */
This comment is really helpful, as the difference for ELS had actually
slipped my mind again...
> } __attribute__((packed)) ReadInfo;
>
> typedef struct ReadCpuInfo {
Reviewed-by: Cornelia Huck <cohuck@xxxxxxxxxx>
