On (Tue) 12 Apr 2011 [16:07:23], Kevin Wolf wrote:
> Am 12.04.2011 15:59, schrieb Amit Shah:
> > On (Tue) 12 Apr 2011 [15:09:33], Kevin Wolf wrote:
> >> Am 12.04.2011 13:43, schrieb Amit Shah:
> >>> Handle GET_EVENT_STATUS_NOTIFICATION's No Event Available response in a
> >>> generic way so that future additions to the code to handle other
> >>> response types is easier.
> >>>
> >>> Signed-off-by: Amit Shah <amit.s...@redhat.com>
> >>> ---
> >>> hw/ide/core.c | 24 +++++++++++++++++++-----
> >>> 1 files changed, 19 insertions(+), 5 deletions(-)
> >>>
> >>> diff --git a/hw/ide/core.c b/hw/ide/core.c
> >>> index e838990..a9bc1e3 100644
> >>> --- a/hw/ide/core.c
> >>> +++ b/hw/ide/core.c
> >>> @@ -1098,9 +1098,17 @@ static void
> >>> handle_get_event_status_notification(IDEState *s,
> >>> uint8_t control;
> >>> } __attribute__((packed)) *gesn_cdb;
> >>>
> >>> + struct {
> >>> + uint16_t len;
> >>> + uint8_t notification_class;
> >>> + uint8_t supported_events;
> >>> + } __attribute((packed)) *gesn_event_header;
> >>> +
> >>> unsigned int max_len, used_len;
> >>>
> >>> gesn_cdb = (void *)packet;
> >>> + gesn_event_header = (void *)packet;
> >>
> >> I think this should be buf, not packet.
> >
> > Ah, right. (Though they're the same.)
>
> Oh, you're right. I wasn't even aware of that. At some point we should
> clean this up, it's an invitation for bugs...
It is. Actually it's surprising gcc didn't tell me that casting from
const uint8_t * to void * and then writing to the pointer is not a
good idea.
Amit
