Re: [Qemu-ppc] [PATCH v4 5/5] adb.c: add power key support

[Programmingkid]

On Aug 15, 2016, at 8:19 AM, David Gibson wrote:

> On Fri, Aug 12, 2016 at 08:10:03PM -0400, John Arbuckle wrote:
>> Add support for the power key.
>> 
>> Signed-off-by: John Arbuckle <address@hidden>
>> Reviewed-by: Peter Maydell <address@hidden>
>> ---
>> v4 changes
>> Removed double-quote from end of comment.
>> Removed debug printf statement.
>> 
>> v3 change
>> Add several suggested comments.
>> Moved the location of an else statement in the adb_keyboard_event() function.
>> 
>> hw/input/adb.c | 43 +++++++++++++++++++++++++++++++++----------
>> 1 file changed, 33 insertions(+), 10 deletions(-)
>> 
>> diff --git a/hw/input/adb.c b/hw/input/adb.c
>> index 2042903..3998490 100644
>> --- a/hw/input/adb.c
>> +++ b/hw/input/adb.c
>> @@ -340,10 +340,25 @@ static int adb_kbd_poll(ADBDevice *d, uint8_t *obuf)
>>         s->rptr = 0;
>>     }
>>     s->count--;
>> -    obuf[0] = keycode;
>> -    /* NOTE: could put a second keycode if needed */
>> -    obuf[1] = 0xff;
>> -    olen = 2;
>> +    /*
>> +     * The power key is the only two byte value key, so it is a special 
>> case.
>> +     * Since 0x7f is not a used keycode for ADB we overload it to indicate 
>> the
>> +     * power button when we're storing keycodes in our internal buffer, and
>> +     * expand it out to two bytes when we send to the guest.
>> +     */
>> +    if (keycode == 0x7f) {
>> +        obuf[0] = 0x7f;
>> +        obuf[1] = 0x7f;
>> +        olen = 2;
>> +    } else {
>> +        obuf[0] = keycode;
>> +        /* NOTE: the power key key-up is the two byte sequence 0xff 0xff;
>> +         * otherwise we could in theory send a second keycode in the second
>> +         * byte, but choose not to bother.
>> +         */
>> +        obuf[1] = 0xff;
>> +        olen = 2;
>> +    }
>> 
>>     return olen;
>> }
>> @@ -421,14 +436,22 @@ static void adb_keyboard_event(DeviceState *dev, 
>> QemuConsole *src,
>>         return;
>>     }
>>     keycode = qcode_to_adb_keycode[qcode];
>> -    if (keycode == NO_KEY) {  /* We don't want to send this to the guest */
>> +
>> +    /* The power button is a special case because it is a 16-bit value */
>> +    if (qcode == Q_KEY_CODE_POWER) {
>> +        if (evt->u.key.data->down == true) { /* Power button pushed keycode 
>> */
>> +            adb_kbd_put_keycode(s, 0x7f);
>> +        } else {                           /* Power button released keycode 
>> */
>> +            adb_kbd_put_keycode(s, 0xff);
>> +        }
>> +    } else if (keycode == NO_KEY) { /* NO_KEY shouldn't be sent to guest */
>>         return;
>> +    } else {   /* For all non-power keys - safe for 8-bit keycodes */
>> +        if (evt->u.key.data->down == false) { /* if key release event */
>> +            keycode = keycode | 0x80;   /* create keyboard break code */
>> +        }
>> +        adb_kbd_put_keycode(s, keycode);
> 
> The logic in the CODE_POWER and else cases doesn't actually seem to be
> different AFAICT.

The power key is a 16 bit value. It needs its own special case or the operating
system would not see it. After disabling the "if (qcode == Q_KEY_CODE_POWER)" 
condition
so the power key would go thru the else statement, the guest stopped detecting 
the
power key.

> 
>>     }
>> -    if (evt->u.key.data->down == false) { /* if key release event */
>> -        keycode = keycode | 0x80;   /* create keyboard break code */
>> -    }
>> -
>> -    adb_kbd_put_keycode(s, keycode);
>> }
>> 
>> static const VMStateDescription vmstate_adb_kbd = {
>