On Wed, Sep 27, 2017 at 4:47 PM, Peter Maydell <peter.mayd...@linaro.org> wrote:
> On 26 September 2017 at 06:32, Eric Blake <ebl...@redhat.com> wrote:
>> On 09/25/2017 07:08 PM, Alistair Francis wrote:
>>> diff --git a/hw/arm/nseries.c b/hw/arm/nseries.c
>>> index 58005b6619..32687afced 100644
>>> --- a/hw/arm/nseries.c
>>> +++ b/hw/arm/nseries.c
>>> @@ -463,7 +463,7 @@ static uint32_t mipid_txrx(void *opaque, uint32_t cmd,
>>> int len)
>>> uint8_t ret;
>>>
>>> if (len > 9) {
>>> - hw_error("%s: FIXME: bad SPI word width %i\n", __FUNCTION__, len);
>>> + hw_error("%s: FIXME: bad SPI word width %i\n", __func__, len);
>>
>> Not this patch's problem, but it would probably be simpler if hw_error()
>> were a macro that automatically prefixed __func__, rather than making
>> every caller have to supply it themselves.
>
> I'm not sure there's a great deal of benefit to that change, because
> use of hw_error() in new code is rarely correct (it does an abort()
> so it should never be used for guest-triggered conditions, which is
> about the only time that you might be interested in a guest register
> dump rather than just asserting). Most of its existing uses are in
> crufty old device models.
Ok, another excuse to leave it out :)
Thanks,
Alistair
>
> thanks
> -- PMM
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
