On 23.04.2013, at 13:45, Christian Borntraeger wrote: > On 23/04/13 13:31, Alexander Graf wrote: >> On 04/23/2013 01:23 PM, Christian Borntraeger wrote: >>> Lets fix this gcc warning: >>> >>> virtio.c: In function ‘vring_send_buf’: >>> virtio.c:125:35: error: operation on ‘vr->next_idx’ may be undefined >>> [-Werror=sequence-point] >>> >>> Signed-off-by: Christian Borntraeger<borntrae...@de.ibm.com> >>> --- >>> pc-bios/s390-ccw/virtio.c | 4 +++- >>> 1 file changed, 3 insertions(+), 1 deletion(-) >>> >>> diff --git a/pc-bios/s390-ccw/virtio.c b/pc-bios/s390-ccw/virtio.c >>> index 79e2941..1968fc6 100644 >>> --- a/pc-bios/s390-ccw/virtio.c >>> +++ b/pc-bios/s390-ccw/virtio.c >>> @@ -122,7 +122,9 @@ static void vring_send_buf(struct vring *vr, void *p, >>> int len, int flags) >>> vr->desc[vr->next_idx].addr = (ulong)p; >>> vr->desc[vr->next_idx].len = len; >>> vr->desc[vr->next_idx].flags = flags& ~VRING_HIDDEN_IS_CHAIN; >>> - vr->desc[vr->next_idx].next = ++vr->next_idx; >>> + vr->desc[vr->next_idx].next = vr->next_idx; >>> + vr->desc[vr->next_idx].next++; >>> + vr->next_idx++; >> >> Hrm. This looks a lot more complicated than it needs to. Why is gcc emitting >> a warning here? Maybe it only wants braces. Please try whether >> >> .next = ++(vr->next_id); >> >> works for you. >> >> >> Alex >> >> > > I already did that, but I get the same error message. I dont understand why > gcc emits that warning, > but it does :-(
*shrug* I'll just add it to the queue. Alex
