On Tue, 2018-07-10 at 13:03 -0600, Jason Gunthorpe wrote:
> In some configurations even gcc 7 cannot unravel this complexity and
> still throws a warning.
>
> Fixes: 4ab39e2f98f2 ("RDMA/cxgb4: Make c4iw_poll_cq_one() easier to analyze")
> Reported-by: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx>
> Signed-off-by: Jason Gunthorpe <jgg@xxxxxxxxxxxx>
> ---
> drivers/infiniband/hw/cxgb4/cq.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> Ideally 'poll one' would be simplified into something that gcc could
> actually analyze..
>
> diff --git a/drivers/infiniband/hw/cxgb4/cq.c b/drivers/infiniband/hw/cxgb4/cq.c
> index a5280d8d002fb2..a3a829951ac4a5 100644
> --- a/drivers/infiniband/hw/cxgb4/cq.c
> +++ b/drivers/infiniband/hw/cxgb4/cq.c
> @@ -671,7 +671,7 @@ static int poll_cq(struct t4_wq *wq, struct t4_cq *cq, struct t4_cqe *cqe,
> static int __c4iw_poll_cq_one(struct c4iw_cq *chp, struct c4iw_qp *qhp,
> struct ib_wc *wc)
> {
> - struct t4_cqe cqe;
> + struct t4_cqe uninitialized_var(cqe);
> struct t4_wq *wq = qhp ? &qhp->wq : NULL;
> u32 credit = 0;
> u8 cqe_flushed;
Reviewed-by: Bart Van Assche <bart.vanassche@xxxxxxx>
��.n��������+%������w��{.n�����{���fk��ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
