| Message ID | 20210514222402.295157-2-kuba@kernel.org |
|---|---|
| State | New |
| Headers | show |
| Series | [net-next,1/2] net: add a napi variant for RT-well-behaved drivers | expand |
On Fri, May 14, 2021 at 03:24:02PM -0700, Jakub Kicinski wrote: > NFP uses MSI-X and has the most trivial IRQ handler possible. > Perfect candidate for napi_schedule_irq(). > > Signed-off-by: Jakub Kicinski <kuba@kernel.org> > --- > drivers/net/ethernet/netronome/nfp/nfp_net_common.c | 6 +++--- > 1 file changed, 3 insertions(+), 3 deletions(-) I understand there is some discussion around the merit of PATCH 1/2. But if that approach is accepted then this patch looks good to me. Reviewed-by: Simon Horman <simon.horman@netronome.com>
diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c index eeb30680b4dc..12222a6bb08a 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c @@ -474,7 +474,7 @@ static irqreturn_t nfp_net_irq_rxtx(int irq, void *data) { struct nfp_net_r_vector *r_vec = data; - napi_schedule_irqoff(&r_vec->napi); + napi_schedule_irq(&r_vec->napi); /* The FW auto-masks any interrupt, either via the MASK bit in * the MSI-X table or via the per entry ICR field. So there @@ -2631,8 +2631,8 @@ nfp_net_prepare_vector(struct nfp_net *nn, struct nfp_net_r_vector *r_vec, snprintf(r_vec->name, sizeof(r_vec->name), "%s-rxtx-%d", nfp_net_name(nn), idx); - err = request_irq(r_vec->irq_vector, r_vec->handler, 0, r_vec->name, - r_vec); + err = request_irq(r_vec->irq_vector, r_vec->handler, IRQF_NO_THREAD, + r_vec->name, r_vec); if (err) { if (nn->dp.netdev) netif_napi_del(&r_vec->napi);
NFP uses MSI-X and has the most trivial IRQ handler possible. Perfect candidate for napi_schedule_irq(). Signed-off-by: Jakub Kicinski <kuba@kernel.org> --- drivers/net/ethernet/netronome/nfp/nfp_net_common.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)