On 07.12.22 11:24, Paolo Abeni wrote:
On Sat, 2022-12-03 at 22:54 +0100, Gerhard Engleder wrote: [...]+/* This function requires __netif_tx_lock is held by the caller. */ +static int tsnep_xdp_xmit_frame_ring(struct xdp_frame *xdpf, + struct tsnep_tx *tx, bool dma_map) +{ + struct skb_shared_info *shinfo = xdp_get_shared_info_from_frame(xdpf); + unsigned long flags; + int count = 1; + struct tsnep_tx_entry *entry; + int length; + int i; + int retval; + + if (unlikely(xdp_frame_has_frags(xdpf))) + count += shinfo->nr_frags; + + spin_lock_irqsave(&tx->lock, flags);Not strictily related to this patch, but why are you using the _irqsafe variant? it looks like all the locak users are either in process or BH context.
You are right. I will check that and would post a patch later. Thank you! Gerhard
