<div dir="ltr"><br><div class="gmail_extra"><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
where the tx and rx rings are cleaned up in the same thread and there<br>
is only one interrupt line for both.<br>
<br>
51: 18 59244 253350 314273 PCI-MSI<br>
1572865-edge enp3s0-TxRx-0<br>
52: 5 484274 141746 197260 PCI-MSI<br>
1572866-edge enp3s0-TxRx-1<br>
53: 9 152225 29943 436749 PCI-MSI<br>
1572867-edge enp3s0-TxRx-2<br>
54: 22 54327 299670 360356 PCI-MSI<br>
1572868-edge enp3s0-TxRx-3<br>
56: 525343 513165 2355680 525593 PCI-MSI<br>
2097152-edge ath10k_pci<br>
<br>
and the ath10k only uses one interrupt. Maybe I'm wrong on my<br>
assumptions, I'd think in today's multi-core environment that<br>
processing tx and rx separately might be a win. (?)<br></blockquote><div><br></div><div>The TX interrupt is used to free the SKB after the DMA from memory to the NIC, correct? (hard_start_xmit()?)</div><div><br></div><div>-Aaron</div></div></div></div>