Re: PR/54297 CVS commit: src/sys/dev/usb

[sc.dying%gmail.com@localhost]

On 2019/06/16 14:50, Christos Zoulas wrote:
> The following reply was made to PR kern/54297; it has been noted by GNATS.
> 
> From: "Christos Zoulas" <christos%netbsd.org@localhost>
> To: gnats-bugs%gnats.NetBSD.org@localhost
> Cc: 
> Subject: PR/54297 CVS commit: src/sys/dev/usb
> Date: Sun, 16 Jun 2019 10:47:50 -0400
> 
>  Module Name:	src
>  Committed By:	christos
>  Date:		Sun Jun 16 14:47:49 UTC 2019
>  
>  Modified Files:
>  	src/sys/dev/usb: if_ure.c
>  
>  Log Message:
>  PR/54297: sc dying: opacket of ure(4) is always 0

--- src/sys/dev/usb/if_ure.c	9 Jun 2019 13:35:47 -0000	1.8
+++ src/sys/dev/usb/if_ure.c	16 Jun 2019 14:47:49 -0000	1.9
@@ -1518,8 +1518,10 @@ ure_txeof(struct usbd_xfer *xfer, void *
 
 	ifp->if_flags &= ~IFF_OACTIVE;
 
-	if (IFQ_IS_EMPTY(&ifp->if_snd) == 0)
+	if (!IFQ_IS_EMPTY(&ifp->if_snd)) {
+		ifp->if_opackets++;
 		ure_start(ifp);
+	}
 
 	splx(s);
 }

Doesn't opacket increament if the kernel sends one packet?
The packet that is being sent must be IFQ_DEQUEUE'd in
ure_start _before_ ure_txeof is called back.