| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20250912073522.26c1b04a@kernel.org> Date: Fri, 12 Sep 2025 07:35:22 -0700 From: Jakub Kicinski <kuba@...nel.org> To: "Yeounsu Moon" <yyyynoom@...il.com> Cc: "Andrew Lunn" <andrew+netdev@...n.ch>, "David S. Miller" <davem@...emloft.net>, "Eric Dumazet" <edumazet@...gle.com>, "Paolo Abeni" <pabeni@...hat.com>, <netdev@...r.kernel.org>, <linux-kernel@...r.kernel.org> Subject: Re: [PATCH net-next] net: dlink: count dropped packets on skb allocation failure On Fri, 12 Sep 2025 19:03:46 +0900 Yeounsu Moon wrote: > > I'm not sure that failing to allocate a buffer results in dropping > > one packet in this driver. The statistics have specific meaning, if > > you're just trying to use dropped to mean "buffer allocation failures" > > that's not allowed. If I'm misreading the code please explain in more > > detail in the commit message and repost. > > I think you understand the code better than I do. > Your insights are always surprising to me. > > I believed that when `netdev_alloc_skb()` fails, it leads to dropping packets. > I also found many cases where `rx_dropped` was incremented when > `netdev_alloc_skb()` failed. > > However, I'm not entirely sure whether such a failure actually results > in a misisng packet. I'll resend the patch after verifying whether the packet > is really dropped. There's a ring of outstanding Rx buffers. If the ring becomes completely empty we'll start dropping. But that's not the same as one allocation failure == one packet drop.
Powered by blists - more mailing lists