netdev - Re: [PATCH net-next] eth: bnxt: add netmem TX support

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAHS8izNXHvavBAWyyvwzwFh6CgaBhCnvQvtMsE4B2CHVm206hg@mail.gmail.com>
Date: Tue, 17 Jun 2025 13:14:24 -0700
From: Mina Almasry <almasrymina@...gle.com>
To: Taehee Yoo <ap420073@...il.com>, Pranjal Shrivastava <praan@...gle.com>, 
	Shivaji Kant <shivajikant@...gle.com>, Stanislav Fomichev <sdf@...ichev.me>, 
	Pavel Begunkov <asml.silence@...il.com>
Cc: davem@...emloft.net, kuba@...nel.org, pabeni@...hat.com, 
	edumazet@...gle.com, andrew+netdev@...n.ch, horms@...nel.org, 
	michael.chan@...adcom.com, pavan.chebbi@...adcom.com, netdev@...r.kernel.org
Subject: Re: [PATCH net-next] eth: bnxt: add netmem TX support

On Tue, Jun 17, 2025 at 2:45 AM Taehee Yoo <ap420073@...il.com> wrote:
>
> Use netmem_dma_*() helpers and declare netmem_tx to support netmem TX.
> By this change, all bnxt devices will support the netmem TX.
>
> bnxt_start_xmit() uses memcpy() if a packet is too small. However,

nit: this is slightly inaccurate. memcpy itself is not a problem (via
skb_copy_from_linear_data) is not an issue because I think that's
copying the linear part of the skb. What is really a is
skb_frag_address_safe(). Unreadable skbs have no valid address.

This made me realize that skb_frag_address_safe() is broken :( it
needs this check, similar to skb_frag_address():

```
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index c05057869e08..da03ff71b05e 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -3681,7 +3681,12 @@ static inline void *skb_frag_address(const
skb_frag_t *frag)
  */
 static inline void *skb_frag_address_safe(const skb_frag_t *frag)
 {
-       void *ptr = page_address(skb_frag_page(frag));
+       void *ptr;
+
+       if (!skb_frag_page(frag))
+               return NULL;
+
+       ptr = page_address(skb_frag_page(frag));
        if (unlikely(!ptr))
                return NULL;
```

I guess I'll send this fix to net.

> netmem packets are unreadable, so memcpy() is not allowed.
> It should check whether an skb is readable, and if an SKB is unreadable,
> it is processed by the normal transmission logic.
>
> netmem TX can be tested with ncdevmem.c
>
> Signed-off-by: Taehee Yoo <ap420073@...il.com>

Seems like a straightforward conversion to using the netmem dma
mapping API. I don't see anything concerning/unusualy.

Acked-by: Mina Almasry <almasrymina@...gle.com>

-- 
Thanks,
Mina