| Message ID | 20210228170823.1488-1-dqfext@gmail.com |
|---|---|
| State | New |
| Headers | show |
| Series | [net] net: dsa: tag_rtl4_a: fix egress tags | expand |
On 2/28/2021 9:08 AM, DENG Qingfang wrote: > Commit 86dd9868b878 has several issues, but was accepted too soon > before anyone could take a look. > > - Double free. dsa_slave_xmit() will free the skb if the xmit function > returns NULL, but the skb is already freed by eth_skb_pad(). Use > __skb_put_padto() to avoid that. > - Unnecessary allocation. It has been done by DSA core since commit > a3b0b6479700. > - A u16 pointer points to skb data. It should be __be16 for network > byte order. > - Typo in comments. "numer" -> "number". > > Fixes: 86dd9868b878 ("net: dsa: tag_rtl4_a: Support also egress tags") > Signed-off-by: DENG Qingfang <dqfext@gmail.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
On Sun, Feb 28, 2021 at 6:08 PM DENG Qingfang <dqfext@gmail.com> wrote: > Commit 86dd9868b878 has several issues, but was accepted too soon > before anyone could take a look. > > - Double free. dsa_slave_xmit() will free the skb if the xmit function > returns NULL, but the skb is already freed by eth_skb_pad(). Use > __skb_put_padto() to avoid that. > - Unnecessary allocation. It has been done by DSA core since commit > a3b0b6479700. > - A u16 pointer points to skb data. It should be __be16 for network > byte order. > - Typo in comments. "numer" -> "number". > > Fixes: 86dd9868b878 ("net: dsa: tag_rtl4_a: Support also egress tags") > Signed-off-by: DENG Qingfang <dqfext@gmail.com> Ooops I send patches before properly going through the mailbox. Oh well things like that happen. David: ignore my patches to the same tagger and apply this instead! Reviewed-by: Linus Walleij <linus.walleij@linaro.org> Yours, Linus Walleij
On Mon, Mar 01, 2021 at 02:58:59PM +0100, Linus Walleij wrote: > On Sun, Feb 28, 2021 at 6:08 PM DENG Qingfang <dqfext@gmail.com> wrote: > > > Commit 86dd9868b878 has several issues, but was accepted too soon > > before anyone could take a look. > > > > - Double free. dsa_slave_xmit() will free the skb if the xmit function > > returns NULL, but the skb is already freed by eth_skb_pad(). Use > > __skb_put_padto() to avoid that. > > - Unnecessary allocation. It has been done by DSA core since commit > > a3b0b6479700. > > - A u16 pointer points to skb data. It should be __be16 for network > > byte order. > > - Typo in comments. "numer" -> "number". > > > > Fixes: 86dd9868b878 ("net: dsa: tag_rtl4_a: Support also egress tags") > > Signed-off-by: DENG Qingfang <dqfext@gmail.com> > > Ooops I send patches before properly going through the mailbox. > Oh well things like that happen. > > David: ignore my patches to the same tagger and apply this instead! > > Reviewed-by: Linus Walleij <linus.walleij@linaro.org> > > Yours, > Linus Walleij Last time I checked, performance/timing sensitive code is impacted by netdev_dbg calls even if dynamic debugging isn't turned on. However, neither your patches nor Qingfang's have removed that netdev_dbg line. Is there any good reason to keep it?
diff --git a/net/dsa/tag_rtl4_a.c b/net/dsa/tag_rtl4_a.c index c17d39b4a1a0..e9176475bac8 100644 --- a/net/dsa/tag_rtl4_a.c +++ b/net/dsa/tag_rtl4_a.c @@ -35,14 +35,12 @@ static struct sk_buff *rtl4a_tag_xmit(struct sk_buff *skb, struct net_device *dev) { struct dsa_port *dp = dsa_slave_to_port(dev); + __be16 *p; u8 *tag; - u16 *p; u16 out; /* Pad out to at least 60 bytes */ - if (unlikely(eth_skb_pad(skb))) - return NULL; - if (skb_cow_head(skb, RTL4_A_HDR_LEN) < 0) + if (unlikely(__skb_put_padto(skb, ETH_ZLEN, false))) return NULL; netdev_dbg(dev, "add realtek tag to package to port %d\n", @@ -53,13 +51,13 @@ static struct sk_buff *rtl4a_tag_xmit(struct sk_buff *skb, tag = skb->data + 2 * ETH_ALEN; /* Set Ethertype */ - p = (u16 *)tag; + p = (__be16 *)tag; *p = htons(RTL4_A_ETHERTYPE); out = (RTL4_A_PROTOCOL_RTL8366RB << 12) | (2 << 8); - /* The lower bits is the port numer */ + /* The lower bits is the port number */ out |= (u8)dp->index; - p = (u16 *)(tag + 2); + p = (__be16 *)(tag + 2); *p = htons(out); return skb;
Commit 86dd9868b878 has several issues, but was accepted too soon before anyone could take a look. - Double free. dsa_slave_xmit() will free the skb if the xmit function returns NULL, but the skb is already freed by eth_skb_pad(). Use __skb_put_padto() to avoid that. - Unnecessary allocation. It has been done by DSA core since commit a3b0b6479700. - A u16 pointer points to skb data. It should be __be16 for network byte order. - Typo in comments. "numer" -> "number". Fixes: 86dd9868b878 ("net: dsa: tag_rtl4_a: Support also egress tags") Signed-off-by: DENG Qingfang <dqfext@gmail.com> --- net/dsa/tag_rtl4_a.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-)