netdev - [PATCH v2 net-next 04/12] sfc: add ability for extra channels to receive raw RX buffers

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: <15d47ffdd878dedb726abbc27a0932b6080191e9.1668430870.git.ecree.xilinx@gmail.com>
Date:   Mon, 14 Nov 2022 13:15:53 +0000
From:   <edward.cree@....com>
To:     <netdev@...r.kernel.org>, <linux-net-drivers@....com>
CC:     <davem@...emloft.net>, <kuba@...nel.org>, <pabeni@...hat.com>,
        <edumazet@...gle.com>, <habetsm.xilinx@...il.com>,
        Edward Cree <ecree.xilinx@...il.com>
Subject: [PATCH v2 net-next 04/12] sfc: add ability for extra channels to receive raw RX buffers

From: Edward Cree <ecree.xilinx@...il.com>

The TC extra channel will need its own special RX handling, which must
 operate before any code that expects the RX buffer to contain a network
 packet; buffers on this RX queue contain MAE counter packets in a
 special format that does not resemble an Ethernet frame, and many fields
 of the RX packet prefix are not populated.
The USER_MARK field, however, is populated with the generation count from
 the counter subsystem, which needs to be passed on to the RX handler.

Signed-off-by: Edward Cree <ecree.xilinx@...il.com>
---
 drivers/net/ethernet/sfc/ef100_rx.c   | 7 +++++++
 drivers/net/ethernet/sfc/net_driver.h | 2 ++
 2 files changed, 9 insertions(+)

diff --git a/drivers/net/ethernet/sfc/ef100_rx.c b/drivers/net/ethernet/sfc/ef100_rx.c
index 735f50385919..83d9db71d7d7 100644
--- a/drivers/net/ethernet/sfc/ef100_rx.c
+++ b/drivers/net/ethernet/sfc/ef100_rx.c
@@ -67,6 +67,13 @@ void __ef100_rx_packet(struct efx_channel *channel)
 
 	prefix = (u32 *)(eh - ESE_GZ_RX_PKT_PREFIX_LEN);
 
+	if (channel->type->receive_raw) {
+		u32 mark = PREFIX_FIELD(prefix, USER_MARK);
+
+		if (channel->type->receive_raw(rx_queue, mark))
+			return; /* packet was consumed */
+	}
+
 	if (ef100_has_fcs_error(channel, prefix) &&
 	    unlikely(!(efx->net_dev->features & NETIF_F_RXALL)))
 		goto out;
diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h
index b3d413896230..1e42f3447b24 100644
--- a/drivers/net/ethernet/sfc/net_driver.h
+++ b/drivers/net/ethernet/sfc/net_driver.h
@@ -593,6 +593,7 @@ struct efx_msi_context {
  * @copy: Copy the channel state prior to reallocation.  May be %NULL if
  *	reallocation is not supported.
  * @receive_skb: Handle an skb ready to be passed to netif_receive_skb()
+ * @receive_raw: Handle an RX buffer ready to be passed to __efx_rx_packet()
  * @want_txqs: Determine whether this channel should have TX queues
  *	created.  If %NULL, TX queues are not created.
  * @keep_eventq: Flag for whether event queue should be kept initialised
@@ -609,6 +610,7 @@ struct efx_channel_type {
 	void (*get_name)(struct efx_channel *, char *buf, size_t len);
 	struct efx_channel *(*copy)(const struct efx_channel *);
 	bool (*receive_skb)(struct efx_channel *, struct sk_buff *);
+	bool (*receive_raw)(struct efx_rx_queue *, u32);
 	bool (*want_txqs)(struct efx_channel *);
 	bool keep_eventq;
 	bool want_pio;