* [PATCH net-next v6 1/3] net: ti: icssm-prueth: Add HSR and PRP HW offload mode support for AM57xx, AM437x and AM335x
2026-11-13 14:28 [PATCH net-next v6 0/3] Introduce HSR/PRP HW offload support for PRU-ICSSM Ethernet driver Sai Varun Mudepally
@ 2026-11-13 14:28 ` Sai Varun Mudepally
2026-11-17 15:05 ` netdev-bot+sashiko
2026-11-13 14:28 ` [PATCH net-next v6 2/3] net: ti: icssm-prueth: Add priority based RX IRQ handlers Sai Varun Mudepally
2026-11-13 14:28 ` [PATCH net-next v6 3/3] net: ti: icssm-prueth: Support duplicate HW offload feature for HSR and PRP Sai Varun Mudepally
2 siblings, 1 reply; 7+ messages in thread
From: Sai Varun Mudepally @ 2026-11-13 14:28 UTC (permalink / raw)
To: danishanwar, parvathi, rogerq, pmohan, vadim.fedorenko,
andrew+netdev, davem, edumazet, kuba, pabeni, afd, basharath,
arnd, haokexin
Cc: linux-arm-kernel, netdev, linux-kernel, pratheesh, vigneshr,
praneeth, srk, j-rameshbabu, m-malladi, rogerq, krishna, mohan,
saivarun
From: Roger Quadros <rogerq@ti.com>
The PRU-ICSS subsystem on AM335x, AM437x and AM57xx SoCs supports dedicated
firmware implementing the IEC 62439-3 redundancy protocols: HSR and PRP.
Extend the ICSSM PRUETH driver to enable these operating modes in addition
to the existing dual-EMAC and RSTP switch configurations.
In both HSR and PRP modes, the two PRU Ethernet ports operate as LRE (Link
Redundancy Entity) slave ports, while the host port acts as the master.
In case of HW offload mode, Frame duplicate detection/discard for both HSR
and PRP and L2 forwarding in case of HSR are handled entirely by firmware
within the PRU cores.
For HSR, frames received on one PRU port are forwarded to the host and to
the peer PRU port (store-and-forward or cut-through), providing a redundant
ring path. For PRP, any one of the PRU port forwards received frames to the
host after firmware discards the duplicates.
The PRU-ICSS subsystem loads the Dual EMAC firmware by default. To enable
HSR or PRP functionality, the firmware must be changed accordingly. The
required reconfiguration steps are detailed below.
To switch from dual-EMAC to HSR (example: eth2 and eth3 as slave raw
ports):
$ ip link set eth2 down && ip link set eth3 down
$ ip link set eth2 address <mac-addr>
$ ip link set eth3 address <mac-addr>
$ ethtool -K eth2 hsr-tag-rm-offload on
$ ethtool -K eth2 hsr-fwd-offload on
$ ethtool -K eth3 hsr-tag-rm-offload on
$ ethtool -K eth3 hsr-fwd-offload on
$ ip link add name hsr0 type hsr slave1 eth2 slave2 eth3 supervision 45
version 1
$ ip link set eth2 up
$ ip link set eth3 up
To switch from dual-EMAC to PRP (example: eth2 and eth3 as slave raw
ports):
$ ip link set eth2 down && ip link set eth3 down
$ ip link set eth2 address <mac-addr>
$ ip link set eth3 address <mac-addr>
$ ethtool -K eth2 hsr-tag-rm-offload on
$ ethtool -K eth3 hsr-tag-rm-offload on
$ ip link add name prp0 type hsr slave1 eth2 slave2 eth3 supervision 45
proto 1
$ ip link set eth2 up
$ ip link set eth3 up
To revert back to dual-EMAC:
$ ip link set eth2 down && ip link set eth3 down
$ ip link delete hsr0 or prp0
$ ethtool -K eth2 hsr-tag-rm-offload off
$ ethtool -K eth2 hsr-fwd-offload off
$ ethtool -K eth3 hsr-tag-rm-offload off
$ ethtool -K eth3 hsr-fwd-offload off
Signed-off-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Parvathi Pudi <parvathi@couthit.com>
Signed-off-by: Sai Varun Mudepally <saivarun@couthit.com>
---
drivers/net/ethernet/ti/Kconfig | 1 +
drivers/net/ethernet/ti/Makefile | 2 +-
.../ethernet/ti/icssm/icssm_lre_firmware.h | 141 ++++++
drivers/net/ethernet/ti/icssm/icssm_prueth.c | 410 +++++++++++++++++-
drivers/net/ethernet/ti/icssm/icssm_prueth.h | 32 +-
.../net/ethernet/ti/icssm/icssm_prueth_lre.c | 216 +++++++++
.../net/ethernet/ti/icssm/icssm_prueth_lre.h | 19 +
7 files changed, 797 insertions(+), 24 deletions(-)
create mode 100644 drivers/net/ethernet/ti/icssm/icssm_lre_firmware.h
create mode 100644 drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c
create mode 100644 drivers/net/ethernet/ti/icssm/icssm_prueth_lre.h
diff --git a/drivers/net/ethernet/ti/Kconfig b/drivers/net/ethernet/ti/Kconfig
index c60b04921c62..afed42572fd1 100644
--- a/drivers/net/ethernet/ti/Kconfig
+++ b/drivers/net/ethernet/ti/Kconfig
@@ -237,6 +237,7 @@ config TI_PRUETH
depends on NET_SWITCHDEV
select TI_ICSS_IEP
depends on PTP_1588_CLOCK_OPTIONAL
+ depends on HSR || !HSR
help
Some TI SoCs has Programmable Realtime Unit (PRU) cores which can
support Single or Dual Ethernet ports with the help of firmware code
diff --git a/drivers/net/ethernet/ti/Makefile b/drivers/net/ethernet/ti/Makefile
index f4276c9a7762..e4a10d60e1a6 100644
--- a/drivers/net/ethernet/ti/Makefile
+++ b/drivers/net/ethernet/ti/Makefile
@@ -4,7 +4,7 @@
#
obj-$(CONFIG_TI_PRUETH) += icssm-prueth.o
-icssm-prueth-y := icssm/icssm_prueth.o icssm/icssm_prueth_switch.o icssm/icssm_switchdev.o
+icssm-prueth-y := icssm/icssm_prueth.o icssm/icssm_prueth_switch.o icssm/icssm_switchdev.o icssm/icssm_prueth_lre.o
ti-cpsw-common-y += cpsw-common.o davinci_cpdma.o
ti-cpsw-priv-y += cpsw_priv.o cpsw_ethtool.o
diff --git a/drivers/net/ethernet/ti/icssm/icssm_lre_firmware.h b/drivers/net/ethernet/ti/icssm/icssm_lre_firmware.h
new file mode 100644
index 000000000000..b5ab0ec87c5f
--- /dev/null
+++ b/drivers/net/ethernet/ti/icssm/icssm_lre_firmware.h
@@ -0,0 +1,141 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (C) 2017-2020 Texas Instruments Incorporated - http://www.ti.com */
+#ifndef __ICSS_LRE_FIRMWARE_H
+#define __ICSS_LRE_FIRMWARE_H
+
+#define ICSS_LRE_HSR_MODE_OFFSET 0x1E76
+#define ICSS_LRE_MODEH 0x01
+
+/* PRU0 DMEM */
+#define ICSS_LRE_DBG_START 0x1E00
+
+#define ICSS_LRE_DUPLICATE_HOST_TABLE 0x0200
+
+/* PRU1 DMEM */
+#define ICSS_LRE_DUPLICATE_PORT_TABLE_PRU0 0x0200
+#define ICSS_LRE_DUPLICATE_PORT_TABLE_PRU1 0x0E00
+
+/* Size and setup (N and M) of duplicate host table */
+#define ICSS_LRE_DUPLICATE_HOST_TABLE_SIZE 0x1C08
+/* Size and setup (N and M) of duplicate port table (HSR Only) */
+#define ICSS_LRE_DUPLICATE_PORT_TABLE_SIZE 0x1C1C
+/* Time after which an entry is removed from the duplicate
+ * table (10 ms resolution)
+ */
+#define ICSS_LRE_DUPLI_FORGET_TIME 0x1C24
+/* Time interval to check the port duplicate table */
+#define ICSS_LRE_DUPLI_PORT_CHECK_RESO 0x1C2C
+/* Time interval to check the host duplicate table */
+#define ICSS_LRE_DUPLI_HOST_CHECK_RESO 0x1C30
+/* NodeTable | Host | Port */
+#define ICSS_LRE_HOST_TIMER_CHECK_FLAGS 0x1C38
+/* Arbitration flag for the host duplicate table */
+#define ICSS_LRE_HOST_DUPLICATE_ARBITRATION 0x1C3C
+/* Supervision address in LRE */
+#define ICSS_LRE_SUP_ADDR 0x1C4C
+#define ICSS_LRE_SUP_ADDR_LOW 0x1C50
+
+/* Time in TimeTicks (1/100s) */
+#define ICSS_LRE_DUPLICATE_FORGET_TIME_400_MS 40
+#define ICSS_LRE_NODE_FORGET_TIME_60000_MS 6000
+#define ICSS_LRE_MAX_FORGET_TIME 0xFFDF
+
+#define ICSS_LRE_DUPLICATE_PORT_TABLE_DMEM_SIZE 0x0C00
+#define ICSS_LRE_DUPLICATE_HOST_TABLE_DMEM_SIZE 0x1800
+#define ICSS_LRE_STATS_DMEM_SIZE 0x0080
+#define ICSS_LRE_DEBUG_COUNTER_DMEM_SIZE 0x0050
+
+#define ICSS_LRE_DUPLICATE_HOST_TABLE_SIZE_INIT 0x800004 /* N = 128, M = 4 */
+#define ICSS_LRE_DUPLICATE_PORT_TABLE_SIZE_INIT 0x400004 /* N = 64, M = 4 */
+#define ICSS_LRE_MASTER_SLAVE_BUSY_BITS_CLEAR 0x0
+#define ICSS_LRE_TABLE_CHECK_RESOLUTION_10_MS 0xA
+#define ICSS_LRE_SUP_ADDRESS_INIT_OCTETS_HIGH 0x4E1501
+#define ICSS_LRE_SUP_ADDRESS_INIT_OCTETS_LOW 0x1
+
+/* SHARED RAM */
+
+/* 8 bytes of VLAN PCP to RX QUEUE MAPPING */
+#define ICSS_LRE_QUEUE_2_PCP_MAP_OFFSET 0x124
+#define ICSS_LRE_START 0x140
+
+/* Count of HSR/PRP tagged frames successfully transmitted on port A/B */
+#define ICSS_LRE_CNT_TX_A (ICSS_LRE_START + 4)
+#define ICSS_LRE_DUPLICATE_DISCARD (ICSS_LRE_START + 104)
+#define ICSS_LRE_TRANSPARENT_RECEPTION (ICSS_LRE_START + 108)
+#define ICSS_LRE_CNT_NODES (ICSS_LRE_START + 52)
+
+/* SRAM */
+#define ICSS_LRE_IEC62439_CONST_DUPLICATE_ACCEPT 0x01
+#define ICSS_LRE_IEC62439_CONST_DUPLICATE_DISCARD 0x02
+#define ICSS_LRE_IEC62439_CONST_TRANSP_RECEPTION_REMOVE_RCT 0x01
+#define ICSS_LRE_IEC62439_CONST_TRANSP_RECEPTION_PASS_RCT 0x02
+
+/* Enable/disable interrupts for high/low priority instead of per port.
+ * 0 = disabled (default), 1 = enabled
+ */
+#define ICSS_LRE_PRIORITY_INTRS_STATUS_OFFSET 0x1FAA
+/* Enable/disable timestamping of packets. 0 = disabled (default) 1 = enabled */
+#define ICSS_LRE_TIMESTAMP_PKTS_STATUS_OFFSET 0x1FAB
+#define ICSS_LRE_TIMESTAMP_ARRAY_OFFSET 0xC200
+
+/* HOST_TIMER_CHECK_FLAGS bits */
+#define ICSS_LRE_HOST_TIMER_NODE_TABLE_CHECK_BIT BIT(0)
+#define ICSS_LRE_HOST_TIMER_NODE_TABLE_CLEAR_BIT BIT(4)
+#define ICSS_LRE_HOST_TIMER_HOST_TABLE_CHECK_BIT BIT(8)
+#define ICSS_LRE_HOST_TIMER_P1_TABLE_CHECK_BIT BIT(16)
+#define ICSS_LRE_HOST_TIMER_P2_TABLE_CHECK_BIT BIT(24)
+#define ICSS_LRE_HOST_TIMER_PORT_TABLE_CHECK_BITS \
+ (ICSS_LRE_HOST_TIMER_P1_TABLE_CHECK_BIT | \
+ ICSS_LRE_HOST_TIMER_P2_TABLE_CHECK_BIT)
+
+/* PRU1 DMEM */
+/* Node table offsets are different for AM3/4 vs AM57/K2G, set by firmware */
+#define ICSS_LRE_V1_0_HASH_MASK 0x3F
+#define ICSS_LRE_V1_0_INDEX_ARRAY_NT 0x60
+#define ICSS_LRE_V1_0_BIN_ARRAY 0x1A00
+#define ICSS_LRE_V1_0_NODE_TABLE_NEW 0x1FC0
+#define ICSS_LRE_V1_0_INDEX_ARRAY_LOC PRUETH_MEM_DRAM0
+#define ICSS_LRE_V1_0_BIN_ARRAY_LOC PRUETH_MEM_DRAM0
+#define ICSS_LRE_V1_0_NODE_TABLE_LOC PRUETH_MEM_SHARED_RAM
+#define ICSS_LRE_V1_0_INDEX_TBL_MAX_ENTRIES 64
+#define ICSS_LRE_V1_0_BIN_TBL_MAX_ENTRIES 128
+#define ICSS_LRE_V1_0_NODE_TBL_MAX_ENTRIES 128
+
+#define ICSS_LRE_V2_1_HASH_MASK 0xFF
+#define ICSS_LRE_V2_1_INDEX_ARRAY_NT 0x3000
+#define ICSS_LRE_V2_1_BIN_ARRAY \
+ (ICSS_LRE_V2_1_INDEX_ARRAY_NT + \
+ (ICSS_LRE_V2_1_INDEX_TBL_MAX_ENTRIES * 6))
+#define ICSS_LRE_V2_1_NODE_TABLE_NEW \
+ (ICSS_LRE_V2_1_BIN_ARRAY + \
+ (ICSS_LRE_V2_1_BIN_TBL_MAX_ENTRIES * 8))
+#define ICSS_LRE_V2_1_INDEX_ARRAY_LOC PRUETH_MEM_SHARED_RAM
+#define ICSS_LRE_V2_1_BIN_ARRAY_LOC PRUETH_MEM_SHARED_RAM
+#define ICSS_LRE_V2_1_NODE_TABLE_LOC PRUETH_MEM_SHARED_RAM
+#define ICSS_LRE_V2_1_INDEX_TBL_MAX_ENTRIES 256
+#define ICSS_LRE_V2_1_BIN_TBL_MAX_ENTRIES 256
+#define ICSS_LRE_V2_1_NODE_TBL_MAX_ENTRIES 256
+
+#define ICSS_LRE_NODE_FREE 0x10
+#define ICSS_LRE_NODE_TAKEN 0x01
+#define ICSS_LRE_NT_REM_NODE_TYPE_MASK 0x1F
+#define ICSS_LRE_NT_REM_NODE_TYPE_SHIFT 0x00
+
+#define ICSS_LRE_NT_REM_NODE_TYPE_SANA 0x01
+#define ICSS_LRE_NT_REM_NODE_TYPE_SANB 0x02
+#define ICSS_LRE_NT_REM_NODE_TYPE_SANAB 0x03
+#define ICSS_LRE_NT_REM_NODE_TYPE_DAN 0x04
+#define ICSS_LRE_NT_REM_NODE_TYPE_REDBOX 0x08
+#define ICSS_LRE_NT_REM_NODE_TYPE_VDAN 0x10
+
+#define ICSS_LRE_NT_REM_NODE_HSR_BIT 0x20 /* if set node is HSR */
+
+#define ICSS_LRE_NT_REM_NODE_DUP_MASK 0xC0
+#define ICSS_LRE_NT_REM_NODE_DUP_SHIFT 0x06
+
+/* Node entry duplicate type: DupAccept */
+#define ICSS_LRE_NT_REM_NODE_DUP_ACCEPT 0x40
+/* Node entry duplicate type: DupDiscard */
+#define ICSS_LRE_NT_REM_NODE_DUP_DISCARD 0x80
+
+#endif /* __ICSS_LRE_FIRMWARE_H */
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.c b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
index 1aadef061e92..4f631dcc4d50 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth.c
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
@@ -30,6 +30,7 @@
#include "icssm_prueth.h"
#include "icssm_prueth_switch.h"
+#include "icssm_prueth_lre.h"
#include "icssm_vlan_mcast_filter_mmap.h"
#include "../icssg/icssg_mii_rt.h"
#include "../icssg/icss_iep.h"
@@ -40,6 +41,29 @@
#define TX_CLK_DELAY_100M 0x6
#define HR_TIMER_TX_DELAY_US 100
+#define NETIF_PRUETH_LRE_OFFLOAD_FEATURES (NETIF_F_HW_HSR_FWD | \
+ NETIF_F_HW_HSR_TAG_RM)
+
+static void icssm_prueth_set_fw_offsets(struct prueth *prueth)
+{
+ /* Set Multicast filter control and table offsets */
+ if (PRUETH_IS_EMAC(prueth) || PRUETH_IS_SWITCH(prueth)) {
+ prueth->fw_offsets.mc_ctrl_offset =
+ ICSS_EMAC_FW_MULTICAST_FILTER_CTRL_OFFSET;
+ prueth->fw_offsets.mc_filter_mask =
+ ICSS_EMAC_FW_MULTICAST_FILTER_MASK_OFFSET;
+ prueth->fw_offsets.mc_filter_tbl =
+ ICSS_EMAC_FW_MULTICAST_FILTER_TABLE;
+ } else {
+ prueth->fw_offsets.mc_ctrl_offset =
+ ICSS_LRE_FW_MULTICAST_TABLE_SEARCH_OP_CONTROL_BIT;
+ prueth->fw_offsets.mc_filter_mask =
+ ICSS_LRE_FW_MULTICAST_FILTER_MASK;
+ prueth->fw_offsets.mc_filter_tbl =
+ ICSS_LRE_FW_MULTICAST_FILTER_TABLE;
+ }
+}
+
static void icssm_prueth_write_reg(struct prueth *prueth,
enum prueth_mem region,
unsigned int reg, u32 val)
@@ -309,12 +333,15 @@ static void icssm_prueth_hostinit(struct prueth *prueth)
icssm_prueth_mii_init(prueth);
}
-/* This function initialize the driver in EMAC mode
+/* Initialize the driver in EMAC, HSR or PRP mode
* based on eth_type
*/
static void icssm_prueth_init_ethernet_mode(struct prueth *prueth)
{
+ icssm_prueth_set_fw_offsets(prueth);
icssm_prueth_hostinit(prueth);
+ if (prueth_is_lre(prueth))
+ icssm_prueth_lre_config(prueth);
}
static void icssm_prueth_port_enable(struct prueth_emac *emac, bool enable)
@@ -609,6 +636,9 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
/* update first buffer descriptor */
wr_buf_desc = (pktlen << PRUETH_BD_LENGTH_SHIFT) &
PRUETH_BD_LENGTH_MASK;
+ if (PRUETH_IS_HSR(prueth))
+ wr_buf_desc |= BIT(PRUETH_BD_HSR_FRAME_SHIFT);
+
sram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
if (!PRUETH_IS_EMAC(prueth))
writel(wr_buf_desc, sram + readw(&queue_desc->wr_ptr));
@@ -627,6 +657,12 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
void icssm_parse_packet_info(struct prueth *prueth, u32 buffer_descriptor,
struct prueth_packet_info *pkt_info)
{
+ if (prueth_is_lre(prueth))
+ pkt_info->start_offset = !!(buffer_descriptor &
+ PRUETH_BD_START_FLAG_MASK);
+ else
+ pkt_info->start_offset = false;
+
pkt_info->port = (buffer_descriptor & PRUETH_BD_PORT_MASK) >>
PRUETH_BD_PORT_SHIFT;
pkt_info->length = (buffer_descriptor & PRUETH_BD_LENGTH_MASK) >>
@@ -661,10 +697,14 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
unsigned int actual_pkt_len;
bool buffer_wrapped = false;
void *src_addr, *dst_addr;
+ u16 start_offset = 0;
struct sk_buff *skb;
int pkt_block_size;
void *ocmc_ram;
+ if (PRUETH_IS_HSR(emac->prueth))
+ start_offset = (pkt_info->start_offset ?
+ ICSSM_LRE_TAG_SIZE : 0);
/* the PRU firmware deals mostly in pointers already
* offset into ram, we would like to deal in indexes
* within the queue we are working with for code
@@ -687,7 +727,8 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
/* calculate new pointer in ram */
*bd_rd_ptr = rxqueue->buffer_desc_offset + (update_block * BD_SIZE);
- actual_pkt_len = pkt_info->length;
+ /* Exclude the HSR tag bytes already stripped by firmware, if any. */
+ actual_pkt_len = pkt_info->length - start_offset;
/* Allocate a socket buffer for this packet */
skb = netdev_alloc_skb_ip_align(ndev, actual_pkt_len);
@@ -707,6 +748,7 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
*/
src_addr = ocmc_ram + rxqueue->buffer_offset +
(read_block * ICSS_BLOCK_SIZE);
+ src_addr += start_offset;
/* Copy the data from PRU buffers(OCMC) to socket buffer(DRAM) */
if (buffer_wrapped) { /* wrapped around buffer */
@@ -720,6 +762,9 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
if (pkt_info->length < bytes)
bytes = pkt_info->length;
+ /* If applicable, account for the HSR tag removed */
+ bytes -= start_offset;
+
/* copy non-wrapped part */
memcpy(dst_addr, src_addr, bytes);
@@ -741,6 +786,12 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
icssm_prueth_sw_learn_fdb(emac, skb->data + ETH_ALEN);
}
+ /* For PRP, the RCT trailer is at the frame tail, exclude it from
+ * the length to avoid passing it up the stack.
+ */
+ if (PRUETH_IS_PRP(emac->prueth) && pkt_info->start_offset)
+ actual_pkt_len -= ICSSM_LRE_TAG_SIZE;
+
skb_put(skb, actual_pkt_len);
/* send packet up the stack */
@@ -804,13 +855,12 @@ static int icssm_emac_rx_packets(struct prueth_emac *emac, int budget)
rd_buf_desc = readl(shared_ram + bd_rd_ptr);
icssm_parse_packet_info(prueth, rd_buf_desc, &pkt_info);
- if (pkt_info.length <= 0) {
- /* a packet length of zero will cause us to
- * never move the read pointer ahead, locking
- * the driver, so we manually have to move it
- * to the write pointer, discarding all
- * remaining packets in this queue. This should
- * never happen.
+ if (pkt_info.length < EMAC_MIN_PKTLEN) {
+ /* if the packet is too small we skip it but we
+ * still need to move the read pointer ahead
+ * and assume something is wrong with the read
+ * pointer as the firmware should be filtering
+ * these packets
*/
update_rd_ptr = bd_wr_ptr;
emac->stats.rx_length_errors++;
@@ -982,14 +1032,20 @@ static int icssm_emac_ndo_open(struct net_device *ndev)
icssm_prueth_init_ethernet_mode(prueth);
/* reset and start PRU firmware */
- if (PRUETH_IS_SWITCH(prueth)) {
+ if (!PRUETH_IS_EMAC(prueth)) {
+ /* Switch, HSR and PRP protocols share same queue structure */
ret = icssm_prueth_sw_emac_config(emac);
if (ret)
- return ret;
+ goto free_hrtimer;
- ret = icssm_prueth_sw_init_fdb_table(prueth);
- if (ret)
- return ret;
+ if (PRUETH_IS_SWITCH(prueth)) {
+ ret = icssm_prueth_sw_init_fdb_table(prueth);
+ if (ret)
+ goto free_hrtimer;
+ } else {
+ /* LRE mode: set up duplicate-table check flags */
+ icssm_prueth_lre_config_check_flags(prueth);
+ }
} else {
icssm_prueth_emac_config(emac);
}
@@ -1046,6 +1102,9 @@ static int icssm_emac_ndo_open(struct net_device *ndev)
icss_iep_exit(prueth->iep);
free_mem:
icssm_prueth_free_memory(emac->prueth);
+free_hrtimer:
+ if (prueth_is_lre(prueth) && !prueth->emac_configured)
+ icssm_prueth_lre_cleanup(prueth);
return ret;
}
@@ -1079,6 +1138,9 @@ static int icssm_emac_ndo_stop(struct net_device *ndev)
else
rproc_shutdown(emac->pru);
+ if (prueth_is_lre(prueth) && !prueth->emac_configured)
+ icssm_prueth_lre_cleanup(prueth);
+
/* free rx interrupts */
free_irq(emac->rx_irq, ndev);
@@ -1122,7 +1184,8 @@ static int icssm_prueth_change_mode(struct prueth *prueth,
}
}
- if (mode == PRUSS_ETHTYPE_EMAC || mode == PRUSS_ETHTYPE_SWITCH) {
+ if (mode == PRUSS_ETHTYPE_EMAC || mode == PRUSS_ETHTYPE_SWITCH ||
+ mode == PRUSS_ETHTYPE_HSR || mode == PRUSS_ETHTYPE_PRP) {
prueth->eth_type = mode;
} else {
dev_err(prueth->dev, "unknown mode\n");
@@ -1266,11 +1329,16 @@ static void icssm_emac_mc_filter_ctrl(struct prueth_emac *emac, bool enable)
{
struct prueth *prueth = emac->prueth;
void __iomem *mc_filter_ctrl;
+ u32 mc_ctrl_offset;
void __iomem *ram;
u32 reg;
ram = prueth->mem[emac->dram].va;
- mc_filter_ctrl = ram + ICSS_EMAC_FW_MULTICAST_FILTER_CTRL_OFFSET;
+ if (prueth_is_lre(prueth))
+ ram = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ mc_ctrl_offset = prueth->fw_offsets.mc_ctrl_offset;
+ mc_filter_ctrl = ram + mc_ctrl_offset;
if (enable)
reg = ICSS_EMAC_FW_MULTICAST_FILTER_CTRL_ENABLED;
@@ -1289,7 +1357,10 @@ static void icssm_emac_mc_filter_reset(struct prueth_emac *emac)
void __iomem *ram;
ram = prueth->mem[emac->dram].va;
- mc_filter_tbl_base = ICSS_EMAC_FW_MULTICAST_FILTER_TABLE;
+ if (prueth_is_lre(prueth))
+ ram = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ mc_filter_tbl_base = prueth->fw_offsets.mc_filter_tbl;
mc_filter_tbl = ram + mc_filter_tbl_base;
memset_io(mc_filter_tbl, 0, ICSS_EMAC_FW_MULTICAST_TABLE_SIZE_BYTES);
@@ -1302,11 +1373,16 @@ static void icssm_emac_mc_filter_hashmask
{
struct prueth *prueth = emac->prueth;
void __iomem *mc_filter_mask;
+ u32 mc_filter_mask_base;
void __iomem *ram;
ram = prueth->mem[emac->dram].va;
+ if (prueth_is_lre(prueth))
+ ram = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ mc_filter_mask_base = prueth->fw_offsets.mc_filter_mask;
- mc_filter_mask = ram + ICSS_EMAC_FW_MULTICAST_FILTER_MASK_OFFSET;
+ mc_filter_mask = ram + mc_filter_mask_base;
memcpy_toio(mc_filter_mask, mask,
ICSS_EMAC_FW_MULTICAST_FILTER_MASK_SIZE_BYTES);
}
@@ -1316,11 +1392,16 @@ static void icssm_emac_mc_filter_bin_update(struct prueth_emac *emac, u8 hash,
{
struct prueth *prueth = emac->prueth;
void __iomem *mc_filter_tbl;
+ u32 mc_filter_tbl_base;
void __iomem *ram;
ram = prueth->mem[emac->dram].va;
+ if (prueth_is_lre(prueth))
+ ram = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ mc_filter_tbl_base = prueth->fw_offsets.mc_filter_tbl;
- mc_filter_tbl = ram + ICSS_EMAC_FW_MULTICAST_FILTER_TABLE;
+ mc_filter_tbl = ram + mc_filter_tbl_base;
writeb(val, mc_filter_tbl + hash);
}
@@ -1360,6 +1441,8 @@ static void icssm_emac_ndo_set_rx_mode(struct net_device *ndev)
{
struct prueth_emac *emac = netdev_priv(ndev);
bool promisc = ndev->flags & IFF_PROMISC;
+ /* Spinlock for multicast filter table */
+ spinlock_t *mc_filter_tbl_lock;
struct netdev_hw_addr *ha;
struct prueth *prueth;
unsigned long flags;
@@ -1371,8 +1454,18 @@ static void icssm_emac_ndo_set_rx_mode(struct net_device *ndev)
sram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
reg = readl(sram + EMAC_PROMISCUOUS_MODE_OFFSET);
+ /* In LRE mode, the HSR core syncs the master's multicast list and
+ * IFF_ALLMULTI to both slaves. Settings applied to a single slave are
+ * not merged into the shared filter, so configure them on the HSR
+ * master instead.
+ */
+ if (prueth_is_lre(prueth))
+ mc_filter_tbl_lock = &prueth->addr_lock;
+ else
+ mc_filter_tbl_lock = &emac->addr_lock;
+
/* It is a shared table. So lock the access */
- spin_lock_irqsave(&emac->addr_lock, flags);
+ spin_lock_irqsave(mc_filter_tbl_lock, flags);
/* Disable and reset multicast filter, allows allmulti */
icssm_emac_mc_filter_ctrl(emac, false);
@@ -1429,15 +1522,126 @@ static void icssm_emac_ndo_set_rx_mode(struct net_device *ndev)
}
unlock:
- spin_unlock_irqrestore(&emac->addr_lock, flags);
+ spin_unlock_irqrestore(mc_filter_tbl_lock, flags);
+}
+
+static netdev_features_t icssm_emac_ndo_fix_features(struct net_device *ndev,
+ netdev_features_t features)
+{
+ /* hsr tag removal offload and hsr fwd offload are tightly coupled in
+ * firmware implementation. Both these features need to be enabled /
+ * disabled together.
+ */
+ if (!(ndev->features & NETIF_PRUETH_LRE_OFFLOAD_FEATURES)) {
+ if ((features & NETIF_F_HW_HSR_FWD) ||
+ (features & NETIF_F_HW_HSR_TAG_RM)) {
+ features |= NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
+ ndev->wanted_features |=
+ NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
+ }
+ }
+
+ if ((ndev->features & NETIF_F_HW_HSR_FWD) ||
+ (ndev->features & NETIF_F_HW_HSR_TAG_RM)) {
+ if (!(features & NETIF_F_HW_HSR_FWD) ||
+ !(features & NETIF_F_HW_HSR_TAG_RM)) {
+ features &= ~NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
+ ndev->wanted_features &=
+ ~NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
+ }
+ }
+
+ return features;
+}
+
+/**
+ * icssm_emac_ndo_set_features - Configure HSR/PRP offload features
+ * @ndev: network device
+ * @features: Requested feature set
+ *
+ * Called by ethtool -K to configure HSR/PRP offload features. The request
+ * is rejected if this interface or its paired interface is running.
+ *
+ * Return: 0 on success, -EINVAL or -EBUSY on error.
+ */
+static int icssm_emac_ndo_set_features(struct net_device *ndev,
+ netdev_features_t features)
+{
+ struct prueth_emac *emac, *other_emac;
+ netdev_features_t have, wanted;
+ struct net_device *upper;
+ struct prueth *prueth;
+ bool change_request;
+ int ret = -EBUSY;
+
+ emac = netdev_priv(ndev);
+ prueth = emac->prueth;
+ /* MAC instance index starts from 0. So index by port_id - 1 */
+ other_emac = emac->prueth->emac[(emac->port_id == PRUETH_PORT_MII0) ?
+ PRUETH_PORT_MII1 - 1 : PRUETH_PORT_MII0 - 1];
+ wanted = features & NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
+ have = ndev->features & NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
+ change_request = ((wanted ^ have) != 0);
+
+ if (!prueth->fw_data->support_lre)
+ return 0;
+
+ if (PRUETH_IS_SWITCH(prueth) && change_request) {
+ /* LRE offload cannot be enabled in switch mode, remove the
+ * bridge first to revert to EMAC mode.
+ */
+ netdev_err(ndev,
+ "Switch to HSR/PRP not allowed\n");
+ ret = -EINVAL;
+ goto err_set_features;
+ }
+
+ upper = netdev_master_upper_dev_get(ndev);
+ if (upper && is_hsr_master(upper) && change_request) {
+ /* Firmware mode is switched to HSR/PRP based on the offload
+ * features set at the time this port joins the hsr device, so
+ * changing the features afterwards would desync ndev->features
+ * from the firmware mode without the hsr core noticing.
+ * Require unlinking from the hsr device first.
+ */
+ netdev_err(ndev,
+ "Can't change offload features while port is an HSR/PRP slave\n");
+ ret = -EBUSY;
+ goto err_set_features;
+ }
+
+ if (netif_running(ndev) && change_request) {
+ netdev_err(ndev,
+ "Can't change feature when device runs\n");
+ goto err_set_features;
+ }
+
+ if (other_emac && netif_running(other_emac->ndev) && change_request) {
+ netdev_err(ndev,
+ "Can't change feature when other device runs\n");
+ goto err_set_features;
+ }
+
+ return 0;
+
+err_set_features:
+ /* Restore the LRE bits in wanted_features to the current state */
+ ndev->wanted_features &= ~NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
+ ndev->wanted_features |= have;
+
+ return ret;
}
static const struct net_device_ops emac_netdev_ops = {
.ndo_open = icssm_emac_ndo_open,
.ndo_stop = icssm_emac_ndo_stop,
.ndo_start_xmit = icssm_emac_ndo_start_xmit,
+ .ndo_set_mac_address = eth_mac_addr,
+ .ndo_validate_addr = eth_validate_addr,
.ndo_get_stats64 = icssm_emac_ndo_get_stats64,
.ndo_set_rx_mode = icssm_emac_ndo_set_rx_mode,
+ .ndo_set_features = icssm_emac_ndo_set_features,
+ .ndo_fix_features = icssm_emac_ndo_fix_features,
};
/* get emac_port corresponding to eth_node name */
@@ -1591,6 +1795,9 @@ static int icssm_prueth_netdev_init(struct prueth *prueth,
ndev->hw_features |= NETIF_F_HW_L2FW_DOFFLOAD;
}
+ if (prueth->support_lre)
+ ndev->hw_features |= NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
+
ndev->dev.of_node = eth_node;
netif_napi_add(ndev, &emac->napi, icssm_emac_napi_poll);
@@ -1742,6 +1949,111 @@ static int icssm_prueth_ndev_port_unlink(struct net_device *ndev)
return ret;
}
+static int icssm_prueth_hsr_port_link(struct net_device *ndev,
+ struct net_device *hsr_ndev)
+{
+ struct prueth_emac *emac = netdev_priv(ndev), *other_emac;
+ struct prueth *prueth = emac->prueth;
+ enum pruss_ethtype mode;
+ enum hsr_version ver;
+ unsigned long flags;
+ u8 all_slaves;
+ int ret = 0;
+
+ if (PRUETH_IS_SWITCH(prueth))
+ return -EOPNOTSUPP;
+
+ hsr_get_version(hsr_ndev, &ver);
+
+ if (ver == HSR_V1)
+ mode = PRUSS_ETHTYPE_HSR;
+ else if (ver == PRP_V1)
+ mode = PRUSS_ETHTYPE_PRP;
+ else
+ return -EOPNOTSUPP;
+
+ all_slaves = BIT(PRUETH_PORT_MII0) | BIT(PRUETH_PORT_MII1);
+ other_emac = emac->prueth->emac[(emac->port_id == PRUETH_PORT_MII0) ?
+ PRUETH_PORT_MII1 - 1 : PRUETH_PORT_MII0 - 1];
+
+ /* Reject HSR/PRP offload if the peer port does not have the offload
+ * features enabled.
+ */
+ if (other_emac &&
+ !(other_emac->ndev->features & NETIF_PRUETH_LRE_OFFLOAD_FEATURES))
+ return -EOPNOTSUPP;
+
+ spin_lock_irqsave(&prueth->addr_lock, flags);
+
+ if (!prueth->hsr_members) {
+ prueth->hsr_dev = hsr_ndev;
+ } else {
+ /* Adding the port to a second master is not supported */
+ if (prueth->hsr_dev != hsr_ndev) {
+ spin_unlock_irqrestore(&prueth->addr_lock, flags);
+ return -EOPNOTSUPP;
+ }
+ }
+
+ prueth->hsr_members |= BIT(emac->port_id);
+
+ spin_unlock_irqrestore(&prueth->addr_lock, flags);
+
+ if (!prueth_is_lre(prueth) && prueth->hsr_members == all_slaves) {
+ ret = icssm_prueth_change_mode(prueth, mode);
+ if (ret < 0) {
+ dev_err(prueth->dev, "Failed to enable %s mode\n",
+ (mode == PRUSS_ETHTYPE_HSR) ?
+ "HSR" : "PRP");
+ goto free_hsr;
+ } else {
+ dev_info(prueth->dev,
+ "TI PRU ethernet now in %s mode\n",
+ (mode == PRUSS_ETHTYPE_HSR) ?
+ "HSR" : "PRP");
+ }
+ }
+
+ return 0;
+
+free_hsr:
+ spin_lock_irqsave(&prueth->addr_lock, flags);
+
+ prueth->hsr_members &= ~BIT(emac->port_id);
+
+ spin_unlock_irqrestore(&prueth->addr_lock, flags);
+ return ret;
+}
+
+static int icssm_prueth_hsr_port_unlink(struct net_device *ndev)
+{
+ struct prueth_emac *emac = netdev_priv(ndev);
+ struct prueth *prueth = emac->prueth;
+ unsigned long flags;
+ int ret = 0;
+
+ spin_lock_irqsave(&prueth->addr_lock, flags);
+
+ prueth->hsr_members &= ~BIT(emac->port_id);
+
+ spin_unlock_irqrestore(&prueth->addr_lock, flags);
+
+ if (prueth_is_lre(prueth) && !prueth->hsr_members) {
+ ret = icssm_prueth_change_mode(prueth, PRUSS_ETHTYPE_EMAC);
+ if (ret < 0)
+ dev_err(prueth->dev, "Failed to enable dual EMAC mode\n");
+ }
+
+ spin_lock_irqsave(&prueth->addr_lock, flags);
+
+ if (!prueth->hsr_members)
+ prueth->hsr_dev = NULL;
+
+ spin_unlock_irqrestore(&prueth->addr_lock, flags);
+
+ return ret;
+}
+
static int icssm_prueth_ndev_event(struct notifier_block *unused,
unsigned long event, void *ptr)
{
@@ -1749,13 +2061,27 @@ static int icssm_prueth_ndev_event(struct notifier_block *unused,
struct netdev_notifier_changeupper_info *info;
int ret = NOTIFY_DONE;
- if (!icssm_prueth_sw_port_dev_check(ndev))
+ if (ndev->netdev_ops != &emac_netdev_ops)
return NOTIFY_DONE;
switch (event) {
case NETDEV_CHANGEUPPER:
info = ptr;
+ if (is_hsr_master(info->upper_dev)) {
+ if (info->linking) {
+ if (ndev->features &
+ NETIF_PRUETH_LRE_OFFLOAD_FEATURES)
+ ret = icssm_prueth_hsr_port_link
+ (ndev, info->upper_dev);
+ } else {
+ ret = icssm_prueth_hsr_port_unlink(ndev);
+ }
+ }
+
if (netif_is_bridge_master(info->upper_dev)) {
+ if (!icssm_prueth_sw_port_dev_check(ndev))
+ return NOTIFY_DONE;
+
if (info->linking)
ret = icssm_prueth_ndev_port_link
(ndev, info->upper_dev);
@@ -1797,6 +2123,7 @@ static int icssm_prueth_probe(struct platform_device *pdev)
struct device *dev = &pdev->dev;
struct device_node *np;
struct prueth *prueth;
+ bool has_lre = false;
struct pruss *pruss;
int i, ret;
@@ -1811,6 +2138,7 @@ static int icssm_prueth_probe(struct platform_device *pdev)
platform_set_drvdata(pdev, prueth);
prueth->dev = dev;
prueth->fw_data = device_get_match_data(dev);
+ icssm_prueth_set_fw_offsets(prueth);
eth_ports_node = of_get_child_by_name(np, "ethernet-ports");
if (!eth_ports_node)
@@ -1956,6 +2284,17 @@ static int icssm_prueth_probe(struct platform_device *pdev)
prueth->mem[PRUETH_MEM_OCMC].va,
prueth->mem[PRUETH_MEM_OCMC].size);
+ if (IS_ENABLED(CONFIG_HSR) && prueth->fw_data->support_lre)
+ has_lre = true;
+
+ /* LRE requires both ethernet nodes to be present in
+ * DT, otherwise clear the support flag
+ */
+ if (has_lre && (!eth0_node || !eth1_node))
+ has_lre = false;
+
+ prueth->support_lre = has_lre;
+ spin_lock_init(&prueth->addr_lock);
/* setup netdev interfaces */
if (eth0_node) {
ret = icssm_prueth_netdev_init(prueth, eth0_node);
@@ -2177,15 +2516,24 @@ static struct prueth_private_data am335x_prueth_pdata = {
.fw_pru[PRUSS_PRU0] = {
.fw_name[PRUSS_ETHTYPE_EMAC] =
"ti-pruss/am335x-pru0-prueth-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_HSR] =
+ "ti-pruss/am335x-pru0-pruhsr-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_PRP] =
+ "ti-pruss/am335x-pru0-pruprp-fw.elf",
.fw_name[PRUSS_ETHTYPE_SWITCH] =
"ti-pruss/am335x-pru0-prusw-fw.elf",
},
.fw_pru[PRUSS_PRU1] = {
.fw_name[PRUSS_ETHTYPE_EMAC] =
"ti-pruss/am335x-pru1-prueth-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_HSR] =
+ "ti-pruss/am335x-pru1-pruhsr-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_PRP] =
+ "ti-pruss/am335x-pru1-pruprp-fw.elf",
.fw_name[PRUSS_ETHTYPE_SWITCH] =
"ti-pruss/am335x-pru1-prusw-fw.elf",
},
+ .support_lre = true,
.support_switch = true,
};
@@ -2195,15 +2543,24 @@ static struct prueth_private_data am437x_prueth_pdata = {
.fw_pru[PRUSS_PRU0] = {
.fw_name[PRUSS_ETHTYPE_EMAC] =
"ti-pruss/am437x-pru0-prueth-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_HSR] =
+ "ti-pruss/am437x-pru0-pruhsr-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_PRP] =
+ "ti-pruss/am437x-pru0-pruprp-fw.elf",
.fw_name[PRUSS_ETHTYPE_SWITCH] =
"ti-pruss/am437x-pru0-prusw-fw.elf",
},
.fw_pru[PRUSS_PRU1] = {
.fw_name[PRUSS_ETHTYPE_EMAC] =
"ti-pruss/am437x-pru1-prueth-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_HSR] =
+ "ti-pruss/am437x-pru1-pruhsr-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_PRP] =
+ "ti-pruss/am437x-pru1-pruprp-fw.elf",
.fw_name[PRUSS_ETHTYPE_SWITCH] =
"ti-pruss/am437x-pru1-prusw-fw.elf",
},
+ .support_lre = true,
.support_switch = true,
};
@@ -2213,16 +2570,25 @@ static struct prueth_private_data am57xx_prueth_pdata = {
.fw_pru[PRUSS_PRU0] = {
.fw_name[PRUSS_ETHTYPE_EMAC] =
"ti-pruss/am57xx-pru0-prueth-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_HSR] =
+ "ti-pruss/am57xx-pru0-pruhsr-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_PRP] =
+ "ti-pruss/am57xx-pru0-pruprp-fw.elf",
.fw_name[PRUSS_ETHTYPE_SWITCH] =
"ti-pruss/am57xx-pru0-prusw-fw.elf",
},
.fw_pru[PRUSS_PRU1] = {
.fw_name[PRUSS_ETHTYPE_EMAC] =
"ti-pruss/am57xx-pru1-prueth-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_HSR] =
+ "ti-pruss/am57xx-pru1-pruhsr-fw.elf",
+ .fw_name[PRUSS_ETHTYPE_PRP] =
+ "ti-pruss/am57xx-pru1-pruprp-fw.elf",
.fw_name[PRUSS_ETHTYPE_SWITCH] =
"ti-pruss/am57xx-pru1-prusw-fw.elf",
},
+ .support_lre = true,
.support_switch = true,
};
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.h b/drivers/net/ethernet/ti/icssm/icssm_prueth.h
index d5b49b462c24..a5d5bcd08bcd 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth.h
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.h
@@ -16,13 +16,17 @@
#include "icssm_switch.h"
#include "icssm_prueth_ptp.h"
#include "icssm_prueth_fdb_tbl.h"
+#include "icssm_lre_firmware.h"
/* ICSSM size of redundancy tag */
#define ICSSM_LRE_TAG_SIZE 6
+#define PRUETH_TIMER_MS (10)
+
/* PRUSS local memory map */
#define ICSS_LOCAL_SHARED_RAM 0x00010000
#define EMAC_MAX_PKTLEN (ETH_HLEN + VLAN_HLEN + ETH_DATA_LEN)
+#define EMAC_MIN_PKTLEN ETH_ZLEN
/* Below macro is for 1528 Byte Frame support, to Allow even with
* Redundancy tag
*/
@@ -42,6 +46,8 @@ enum pruss_ethtype {
#define PRUETH_IS_EMAC(p) ((p)->eth_type == PRUSS_ETHTYPE_EMAC)
#define PRUETH_IS_SWITCH(p) ((p)->eth_type == PRUSS_ETHTYPE_SWITCH)
+#define PRUETH_IS_HSR(p) ((p)->eth_type == PRUSS_ETHTYPE_HSR)
+#define PRUETH_IS_PRP(p) ((p)->eth_type == PRUSS_ETHTYPE_PRP)
/**
* struct prueth_queue_desc - Queue descriptor
@@ -86,6 +92,7 @@ struct prueth_queue_info {
/**
* struct prueth_packet_info - Info about a packet in buffer
+ * @start_offset: true if frame carries an HSR/PRP start offset
* @shadow: this packet is stored in the collision queue
* @port: port packet is on
* @length: length of packet
@@ -96,6 +103,7 @@ struct prueth_queue_info {
* @timestamp: Specifies if timestamp is appended to the packet
*/
struct prueth_packet_info {
+ bool start_offset;
bool shadow;
unsigned int port;
unsigned int length;
@@ -171,6 +179,13 @@ enum prueth_mem {
PRUETH_MEM_MAX,
};
+/* Firmware offsets/size information */
+struct prueth_fw_offsets {
+ u32 mc_ctrl_offset;
+ u32 mc_filter_mask;
+ u32 mc_filter_tbl;
+};
+
enum pruss_device {
PRUSS_AM57XX = 0,
PRUSS_AM43XX,
@@ -183,11 +198,13 @@ enum pruss_device {
* @driver_data: PRU Ethernet device name
* @fw_pru: firmware names to be used for PRUSS ethernet usecases
* @support_switch: boolean to indicate if switch is enabled
+ * @support_lre: boolean to indicate if LRE is enabled
*/
struct prueth_private_data {
enum pruss_device driver_data;
const struct prueth_firmware fw_pru[PRUSS_NUM_PRUS];
bool support_switch;
+ bool support_lre;
};
struct prueth_emac_stats {
@@ -248,13 +265,20 @@ struct prueth {
struct icss_iep *iep;
const struct prueth_private_data *fw_data;
- struct prueth_fw_offsets *fw_offsets;
+ struct prueth_fw_offsets fw_offsets;
struct device_node *eth_node[PRUETH_NUM_MACS];
struct prueth_emac *emac[PRUETH_NUM_MACS];
struct net_device *registered_netdevs[PRUETH_NUM_MACS];
+ bool support_lre;
+ unsigned int tbl_check_mask;
+ struct hrtimer tbl_check_timer;
+ /* serialize access to LRE VLAN/MC filter table */
+ spinlock_t addr_lock;
+
struct net_device *hw_bridge_dev;
+ struct net_device *hsr_dev;
struct fdb_tbl *fdb_tbl;
struct notifier_block prueth_netdevice_nb;
@@ -264,6 +288,7 @@ struct prueth {
unsigned int eth_type;
size_t ocmc_ram_size;
u8 emac_configured;
+ u8 hsr_members;
u8 br_members;
};
@@ -277,4 +302,9 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
void icssm_emac_mc_filter_bin_allow(struct prueth_emac *emac, u8 hash);
void icssm_emac_mc_filter_bin_disallow(struct prueth_emac *emac, u8 hash);
u8 icssm_emac_get_mc_hash(u8 *mac, u8 *mask);
+
+static inline bool prueth_is_lre(struct prueth *prueth)
+{
+ return PRUETH_IS_HSR(prueth) || PRUETH_IS_PRP(prueth);
+}
#endif /* __NET_TI_PRUETH_H */
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c
new file mode 100644
index 000000000000..9e84244e0649
--- /dev/null
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c
@@ -0,0 +1,216 @@
+// SPDX-License-Identifier: GPL-2.0
+
+/* Texas Instruments PRUETH hsr/prp Link Redundancy Entity (LRE) Driver.
+ *
+ * Copyright (C) 2020 Texas Instruments Incorporated - http://www.ti.com
+ */
+
+#include <linux/kernel.h>
+#include <linux/string.h>
+
+#include "icssm_lre_firmware.h"
+#include "icssm_prueth_lre.h"
+#include "icssm_prueth.h"
+#include "icssm_prueth_switch.h"
+
+void icssm_prueth_lre_config_check_flags(struct prueth *prueth)
+{
+ void __iomem *dram1 = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ /* HSR/PRP: initialize check table when first port is up */
+ if (prueth->emac_configured)
+ return;
+
+ prueth->tbl_check_mask = ICSS_LRE_HOST_TIMER_HOST_TABLE_CHECK_BIT;
+ if (PRUETH_IS_HSR(prueth))
+ prueth->tbl_check_mask |=
+ ICSS_LRE_HOST_TIMER_PORT_TABLE_CHECK_BITS;
+ writel(prueth->tbl_check_mask, dram1 + ICSS_LRE_HOST_TIMER_CHECK_FLAGS);
+}
+
+/* A group of PCPs are mapped to a Queue. This is the size of firmware
+ * array in shared memory
+ */
+#define PCP_GROUP_TO_QUEUE_MAP_SIZE 4
+
+/* PRU firmware default PCP to priority Queue map for ingress & egress
+ *
+ * At ingress to Host
+ * ==================
+ * byte 0 => PRU 1, PCP 0-3 => Q3
+ * byte 1 => PRU 1, PCP 4-7 => Q2
+ * byte 2 => PRU 0, PCP 0-3 => Q1
+ * byte 3 => PRU 0, PCP 4-7 => Q0
+ *
+ * At egress to wire/network on PRU-0 and PRU-1
+ * ============================================
+ * byte 0 => Host, PCP 0-3 => Q3
+ * byte 1 => Host, PCP 4-7 => Q2
+ *
+ * PRU-0
+ * -----
+ * byte 2 => PRU-1, PCP 0-3 => Q1
+ * byte 3 => PRU-1, PCP 4-7 => Q0
+ *
+ * PRU-1
+ * -----
+ * byte 2 => PRU-0, PCP 0-3 => Q1
+ * byte 3 => PRU-0, PCP 4-7 => Q0
+ *
+ * queue names below are named 1 based. i.e PRUETH_QUEUE1 is Q0,
+ * PRUETH_QUEUE2 is Q1 and so forth. Firmware convention is that
+ * a lower queue number has higher priority than a higher queue
+ * number.
+ */
+static u8 fw_pcp_default_priority_queue_map[PCP_GROUP_TO_QUEUE_MAP_SIZE] = {
+ /* port 2 or PRU 1 */
+ PRUETH_QUEUE4, PRUETH_QUEUE3,
+ /* port 1 or PRU 0 */
+ PRUETH_QUEUE2, PRUETH_QUEUE1,
+};
+
+static void icssm_prueth_lre_pcp_queue_map_config(struct prueth *prueth)
+{
+ void __iomem *sram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
+
+ memcpy_toio(sram + ICSS_LRE_QUEUE_2_PCP_MAP_OFFSET,
+ &fw_pcp_default_priority_queue_map[0],
+ PCP_GROUP_TO_QUEUE_MAP_SIZE);
+}
+
+static void icssm_prueth_lre_host_table_init(struct prueth *prueth)
+{
+ void __iomem *dram0 = prueth->mem[PRUETH_MEM_DRAM0].va;
+ void __iomem *dram1 = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ memset_io(dram0 + ICSS_LRE_DUPLICATE_HOST_TABLE, 0,
+ ICSS_LRE_DUPLICATE_HOST_TABLE_DMEM_SIZE);
+
+ writel(ICSS_LRE_DUPLICATE_HOST_TABLE_SIZE_INIT,
+ dram1 + ICSS_LRE_DUPLICATE_HOST_TABLE_SIZE);
+
+ writel(ICSS_LRE_TABLE_CHECK_RESOLUTION_10_MS,
+ dram1 + ICSS_LRE_DUPLI_HOST_CHECK_RESO);
+
+ writel(ICSS_LRE_MASTER_SLAVE_BUSY_BITS_CLEAR,
+ dram1 + ICSS_LRE_HOST_DUPLICATE_ARBITRATION);
+}
+
+static void icssm_prueth_lre_port_table_init(struct prueth *prueth)
+{
+ void __iomem *dram1 = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ if (PRUETH_IS_HSR(prueth)) {
+ memset_io(dram1 + ICSS_LRE_DUPLICATE_PORT_TABLE_PRU0, 0,
+ ICSS_LRE_DUPLICATE_PORT_TABLE_DMEM_SIZE);
+ memset_io(dram1 + ICSS_LRE_DUPLICATE_PORT_TABLE_PRU1, 0,
+ ICSS_LRE_DUPLICATE_PORT_TABLE_DMEM_SIZE);
+
+ writel(ICSS_LRE_DUPLICATE_PORT_TABLE_SIZE_INIT,
+ dram1 + ICSS_LRE_DUPLICATE_PORT_TABLE_SIZE);
+ } else {
+ writel(0, dram1 + ICSS_LRE_DUPLICATE_PORT_TABLE_SIZE);
+ }
+
+ writel(ICSS_LRE_TABLE_CHECK_RESOLUTION_10_MS,
+ dram1 + ICSS_LRE_DUPLI_PORT_CHECK_RESO);
+}
+
+static void icssm_prueth_lre_init(struct prueth *prueth)
+{
+ void __iomem *sram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
+
+ memset_io(sram + ICSS_LRE_START, 0, ICSS_LRE_STATS_DMEM_SIZE);
+
+ writel(ICSS_LRE_IEC62439_CONST_DUPLICATE_DISCARD,
+ sram + ICSS_LRE_DUPLICATE_DISCARD);
+ writel(ICSS_LRE_IEC62439_CONST_TRANSP_RECEPTION_REMOVE_RCT,
+ sram + ICSS_LRE_TRANSPARENT_RECEPTION);
+}
+
+static void icssm_prueth_lre_dbg_init(struct prueth *prueth)
+{
+ void __iomem *dram0 = prueth->mem[PRUETH_MEM_DRAM0].va;
+
+ memset_io(dram0 + ICSS_LRE_DBG_START, 0,
+ ICSS_LRE_DEBUG_COUNTER_DMEM_SIZE);
+}
+
+static void icssm_prueth_lre_protocol_init(struct prueth *prueth)
+{
+ void __iomem *dram0 = prueth->mem[PRUETH_MEM_DRAM0].va;
+ void __iomem *dram1 = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ if (PRUETH_IS_HSR(prueth))
+ writew(ICSS_LRE_MODEH, dram0 + ICSS_LRE_HSR_MODE_OFFSET);
+
+ writel(ICSS_LRE_DUPLICATE_FORGET_TIME_400_MS,
+ dram1 + ICSS_LRE_DUPLI_FORGET_TIME);
+
+ /* PRU firmware matches supervision frames on 01:15:4E:00:01 only
+ * and forwards them to the host, bypassing the multicast filter.
+ */
+ writel(ICSS_LRE_SUP_ADDRESS_INIT_OCTETS_HIGH,
+ dram1 + ICSS_LRE_SUP_ADDR);
+ writel(ICSS_LRE_SUP_ADDRESS_INIT_OCTETS_LOW,
+ dram1 + ICSS_LRE_SUP_ADDR_LOW);
+}
+
+static enum hrtimer_restart icssm_prueth_lre_timer(struct hrtimer *timer)
+{
+ struct prueth *prueth;
+ unsigned int timeout;
+ void __iomem *dram;
+
+ prueth = container_of(timer, struct prueth, tbl_check_timer);
+ dram = prueth->mem[PRUETH_MEM_DRAM1].va;
+ timeout = PRUETH_TIMER_MS;
+
+ hrtimer_forward_now(timer, ms_to_ktime(timeout));
+ /* HSR/PRP requires both ports to be up, regardless of link state. */
+ if (prueth->emac_configured !=
+ (BIT(PRUETH_PORT_MII0) | BIT(PRUETH_PORT_MII1)))
+ return HRTIMER_RESTART;
+
+ /* Set the flags for duplicate tables so the firmware checks and
+ * updates them every 10 milliseconds.
+ */
+ writel(prueth->tbl_check_mask, dram + ICSS_LRE_HOST_TIMER_CHECK_FLAGS);
+
+ return HRTIMER_RESTART;
+}
+
+static void icssm_prueth_lre_init_timer(struct prueth *prueth)
+{
+ hrtimer_setup(&prueth->tbl_check_timer, &icssm_prueth_lre_timer,
+ CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+}
+
+static void icssm_prueth_lre_start_timer(struct prueth *prueth)
+{
+ unsigned int timeout = PRUETH_TIMER_MS;
+
+ if (hrtimer_active(&prueth->tbl_check_timer))
+ return;
+
+ hrtimer_start(&prueth->tbl_check_timer, ms_to_ktime(timeout),
+ HRTIMER_MODE_REL);
+}
+
+void icssm_prueth_lre_config(struct prueth *prueth)
+{
+ icssm_prueth_lre_init_timer(prueth);
+ icssm_prueth_lre_start_timer(prueth);
+ icssm_prueth_lre_pcp_queue_map_config(prueth);
+ icssm_prueth_lre_host_table_init(prueth);
+ icssm_prueth_lre_port_table_init(prueth);
+ icssm_prueth_lre_init(prueth);
+ icssm_prueth_lre_dbg_init(prueth);
+ icssm_prueth_lre_protocol_init(prueth);
+}
+
+void icssm_prueth_lre_cleanup(struct prueth *prueth)
+{
+ if (hrtimer_active(&prueth->tbl_check_timer))
+ hrtimer_cancel(&prueth->tbl_check_timer);
+}
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.h b/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.h
new file mode 100644
index 000000000000..0fe4d1ae5823
--- /dev/null
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.h
@@ -0,0 +1,19 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (C) 2020 Texas Instruments Incorporated - http://www.ti.com
+ */
+
+#ifndef __NET_TI_PRUETH_LRE_H
+#define __NET_TI_PRUETH_LRE_H
+
+#include <linux/etherdevice.h>
+#include <linux/interrupt.h>
+#include <linux/if_vlan.h>
+
+#include "icssm_prueth.h"
+#include "icssm_lre_firmware.h"
+
+void icssm_prueth_lre_config(struct prueth *prueth);
+void icssm_prueth_lre_cleanup(struct prueth *prueth);
+void icssm_prueth_lre_config_check_flags(struct prueth *prueth);
+
+#endif /* __NET_TI_PRUETH_LRE_H */
--
2.43.0
^ permalink raw reply related [flat|nested] 7+ messages in thread* Re: [PATCH net-next v6 1/3] net: ti: icssm-prueth: Add HSR and PRP HW offload mode support for AM57xx, AM437x and AM335x
2026-11-13 14:28 ` [PATCH net-next v6 1/3] net: ti: icssm-prueth: Add HSR and PRP HW offload mode support for AM57xx, AM437x and AM335x Sai Varun Mudepally
@ 2026-11-17 15:05 ` netdev-bot+sashiko
0 siblings, 0 replies; 7+ messages in thread
From: netdev-bot+sashiko @ 2026-11-17 15:05 UTC (permalink / raw)
To: saivarun
Cc: danishanwar, parvathi, rogerq, pmohan, vadim.fedorenko,
andrew+netdev, davem, edumazet, kuba, pabeni, afd, basharath,
arnd, haokexin, linux-arm-kernel, netdev, linux-kernel, pratheesh,
vigneshr, praneeth, srk, j-rameshbabu, m-malladi, rogerq, krishna,
mohan
Thank you for your contribution! Sashiko AI review found 8 potential
issue(s) to consider.
Critical: 0 · High: 0 (1) · Medium: 5 · Low: 3
If you believe none of the issues are legitimate, please reply to each
issue of Medium severity or higher and explain why the AI is wrong.
If any of them are legitimate and you plan to respin, please let patchwork
know by including "pw-bot: cr" as a separate line at the end of your reply
(one such reply per series is enough).
- [Medium] In PRP mode with ports running, the firmware MAC address for
slave B gets out of sync with ndev->dev_addr, and…
- [Medium] icssm_prueth_hsr_port_link() counts every prueth port linked
under an HSR master into hsr_members, whatever its HSR port type.
- [Medium] icssm_emac_ndo_set_rx_mode() picks the multicast-table lock
from prueth->eth_type without synchronization: prueth->addr_lock if…
- [Medium] The PRU firmware supervision address is hard-coded to
01:15:4E:00:01:00.
- [Medium] net: ti: icssm-prueth: broken rollback in
icssm_prueth_change_mode()
- [Low] icssm_emac_ndo_fix_features() writes ndev->wanted_features
(`ndev->wanted_features |= NETIF_PRUETH_LRE_OFFLOAD_FEATURES` and `&=…
- [Low] The patch adds `.ndo_set_mac_address = eth_mac_addr` and
`.ndo_validate_addr = eth_validate_addr` to emac_netdev_ops, which…
- [Low] The commit message does not match the feature semantics in a few
places.
Pre-existing issues:
- [High] icssm_prueth_change_mode() calls ndo_stop() and then ndo_open()
directly on ports that are netif_running().
--
net: ti: icssm-prueth: Add HSR and PRP HW offload mode support for AM57xx, AM437x and AM335x
This commit adds HSR and PRP firmware offload modes to the ICSSM PRUETH
driver on AM335x, AM437x and AM57xx. When both PRU ports join an hsr
master with the LRE offload features enabled, the driver switches the PRU
firmware to HSR or PRP mode. It also adds LRE table and timer setup, RX
tag handling and feature negotiation callbacks.
> To switch from dual-EMAC to PRP (example: eth2 and eth3 as slave raw
> ports):
>
> $ ip link set eth2 down && ip link set eth3 down
> $ ip link set eth2 address <mac-addr>
> $ ip link set eth3 address <mac-addr>
> $ ethtool -K eth2 hsr-tag-rm-offload on
> $ ethtool -K eth3 hsr-tag-rm-offload on
[Severity: Low]
This isn't a bug, but does the PRP recipe match what the driver does?
icssm_emac_ndo_fix_features() couples the LRE bits. Requesting only
hsr-tag-rm-offload also forces NETIF_F_HW_HSR_FWD on and writes it into
wanted_features. Later in the series NETIF_F_HW_HSR_DUP is forced on too.
So PRP ports will always report hsr-fwd-offload as enabled.
The changelog also doesn't mention two cases where, with offload enabled,
icssm_prueth_hsr_port_link() returns -EOPNOTSUPP from NETDEV_CHANGEUPPER:
an hsr version other than HSR_V1/PRP_V1, or a peer port without the
features. In both cases creating the hsr device fails outright instead of
falling back to software HSR.
Could the commit message describe this behaviour?
> diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.c b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
> index 1aadef061e923..4f631dcc4d509 100644
> --- a/drivers/net/ethernet/ti/icssm/icssm_prueth.c
> +++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
[ ... ]
> @@ -1371,8 +1454,18 @@ static void icssm_emac_ndo_set_rx_mode(struct net_device *ndev)
> sram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
> reg = readl(sram + EMAC_PROMISCUOUS_MODE_OFFSET);
>
> + /* In LRE mode, the HSR core syncs the master's multicast list and
> + * IFF_ALLMULTI to both slaves. Settings applied to a single slave are
> + * not merged into the shared filter, so configure them on the HSR
> + * master instead.
> + */
> + if (prueth_is_lre(prueth))
> + mc_filter_tbl_lock = &prueth->addr_lock;
> + else
> + mc_filter_tbl_lock = &emac->addr_lock;
> +
> /* It is a shared table. So lock the access */
> - spin_lock_irqsave(&emac->addr_lock, flags);
> + spin_lock_irqsave(mc_filter_tbl_lock, flags);
[Severity: Medium]
Is prueth->eth_type stable at this point? icssm_prueth_change_mode()
writes eth_type, and ndo_open rewrites prueth->fw_offsets through
icssm_prueth_set_fw_offsets(). Both run under RTNL only and take neither
addr_lock.
This driver uses the synchronous __dev_set_rx_mode() path. That means
ndo_set_rx_mode can run from dev_mc_add() callers that don't hold RTNL,
for example igmp6_group_added() reached through IPV6_ADD_MEMBERSHIP on a
slave.
During an EMAC to HSR/PRP transition, could one slave hold
emac->addr_lock while it writes the shared DRAM1 table, while the other
slave holds prueth->addr_lock?
The helpers icssm_emac_mc_filter_ctrl(), icssm_emac_mc_filter_reset(),
icssm_emac_mc_filter_hashmask() and icssm_emac_mc_filter_bin_update()
also each re-read prueth_is_lre() and prueth->fw_offsets. Consider the
window between the eth_type update in change_mode and
icssm_prueth_set_fw_offsets() in ndo_open. Could a call in that window
write the EMAC filter layout into DRAM1 while the LRE firmware, whose
filter is at different offsets, is loaded?
[ ... ]
> @@ -1429,15 +1522,126 @@ static void icssm_emac_ndo_set_rx_mode(struct net_device *ndev)
> }
>
> unlock:
> - spin_unlock_irqrestore(&emac->addr_lock, flags);
> + spin_unlock_irqrestore(mc_filter_tbl_lock, flags);
> +}
> +
> +static netdev_features_t icssm_emac_ndo_fix_features(struct net_device *ndev,
> + netdev_features_t features)
> +{
> + /* hsr tag removal offload and hsr fwd offload are tightly coupled in
> + * firmware implementation. Both these features need to be enabled /
> + * disabled together.
> + */
> + if (!(ndev->features & NETIF_PRUETH_LRE_OFFLOAD_FEATURES)) {
> + if ((features & NETIF_F_HW_HSR_FWD) ||
> + (features & NETIF_F_HW_HSR_TAG_RM)) {
> + features |= NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
> + ndev->wanted_features |=
> + NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
> + }
> + }
[Severity: Low]
Documentation/networking/netdev-features.rst says ndo_fix_features
should not modify hardware or driver state (it should be stateless). It
also says it is safer to disable a feature whose dependencies are not
met than to force the dependency on.
This callback writes ndev->wanted_features, forces the other LRE bit on,
and its result depends on the current ndev->features.
icssm_emac_ndo_set_features() then has to undo the wanted_features
change on its error path.
Could the coupling be done by clearing the unmet bit, without touching
wanted_features?
[ ... ]
> static const struct net_device_ops emac_netdev_ops = {
> .ndo_open = icssm_emac_ndo_open,
> .ndo_stop = icssm_emac_ndo_stop,
> .ndo_start_xmit = icssm_emac_ndo_start_xmit,
> + .ndo_set_mac_address = eth_mac_addr,
> + .ndo_validate_addr = eth_validate_addr,
[Severity: Low]
This isn't a bug, but every operating mode shares these ops. Changing
the MAC on a stopped port now also succeeds in plain dual-EMAC and
switch modes, where it used to return -EOPNOTSUPP.
Should the commit message mention this new capability? Should it also
say whether the two slaves need the same address for HSR/PRP?
[ ... ]
> @@ -1742,6 +1949,111 @@ static int icssm_prueth_ndev_port_unlink(struct net_device *ndev)
> return ret;
> }
>
> +static int icssm_prueth_hsr_port_link(struct net_device *ndev,
> + struct net_device *hsr_ndev)
> +{
[ ... ]
> + /* Reject HSR/PRP offload if the peer port does not have the offload
> + * features enabled.
> + */
> + if (other_emac &&
> + !(other_emac->ndev->features & NETIF_PRUETH_LRE_OFFLOAD_FEATURES))
> + return -EOPNOTSUPP;
> +
> + spin_lock_irqsave(&prueth->addr_lock, flags);
> +
> + if (!prueth->hsr_members) {
> + prueth->hsr_dev = hsr_ndev;
> + } else {
> + /* Adding the port to a second master is not supported */
> + if (prueth->hsr_dev != hsr_ndev) {
> + spin_unlock_irqrestore(&prueth->addr_lock, flags);
> + return -EOPNOTSUPP;
> + }
> + }
> +
> + prueth->hsr_members |= BIT(emac->port_id);
> +
> + spin_unlock_irqrestore(&prueth->addr_lock, flags);
> +
> + if (!prueth_is_lre(prueth) && prueth->hsr_members == all_slaves) {
> + ret = icssm_prueth_change_mode(prueth, mode);
[Severity: Medium]
What state is prueth left in when icssm_prueth_change_mode() fails here?
By the time change_mode() starts calling ndo_open(), it has already
called ndo_stop() on every running port and set prueth->eth_type to HSR
or PRP. A missing pruhsr or pruprp firmware, for example, makes
icssm_prueth_sw_boot_prus() fail, and change_mode() then returns
straight away. At that point:
- eth_type is not restored.
- A port that was already re-opened keeps running the LRE firmware.
- The failed port has its NAPI disabled and its rx_irq freed, but
netif_running() is still true.
The free_hsr path only clears this port's bit in hsr_members. It does
not touch eth_type or hsr_dev, so prueth_is_lre() stays true.
The hsr core then unwinds hsr_dev_finalize() by unlinking the first
slave. icssm_prueth_hsr_port_unlink() sees prueth_is_lre() with
hsr_members == 0 and calls icssm_prueth_change_mode(EMAC). That runs
icssm_emac_ndo_stop() a second time on the port that was never
re-opened.
Wouldn't that block forever in napi_disable(&emac->napi) with RTNL
held? It would also call free_irq(emac->rx_irq, ndev) on an IRQ that
was already freed. Even without the unwind, a later dev_close() of
that port would hit the same double stop.
Should change_mode() restore the previous eth_type and re-open the
ports in the old mode when ndo_open() fails? Or should the free_hsr
path here revert to dual-EMAC itself, so that hsr_port_unlink() does
not attempt a second mode change on half-opened ports?
[Severity: Medium]
Does this also count a port that is linked as the RedBox interlink?
hsr_portdev_setup() links interlink ports with
netdev_master_upper_dev_link() too. So this notifier sees
is_hsr_master() for HSR_PT_INTERLINK ports as well.
Take a configuration like this one, where eth2 and eth3 are both prueth
ports with offload enabled:
ip link add name hsr0 type hsr slave1 eth2 slave2 Y interlink eth3
hsr_members would reach all_slaves when hsr_dev_finalize() adds the
interlink, and change_mode(HSR) would run. The firmware would then run
an HSR ring between eth2 and the interlink port. It would forward and
duplicate HSR-tagged frames onto the SAN side. At the same time the hsr
core keeps doing software forwarding, because fwd_offloaded is false
for Y.
The peer-feature check above pushes users to enable offload on both
ports. Should the hsr port type be checked before updating hsr_members?
[ ... ]
> +static int icssm_prueth_hsr_port_unlink(struct net_device *ndev)
> +{
> + struct prueth_emac *emac = netdev_priv(ndev);
> + struct prueth *prueth = emac->prueth;
> + unsigned long flags;
> + int ret = 0;
> +
> + spin_lock_irqsave(&prueth->addr_lock, flags);
> +
> + prueth->hsr_members &= ~BIT(emac->port_id);
> +
> + spin_unlock_irqrestore(&prueth->addr_lock, flags);
> +
> + if (prueth_is_lre(prueth) && !prueth->hsr_members) {
> + ret = icssm_prueth_change_mode(prueth, PRUSS_ETHTYPE_EMAC);
> + if (ret < 0)
> + dev_err(prueth->dev, "Failed to enable dual EMAC mode\n");
> + }
[Severity: High]
This is a pre-existing issue in icssm_prueth_change_mode(), but this
patch adds a new way to reach it, inside the hsr device creation
rollback.
change_mode() calls ndo_stop() and ndo_open() directly on running ports
and returns on the first ndo_open() failure. The failed port, and any
port not yet re-opened, keeps IFF_UP but has no resources: no IRQ, NAPI
disabled, PHY stopped and PRUs not booted.
Consider eth2 and eth3 up with LRE offload enabled and the pruhsr
firmware missing:
hsr_dev_finalize()
hsr_add_port(slave[1])
icssm_prueth_hsr_port_link()
icssm_prueth_change_mode(HSR)
ndo_stop(eth2), ndo_stop(eth3)
eth_type = HSR
ndo_open(eth2) fails in icssm_prueth_sw_boot_prus()
err_unregister:
hsr_del_ports()
hsr_del_port(slave A)
netdev_upper_dev_unlink()
icssm_prueth_hsr_port_unlink()
icssm_prueth_change_mode(EMAC)
ndo_stop(eth2) <-- nothing left to tear down
Can this second icssm_emac_ndo_stop() spin forever with RTNL held? Its
napi_disable(&emac->napi) runs on a NAPI that is already disabled.
phy_stop() would also warn on this path.
Could change_mode() track which ports it actually re-opened, or could
ndo_stop() cope with a port whose open failed?
[ ... ]
> @@ -1749,13 +2061,27 @@ static int icssm_prueth_ndev_event(struct notifier_block *unused,
> struct netdev_notifier_changeupper_info *info;
> int ret = NOTIFY_DONE;
>
> - if (!icssm_prueth_sw_port_dev_check(ndev))
> + if (ndev->netdev_ops != &emac_netdev_ops)
> return NOTIFY_DONE;
>
> switch (event) {
> case NETDEV_CHANGEUPPER:
> info = ptr;
> + if (is_hsr_master(info->upper_dev)) {
> + if (info->linking) {
> + if (ndev->features &
> + NETIF_PRUETH_LRE_OFFLOAD_FEATURES)
> + ret = icssm_prueth_hsr_port_link
> + (ndev, info->upper_dev);
> + } else {
> + ret = icssm_prueth_hsr_port_unlink(ndev);
> + }
> + }
[Severity: Medium]
In PRP mode with the ports running, can the firmware MAC for slave B
drift from ndev->dev_addr? The default case of this switch ignores
NETDEV_CHANGEADDR.
At creation, hsr_dev_finalize() adds slave B first. That runs
icssm_prueth_hsr_port_link()->icssm_prueth_change_mode(PRP), and
ndo_open copies the old dev_addr into emac->mac_addr and PORT_MAC_ADDR.
Only after that does hsr_dev_finalize() run:
eth_hw_addr_set(slave[1], slave[0]->dev_addr);
call_netdevice_notifiers(NETDEV_CHANGEADDR, slave[1]);
At deletion, hsr_del_port() unlinks slave B first. So
icssm_prueth_hsr_port_unlink()->icssm_prueth_change_mode(EMAC) re-opens
the ports while slave B still has slave A's MAC. Only then does it
restore port->original_macaddress and send NETDEV_CHANGEADDR.
After a PRP device is deleted, wouldn't the port B dual-EMAC firmware
keep filtering unicast on slave A's MAC? Frames to eth3's real address
would then be dropped until the interface is bounced. The recipe in the
commit message brings the ports down first, but nothing enforces that.
[ ... ]
> diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c
> new file mode 100644
> index 0000000000000..9e84244e0649d
> --- /dev/null
> +++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c
[ ... ]
> +static void icssm_prueth_lre_protocol_init(struct prueth *prueth)
> +{
[ ... ]
> + /* PRU firmware matches supervision frames on 01:15:4E:00:01 only
> + * and forwards them to the host, bypassing the multicast filter.
> + */
> + writel(ICSS_LRE_SUP_ADDRESS_INIT_OCTETS_HIGH,
> + dram1 + ICSS_LRE_SUP_ADDR);
> + writel(ICSS_LRE_SUP_ADDRESS_INIT_OCTETS_LOW,
> + dram1 + ICSS_LRE_SUP_ADDR_LOW);
[Severity: Medium]
This writes a fixed 01:15:4E:00:01:00 whatever the hsr device's
supervision address is. The last byte of that address comes from the
supervision option. The commit message example uses supervision 45,
which gives 01:15:4E:00:01:2D.
In offload mode the hsr core doesn't make the slaves promiscuous and
doesn't add the supervision address to the slave mc lists. The driver
also ignores IFF_PROMISC in LRE mode.
The comment says the firmware matches 5 bytes. If it actually compares
the full 6-byte value written here, would supervision frames for any
non-zero supervision value be dropped by the LRE multicast filter?
Should the configured supervision address be programmed here, or could
the comment confirm how many bytes the firmware compares?
--
Sashiko AI review · https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20261113143127.282466-1-saivarun%40couthit.com
^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH net-next v6 2/3] net: ti: icssm-prueth: Add priority based RX IRQ handlers
2026-11-13 14:28 [PATCH net-next v6 0/3] Introduce HSR/PRP HW offload support for PRU-ICSSM Ethernet driver Sai Varun Mudepally
2026-11-13 14:28 ` [PATCH net-next v6 1/3] net: ti: icssm-prueth: Add HSR and PRP HW offload mode support for AM57xx, AM437x and AM335x Sai Varun Mudepally
@ 2026-11-13 14:28 ` Sai Varun Mudepally
2026-11-17 15:05 ` netdev-bot+sashiko
2026-11-13 14:28 ` [PATCH net-next v6 3/3] net: ti: icssm-prueth: Support duplicate HW offload feature for HSR and PRP Sai Varun Mudepally
2 siblings, 1 reply; 7+ messages in thread
From: Sai Varun Mudepally @ 2026-11-13 14:28 UTC (permalink / raw)
To: danishanwar, parvathi, rogerq, pmohan, vadim.fedorenko,
andrew+netdev, davem, edumazet, kuba, pabeni, afd, basharath,
arnd, haokexin
Cc: linux-arm-kernel, netdev, linux-kernel, pratheesh, vigneshr,
praneeth, srk, j-rameshbabu, m-malladi, rogerq, krishna, mohan,
saivarun
From: Roger Quadros <rogerq@ti.com>
This patch adds support for priority based interrupt handling for the STP/
RSTP Switch, HSR and PRP protocols along with extra logic to address first
come first served to avoid port dominance.
In RSTP switch, HSR, and PRP modes the host port can receive frames from
any one of PRU ports. Servicing RX interrupts in arrival order does not
guarantee the frames are delivered to the stack in wire-arrival order due
to port dominance.
In order to achieve that, each PRU records an IEP (Industrial Ethernet
Peripheral) arrival HW timestamp into the receive buffer and pass this
information to driver. The driver will read the RX HW timestamp from frame
and process the frame which has arrived first among the two ports, giving
the stack wire-arrival ordering.
Dual-EMAC mode continues to use per-port interrupts.
Also convert struct prueth_emac_stats counters from u64 to atomic64_t,
since napi_hpq and napi_lpq can now run concurrently on different CPUs
and update the same counters.
The IEP wrap value used in the timestamp comparison differs between
firmware revisions, so this patch adds an fw_rev field to the platform
data and an enum fw_revision to identify it. icssm_prueth_probe() now
fails with -EINVAL if a platform data entry does not set fw_rev.
icssm_emac_ndo_stop() now stops TX with netif_tx_disable() before
cancelling tx_hrtimer, and the timer and link-up handlers only wake the
queue on a started port. icssm_prueth_change_mode() calls ndo_stop()
directly while the device stays running, so TX must not be restarted
while the shared RX/TX resources are reconfigured or torn down.
Signed-off-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Basharath Hussain Khaja <basharath@couthit.com>
Signed-off-by: Parvathi Pudi <parvathi@couthit.com>
Signed-off-by: Sai Varun Mudepally <saivarun@couthit.com>
---
drivers/net/ethernet/ti/Makefile | 2 +-
drivers/net/ethernet/ti/icssm/icssm_prueth.c | 183 ++++++++--
drivers/net/ethernet/ti/icssm/icssm_prueth.h | 44 ++-
.../ethernet/ti/icssm/icssm_prueth_common.c | 337 ++++++++++++++++++
.../net/ethernet/ti/icssm/icssm_prueth_lre.c | 13 +
.../ethernet/ti/icssm/icssm_prueth_switch.c | 8 +
.../ethernet/ti/icssm/icssm_prueth_switch.h | 1 +
drivers/net/ethernet/ti/icssm/icssm_switch.h | 6 +
8 files changed, 554 insertions(+), 40 deletions(-)
create mode 100644 drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
diff --git a/drivers/net/ethernet/ti/Makefile b/drivers/net/ethernet/ti/Makefile
index e4a10d60e1a6..b6651fe73afd 100644
--- a/drivers/net/ethernet/ti/Makefile
+++ b/drivers/net/ethernet/ti/Makefile
@@ -4,7 +4,7 @@
#
obj-$(CONFIG_TI_PRUETH) += icssm-prueth.o
-icssm-prueth-y := icssm/icssm_prueth.o icssm/icssm_prueth_switch.o icssm/icssm_switchdev.o icssm/icssm_prueth_lre.o
+icssm-prueth-y := icssm/icssm_prueth.o icssm/icssm_prueth_switch.o icssm/icssm_switchdev.o icssm/icssm_prueth_lre.o icssm/icssm_prueth_common.o
ti-cpsw-common-y += cpsw-common.o davinci_cpdma.o
ti-cpsw-priv-y += cpsw_priv.o cpsw_ethtool.o
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.c b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
index 4f631dcc4d50..0536b0965c22 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth.c
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
@@ -44,6 +44,21 @@
#define NETIF_PRUETH_LRE_OFFLOAD_FEATURES (NETIF_F_HW_HSR_FWD | \
NETIF_F_HW_HSR_TAG_RM)
+/* ICSSM (v2.1) - supports 64-bit IEP counter.
+ * Firmware stores packet timestamps using lower 32 bits
+ * which wraps at 0xffffffff.
+ */
+static const struct prueth_fw_offsets fw_offsets_v2_1 = {
+ .iep_wrap = 0xffffffff,
+};
+
+/* ICSSM (v1.0) - supports 32-bit IEP counter, which resets the
+ * counter every one second (nanosecond resolution).
+ */
+static const struct prueth_fw_offsets fw_offsets_v1_0 = {
+ .iep_wrap = NSEC_PER_SEC,
+};
+
static void icssm_prueth_set_fw_offsets(struct prueth *prueth)
{
/* Set Multicast filter control and table offsets */
@@ -342,6 +357,8 @@ static void icssm_prueth_init_ethernet_mode(struct prueth *prueth)
icssm_prueth_hostinit(prueth);
if (prueth_is_lre(prueth))
icssm_prueth_lre_config(prueth);
+ else if (PRUETH_IS_SWITCH(prueth))
+ icssm_prueth_sw_config_packet_timestamping(prueth);
}
static void icssm_prueth_port_enable(struct prueth_emac *emac, bool enable)
@@ -464,7 +481,8 @@ static void icssm_emac_adjust_link(struct net_device *ndev)
if (emac->link) {
/* reactivate the transmit queue if it is stopped */
- if (netif_running(ndev) && netif_queue_stopped(ndev))
+ if (netif_running(ndev) && netif_queue_stopped(ndev) &&
+ (prueth->emac_configured & BIT(emac->port_id)))
netif_wake_queue(ndev);
} else {
if (!netif_queue_stopped(ndev))
@@ -801,8 +819,8 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
local_bh_enable();
/* update stats */
- emac->stats.rx_bytes += actual_pkt_len;
- emac->stats.rx_packets++;
+ atomic64_add(actual_pkt_len, &emac->stats.rx_bytes);
+ atomic64_inc(&emac->stats.rx_packets);
return 0;
}
@@ -824,7 +842,6 @@ static int icssm_emac_rx_packets(struct prueth_emac *emac, int budget)
shared_ram = emac->prueth->mem[PRUETH_MEM_SHARED_RAM].va;
- /* Start and end queue is made common for EMAC, RSTP */
start_queue = emac->rx_queue_start;
end_queue = emac->rx_queue_end;
@@ -835,13 +852,10 @@ static int icssm_emac_rx_packets(struct prueth_emac *emac, int budget)
/* search host queues for packets */
for (i = start_queue; i <= end_queue; i++) {
queue_desc = emac->rx_queue_descs + i;
- if (PRUETH_IS_SWITCH(emac->prueth))
- rxqueue = &sw_queue_infos[PRUETH_PORT_HOST][i];
- else
- rxqueue = &queue_infos[PRUETH_PORT_HOST][i];
+ rxqueue = &queue_infos[PRUETH_PORT_HOST][i];
overflow_cnt = readb(&queue_desc->overflow_cnt);
if (overflow_cnt > 0) {
- emac->stats.rx_over_errors += overflow_cnt;
+ atomic64_add(overflow_cnt, &emac->stats.rx_over_errors);
/* reset to zero */
writeb(0, &queue_desc->overflow_cnt);
}
@@ -863,7 +877,7 @@ static int icssm_emac_rx_packets(struct prueth_emac *emac, int budget)
* these packets
*/
update_rd_ptr = bd_wr_ptr;
- emac->stats.rx_length_errors++;
+ atomic64_inc(&emac->stats.rx_length_errors);
} else if (pkt_info.length > EMAC_MAX_FRM_SUPPORT) {
/* if the packet is too large we skip it but we
* still need to move the read pointer ahead
@@ -872,7 +886,7 @@ static int icssm_emac_rx_packets(struct prueth_emac *emac, int budget)
* these packets
*/
update_rd_ptr = bd_wr_ptr;
- emac->stats.rx_length_errors++;
+ atomic64_inc(&emac->stats.rx_length_errors);
} else {
update_rd_ptr = bd_rd_ptr;
ret = icssm_emac_rx_packet(emac, &update_rd_ptr,
@@ -1070,19 +1084,33 @@ static int icssm_emac_ndo_open(struct net_device *ndev)
goto iep_exit;
}
- ret = icssm_emac_request_irqs(emac);
- if (ret)
- goto rproc_shutdown;
+ if (PRUETH_IS_EMAC(prueth)) {
+ napi_enable(&emac->napi);
+ } else {
+ if (!prueth->emac_configured &&
+ (PRUETH_IS_SWITCH(prueth) || prueth_is_lre(prueth))) {
+ napi_enable(&prueth->napi_hpq);
+ napi_enable(&prueth->napi_lpq);
+ }
+ }
- napi_enable(&emac->napi);
+ /* In switch and LRE modes the shared HPQ/LPQ IRQs are used,
+ * register them here and reuse for both modes.
+ */
+ if (PRUETH_IS_EMAC(prueth))
+ ret = icssm_emac_request_irqs(emac);
+ else
+ ret = icssm_prueth_common_request_irqs(emac);
+ if (ret)
+ goto disable_napi;
+ prueth->emac_configured |= BIT(emac->port_id);
/* start PHY */
phy_start(emac->phydev);
/* enable the port and vlan */
icssm_prueth_port_enable(emac, true);
- prueth->emac_configured |= BIT(emac->port_id);
if (PRUETH_IS_SWITCH(prueth))
icssm_prueth_sw_set_stp_state(prueth, emac->port_id,
BR_STATE_LEARNING);
@@ -1091,7 +1119,10 @@ static int icssm_emac_ndo_open(struct net_device *ndev)
return 0;
-rproc_shutdown:
+disable_napi:
+ if (PRUETH_IS_EMAC(prueth))
+ napi_disable(&emac->napi);
+
if (!PRUETH_IS_EMAC(prueth))
icssm_prueth_sw_shutdown_prus(emac, ndev);
else
@@ -1126,11 +1157,28 @@ static int icssm_emac_ndo_stop(struct net_device *ndev)
/* disable the mac port */
icssm_prueth_port_enable(emac, false);
+ /* Stop TX first. netif_tx_disable() also waits for an xmit that is
+ * already running. Then cancel any tx_hrtimer that xmit may have
+ * armed.
+ */
+ netif_tx_disable(ndev);
+ hrtimer_cancel(&emac->tx_hrtimer);
+
/* stop PHY */
phy_stop(emac->phydev);
- napi_disable(&emac->napi);
- hrtimer_cancel(&emac->tx_hrtimer);
+ if (PRUETH_IS_EMAC(prueth)) {
+ napi_disable(&emac->napi);
+ free_irq(emac->rx_irq, ndev);
+ } else {
+ if (!prueth->emac_configured &&
+ (PRUETH_IS_SWITCH(prueth) || prueth_is_lre(prueth))) {
+ napi_disable(&prueth->napi_lpq);
+ napi_disable(&prueth->napi_hpq);
+ }
+ /* Free IRQs on last port before halting PRU */
+ icssm_prueth_common_free_irqs(emac);
+ }
/* stop the PRU */
if (!PRUETH_IS_EMAC(prueth))
@@ -1141,9 +1189,6 @@ static int icssm_emac_ndo_stop(struct net_device *ndev)
if (prueth_is_lre(prueth) && !prueth->emac_configured)
icssm_prueth_lre_cleanup(prueth);
- /* free rx interrupts */
- free_irq(emac->rx_irq, ndev);
-
/* free memory related to sw */
icssm_prueth_free_memory(emac->prueth);
@@ -1280,8 +1325,8 @@ static enum netdev_tx icssm_emac_ndo_start_xmit(struct sk_buff *skb,
goto fail_tx;
}
- emac->stats.tx_packets++;
- emac->stats.tx_bytes += skb->len;
+ atomic64_inc(&emac->stats.tx_packets);
+ atomic64_add(skb->len, &emac->stats.tx_bytes);
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
@@ -1295,7 +1340,7 @@ static enum netdev_tx icssm_emac_ndo_start_xmit(struct sk_buff *skb,
ret = NETDEV_TX_BUSY;
} else {
/* error */
- emac->stats.tx_dropped++;
+ atomic64_inc(&emac->stats.tx_dropped);
ret = NET_XMIT_DROP;
}
@@ -1315,13 +1360,13 @@ static void icssm_emac_ndo_get_stats64(struct net_device *ndev,
{
struct prueth_emac *emac = netdev_priv(ndev);
- stats->rx_packets = emac->stats.rx_packets;
- stats->rx_bytes = emac->stats.rx_bytes;
- stats->tx_packets = emac->stats.tx_packets;
- stats->tx_bytes = emac->stats.tx_bytes;
- stats->tx_dropped = emac->stats.tx_dropped;
- stats->rx_over_errors = emac->stats.rx_over_errors;
- stats->rx_length_errors = emac->stats.rx_length_errors;
+ stats->rx_packets = atomic64_read(&emac->stats.rx_packets);
+ stats->rx_bytes = atomic64_read(&emac->stats.rx_bytes);
+ stats->tx_packets = atomic64_read(&emac->stats.tx_packets);
+ stats->tx_bytes = atomic64_read(&emac->stats.tx_bytes);
+ stats->tx_dropped = atomic64_read(&emac->stats.tx_dropped);
+ stats->rx_over_errors = atomic64_read(&emac->stats.rx_over_errors);
+ stats->rx_length_errors = atomic64_read(&emac->stats.rx_length_errors);
}
/* enable/disable MC filter */
@@ -1684,6 +1729,11 @@ static enum hrtimer_restart icssm_emac_tx_timer_callback(struct hrtimer *timer)
{
struct prueth_emac *emac =
container_of(timer, struct prueth_emac, tx_hrtimer);
+ struct prueth *prueth = emac->prueth;
+
+ /* Don't restart TX on a port that ndo_stop() is tearing down */
+ if (!(READ_ONCE(prueth->emac_configured) & BIT(emac->port_id)))
+ return HRTIMER_NORESTART;
if (netif_queue_stopped(emac->ndev))
netif_wake_queue(emac->ndev);
@@ -1802,9 +1852,25 @@ static int icssm_prueth_netdev_init(struct prueth *prueth,
netif_napi_add(ndev, &emac->napi, icssm_emac_napi_poll);
+ if ((prueth->support_lre || fw_data->support_switch) &&
+ emac->port_id == PRUETH_PORT_MII0) {
+ netif_napi_add(ndev, &prueth->napi_hpq,
+ icssm_prueth_common_napi_poll_hpq);
+ netif_napi_add(ndev, &prueth->napi_lpq,
+ icssm_prueth_common_napi_poll_lpq);
+ }
+
hrtimer_setup(&emac->tx_hrtimer, &icssm_emac_tx_timer_callback,
CLOCK_MONOTONIC, HRTIMER_MODE_REL_PINNED);
+ if ((prueth->support_lre || fw_data->support_switch) &&
+ emac->port_id == PRUETH_PORT_MII0) {
+ prueth->hp->ndev = ndev;
+ prueth->hp->priority = 0;
+ prueth->lp->ndev = ndev;
+ prueth->lp->priority = 1;
+ }
+
return 0;
free:
emac->ndev = NULL;
@@ -1816,6 +1882,7 @@ static int icssm_prueth_netdev_init(struct prueth *prueth,
static void icssm_prueth_netdev_exit(struct prueth *prueth,
struct device_node *eth_node)
{
+ const struct prueth_private_data *fw_data = prueth->fw_data;
struct prueth_emac *emac;
enum prueth_mac mac;
@@ -1830,6 +1897,13 @@ static void icssm_prueth_netdev_exit(struct prueth *prueth,
phy_disconnect(emac->phydev);
netif_napi_del(&emac->napi);
+
+ if ((prueth->support_lre || fw_data->support_switch) &&
+ emac->port_id == PRUETH_PORT_MII0) {
+ netif_napi_del(&prueth->napi_hpq);
+ netif_napi_del(&prueth->napi_lpq);
+ }
+
prueth->emac[mac] = NULL;
}
@@ -2138,6 +2212,13 @@ static int icssm_prueth_probe(struct platform_device *pdev)
platform_set_drvdata(pdev, prueth);
prueth->dev = dev;
prueth->fw_data = device_get_match_data(dev);
+ if (prueth->fw_data->fw_rev == FW_REV_V1_0)
+ prueth->fw_offsets = fw_offsets_v1_0;
+ else if (prueth->fw_data->fw_rev == FW_REV_V2_1)
+ prueth->fw_offsets = fw_offsets_v2_1;
+ else
+ return -EINVAL;
+
icssm_prueth_set_fw_offsets(prueth);
eth_ports_node = of_get_child_by_name(np, "ethernet-ports");
@@ -2293,6 +2374,41 @@ static int icssm_prueth_probe(struct platform_device *pdev)
if (has_lre && (!eth0_node || !eth1_node))
has_lre = false;
+ /* Switch and LRE share HPQ/LPQ IRQs across both ports,
+ * allocate the shared priority structures once here
+ */
+ if (prueth->fw_data->support_switch || has_lre) {
+ prueth->hp = devm_kzalloc(dev,
+ sizeof(struct prueth_ndev_priority),
+ GFP_KERNEL);
+ if (!prueth->hp) {
+ ret = -ENOMEM;
+ goto free_pool;
+ }
+ prueth->lp = devm_kzalloc(dev,
+ sizeof(struct prueth_ndev_priority),
+ GFP_KERNEL);
+ if (!prueth->lp) {
+ ret = -ENOMEM;
+ goto free_pool;
+ }
+
+ prueth->rx_lpq_irq = of_irq_get_byname(np, "rx_lp");
+ if (prueth->rx_lpq_irq < 0) {
+ ret = prueth->rx_lpq_irq;
+ if (ret != -EPROBE_DEFER)
+ dev_err(prueth->dev, "could not get rx_lp irq\n");
+ goto free_pool;
+ }
+ prueth->rx_hpq_irq = of_irq_get_byname(np, "rx_hp");
+ if (prueth->rx_hpq_irq < 0) {
+ ret = prueth->rx_hpq_irq;
+ if (ret != -EPROBE_DEFER)
+ dev_err(prueth->dev, "could not get rx_hp irq\n");
+ goto free_pool;
+ }
+ }
+
prueth->support_lre = has_lre;
spin_lock_init(&prueth->addr_lock);
/* setup netdev interfaces */
@@ -2533,6 +2649,7 @@ static struct prueth_private_data am335x_prueth_pdata = {
.fw_name[PRUSS_ETHTYPE_SWITCH] =
"ti-pruss/am335x-pru1-prusw-fw.elf",
},
+ .fw_rev = FW_REV_V1_0,
.support_lre = true,
.support_switch = true,
};
@@ -2560,6 +2677,7 @@ static struct prueth_private_data am437x_prueth_pdata = {
.fw_name[PRUSS_ETHTYPE_SWITCH] =
"ti-pruss/am437x-pru1-prusw-fw.elf",
},
+ .fw_rev = FW_REV_V1_0,
.support_lre = true,
.support_switch = true,
};
@@ -2588,6 +2706,7 @@ static struct prueth_private_data am57xx_prueth_pdata = {
"ti-pruss/am57xx-pru1-prusw-fw.elf",
},
+ .fw_rev = FW_REV_V2_1,
.support_lre = true,
.support_switch = true,
};
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.h b/drivers/net/ethernet/ti/icssm/icssm_prueth.h
index a5d5bcd08bcd..8cca5b87d4ff 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth.h
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.h
@@ -179,11 +179,22 @@ enum prueth_mem {
PRUETH_MEM_MAX,
};
+/* PRU firmware revision */
+enum fw_revision {
+ FW_REV_INVALID = 0,
+ FW_REV_V1_0,
+ FW_REV_V2_1
+};
+
/* Firmware offsets/size information */
struct prueth_fw_offsets {
u32 mc_ctrl_offset;
u32 mc_filter_mask;
u32 mc_filter_tbl;
+ /* IEP wrap is used in the rx packet ordering logic and
+ * is different for ICSSM v1.0 vs 2.1
+ */
+ u32 iep_wrap;
};
enum pruss_device {
@@ -197,24 +208,26 @@ enum pruss_device {
* struct prueth_private_data - PRU Ethernet private data
* @driver_data: PRU Ethernet device name
* @fw_pru: firmware names to be used for PRUSS ethernet usecases
+ * @fw_rev: Firmware revision identifier
* @support_switch: boolean to indicate if switch is enabled
* @support_lre: boolean to indicate if LRE is enabled
*/
struct prueth_private_data {
enum pruss_device driver_data;
const struct prueth_firmware fw_pru[PRUSS_NUM_PRUS];
+ enum fw_revision fw_rev;
bool support_switch;
bool support_lre;
};
struct prueth_emac_stats {
- u64 tx_packets;
- u64 tx_dropped;
- u64 tx_bytes;
- u64 rx_packets;
- u64 rx_bytes;
- u64 rx_length_errors;
- u64 rx_over_errors;
+ atomic64_t tx_packets;
+ atomic64_t tx_dropped;
+ atomic64_t tx_bytes;
+ atomic64_t rx_packets;
+ atomic64_t rx_bytes;
+ atomic64_t rx_length_errors;
+ atomic64_t rx_over_errors;
};
/* data for each emac port */
@@ -255,6 +268,11 @@ struct prueth_emac {
int offload_fwd_mark;
};
+struct prueth_ndev_priority {
+ struct net_device *ndev;
+ int priority;
+};
+
struct prueth {
struct device *dev;
struct pruss *pruss;
@@ -270,6 +288,12 @@ struct prueth {
struct device_node *eth_node[PRUETH_NUM_MACS];
struct prueth_emac *emac[PRUETH_NUM_MACS];
struct net_device *registered_netdevs[PRUETH_NUM_MACS];
+ struct prueth_ndev_priority *hp, *lp;
+ /* NAPI for lp and hp queue scans */
+ struct napi_struct napi_lpq;
+ struct napi_struct napi_hpq;
+ int rx_lpq_irq;
+ int rx_hpq_irq;
bool support_lre;
unsigned int tbl_check_mask;
@@ -303,6 +327,12 @@ void icssm_emac_mc_filter_bin_allow(struct prueth_emac *emac, u8 hash);
void icssm_emac_mc_filter_bin_disallow(struct prueth_emac *emac, u8 hash);
u8 icssm_emac_get_mc_hash(u8 *mac, u8 *mask);
+int icssm_prueth_common_napi_poll_lpq(struct napi_struct *napi, int budget);
+int icssm_prueth_common_napi_poll_hpq(struct napi_struct *napi, int budget);
+
+int icssm_prueth_common_request_irqs(struct prueth_emac *emac);
+void icssm_prueth_common_free_irqs(struct prueth_emac *emac);
+
static inline bool prueth_is_lre(struct prueth *prueth)
{
return PRUETH_IS_HSR(prueth) || PRUETH_IS_PRP(prueth);
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
new file mode 100644
index 000000000000..992fadc1521f
--- /dev/null
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
@@ -0,0 +1,337 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Texas Instruments ICSSM Ethernet Driver
+ *
+ * Copyright (C) 2018-2022 Texas Instruments Incorporated - https://www.ti.com/
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/if_vlan.h>
+
+#include "icssm_prueth.h"
+#include "icssm_prueth_switch.h"
+
+static int icssm_prueth_common_emac_rx_packets(struct prueth_emac *emac,
+ int quota, u8 qid1, u8 qid2)
+{
+ u16 bd_rd_ptr, bd_wr_ptr, update_rd_ptr, bd_rd_ptr_o, bd_wr_ptr_o;
+ const struct prueth_queue_info *rxqueue, *rxqueue_o, *rxqueue_p;
+ struct prueth_packet_info pkt_info, pkt_info_o, *pkt_info_p;
+ u32 rd_buf_desc, rd_buf_desc_o, pkt_ts, pkt_ts_o, iep_wrap;
+ int ret, used = 0, port, port0_q_empty, port1_q_empty;
+ struct prueth_queue_desc __iomem *queue_desc_o = NULL;
+ struct prueth_queue_desc __iomem *queue_desc = NULL;
+ struct prueth_emac *emac_p, *other_emac;
+ void __iomem *shared_ram, *ocmc_ram;
+ bool port0_configured, port1_configured;
+ u8 overflow_cnt, overflow_cnt_o;
+ u16 *bd_rd_ptr_p, *bd_wr_ptr_p;
+ struct prueth *prueth;
+
+ prueth = emac->prueth;
+ ocmc_ram = prueth->mem[PRUETH_MEM_OCMC].va;
+ shared_ram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
+ other_emac = prueth->emac[(emac->port_id == PRUETH_PORT_MII0) ?
+ PRUETH_PORT_MII1 - 1 : PRUETH_PORT_MII0 - 1];
+
+ iep_wrap = prueth->fw_offsets.iep_wrap;
+ rxqueue = &sw_queue_infos[PRUETH_PORT_HOST][qid1];
+ rxqueue_o = &sw_queue_infos[PRUETH_PORT_HOST][qid2];
+
+ /* skip Rx if budget is 0 */
+ if (!quota)
+ return 0;
+
+ /* A port's rx_queue_descs is only valid once that port has
+ * completed its own ndo_open() for the currently active mode:
+ * before that it is NULL (never opened) or stale (opened in a
+ * different mode), since the shared HP/LP IRQs and PRUs are
+ * started as soon as the first port opens, independent of the
+ * peer port's state. Treat an unconfigured port's queue as
+ * permanently empty instead of dereferencing rx_queue_descs.
+ */
+ port0_configured = READ_ONCE(prueth->emac_configured) &
+ BIT(emac->port_id);
+ port1_configured = other_emac &&
+ (READ_ONCE(prueth->emac_configured) &
+ BIT(other_emac->port_id));
+
+ if (port0_configured) {
+ queue_desc = emac->rx_queue_descs + qid1;
+
+ overflow_cnt = readb(&queue_desc->overflow_cnt);
+ if (overflow_cnt > 0) {
+ atomic64_add(overflow_cnt, &emac->stats.rx_over_errors);
+ writeb(0, &queue_desc->overflow_cnt);
+ }
+
+ bd_rd_ptr = readw(&queue_desc->rd_ptr);
+ bd_wr_ptr = readw(&queue_desc->wr_ptr);
+ port0_q_empty = (bd_rd_ptr == bd_wr_ptr);
+ } else {
+ bd_rd_ptr = 0;
+ bd_wr_ptr = 0;
+ port0_q_empty = 1;
+ }
+
+ if (port1_configured) {
+ queue_desc_o = other_emac->rx_queue_descs + qid2;
+
+ overflow_cnt_o = readb(&queue_desc_o->overflow_cnt);
+ if (overflow_cnt_o > 0) {
+ atomic64_add(overflow_cnt_o,
+ &other_emac->stats.rx_over_errors);
+ writeb(0, &queue_desc_o->overflow_cnt);
+ }
+
+ bd_rd_ptr_o = readw(&queue_desc_o->rd_ptr);
+ bd_wr_ptr_o = readw(&queue_desc_o->wr_ptr);
+ port1_q_empty = (bd_rd_ptr_o == bd_wr_ptr_o);
+ } else {
+ bd_rd_ptr_o = 0;
+ bd_wr_ptr_o = 0;
+ port1_q_empty = 1;
+ }
+
+ while (!port0_q_empty || !port1_q_empty) {
+ /* Before taking a frame from one queue without a timestamp
+ * comparison, read the empty queue's wr_ptr again. This
+ * read is then newer than the peer's, so any frame that
+ * completed before the peer's frames takes part in the
+ * comparison instead of being overtaken.
+ */
+ if (port0_q_empty && port0_configured) {
+ bd_wr_ptr = readw(&queue_desc->wr_ptr);
+ port0_q_empty = (bd_rd_ptr == bd_wr_ptr);
+ } else if (port1_q_empty && port1_configured) {
+ bd_wr_ptr_o = readw(&queue_desc_o->wr_ptr);
+ port1_q_empty = (bd_rd_ptr_o == bd_wr_ptr_o);
+ }
+
+ rd_buf_desc = readl(shared_ram + bd_rd_ptr);
+ rd_buf_desc_o = readl(shared_ram + bd_rd_ptr_o);
+
+ icssm_parse_packet_info(prueth, rd_buf_desc, &pkt_info);
+ icssm_parse_packet_info(prueth, rd_buf_desc_o, &pkt_info_o);
+
+ pkt_ts = readl(ocmc_ram + ICSS_LRE_TIMESTAMP_ARRAY_OFFSET +
+ bd_rd_ptr - SRAM_START_OFFSET);
+ pkt_ts_o = readl(ocmc_ram + ICSS_LRE_TIMESTAMP_ARRAY_OFFSET +
+ bd_rd_ptr_o - SRAM_START_OFFSET);
+
+ if (!port0_q_empty && !port1_q_empty) {
+ /* Both ports have a pending frame, pick the
+ * earlier one by comparing timestamps and
+ * account for wraparound.
+ */
+ if (pkt_ts > pkt_ts_o)
+ port = (pkt_ts - pkt_ts_o) > (iep_wrap / 2) ?
+ 0 : 1;
+ else
+ port = (pkt_ts_o - pkt_ts) > (iep_wrap / 2) ?
+ 1 : 0;
+
+ } else if (!port0_q_empty) {
+ /* Packet(s) in port0 queue only */
+ port = 0;
+ } else {
+ /* Packet(s) in port1 queue only */
+ port = 1;
+ }
+
+ /* Select correct data structures for queue/packet selected */
+ if (port == 0) {
+ pkt_info_p = &pkt_info;
+ bd_wr_ptr_p = &bd_wr_ptr;
+ bd_rd_ptr_p = &bd_rd_ptr;
+ emac_p = emac;
+ rxqueue_p = rxqueue;
+ } else {
+ pkt_info_p = &pkt_info_o;
+ bd_wr_ptr_p = &bd_wr_ptr_o;
+ bd_rd_ptr_p = &bd_rd_ptr_o;
+ emac_p = other_emac;
+ rxqueue_p = rxqueue_o;
+ }
+
+ if ((*pkt_info_p).length < EMAC_MIN_PKTLEN) {
+ /* Undersized frame: firmware should have filtered
+ * it, so the read pointer can't be trusted. Drop
+ * everything pending in this queue by moving the
+ * read pointer to the write pointer.
+ */
+ update_rd_ptr = *bd_wr_ptr_p;
+ atomic64_inc(&emac_p->stats.rx_length_errors);
+ } else if ((*pkt_info_p).length > EMAC_MAX_FRM_SUPPORT) {
+ /* Oversized frame: same handling as above */
+ update_rd_ptr = *bd_wr_ptr_p;
+ atomic64_inc(&emac_p->stats.rx_length_errors);
+ } else {
+ update_rd_ptr = *bd_rd_ptr_p;
+ ret = icssm_emac_rx_packet(emac_p, &update_rd_ptr,
+ pkt_info_p, rxqueue_p);
+ if (ret)
+ return used;
+
+ used++;
+ }
+
+ /* Zero the BD after consuming it, a misaligned rd_ptr
+ * would otherwise mistake stale data for a valid incoming
+ * frame.
+ */
+ if (port == 0) {
+ writel(0, shared_ram + bd_rd_ptr);
+ writew(update_rd_ptr, &queue_desc->rd_ptr);
+ bd_rd_ptr = update_rd_ptr;
+ } else {
+ writel(0, shared_ram + bd_rd_ptr_o);
+ writew(update_rd_ptr, &queue_desc_o->rd_ptr);
+ bd_rd_ptr_o = update_rd_ptr;
+ }
+
+ /* Refresh wr_ptr of any queue that looks empty, so a frame
+ * that arrived during this poll takes part in the timestamp
+ * comparison instead of being overtaken by the peer queue.
+ */
+ port0_q_empty = (bd_rd_ptr == bd_wr_ptr) ? 1 : 0;
+ if (port0_q_empty && port0_configured) {
+ bd_wr_ptr = readw(&queue_desc->wr_ptr);
+ port0_q_empty = (bd_rd_ptr == bd_wr_ptr) ? 1 : 0;
+ }
+ port1_q_empty = (bd_rd_ptr_o == bd_wr_ptr_o) ? 1 : 0;
+ if (port1_q_empty && port1_configured) {
+ bd_wr_ptr_o = readw(&queue_desc_o->wr_ptr);
+ port1_q_empty = (bd_rd_ptr_o == bd_wr_ptr_o) ? 1 : 0;
+ }
+
+ if (used >= quota)
+ return used;
+ }
+
+ return used;
+}
+
+int icssm_prueth_common_napi_poll_lpq(struct napi_struct *napi, int budget)
+{
+ struct prueth_emac *emac;
+ struct net_device *ndev;
+ struct prueth *prueth;
+ int num_rx_packets;
+ u8 qid1, qid2;
+
+ prueth = container_of(napi, struct prueth, napi_lpq);
+ ndev = prueth->lp->ndev;
+ emac = netdev_priv(ndev);
+ qid1 = PRUETH_QUEUE2;
+ qid2 = PRUETH_QUEUE4;
+
+ num_rx_packets = icssm_prueth_common_emac_rx_packets(emac, budget,
+ qid1, qid2);
+ if (num_rx_packets < budget && napi_complete_done(napi, num_rx_packets))
+ enable_irq(prueth->rx_lpq_irq);
+
+ return num_rx_packets;
+}
+
+int icssm_prueth_common_napi_poll_hpq(struct napi_struct *napi, int budget)
+{
+ struct prueth_emac *emac;
+ struct net_device *ndev;
+ struct prueth *prueth;
+ int num_rx_packets;
+ u8 qid1, qid2;
+
+ prueth = container_of(napi, struct prueth, napi_hpq);
+ ndev = prueth->hp->ndev;
+ emac = netdev_priv(ndev);
+ qid1 = PRUETH_QUEUE1;
+ qid2 = PRUETH_QUEUE3;
+
+ num_rx_packets = icssm_prueth_common_emac_rx_packets(emac, budget,
+ qid1, qid2);
+ if (num_rx_packets < budget && napi_complete_done(napi, num_rx_packets))
+ enable_irq(prueth->rx_hpq_irq);
+
+ return num_rx_packets;
+}
+
+static irqreturn_t icssm_prueth_common_emac_rx_hardirq(int irq, void *dev_id)
+{
+ struct prueth_ndev_priority *ndev_prio;
+ struct prueth_emac *emac;
+ struct net_device *ndev;
+ struct prueth *prueth;
+
+ ndev_prio = (struct prueth_ndev_priority *)dev_id;
+ ndev = ndev_prio->ndev;
+ emac = netdev_priv(ndev);
+ prueth = emac->prueth;
+
+ /* disable Rx system event */
+ if (ndev_prio->priority == 1) {
+ disable_irq_nosync(prueth->rx_lpq_irq);
+ napi_schedule(&prueth->napi_lpq);
+ } else {
+ disable_irq_nosync(prueth->rx_hpq_irq);
+ napi_schedule(&prueth->napi_hpq);
+ }
+
+ return IRQ_HANDLED;
+}
+
+int icssm_prueth_common_request_irqs(struct prueth_emac *emac)
+{
+ struct prueth *prueth = emac->prueth;
+ int ret;
+
+ /* Request irq when first port is initialized */
+ if (prueth->emac_configured)
+ return 0;
+
+ ret = request_irq(prueth->rx_hpq_irq,
+ icssm_prueth_common_emac_rx_hardirq,
+ IRQF_TRIGGER_HIGH, "eth_hp_int", prueth->hp);
+ if (ret) {
+ netdev_err(emac->ndev, "unable to request RX HPQ IRQ\n");
+ goto free_napi_hpq_lpq;
+ }
+
+ ret = request_irq(prueth->rx_lpq_irq,
+ icssm_prueth_common_emac_rx_hardirq,
+ IRQF_TRIGGER_HIGH, "eth_lp_int", prueth->lp);
+ if (ret) {
+ netdev_err(emac->ndev, "unable to request RX LPQ IRQ\n");
+ goto free_rx_hpq_irq;
+ }
+
+ return 0;
+
+free_rx_hpq_irq:
+ free_irq(prueth->rx_hpq_irq, prueth->hp);
+
+free_napi_hpq_lpq:
+ napi_disable(&prueth->napi_lpq);
+ napi_disable(&prueth->napi_hpq);
+
+ return ret;
+}
+
+/**
+ * icssm_prueth_common_free_irqs - free irq
+ *
+ * @emac: EMAC data structure
+ *
+ */
+void icssm_prueth_common_free_irqs(struct prueth_emac *emac)
+{
+ struct prueth *prueth = emac->prueth;
+
+ /* HSR/PRP/Switch: free irqs when last port is down */
+ if (prueth->emac_configured)
+ return;
+
+ free_irq(prueth->rx_lpq_irq, prueth->lp);
+ free_irq(prueth->rx_hpq_irq, prueth->hp);
+}
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c
index 9e84244e0649..967207e10799 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_lre.c
@@ -156,6 +156,14 @@ static void icssm_prueth_lre_protocol_init(struct prueth *prueth)
dram1 + ICSS_LRE_SUP_ADDR_LOW);
}
+static void icssm_prueth_lre_config_packet_timestamping(struct prueth *prueth)
+{
+ void __iomem *sram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
+
+ writeb(1, sram + ICSS_LRE_PRIORITY_INTRS_STATUS_OFFSET);
+ writeb(1, sram + ICSS_LRE_TIMESTAMP_PKTS_STATUS_OFFSET);
+}
+
static enum hrtimer_restart icssm_prueth_lre_timer(struct hrtimer *timer)
{
struct prueth *prueth;
@@ -207,6 +215,11 @@ void icssm_prueth_lre_config(struct prueth *prueth)
icssm_prueth_lre_init(prueth);
icssm_prueth_lre_dbg_init(prueth);
icssm_prueth_lre_protocol_init(prueth);
+ /* Enable per-packet timestamping so the driver can order
+ * received frames by arrival time across the two slave ports.
+ */
+ icssm_prueth_lre_config_packet_timestamping(prueth);
+
}
void icssm_prueth_lre_cleanup(struct prueth *prueth)
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
index 66866ea37913..9d67fc7c23ac 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
@@ -886,6 +886,14 @@ void icssm_prueth_sw_hostconfig(struct prueth *prueth)
sizeof(queue_descs[PRUETH_PORT_QUEUE_HOST]));
}
+void icssm_prueth_sw_config_packet_timestamping(struct prueth *prueth)
+{
+ void __iomem *sram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
+
+ writeb(1, sram + SWITCH_PRIORITY_INTRS_STATUS_OFFSET);
+ writeb(1, sram + SWITCH_TIMESTAMP_PKTS_STATUS_OFFSET);
+}
+
static int icssm_prueth_sw_port_config(struct prueth *prueth,
enum prueth_port port_id)
{
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.h b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.h
index e6111bba166e..fe7197f9a6da 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.h
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.h
@@ -29,6 +29,7 @@ void icssm_prueth_sw_fdb_del(struct prueth_emac *emac,
int icssm_prueth_sw_learn_fdb(struct prueth_emac *emac, u8 *src_mac);
int icssm_prueth_sw_purge_fdb(struct prueth_emac *emac);
void icssm_prueth_sw_hostconfig(struct prueth *prueth);
+void icssm_prueth_sw_config_packet_timestamping(struct prueth *prueth);
int icssm_prueth_sw_emac_config(struct prueth_emac *emac);
int icssm_prueth_sw_boot_prus(struct prueth *prueth, struct net_device *ndev);
int icssm_prueth_sw_shutdown_prus(struct prueth_emac *emac,
diff --git a/drivers/net/ethernet/ti/icssm/icssm_switch.h b/drivers/net/ethernet/ti/icssm/icssm_switch.h
index 5ba9ce14da44..b4d6b8660539 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_switch.h
+++ b/drivers/net/ethernet/ti/icssm/icssm_switch.h
@@ -301,6 +301,12 @@
#define P0_Q1_BD_OFFSET P0_BUFFER_DESC_OFFSET
#define P0_BUFFER_DESC_OFFSET SRAM_START_OFFSET
+/* Enable/disable interrupts for high/low priority instead of per port.
+ * 0 = disabled (default), 1 = enabled
+ */
+#define SWITCH_PRIORITY_INTRS_STATUS_OFFSET 0x1FAA
+#define SWITCH_TIMESTAMP_PKTS_STATUS_OFFSET 0x1FAB
+
/* Memory Usage of L3 OCMC RAM */
/* L3 64KB Memory - mainly buffer Pool */
--
2.43.0
^ permalink raw reply related [flat|nested] 7+ messages in thread* Re: [PATCH net-next v6 2/3] net: ti: icssm-prueth: Add priority based RX IRQ handlers
2026-11-13 14:28 ` [PATCH net-next v6 2/3] net: ti: icssm-prueth: Add priority based RX IRQ handlers Sai Varun Mudepally
@ 2026-11-17 15:05 ` netdev-bot+sashiko
0 siblings, 0 replies; 7+ messages in thread
From: netdev-bot+sashiko @ 2026-11-17 15:05 UTC (permalink / raw)
To: saivarun
Cc: danishanwar, parvathi, rogerq, pmohan, vadim.fedorenko,
andrew+netdev, davem, edumazet, kuba, pabeni, afd, basharath,
arnd, haokexin, linux-arm-kernel, netdev, linux-kernel, pratheesh,
vigneshr, praneeth, srk, j-rameshbabu, m-malladi, rogerq, krishna,
mohan
Thank you for your contribution! Sashiko AI review found 8 potential
issue(s) to consider.
Critical: 0 · High: 0 (1) · Medium: 4 · Low: 4
If you believe none of the issues are legitimate, please reply to each
issue of Medium severity or higher and explain why the AI is wrong.
If any of them are legitimate and you plan to respin, please let patchwork
know by including "pw-bot: cr" as a separate line at the end of your reply
(one such reply per series is enough).
- [Medium] The commit message says the timer and link-up handlers 'only
wake the queue on a started port' and that 'TX must not be restarted…
- [Medium] icssm_prueth_common_emac_rx_packets() reads emac_configured
once per poll into port0_configured/port1_configured.
- [Medium] In the new shared HP/LP RX path,
icssm_prueth_common_emac_rx_packets() returns early ('if (ret) return…
- [Medium] net: ti: icssm-prueth: of_irq_get_byname() check misses 0 in
probe
- [Low] The patch fixes a TX-teardown bug that already exists in the
tree, but the fix is buried in a feature commit with no Fixes: tag.
- [Low] fw_offsets_v1_0 sets .iep_wrap = NSEC_PER_SEC, and its comment
says the v1.0 32-bit IEP counter 'resets the counter every one second'.
- [Low] The loop in icssm_prueth_common_emac_rx_packets() runs while
either queue is non-empty, and it re-reads the hardware wr_ptr of any…
- [Low] icssm_emac_adjust_link() now wakes the TX queue only if
'(prueth->emac_configured & BIT(emac->port_id))', and it reads that…
Pre-existing issues:
- [High] This problem wasn't introduced by this patch, but the patch
extends it to the new shared resources.
--
net: ti: icssm-prueth: Add priority based RX IRQ handlers
This commit adds shared high and low priority RX interrupts and NAPI
instances for switch, HSR and PRP modes. Frames from the two PRU ports are
ordered by their IEP arrival timestamps. It also converts the per-port
stats to atomic64_t, adds a firmware revision field that selects the IEP
wrap value, and reorders TX teardown in icssm_emac_ndo_stop().
> diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.c b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
> index 4f631dcc4d509..0536b0965c222 100644
> --- a/drivers/net/ethernet/ti/icssm/icssm_prueth.c
> +++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
> @@ -44,6 +44,21 @@
> #define NETIF_PRUETH_LRE_OFFLOAD_FEATURES (NETIF_F_HW_HSR_FWD | \
> NETIF_F_HW_HSR_TAG_RM)
>
> +/* ICSSM (v2.1) - supports 64-bit IEP counter.
> + * Firmware stores packet timestamps using lower 32 bits
> + * which wraps at 0xffffffff.
> + */
> +static const struct prueth_fw_offsets fw_offsets_v2_1 = {
> + .iep_wrap = 0xffffffff,
> +};
> +
> +/* ICSSM (v1.0) - supports 32-bit IEP counter, which resets the
> + * counter every one second (nanosecond resolution).
> + */
> +static const struct prueth_fw_offsets fw_offsets_v1_0 = {
> + .iep_wrap = NSEC_PER_SEC,
> +};
[Severity: Low]
Where does the one second reset get configured? icssm_emac_ndo_open()
calls icss_iep_init(prueth->iep, NULL, NULL, 0). With cycle_time_ns == 0,
icss_iep_init() takes the plain enable path:
if (cycle_time_ns)
icss_iep_enable_shadow_mode(iep);
else
icss_iep_enable(iep);
icss_iep_settime(iep, ktime_get_real_ns());
Only the shadow mode path sets IEP_CMP_CFG_CMP0_RST_CNT_EN and programs
CMP0. The am335x/am437x IEP plat_data also has .flags = 0. Unless the
v1.0 PRU firmware reprograms CMP0 itself, the 32-bit counter would run
free and wrap at 2^32 ns (about 4.29s).
In that case, when both queues are non-empty,
icssm_prueth_common_emac_rx_packets() would treat any real timestamp
difference between 500ms and 2^31 ns as a wrap. It would then deliver
the later frame first.
Should v1.0 use the same 0xffffffff value as v2.1, or does the v1.0
firmware set up the reset on its own?
[ ... ]
> @@ -464,7 +481,8 @@ static void icssm_emac_adjust_link(struct net_device *ndev)
>
> if (emac->link) {
> /* reactivate the transmit queue if it is stopped */
> - if (netif_running(ndev) && netif_queue_stopped(ndev))
> + if (netif_running(ndev) && netif_queue_stopped(ndev) &&
> + (prueth->emac_configured & BIT(emac->port_id)))
> netif_wake_queue(ndev);
[Severity: Low]
This reads emac_configured under emac->lock. However,
icssm_emac_ndo_stop() clears the bit without taking emac->lock, and it
only calls phy_stop() after netif_tx_disable() and hrtimer_cancel().
Could this interleaving happen?
phylib state machine change_mode()->ndo_stop()
icssm_emac_adjust_link()
sees bit set, queue stopped
emac_configured &= ~BIT(port_id)
netif_tx_disable(ndev)
hrtimer_cancel(&emac->tx_hrtimer)
netif_wake_queue(ndev)
phy_stop(emac->phydev)
If so, the queue stays awake until phy_stop()'s link-down callback stops
it again. An xmit in that window could re-arm tx_hrtimer after
hrtimer_cancel() has already returned.
[ ... ]
> @@ -1126,11 +1157,28 @@ static int icssm_emac_ndo_stop(struct net_device *ndev)
> /* disable the mac port */
> icssm_prueth_port_enable(emac, false);
>
> + /* Stop TX first. netif_tx_disable() also waits for an xmit that is
> + * already running. Then cancel any tx_hrtimer that xmit may have
> + * armed.
> + */
> + netif_tx_disable(ndev);
> + hrtimer_cancel(&emac->tx_hrtimer);
[Severity: Low]
This isn't a bug, but this part of the patch looks like a fix for an
existing problem. It covers three changes:
- netif_tx_disable() before hrtimer_cancel()
- the emac_configured checks in icssm_emac_tx_timer_callback() and
icssm_emac_adjust_link()
- moving the emac_configured assignment in icssm_emac_ndo_open()
Before this patch, icssm_emac_ndo_stop() never stopped the TX queue, and
the timer callback woke the queue unconditionally. This matters because
icssm_prueth_change_mode() calls ndo_stop()/ndo_open() while the netdev
stays running. icssm_prueth_change_mode() was added by commit
eca327ae9aa6.
None of this depends on the priority IRQ rework. Would it make sense to
split it into a separate patch with a Fixes: eca327ae9aa6 tag, so stable
trees can pick it up?
>
> /* stop PHY */
> phy_stop(emac->phydev);
>
> - napi_disable(&emac->napi);
> - hrtimer_cancel(&emac->tx_hrtimer);
> + if (PRUETH_IS_EMAC(prueth)) {
> + napi_disable(&emac->napi);
> + free_irq(emac->rx_irq, ndev);
> + } else {
> + if (!prueth->emac_configured &&
> + (PRUETH_IS_SWITCH(prueth) || prueth_is_lre(prueth))) {
> + napi_disable(&prueth->napi_lpq);
> + napi_disable(&prueth->napi_hpq);
> + }
> + /* Free IRQs on last port before halting PRU */
> + icssm_prueth_common_free_irqs(emac);
> + }
[Severity: High]
This is a pre-existing issue, but this patch extends it to the new shared
NAPIs and IRQs. What happens here if an earlier ndo_open() call from
icssm_prueth_change_mode() or icssm_prueth_resume() failed?
Both callers return the error, and the netdev stays netif_running():
icssm_prueth_change_mode() {
...
ret = ndev->netdev_ops->ndo_open(ndev);
if (ret < 0) {
netdev_err(ndev, "failed to start: %d", ret);
return ret;
}
...
}
ndo_open() can fail in two ways here:
- icssm_prueth_sw_boot_prus() fails, for example because the
switch/HSR/PRP firmware file is missing. napi_hpq and napi_lpq are
then never enabled.
- icssm_prueth_common_request_irqs() fails. Its error path disables
them again.
Either way, emac_configured stays 0 and the rx_hp/rx_lp IRQs are not
held.
Later, dev_close or unregister_netdev() calls icssm_emac_ndo_stop(). It
decides only from !prueth->emac_configured, so it calls napi_disable() on
a NAPI that is already disabled.
Wouldn't napi_disable_locked() then spin forever with RTNL held? A
disabled NAPI keeps NAPIF_STATE_SCHED and NAPIF_STATE_NPSVC set.
If that loop were ever passed, icssm_prueth_common_free_irqs() would free
IRQs that were never requested. The PRU shutdown and icss_iep_exit()
would also run a second time.
Before this patch, the same problem existed for emac->napi and
emac->rx_irq.
[ ... ]
> @@ -1684,6 +1729,11 @@ static enum hrtimer_restart icssm_emac_tx_timer_callback(struct hrtimer *timer)
> {
> struct prueth_emac *emac =
> container_of(timer, struct prueth_emac, tx_hrtimer);
> + struct prueth *prueth = emac->prueth;
> +
> + /* Don't restart TX on a port that ndo_stop() is tearing down */
> + if (!(READ_ONCE(prueth->emac_configured) & BIT(emac->port_id)))
> + return HRTIMER_NORESTART;
>
> if (netif_queue_stopped(emac->ndev))
> netif_wake_queue(emac->ndev);
[Severity: Medium]
The commit message says "the timer and link-up handlers only wake the
queue on a started port". Is this check enough to guarantee that?
The check and the wake are not atomic with respect to
icssm_emac_ndo_stop(). That function clears the bit with a plain
read-modify-write and calls netif_tx_disable() before hrtimer_cancel():
CPU1 (tx_hrtimer callback) CPU2 (change_mode()->ndo_stop())
reads bit as set
emac_configured &= ~BIT(port_id)
netif_tx_disable(ndev)
netif_wake_queue()
hrtimer_cancel() waits, returns
In the change_mode path the qdisc stays active. So
icssm_emac_ndo_start_xmit() could then write into the PRU TX queues at
two bad moments:
- while the PRUs are shut down
- while the following ndo_open() clears shared RAM/OCMC in
icssm_prueth_hostinit()
On -ENOBUFS, xmit could also re-arm tx_hrtimer after hrtimer_cancel() has
returned.
The plain "|=" and "&= ~" writes racing with READ_ONCE() readers in
hardirq context also look like a data race KCSAN would report.
Would any of these close the window?
- clear the bit with WRITE_ONCE()
- call hrtimer_cancel() before netif_tx_disable() (and again after)
- do the check and the wake under a lock that ndo_stop() also takes
[ ... ]
> diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
> new file mode 100644
> index 0000000000000..992fadc1521f9
> --- /dev/null
> +++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
> @@ -0,0 +1,337 @@
[ ... ]
> + port0_configured = READ_ONCE(prueth->emac_configured) &
> + BIT(emac->port_id);
> + port1_configured = other_emac &&
> + (READ_ONCE(prueth->emac_configured) &
> + BIT(other_emac->port_id));
[Severity: Medium]
This snapshot is taken once per poll and then gates the whole loop, up to
the full budget. It covers access to rx_queue_descs, the rd_ptr writes,
BD zeroing, and icssm_emac_rx_packet()->netif_receive_skb() on that
port's netdev.
When one port is stopped while the peer stays up, icssm_emac_ndo_stop()
clears the bit. It does not disable or synchronise napi_hpq/napi_lpq,
because it only does that once emac_configured reaches 0.
Can a poll already running on another CPU keep dequeuing the stopped
port's host queue after ndo_stop() has returned? If so, it would:
- deliver skbs to a netdev that is down
- update that port's stats
- in switch mode, do FDB learning for the stopped port
On the open side, rx_queue_descs is published only by a plain
"emac_configured |= BIT(...)" in icssm_emac_ndo_open(). Here it is read
back through READ_ONCE() plus a control dependency, which does not order
the later load of emac->rx_queue_descs.
Should the single-port stop path wait for an in-flight hpq/lpq poll
rather than relying on the flag?
[ ... ]
> + if ((*pkt_info_p).length < EMAC_MIN_PKTLEN) {
> + /* Undersized frame: firmware should have filtered
> + * it, so the read pointer can't be trusted. Drop
> + * everything pending in this queue by moving the
> + * read pointer to the write pointer.
> + */
> + update_rd_ptr = *bd_wr_ptr_p;
> + atomic64_inc(&emac_p->stats.rx_length_errors);
> + } else if ((*pkt_info_p).length > EMAC_MAX_FRM_SUPPORT) {
> + /* Oversized frame: same handling as above */
> + update_rd_ptr = *bd_wr_ptr_p;
> + atomic64_inc(&emac_p->stats.rx_length_errors);
> + } else {
> + update_rd_ptr = *bd_rd_ptr_p;
> + ret = icssm_emac_rx_packet(emac_p, &update_rd_ptr,
> + pkt_info_p, rxqueue_p);
> + if (ret)
> + return used;
[Severity: Medium]
Can RX on this priority stall here? If icssm_emac_rx_packet() fails, for
example because netdev_alloc_skb_ip_align() returns NULL, frames are
still pending behind rd_ptr.
The poll callbacks then see num_rx_packets < budget, complete NAPI and
unmask the IRQ:
if (num_rx_packets < budget && napi_complete_done(napi, num_rx_packets))
enable_irq(prueth->rx_hpq_irq);
The rx_hp/rx_lp lines are PRUSS INTC system events. pruss_intc_irq_ack()
writes PRU_INTC_SICR when the IRQ is handled, which clears the latched
status. So nothing signals the pending frames again until the firmware
raises a new event for a later frame on the same priority. Until then
the frames stay in OCMC and the queue can overflow.
[ ... ]
> + port0_q_empty = (bd_rd_ptr == bd_wr_ptr) ? 1 : 0;
> + if (port0_q_empty && port0_configured) {
> + bd_wr_ptr = readw(&queue_desc->wr_ptr);
> + port0_q_empty = (bd_rd_ptr == bd_wr_ptr) ? 1 : 0;
> + }
> + port1_q_empty = (bd_rd_ptr_o == bd_wr_ptr_o) ? 1 : 0;
> + if (port1_q_empty && port1_configured) {
> + bd_wr_ptr_o = readw(&queue_desc_o->wr_ptr);
> + port1_q_empty = (bd_rd_ptr_o == bd_wr_ptr_o) ? 1 : 0;
> + }
> +
> + if (used >= quota)
> + return used;
[Severity: Low]
Is this loop bounded by the NAPI budget? Only frames delivered by
icssm_emac_rx_packet() increment used. The undersized and oversized
branches above only move rd_ptr to wr_ptr and bump rx_length_errors.
The hardware wr_ptr is also re-read on every pass. So while firmware
keeps producing BDs with an invalid length, used >= quota is never
reached and the loop keeps running in softirq context.
icssm_emac_rx_packets() in the EMAC path takes one wr_ptr snapshot per
queue, so it is bounded by the queue size. Should dropped frames count
toward quota here as well?
[ ... ]
> @@ -2293,6 +2374,41 @@ static int icssm_prueth_probe(struct platform_device *pdev)
> if (has_lre && (!eth0_node || !eth1_node))
> has_lre = false;
>
> + /* Switch and LRE share HPQ/LPQ IRQs across both ports,
> + * allocate the shared priority structures once here
> + */
> + if (prueth->fw_data->support_switch || has_lre) {
> + prueth->hp = devm_kzalloc(dev,
> + sizeof(struct prueth_ndev_priority),
> + GFP_KERNEL);
> + if (!prueth->hp) {
> + ret = -ENOMEM;
> + goto free_pool;
> + }
> + prueth->lp = devm_kzalloc(dev,
> + sizeof(struct prueth_ndev_priority),
> + GFP_KERNEL);
> + if (!prueth->lp) {
> + ret = -ENOMEM;
> + goto free_pool;
> + }
> +
> + prueth->rx_lpq_irq = of_irq_get_byname(np, "rx_lp");
> + if (prueth->rx_lpq_irq < 0) {
> + ret = prueth->rx_lpq_irq;
> + if (ret != -EPROBE_DEFER)
> + dev_err(prueth->dev, "could not get rx_lp irq\n");
> + goto free_pool;
> + }
> + prueth->rx_hpq_irq = of_irq_get_byname(np, "rx_hp");
> + if (prueth->rx_hpq_irq < 0) {
> + ret = prueth->rx_hpq_irq;
> + if (ret != -EPROBE_DEFER)
> + dev_err(prueth->dev, "could not get rx_hp irq\n");
> + goto free_pool;
> + }
> + }
> +
> prueth->support_lre = has_lre;
> spin_lock_init(&prueth->addr_lock);
> /* setup netdev interfaces */
[Severity: Medium]
of_irq_get_byname() returns 0, not a negative errno, when the
interrupt is described in DT but the mapping in
irq_create_of_mapping() fails. Both checks here only test for < 0.
In that case probe succeeds silently and prueth->rx_lpq_irq or
prueth->rx_hpq_irq is left as 0.
The failure then only shows up much later. When the first port is
opened in switch, HSR or PRP mode, icssm_prueth_common_request_irqs()
calls request_irq() on IRQ 0. On the icssm_prueth_change_mode()
path, that means a bridge or HSR join fails while the ports are
being reopened. The only message is "unable to request RX HPQ IRQ",
which points away from the real cause in probe.
Could these be checked with <= 0, with 0 mapped to an error such as
-ENXIO? The lookup is done on dev->of_node of the platform device,
so platform_get_irq_byname() might be simpler. It never returns 0
and already prints an error for anything other than -EPROBE_DEFER.
The existing per-port emac->rx_irq lookup in
icssm_prueth_netdev_init() has the same < 0 check. It uses the
child eth_node, so it would need the explicit <= 0 check rather than
the platform helper.
--
Sashiko AI review · https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20261113143127.282466-1-saivarun%40couthit.com
^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH net-next v6 3/3] net: ti: icssm-prueth: Support duplicate HW offload feature for HSR and PRP
2026-11-13 14:28 [PATCH net-next v6 0/3] Introduce HSR/PRP HW offload support for PRU-ICSSM Ethernet driver Sai Varun Mudepally
2026-11-13 14:28 ` [PATCH net-next v6 1/3] net: ti: icssm-prueth: Add HSR and PRP HW offload mode support for AM57xx, AM437x and AM335x Sai Varun Mudepally
2026-11-13 14:28 ` [PATCH net-next v6 2/3] net: ti: icssm-prueth: Add priority based RX IRQ handlers Sai Varun Mudepally
@ 2026-11-13 14:28 ` Sai Varun Mudepally
2026-11-17 15:05 ` netdev-bot+sashiko
2 siblings, 1 reply; 7+ messages in thread
From: Sai Varun Mudepally @ 2026-11-13 14:28 UTC (permalink / raw)
To: danishanwar, parvathi, rogerq, pmohan, vadim.fedorenko,
andrew+netdev, davem, edumazet, kuba, pabeni, afd, basharath,
arnd, haokexin
Cc: linux-arm-kernel, netdev, linux-kernel, pratheesh, vigneshr,
praneeth, srk, j-rameshbabu, m-malladi, rogerq, krishna, mohan,
saivarun
From: Roger Quadros <rogerq@ti.com>
In HSR and PRP modes each outgoing frame must be sent on both PRU slave
ports.
Previously the driver was writing the frame into each port's transmit queue
independently after updating the tags resulting in performing two OCMC
buffer copy operations.
Frame duplicate offloading is implemented with a common shared queue
between the two ports. The driver writes the frame once into OCMC RAM,
each port reads from the shared queue and replicates the transmission to
both PRU ports, synchronising between PRU ports are maintained within
firmware with appropriate handling.
For PRP the driver checks the 6-byte RCT trailer for the ETH_P_PRP suffix
to identify redundancy-tagged frames. Frames without an RCT are sent on the
originating port only.
NETIF_F_HW_HSR_DUP is now advertised and is coupled with
NETIF_F_HW_HSR_TAG_RM and NETIF_F_HW_HSR_FWD in ndo_fix_features(),
since the firmware implements them together. To enable HSR/PRP
offload, all three features are set on both slave ports, e.g.:
$ ethtool -K eth2 hsr-dup-offload on
$ ethtool -K eth3 hsr-dup-offload on
With duplicate offload enabled the HSR core sends each frame only on
the first slave port and the firmware transmits it on both.
On the receive side, icssm_parse_packet_info() reads the new HostRecv bit
from the firmware instead of the unused BD Port field. This bit tells which
HSR copy should be passed to the network stack.
icssm_emac_rx_packet() ignores the other duplicate without creating an skb,
since receiving duplicate HSR packets is expected and not an error. The
OCMC-to-skb copy is also split into MAC header, VLAN, and payload parts.
This makes it possible to remove the HSR tag from the middle of the packet.
This patch depends on updated HSR and PRP PRU firmware that implements
the shared TX queue, the LAN/RED TX BD bits, the HostRecv RX BD bit and
the unshifted HSR RX frame layout. The firmware file names are unchanged.
With older HSR/PRP firmware images, HSR and PRP offload modes are not
functional.
Signed-off-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Parvathi Pudi <parvathi@couthit.com>
Signed-off-by: Sai Varun Mudepally <saivarun@couthit.com>
---
drivers/net/ethernet/ti/icssm/icssm_prueth.c | 245 ++++++++++---
drivers/net/ethernet/ti/icssm/icssm_prueth.h | 13 +-
.../ethernet/ti/icssm/icssm_prueth_common.c | 11 +-
.../ethernet/ti/icssm/icssm_prueth_switch.c | 325 +++++++++++++++++-
.../ethernet/ti/icssm/icssm_prueth_switch.h | 1 +
drivers/net/ethernet/ti/icssm/icssm_switch.h | 34 +-
6 files changed, 554 insertions(+), 75 deletions(-)
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.c b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
index 0536b0965c22..3ee44fdffcb1 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth.c
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
@@ -36,12 +36,14 @@
#include "../icssg/icss_iep.h"
#define OCMC_RAM_SIZE (SZ_64K)
+#define PRUETH_ETHER_TYPE_OFFSET 12
#define TX_START_DELAY 0x40
#define TX_CLK_DELAY_100M 0x6
#define HR_TIMER_TX_DELAY_US 100
#define NETIF_PRUETH_LRE_OFFLOAD_FEATURES (NETIF_F_HW_HSR_FWD | \
+ NETIF_F_HW_HSR_DUP | \
NETIF_F_HW_HSR_TAG_RM)
/* ICSSM (v2.1) - supports 64-bit IEP counter.
@@ -79,6 +81,32 @@ static void icssm_prueth_set_fw_offsets(struct prueth *prueth)
}
}
+/* Queue Descriptors initialization for HSR PRP */
+const struct prueth_queue_desc hsr_prp_txopt_queue_descs[][NUM_QUEUES] = {
+ [PRUETH_PORT_QUEUE_HOST] = {
+ { .rd_ptr = P0_Q1_BD_OFFSET, .wr_ptr = P0_Q1_BD_OFFSET, },
+ { .rd_ptr = P0_Q2_BD_OFFSET, .wr_ptr = P0_Q2_BD_OFFSET, },
+ { .rd_ptr = P0_Q3_BD_OFFSET, .wr_ptr = P0_Q3_BD_OFFSET, },
+ { .rd_ptr = P0_Q4_BD_OFFSET, .wr_ptr = P0_Q4_BD_OFFSET, },
+ },
+ [PRUETH_PORT_QUEUE_MII0] = {
+ { .rd_ptr = P0_Q3_BD_OFFSET, .wr_ptr = P0_Q3_BD_OFFSET, },
+ { .rd_ptr = P0_Q4_BD_OFFSET, .wr_ptr = P0_Q4_BD_OFFSET, },
+ { .rd_ptr = P1_Q3_TXOPT_BD_OFFSET,
+ .wr_ptr = P1_Q3_TXOPT_BD_OFFSET, },
+ { .rd_ptr = P2_Q1_TXOPT_BD_OFFSET,
+ .wr_ptr = P2_Q1_TXOPT_BD_OFFSET, },
+ },
+ [PRUETH_PORT_QUEUE_MII1] = {
+ { .rd_ptr = P0_Q1_BD_OFFSET, .wr_ptr = P0_Q1_BD_OFFSET, },
+ { .rd_ptr = P0_Q2_BD_OFFSET, .wr_ptr = P0_Q2_BD_OFFSET, },
+ { .rd_ptr = P1_Q3_TXOPT_BD_OFFSET,
+ .wr_ptr = P1_Q3_TXOPT_BD_OFFSET, },
+ { .rd_ptr = P2_Q1_TXOPT_BD_OFFSET,
+ .wr_ptr = P2_Q1_TXOPT_BD_OFFSET, },
+ }
+};
+
static void icssm_prueth_write_reg(struct prueth *prueth,
enum prueth_mem region,
unsigned int reg, u32 val)
@@ -552,32 +580,34 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
struct sk_buff *skb,
enum prueth_queue_id queue_id)
{
+ struct prueth_queue_desc __iomem *queue_desc_other_port = NULL;
struct prueth_queue_desc __iomem *queue_desc;
const struct prueth_queue_info *txqueue;
- struct net_device *ndev = emac->ndev;
struct prueth *prueth = emac->prueth;
unsigned int buffer_desc_count;
int free_blocks, update_block;
+ struct vlan_ethhdr *vlan_hdr;
bool buffer_wrapped = false;
int write_block, read_block;
+ int free_blocks_other_port;
+ int read_block_other_port;
void *src_addr, *dst_addr;
+ u16 bd_rd_ptr_other_port;
+ struct ethhdr *ethhdr;
int pkt_block_size;
void __iomem *sram;
void __iomem *dram;
int txport, pktlen;
u16 update_wr_ptr;
u32 wr_buf_desc;
+ u16 prp_ethtype;
void *ocmc_ram;
+ __be16 proto;
if (!PRUETH_IS_EMAC(prueth))
dram = prueth->mem[PRUETH_MEM_DRAM1].va;
else
dram = emac->prueth->mem[emac->dram].va;
- if (eth_skb_pad(skb)) {
- if (netif_msg_tx_err(emac) && net_ratelimit())
- netdev_err(ndev, "packet pad failed\n");
- return -ENOMEM;
- }
/* which port to tx: MII0 or MII1 */
txport = emac->tx_port_queue;
@@ -585,7 +615,10 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
pktlen = skb->len;
/* Get the tx queue */
queue_desc = emac->tx_queue_descs + queue_id;
- if (!PRUETH_IS_EMAC(prueth))
+ /* Tx queue context */
+ if (prueth_is_lre(prueth))
+ txqueue = &lre_queue_infos[txport][queue_id];
+ else if (PRUETH_IS_SWITCH(prueth))
txqueue = &sw_queue_infos[txport][queue_id];
else
txqueue = &queue_infos[txport][queue_id];
@@ -608,6 +641,34 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
free_blocks = buffer_desc_count;
}
+ /* Fetch queue state for the second LRE port */
+ if (prueth_is_lre(prueth)) {
+ /* Both PRUs keep running while either port is up. The firmware
+ * of a port that is down or has no link does a dummy update of
+ * its rd_ptr (advances it without transmitting), so the peer
+ * cannot stall the shared queue.
+ */
+ queue_desc_other_port = emac->tx_queue_descs_other_port +
+ queue_id;
+ bd_rd_ptr_other_port = readw(&queue_desc_other_port->rd_ptr);
+
+ read_block_other_port = (bd_rd_ptr_other_port -
+ txqueue->buffer_desc_offset) / BD_SIZE;
+
+ if (write_block > read_block_other_port) {
+ free_blocks_other_port = buffer_desc_count -
+ write_block;
+ free_blocks_other_port += read_block_other_port;
+ } else if (write_block < read_block_other_port) {
+ free_blocks_other_port = read_block_other_port -
+ write_block;
+ } else {
+ free_blocks_other_port = buffer_desc_count;
+ }
+
+ if (free_blocks_other_port < free_blocks)
+ free_blocks = free_blocks_other_port;
+ }
pkt_block_size = DIV_ROUND_UP(pktlen, ICSS_BLOCK_SIZE);
if (pkt_block_size > free_blocks) /* out of queue space */
return -ENOBUFS;
@@ -657,6 +718,40 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
if (PRUETH_IS_HSR(prueth))
wr_buf_desc |= BIT(PRUETH_BD_HSR_FRAME_SHIFT);
+ if (prueth_is_lre(prueth)) {
+ /* LRE slave ports are driven only by the HSR/PRP master (no XDP
+ * TX), so the MAC header is always set.
+ */
+ ethhdr = (struct ethhdr *)skb_mac_header(skb);
+ proto = ethhdr->h_proto;
+
+ if (proto == htons(ETH_P_8021Q)) {
+ vlan_hdr = (struct vlan_ethhdr *)ethhdr;
+ proto = vlan_hdr->h_vlan_encapsulated_proto;
+ }
+
+ /* Check if the SKB has HSR tag */
+ if (PRUETH_IS_HSR(prueth) && proto == htons(ETH_P_HSR)) {
+ /* Enable flags to duplicate transmission in firmware */
+ wr_buf_desc |= PRUETH_BD_LAN_INFO_MASK;
+ wr_buf_desc |= PRUETH_BD_RED_PKT_MASK;
+ } else if (PRUETH_IS_PRP(prueth)) {
+ /* Check if the SKB has PRP tag */
+ prp_ethtype = get_unaligned_be16(skb_tail_pointer(skb) -
+ ETH_TLEN);
+
+ if (prp_ethtype == ETH_P_PRP) {
+ wr_buf_desc |= PRUETH_BD_LAN_INFO_MASK;
+ wr_buf_desc |= PRUETH_BD_RED_PKT_MASK;
+ } else {
+ wr_buf_desc |= (txport <<
+ PRUETH_BD_LAN_A_SHIFT);
+ }
+ } else {
+ wr_buf_desc |= (txport << PRUETH_BD_LAN_A_SHIFT);
+ }
+ }
+
sram = prueth->mem[PRUETH_MEM_SHARED_RAM].va;
if (!PRUETH_IS_EMAC(prueth))
writel(wr_buf_desc, sram + readw(&queue_desc->wr_ptr));
@@ -669,6 +764,10 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
update_wr_ptr = txqueue->buffer_desc_offset + (update_block * BD_SIZE);
writew(update_wr_ptr, &queue_desc->wr_ptr);
+ /* update the write pointer in queue descriptor of other port */
+ if (prueth_is_lre(prueth))
+ writew(update_wr_ptr, &queue_desc_other_port->wr_ptr);
+
return 0;
}
@@ -681,8 +780,10 @@ void icssm_parse_packet_info(struct prueth *prueth, u32 buffer_descriptor,
else
pkt_info->start_offset = false;
- pkt_info->port = (buffer_descriptor & PRUETH_BD_PORT_MASK) >>
- PRUETH_BD_PORT_SHIFT;
+ /* Flag from BD to indicate packet is valid for HOST or not. */
+ pkt_info->host_recv_flag = !!(buffer_descriptor &
+ PRUETH_BD_HOST_RECV_MASK);
+
pkt_info->length = (buffer_descriptor & PRUETH_BD_LENGTH_MASK) >>
PRUETH_BD_LENGTH_SHIFT;
pkt_info->broadcast = !!(buffer_descriptor & PRUETH_BD_BROADCAST_MASK);
@@ -712,17 +813,19 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
struct net_device *ndev = emac->ndev;
unsigned int buffer_desc_count;
int read_block, update_block;
+ struct vlan_ethhdr *vlan_hdr;
unsigned int actual_pkt_len;
bool buffer_wrapped = false;
+ int adjust_for_hsr_tag = 0;
void *src_addr, *dst_addr;
- u16 start_offset = 0;
+ bool has_hsr_tag = false;
+ bool has_vlan = false;
+ struct ethhdr *ethhdr;
struct sk_buff *skb;
int pkt_block_size;
void *ocmc_ram;
+ __be16 proto;
- if (PRUETH_IS_HSR(emac->prueth))
- start_offset = (pkt_info->start_offset ?
- ICSSM_LRE_TAG_SIZE : 0);
/* the PRU firmware deals mostly in pointers already
* offset into ram, we would like to deal in indexes
* within the queue we are working with for code
@@ -731,6 +834,8 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
buffer_desc_count = icssm_get_buff_desc_count(rxqueue);
read_block = (*bd_rd_ptr - rxqueue->buffer_desc_offset) / BD_SIZE;
pkt_block_size = DIV_ROUND_UP(pkt_info->length, ICSS_BLOCK_SIZE);
+ /* OCMC RAM is not cached and read order is not important */
+ ocmc_ram = (__force void *)emac->prueth->mem[PRUETH_MEM_OCMC].va;
/* calculate end BD address post read */
update_block = read_block + pkt_block_size;
@@ -742,11 +847,36 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
buffer_wrapped = true;
}
+ /* Get the start address of the first buffer from
+ * the read buffer description
+ */
+ src_addr = ocmc_ram + rxqueue->buffer_offset +
+ (read_block * ICSS_BLOCK_SIZE);
+
/* calculate new pointer in ram */
*bd_rd_ptr = rxqueue->buffer_desc_offset + (update_block * BD_SIZE);
- /* Exclude the HSR tag bytes already stripped by firmware, if any. */
- actual_pkt_len = pkt_info->length - start_offset;
+ if (PRUETH_IS_HSR(emac->prueth)) {
+ if (!pkt_info->host_recv_flag)
+ return 0;
+
+ ethhdr = (struct ethhdr *)src_addr;
+ proto = ethhdr->h_proto;
+
+ if (proto == htons(ETH_P_8021Q)) {
+ has_vlan = true;
+ vlan_hdr = (struct vlan_ethhdr *)ethhdr;
+ proto = vlan_hdr->h_vlan_encapsulated_proto;
+ }
+
+ if (proto == htons(ETH_P_HSR) && !pkt_info->timestamp)
+ has_hsr_tag = true;
+ }
+
+ actual_pkt_len = pkt_info->length;
+
+ if (has_hsr_tag)
+ actual_pkt_len -= ICSSM_LRE_TAG_SIZE;
/* Allocate a socket buffer for this packet */
skb = netdev_alloc_skb_ip_align(ndev, actual_pkt_len);
@@ -758,15 +888,23 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
dst_addr = skb->data;
- /* OCMC RAM is not cached and read order is not important */
- ocmc_ram = (__force void *)emac->prueth->mem[PRUETH_MEM_OCMC].va;
+ /* Copy destination and source MAC address */
+ memcpy(dst_addr, src_addr, PRUETH_ETHER_TYPE_OFFSET);
+ src_addr += PRUETH_ETHER_TYPE_OFFSET;
+ dst_addr += PRUETH_ETHER_TYPE_OFFSET;
- /* Get the start address of the first buffer from
- * the read buffer description
- */
- src_addr = ocmc_ram + rxqueue->buffer_offset +
- (read_block * ICSS_BLOCK_SIZE);
- src_addr += start_offset;
+ adjust_for_hsr_tag += PRUETH_ETHER_TYPE_OFFSET;
+
+ if (has_vlan) {
+ memcpy(dst_addr, src_addr, VLAN_HLEN);
+ src_addr += VLAN_HLEN;
+ dst_addr += VLAN_HLEN;
+ adjust_for_hsr_tag += VLAN_HLEN;
+ }
+
+ /* HSR tag removal handling */
+ if (has_hsr_tag)
+ src_addr += ICSSM_LRE_TAG_SIZE;
/* Copy the data from PRU buffers(OCMC) to socket buffer(DRAM) */
if (buffer_wrapped) { /* wrapped around buffer */
@@ -780,22 +918,23 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
if (pkt_info->length < bytes)
bytes = pkt_info->length;
- /* If applicable, account for the HSR tag removed */
- bytes -= start_offset;
+ if (has_hsr_tag)
+ bytes -= ICSSM_LRE_TAG_SIZE;
/* copy non-wrapped part */
- memcpy(dst_addr, src_addr, bytes);
+ memcpy(dst_addr, src_addr, bytes - adjust_for_hsr_tag);
/* copy wrapped part */
- dst_addr += bytes;
+ dst_addr += (bytes - adjust_for_hsr_tag);
remaining = actual_pkt_len - bytes;
src_addr = ocmc_ram + rxqueue->buffer_offset;
memcpy(dst_addr, src_addr, remaining);
src_addr += remaining;
} else {
- memcpy(dst_addr, src_addr, actual_pkt_len);
- src_addr += actual_pkt_len;
+ memcpy(dst_addr, src_addr, actual_pkt_len -
+ adjust_for_hsr_tag);
+ src_addr += actual_pkt_len - adjust_for_hsr_tag;
}
if (PRUETH_IS_SWITCH(emac->prueth)) {
@@ -1313,17 +1452,31 @@ static enum netdev_tx icssm_emac_ndo_start_xmit(struct sk_buff *skb,
struct net_device *ndev)
{
struct prueth_emac *emac = netdev_priv(ndev);
+ /* Spinlock for Tx Queues */
+ spinlock_t *lock_queue;
+ unsigned long flags;
int ret;
u16 qid;
qid = icssm_prueth_get_tx_queue_id(emac->prueth, skb);
+ /* Select the TX queue spin lock for this queue ID */
+ if (prueth_is_lre(emac->prueth))
+ lock_queue = &emac->prueth->lre_host_queue_lock[qid - 2];
+ else
+ lock_queue = &emac->host_queue_lock[qid - 2];
+
+ if (eth_skb_pad(skb)) {
+ if (netif_msg_tx_err(emac) && net_ratelimit())
+ netdev_err(ndev, "packet pad failed\n");
+ atomic64_inc(&emac->stats.tx_dropped);
+ return NETDEV_TX_OK;
+ }
+
+ spin_lock_irqsave(lock_queue, flags);
ret = icssm_prueth_tx_enqueue(emac, skb, qid);
- if (ret) {
- if (ret != -ENOBUFS && netif_msg_tx_err(emac) &&
- net_ratelimit())
- netdev_err(ndev, "packet queue failed: %d\n", ret);
+ spin_unlock_irqrestore(lock_queue, flags);
+ if (ret)
goto fail_tx;
- }
atomic64_inc(&emac->stats.tx_packets);
atomic64_add(skb->len, &emac->stats.tx_bytes);
@@ -1338,10 +1491,6 @@ static enum netdev_tx icssm_emac_ndo_start_xmit(struct sk_buff *skb,
us_to_ktime(HR_TIMER_TX_DELAY_US),
HRTIMER_MODE_REL_PINNED);
ret = NETDEV_TX_BUSY;
- } else {
- /* error */
- atomic64_inc(&emac->stats.tx_dropped);
- ret = NET_XMIT_DROP;
}
return ret;
@@ -1573,13 +1722,14 @@ static void icssm_emac_ndo_set_rx_mode(struct net_device *ndev)
static netdev_features_t icssm_emac_ndo_fix_features(struct net_device *ndev,
netdev_features_t features)
{
- /* hsr tag removal offload and hsr fwd offload are tightly coupled in
- * firmware implementation. Both these features need to be enabled /
- * disabled together.
+ /* hsr tag removal offload, hsr fwd offload and hsr dup offload are
+ * tightly coupled in firmware implementation. These features
+ * must always be enabled/disabled together.
*/
if (!(ndev->features & NETIF_PRUETH_LRE_OFFLOAD_FEATURES)) {
if ((features & NETIF_F_HW_HSR_FWD) ||
- (features & NETIF_F_HW_HSR_TAG_RM)) {
+ (features & NETIF_F_HW_HSR_TAG_RM) ||
+ (features & NETIF_F_HW_HSR_DUP)) {
features |= NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
ndev->wanted_features |=
NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
@@ -1587,9 +1737,11 @@ static netdev_features_t icssm_emac_ndo_fix_features(struct net_device *ndev,
}
if ((ndev->features & NETIF_F_HW_HSR_FWD) ||
- (ndev->features & NETIF_F_HW_HSR_TAG_RM)) {
+ (ndev->features & NETIF_F_HW_HSR_TAG_RM) ||
+ (ndev->features & NETIF_F_HW_HSR_DUP)) {
if (!(features & NETIF_F_HW_HSR_FWD) ||
- !(features & NETIF_F_HW_HSR_TAG_RM)) {
+ !(features & NETIF_F_HW_HSR_TAG_RM) ||
+ !(features & NETIF_F_HW_HSR_DUP)) {
features &= ~NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
ndev->wanted_features &=
~NETIF_PRUETH_LRE_OFFLOAD_FEATURES;
@@ -1808,6 +1960,9 @@ static int icssm_prueth_netdev_init(struct prueth *prueth,
spin_lock_init(&emac->lock);
spin_lock_init(&emac->addr_lock);
+ spin_lock_init(&emac->host_queue_lock[0]);
+ spin_lock_init(&emac->host_queue_lock[1]);
+
/* get mac address from DT and set private and netdev addr */
ret = of_get_ethdev_address(eth_node, ndev);
if (!is_valid_ether_addr(ndev->dev_addr)) {
@@ -2411,6 +2566,8 @@ static int icssm_prueth_probe(struct platform_device *pdev)
prueth->support_lre = has_lre;
spin_lock_init(&prueth->addr_lock);
+ spin_lock_init(&prueth->lre_host_queue_lock[0]);
+ spin_lock_init(&prueth->lre_host_queue_lock[1]);
/* setup netdev interfaces */
if (eth0_node) {
ret = icssm_prueth_netdev_init(prueth, eth0_node);
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.h b/drivers/net/ethernet/ti/icssm/icssm_prueth.h
index 8cca5b87d4ff..9ed4176b66fa 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth.h
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.h
@@ -94,7 +94,7 @@ struct prueth_queue_info {
* struct prueth_packet_info - Info about a packet in buffer
* @start_offset: true if frame carries an HSR/PRP start offset
* @shadow: this packet is stored in the collision queue
- * @port: port packet is on
+ * @host_recv_flag: this frame should be received by host
* @length: length of packet
* @broadcast: this packet is a broadcast packet
* @error: this packet has an error
@@ -105,7 +105,7 @@ struct prueth_queue_info {
struct prueth_packet_info {
bool start_offset;
bool shadow;
- unsigned int port;
+ bool host_recv_flag;
unsigned int length;
bool broadcast;
bool error;
@@ -240,6 +240,8 @@ struct prueth_emac {
struct phy_device *phydev;
struct prueth_queue_desc __iomem *rx_queue_descs;
struct prueth_queue_desc __iomem *tx_queue_descs;
+ /* LRE duplicates each TX frame to both ports */
+ struct prueth_queue_desc __iomem *tx_queue_descs_other_port;
int link;
int speed;
@@ -263,6 +265,7 @@ struct prueth_emac {
spinlock_t lock;
spinlock_t addr_lock; /* serialize access to VLAN/MC filter table */
+ spinlock_t host_queue_lock[NUM_QUEUES / 2];
struct hrtimer tx_hrtimer;
struct prueth_emac_stats stats;
int offload_fwd_mark;
@@ -314,9 +317,15 @@ struct prueth {
u8 emac_configured;
u8 hsr_members;
u8 br_members;
+
+ /* Per-queue TX lock - LRE uses only two priority queues
+ * one for high priority and the other for low priority.
+ */
+ spinlock_t lre_host_queue_lock[NUM_QUEUES / 2];
};
extern const struct prueth_queue_desc queue_descs[][NUM_QUEUES];
+extern const struct prueth_queue_desc hsr_prp_txopt_queue_descs[][NUM_QUEUES];
void icssm_parse_packet_info(struct prueth *prueth, u32 buffer_descriptor,
struct prueth_packet_info *pkt_info);
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
index 992fadc1521f..df749ca0a7b1 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
@@ -177,16 +177,17 @@ static int icssm_prueth_common_emac_rx_packets(struct prueth_emac *emac,
used++;
}
- /* Zero the BD after consuming it, a misaligned rd_ptr
- * would otherwise mistake stale data for a valid incoming
- * frame.
+ /* Leave the BD intact after reading. Firmware reuses it to
+ * forward the frame to the second LRE port.
*/
if (port == 0) {
- writel(0, shared_ram + bd_rd_ptr);
+ if (PRUETH_IS_SWITCH(prueth))
+ writel(0, shared_ram + bd_rd_ptr);
writew(update_rd_ptr, &queue_desc->rd_ptr);
bd_rd_ptr = update_rd_ptr;
} else {
- writel(0, shared_ram + bd_rd_ptr_o);
+ if (PRUETH_IS_SWITCH(prueth))
+ writel(0, shared_ram + bd_rd_ptr_o);
writew(update_rd_ptr, &queue_desc_o->rd_ptr);
bd_rd_ptr_o = update_rd_ptr;
}
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
index 9d67fc7c23ac..7bc65b2901cb 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
@@ -199,6 +199,189 @@ static const struct prueth_queue_info rx_queue_infos[][NUM_QUEUES] = {
},
};
+/* Tx Queue context for HSR and PRP */
+const struct prueth_queue_info lre_queue_infos[][NUM_QUEUES] = {
+ [PRUETH_PORT_QUEUE_HOST] = {
+ [PRUETH_QUEUE1] = {
+ P0_Q1_BUFFER_OFFSET,
+ P0_QUEUE_DESC_OFFSET,
+ P0_Q1_BD_OFFSET,
+ P0_Q1_BD_OFFSET + ((HOST_QUEUE_1_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE2] = {
+ P0_Q2_BUFFER_OFFSET,
+ P0_QUEUE_DESC_OFFSET + 8,
+ P0_Q2_BD_OFFSET,
+ P0_Q2_BD_OFFSET + ((HOST_QUEUE_2_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE3] = {
+ P0_Q3_BUFFER_OFFSET,
+ P0_QUEUE_DESC_OFFSET + 16,
+ P0_Q3_BD_OFFSET,
+ P0_Q3_BD_OFFSET + ((HOST_QUEUE_3_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE4] = {
+ P0_Q4_BUFFER_OFFSET,
+ P0_QUEUE_DESC_OFFSET + 24,
+ P0_Q4_BD_OFFSET,
+ P0_Q4_BD_OFFSET + ((HOST_QUEUE_4_SIZE - 1) * BD_SIZE),
+ },
+ },
+ [PRUETH_PORT_QUEUE_MII0] = {
+ [PRUETH_QUEUE1] = {
+ P0_Q3_BUFFER_OFFSET,
+ P0_Q3_BUFFER_OFFSET +
+ ((HOST_QUEUE_3_SIZE - 1) * ICSS_BLOCK_SIZE),
+ P0_Q3_BD_OFFSET,
+ P0_Q3_BD_OFFSET + ((HOST_QUEUE_1_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE2] = {
+ P0_Q4_BUFFER_OFFSET,
+ P0_Q4_BUFFER_OFFSET +
+ ((HOST_QUEUE_4_SIZE - 1) * ICSS_BLOCK_SIZE),
+ P0_Q4_BD_OFFSET,
+ P0_Q4_BD_OFFSET + ((HOST_QUEUE_2_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE3] = {
+ P1_Q3_TXOPT_BUFFER_OFFSET,
+ P1_Q3_TXOPT_BUFFER_OFFSET +
+ ((QUEUE_3_TXOPT_SIZE - 1) * ICSS_BLOCK_SIZE),
+ P1_Q3_TXOPT_BD_OFFSET,
+ P1_Q3_TXOPT_BD_OFFSET +
+ ((QUEUE_3_TXOPT_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE4] = {
+ P2_Q1_TXOPT_BUFFER_OFFSET,
+ P2_Q1_TXOPT_BUFFER_OFFSET +
+ ((QUEUE_4_TXOPT_SIZE - 1) * ICSS_BLOCK_SIZE),
+ P2_Q1_TXOPT_BD_OFFSET,
+ P2_Q1_TXOPT_BD_OFFSET +
+ ((QUEUE_4_TXOPT_SIZE - 1) * BD_SIZE),
+ },
+ },
+ [PRUETH_PORT_QUEUE_MII1] = {
+ [PRUETH_QUEUE1] = {
+ P0_Q1_BUFFER_OFFSET,
+ P0_Q1_BUFFER_OFFSET +
+ ((HOST_QUEUE_1_SIZE - 1) * ICSS_BLOCK_SIZE),
+ P0_Q1_BD_OFFSET,
+ P0_Q1_BD_OFFSET +
+ ((HOST_QUEUE_1_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE2] = {
+ P0_Q2_BUFFER_OFFSET,
+ P0_Q2_BUFFER_OFFSET +
+ ((HOST_QUEUE_2_SIZE - 1) * ICSS_BLOCK_SIZE),
+ P0_Q2_BD_OFFSET,
+ P0_Q2_BD_OFFSET +
+ ((HOST_QUEUE_2_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE3] = {
+ P1_Q3_TXOPT_BUFFER_OFFSET,
+ P1_Q3_TXOPT_BUFFER_OFFSET +
+ ((QUEUE_3_TXOPT_SIZE - 1) * ICSS_BLOCK_SIZE),
+ P1_Q3_TXOPT_BD_OFFSET,
+ P1_Q3_TXOPT_BD_OFFSET +
+ ((QUEUE_3_TXOPT_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE4] = {
+ P2_Q1_TXOPT_BUFFER_OFFSET,
+ P2_Q1_TXOPT_BUFFER_OFFSET +
+ ((QUEUE_4_TXOPT_SIZE - 1) * ICSS_BLOCK_SIZE),
+ P2_Q1_TXOPT_BD_OFFSET,
+ P2_Q1_TXOPT_BD_OFFSET +
+ ((QUEUE_4_TXOPT_SIZE - 1) * BD_SIZE),
+ },
+
+ },
+};
+
+/* Rx Queue Context for HSR and PRP */
+static const struct prueth_queue_info lre_rx_queue_infos[][NUM_QUEUES] = {
+ [PRUETH_PORT_QUEUE_HOST] = {
+ [PRUETH_QUEUE1] = {
+ P0_Q1_BUFFER_OFFSET,
+ HOST_QUEUE_DESC_OFFSET,
+ P0_Q1_BD_OFFSET,
+ P0_Q1_BD_OFFSET + ((HOST_QUEUE_1_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE2] = {
+ P0_Q2_BUFFER_OFFSET,
+ HOST_QUEUE_DESC_OFFSET + 8,
+ P0_Q2_BD_OFFSET,
+ P0_Q2_BD_OFFSET + ((HOST_QUEUE_2_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE3] = {
+ P0_Q3_BUFFER_OFFSET,
+ HOST_QUEUE_DESC_OFFSET + 16,
+ P0_Q3_BD_OFFSET,
+ P0_Q3_BD_OFFSET + ((HOST_QUEUE_3_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE4] = {
+ P0_Q4_BUFFER_OFFSET,
+ HOST_QUEUE_DESC_OFFSET + 24,
+ P0_Q4_BD_OFFSET,
+ P0_Q4_BD_OFFSET + ((HOST_QUEUE_4_SIZE - 1) * BD_SIZE),
+ },
+ },
+ [PRUETH_PORT_QUEUE_MII0] = {
+ [PRUETH_QUEUE1] = {
+ P0_Q3_BUFFER_OFFSET,
+ P1_QUEUE_DESC_OFFSET,
+ P0_Q3_BD_OFFSET,
+ P0_Q3_BD_OFFSET + ((HOST_QUEUE_3_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE2] = {
+ P0_Q4_BUFFER_OFFSET,
+ P1_QUEUE_DESC_OFFSET + 8,
+ P0_Q4_BD_OFFSET,
+ P0_Q4_BD_OFFSET + ((HOST_QUEUE_4_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE3] = {
+ P1_Q3_TXOPT_BUFFER_OFFSET,
+ P1_QUEUE_DESC_OFFSET + 16,
+ P1_Q3_TXOPT_BD_OFFSET,
+ P1_Q3_TXOPT_BD_OFFSET +
+ ((QUEUE_3_TXOPT_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE4] = {
+ P2_Q1_TXOPT_BUFFER_OFFSET,
+ P1_QUEUE_DESC_OFFSET + 24,
+ P2_Q1_TXOPT_BD_OFFSET,
+ P2_Q1_TXOPT_BD_OFFSET +
+ ((QUEUE_4_TXOPT_SIZE - 1) * BD_SIZE),
+ },
+ },
+ [PRUETH_PORT_QUEUE_MII1] = {
+ [PRUETH_QUEUE1] = {
+ P0_Q1_BUFFER_OFFSET,
+ P2_QUEUE_DESC_OFFSET,
+ P0_Q1_BD_OFFSET,
+ P0_Q1_BD_OFFSET + ((HOST_QUEUE_1_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE2] = {
+ P0_Q2_BUFFER_OFFSET,
+ P2_QUEUE_DESC_OFFSET + 8,
+ P0_Q2_BD_OFFSET,
+ P0_Q2_BD_OFFSET + ((HOST_QUEUE_2_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE3] = {
+ P1_Q3_TXOPT_BUFFER_OFFSET,
+ P2_QUEUE_DESC_OFFSET + 16,
+ P1_Q3_TXOPT_BD_OFFSET,
+ P1_Q3_TXOPT_BD_OFFSET +
+ ((QUEUE_3_TXOPT_SIZE - 1) * BD_SIZE),
+ },
+ [PRUETH_QUEUE4] = {
+ P2_Q1_TXOPT_BUFFER_OFFSET,
+ P2_QUEUE_DESC_OFFSET + 24,
+ P2_Q1_TXOPT_BD_OFFSET,
+ P2_Q1_TXOPT_BD_OFFSET +
+ ((QUEUE_4_TXOPT_SIZE - 1) * BD_SIZE),
+ },
+ },
+};
+
void icssm_prueth_sw_free_fdb_table(struct prueth *prueth)
{
if (prueth->emac_configured)
@@ -856,8 +1039,12 @@ void icssm_prueth_sw_hostconfig(struct prueth *prueth)
/* queue information table */
dram = dram1_base + P0_Q1_RX_CONTEXT_OFFSET;
- memcpy_toio(dram, sw_queue_infos[PRUETH_PORT_QUEUE_HOST],
- sizeof(sw_queue_infos[PRUETH_PORT_QUEUE_HOST]));
+ if (prueth_is_lre(prueth))
+ memcpy_toio(dram, lre_queue_infos[PRUETH_PORT_QUEUE_HOST],
+ sizeof(lre_queue_infos[PRUETH_PORT_QUEUE_HOST]));
+ else
+ memcpy_toio(dram, sw_queue_infos[PRUETH_PORT_QUEUE_HOST],
+ sizeof(sw_queue_infos[PRUETH_PORT_QUEUE_HOST]));
/* buffer descriptor offset table*/
dram = dram1_base + QUEUE_DESCRIPTOR_OFFSET_ADDR;
@@ -882,8 +1069,15 @@ void icssm_prueth_sw_hostconfig(struct prueth *prueth)
/* queue table */
dram = dram1_base + P0_QUEUE_DESC_OFFSET;
- memcpy_toio(dram, queue_descs[PRUETH_PORT_QUEUE_HOST],
- sizeof(queue_descs[PRUETH_PORT_QUEUE_HOST]));
+ if (prueth_is_lre(prueth))
+ memcpy_toio(dram,
+ hsr_prp_txopt_queue_descs[PRUETH_PORT_QUEUE_HOST],
+ sizeof(hsr_prp_txopt_queue_descs
+ [PRUETH_PORT_QUEUE_HOST]));
+ else
+ memcpy_toio(dram, queue_descs[PRUETH_PORT_QUEUE_HOST],
+ sizeof(queue_descs[PRUETH_PORT_QUEUE_HOST]));
+
}
void icssm_prueth_sw_config_packet_timestamping(struct prueth *prueth)
@@ -898,7 +1092,7 @@ static int icssm_prueth_sw_port_config(struct prueth *prueth,
enum prueth_port port_id)
{
unsigned int tx_context_ofs_addr, rx_context_ofs, queue_desc_ofs;
- void __iomem *dram, *dram_base, *dram_mac;
+ void __iomem *dram, *dram_base;
struct prueth_emac *emac;
void __iomem *dram1_base;
@@ -909,26 +1103,17 @@ static int icssm_prueth_sw_port_config(struct prueth *prueth,
tx_context_ofs_addr = TX_CONTEXT_P1_Q1_OFFSET_ADDR;
rx_context_ofs = P1_Q1_RX_CONTEXT_OFFSET;
queue_desc_ofs = P1_QUEUE_DESC_OFFSET;
-
- /* for switch PORT MII0 mac addr is in DRAM0. */
- dram_mac = prueth->mem[PRUETH_MEM_DRAM0].va;
break;
case PRUETH_PORT_MII1:
tx_context_ofs_addr = TX_CONTEXT_P2_Q1_OFFSET_ADDR;
rx_context_ofs = P2_Q1_RX_CONTEXT_OFFSET;
queue_desc_ofs = P2_QUEUE_DESC_OFFSET;
-
- /* for switch PORT MII1 mac addr is in DRAM1. */
- dram_mac = prueth->mem[PRUETH_MEM_DRAM1].va;
break;
default:
netdev_err(emac->ndev, "invalid port\n");
return -EINVAL;
}
- /* setup mac address */
- memcpy_toio(dram_mac + PORT_MAC_ADDR, emac->mac_addr, 6);
-
/* Remaining switch port configs are in DRAM1 */
dram_base = prueth->mem[PRUETH_MEM_DRAM1].va;
@@ -983,21 +1168,129 @@ static int icssm_prueth_sw_port_config(struct prueth *prueth,
return 0;
}
+/* Configure TX/RX queue contexts and buffer descriptor tables for LRE port */
+static void icssm_prueth_lre_port_config(struct prueth *prueth,
+ enum prueth_port port_id)
+{
+ unsigned int tx_context_ofs_addr, rx_context_ofs, queue_desc_ofs;
+ void __iomem *dram, *dram_base;
+ struct prueth_emac *emac;
+
+ if (port_id == PRUETH_PORT_MII0) {
+ tx_context_ofs_addr = TX_CONTEXT_P1_Q1_OFFSET_ADDR;
+ rx_context_ofs = P1_Q1_RX_CONTEXT_OFFSET;
+ queue_desc_ofs = P1_QUEUE_DESC_OFFSET;
+ } else if (port_id == PRUETH_PORT_MII1) {
+ tx_context_ofs_addr = TX_CONTEXT_P2_Q1_OFFSET_ADDR;
+ rx_context_ofs = P2_Q1_RX_CONTEXT_OFFSET;
+ queue_desc_ofs = P2_QUEUE_DESC_OFFSET;
+ } else {
+ dev_err(prueth->dev, "invalid port %d\n", port_id);
+ return;
+ }
+
+ emac = prueth->emac[port_id - 1];
+ /* Remaining switch port configs are in DRAM1 */
+ dram_base = prueth->mem[PRUETH_MEM_DRAM1].va;
+
+ /* queue information table */
+ memcpy_toio(dram_base + tx_context_ofs_addr,
+ lre_queue_infos[port_id],
+ sizeof(lre_queue_infos[port_id]));
+
+ memcpy_toio(dram_base + rx_context_ofs,
+ lre_rx_queue_infos[port_id],
+ sizeof(lre_rx_queue_infos[port_id]));
+
+ /* buffer descriptor offset table*/
+ dram = dram_base + QUEUE_DESCRIPTOR_OFFSET_ADDR +
+ (port_id * NUM_QUEUES * sizeof(u16));
+ writew(lre_queue_infos[port_id][PRUETH_QUEUE1].buffer_desc_offset,
+ dram);
+ writew(lre_queue_infos[port_id][PRUETH_QUEUE2].buffer_desc_offset,
+ dram + 2);
+ writew(lre_queue_infos[port_id][PRUETH_QUEUE3].buffer_desc_offset,
+ dram + 4);
+ writew(lre_queue_infos[port_id][PRUETH_QUEUE4].buffer_desc_offset,
+ dram + 6);
+
+ /* buffer offset table */
+ dram = dram_base + QUEUE_OFFSET_ADDR +
+ port_id * NUM_QUEUES * sizeof(u16);
+ writew(lre_queue_infos[port_id][PRUETH_QUEUE1].buffer_offset, dram);
+ writew(lre_queue_infos[port_id][PRUETH_QUEUE2].buffer_offset,
+ dram + 2);
+ writew(lre_queue_infos[port_id][PRUETH_QUEUE3].buffer_offset,
+ dram + 4);
+ writew(lre_queue_infos[port_id][PRUETH_QUEUE4].buffer_offset,
+ dram + 6);
+
+ /* queue size lookup table */
+ dram = dram_base + QUEUE_SIZE_ADDR +
+ port_id * NUM_QUEUES * sizeof(u16);
+ writew(HOST_QUEUE_1_SIZE, dram);
+ writew(HOST_QUEUE_2_SIZE, dram + 2);
+ writew(QUEUE_3_TXOPT_SIZE, dram + 4);
+ writew(QUEUE_4_TXOPT_SIZE, dram + 6);
+
+ /* queue table */
+ memcpy_toio(dram_base + queue_desc_ofs,
+ &hsr_prp_txopt_queue_descs[port_id][0],
+ 4 * sizeof(hsr_prp_txopt_queue_descs[port_id][0]));
+
+ /* In HSR/PRP mode both slave ports share the host receive queue
+ * descriptor region (P0_QUEUE_DESC_OFFSET). The firmware arbitrates
+ * ownership; the driver always reads from the same host-side descriptor
+ * base regardless of which physical port the frame arrived on.
+ */
+ emac->rx_queue_descs = dram_base + P0_QUEUE_DESC_OFFSET;
+ emac->tx_queue_descs = dram_base +
+ lre_rx_queue_infos[port_id][PRUETH_QUEUE1].queue_desc_offset;
+
+ if (port_id == PRUETH_PORT_MII0) {
+ emac->tx_queue_descs_other_port = dram_base +
+ lre_rx_queue_infos
+ [port_id + 1][PRUETH_QUEUE1].queue_desc_offset;
+ } else if (port_id == PRUETH_PORT_MII1) {
+ emac->tx_queue_descs_other_port = dram_base +
+ lre_rx_queue_infos
+ [port_id - 1][PRUETH_QUEUE1].queue_desc_offset;
+ }
+}
+
int icssm_prueth_sw_emac_config(struct prueth_emac *emac)
{
struct prueth *prueth = emac->prueth;
u32 sharedramaddr, ocmcaddr;
- int ret;
+ void __iomem *dram_mac;
+ int ret = 0;
/* PRU needs local shared RAM address for C28 */
sharedramaddr = ICSS_LOCAL_SHARED_RAM;
/* PRU needs real global OCMC address for C30*/
ocmcaddr = (u32)prueth->mem[PRUETH_MEM_OCMC].pa;
+ if (emac->port_id == PRUETH_PORT_MII0) {
+ dram_mac = prueth->mem[PRUETH_MEM_DRAM0].va;
+ } else if (emac->port_id == PRUETH_PORT_MII1) {
+ dram_mac = prueth->mem[PRUETH_MEM_DRAM1].va;
+ } else {
+ netdev_err(emac->ndev, "invalid port\n");
+ return -EINVAL;
+ }
+
+ /* setup mac address */
+ memcpy_toio(dram_mac + PORT_MAC_ADDR, emac->mac_addr, ETH_ALEN);
+
if (prueth->emac_configured & BIT(emac->port_id))
return 0;
- ret = icssm_prueth_sw_port_config(prueth, emac->port_id);
+ if (prueth_is_lre(prueth) && !prueth->emac_configured) {
+ icssm_prueth_lre_port_config(prueth, PRUETH_PORT_MII0);
+ icssm_prueth_lre_port_config(prueth, PRUETH_PORT_MII1);
+ } else if (PRUETH_IS_SWITCH(prueth)) {
+ ret = icssm_prueth_sw_port_config(prueth, emac->port_id);
+ }
if (ret)
return ret;
diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.h b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.h
index fe7197f9a6da..a009827b01d7 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.h
+++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.h
@@ -17,6 +17,7 @@ u8 icssm_prueth_sw_get_stp_state(struct prueth *prueth,
enum prueth_port port);
extern const struct prueth_queue_info sw_queue_infos[][4];
+extern const struct prueth_queue_info lre_queue_infos[][4];
void icssm_prueth_sw_fdb_tbl_init(struct prueth *prueth);
int icssm_prueth_sw_init_fdb_table(struct prueth *prueth);
diff --git a/drivers/net/ethernet/ti/icssm/icssm_switch.h b/drivers/net/ethernet/ti/icssm/icssm_switch.h
index b4d6b8660539..ec5925c812c3 100644
--- a/drivers/net/ethernet/ti/icssm/icssm_switch.h
+++ b/drivers/net/ethernet/ti/icssm/icssm_switch.h
@@ -24,6 +24,9 @@
#define QUEUE_3_SIZE 97 /* Protocol specific */
#define QUEUE_4_SIZE 97 /* NRT (IP,ARP, ICMP) */
+#define QUEUE_3_TXOPT_SIZE 194 /* Protocol specific - High Priority */
+#define QUEUE_4_TXOPT_SIZE 194 /* NRT(IP,ARP, ICMP) - Low Priority*/
+
/* Host queue size (number of BDs). Each BD points to data buffer of 32 bytes.
* HOST PORT QUEUES can buffer up to 4 full sized frames per queue
*/
@@ -51,18 +54,19 @@
* address found in FDB). For switch only.
* 8..12 Block_length number of valid bytes in this specific block.
* Will be <=32 bytes on last block of packet
+ * 8 RED_INFO Set if the frame carries an HSR or PRP
+ * redundancy tag
+ * 10 HostRecv Set if the frame is destined for the host port.
+ * For HSR only.
* 13 More "More" bit indicating that there are more blocks
* 14 Shadow indicates that "index" is pointing into shadow
* buffer
* 15 TimeStamp indicates that this packet has time stamp in
* separate buffer - only needed if PTP runs on
* host
- * 16..17 Port different meaning for ingress and egress,
- * Ingress: Port = 0 indicates phy port 1 and
- * Port = 1 indicates phy port 2.
- * Egress: 0 sends on phy port 1 and 1 sends on
- * phy port 2. Port = 2 goes over MAC table
- * look-up
+ * 16..17 LAN Destination LAN for transmission:
+ * bit 16 = LAN A, bit 17 = LAN B, set both to
+ * duplicate to both LANs. For HSR and PRP only.
* 18..28 Length 11 bit of total packet length which is put into
* first BD only so that host access only one BD
* 29 VlanTag indicates that packet has Length/Type field of
@@ -86,14 +90,21 @@
#define PRUETH_BD_SW_FLOOD_MASK BIT(7)
#define PRUETH_BD_SW_FLOOD_SHIFT 7
+#define PRUETH_BD_RED_PKT_MASK BIT(8)
+#define PRUETH_BD_RED_PKT 8
+
+#define PRUETH_BD_HOST_RECV_MASK BIT(10)
+#define PRUETH_BD_HOST_RECV_SHIFT 10
+
#define PRUETH_BD_SHADOW_MASK BIT(14)
#define PRUETH_BD_SHADOW_SHIFT 14
#define PRUETH_BD_TIMESTAMP_MASK BIT(15)
#define PRUETH_BD_TIMESTAMP_SHIFT 15
-#define PRUETH_BD_PORT_MASK GENMASK(17, 16)
-#define PRUETH_BD_PORT_SHIFT 16
+#define PRUETH_BD_LAN_INFO_MASK GENMASK(17, 16)
+#define PRUETH_BD_LAN_A_SHIFT 16
+#define PRUETH_BD_LAN_B_SHIFT 17
#define PRUETH_BD_LENGTH_MASK GENMASK(28, 18)
#define PRUETH_BD_LENGTH_SHIFT 18
@@ -298,6 +309,9 @@
#define P0_Q4_BD_OFFSET (P0_Q3_BD_OFFSET + HOST_QUEUE_3_SIZE * BD_SIZE)
#define P0_Q3_BD_OFFSET (P0_Q2_BD_OFFSET + HOST_QUEUE_2_SIZE * BD_SIZE)
#define P0_Q2_BD_OFFSET (P0_Q1_BD_OFFSET + HOST_QUEUE_1_SIZE * BD_SIZE)
+#define P1_Q3_TXOPT_BD_OFFSET (P0_Q4_BD_OFFSET + HOST_QUEUE_4_SIZE * BD_SIZE)
+#define P2_Q1_TXOPT_BD_OFFSET (P1_Q3_TXOPT_BD_OFFSET + \
+ QUEUE_3_TXOPT_SIZE * BD_SIZE)
#define P0_Q1_BD_OFFSET P0_BUFFER_DESC_OFFSET
#define P0_BUFFER_DESC_OFFSET SRAM_START_OFFSET
@@ -334,6 +348,10 @@
ICSS_BLOCK_SIZE)
#define P0_Q2_BUFFER_OFFSET (P0_Q1_BUFFER_OFFSET + HOST_QUEUE_1_SIZE * \
ICSS_BLOCK_SIZE)
+#define P1_Q3_TXOPT_BUFFER_OFFSET (P0_Q4_BUFFER_OFFSET + \
+ HOST_QUEUE_4_SIZE * ICSS_BLOCK_SIZE)
+#define P2_Q1_TXOPT_BUFFER_OFFSET (P1_Q3_TXOPT_BUFFER_OFFSET + \
+ QUEUE_3_TXOPT_SIZE * ICSS_BLOCK_SIZE)
#define P0_COL_BUFFER_OFFSET 0xEE00
#define P0_Q1_BUFFER_OFFSET 0x0000
--
2.43.0
^ permalink raw reply related [flat|nested] 7+ messages in thread* Re: [PATCH net-next v6 3/3] net: ti: icssm-prueth: Support duplicate HW offload feature for HSR and PRP
2026-11-13 14:28 ` [PATCH net-next v6 3/3] net: ti: icssm-prueth: Support duplicate HW offload feature for HSR and PRP Sai Varun Mudepally
@ 2026-11-17 15:05 ` netdev-bot+sashiko
0 siblings, 0 replies; 7+ messages in thread
From: netdev-bot+sashiko @ 2026-11-17 15:05 UTC (permalink / raw)
To: saivarun
Cc: danishanwar, parvathi, rogerq, pmohan, vadim.fedorenko,
andrew+netdev, davem, edumazet, kuba, pabeni, afd, basharath,
arnd, haokexin, linux-arm-kernel, netdev, linux-kernel, pratheesh,
vigneshr, praneeth, srk, j-rameshbabu, m-malladi, rogerq, krishna,
mohan
Thank you for your contribution! Sashiko AI review found 9 potential
issue(s) to consider.
Critical: 0 · High: 0 · Medium: 4 · Low: 5
If you believe none of the issues are legitimate, please reply to each
issue of Medium severity or higher and explain why the AI is wrong.
If any of them are legitimate and you plan to respin, please let patchwork
know by including "pw-bot: cr" as a separate line at the end of your reply
(one such reply per series is enough).
- [Medium] This patch changes the HSR/PRP driver/firmware contract but
keeps the same firmware file names and adds no version or capability…
- [Medium] In PRP mode with duplicate offload, icssm_prueth_tx_enqueue()
treats a frame as redundancy-tagged (sets PRUETH_BD_LAN_INFO_MASK and…
- [Medium] The full/empty check on the new shared HSR/PRP TX ring is off
by one.
- [Medium] In LRE mode, icssm_prueth_tx_enqueue() classifies the frame
through skb_mac_header(skb) (`ethhdr = (struct ethhdr…
- [Low] The patch contains several behavioural changes that the commit
message does not mention.
- [Low] The BD layout comment in icssm_switch.h still documents bits
'8..12 Block_length number of valid bytes in this specific block'.
- [Low] The new LRE queue tables describe the same rings with different
size constants.
- [Low] icssm_emac_ndo_start_xmit() takes the TX queue lock with
spin_lock_irqsave() around the whole of icssm_prueth_tx_enqueue(),…
- [Low] The patch now zeroes consumed host RX BDs only when
PRUETH_IS_SWITCH(prueth), so PRP loses the stale-BD safeguard as well…
--
net: ti: icssm-prueth: Support duplicate HW offload feature for HSR and PRP
This commit adds HSR/PRP duplicate offload. The driver writes each TX
frame once into a shared OCMC queue, both PRU ports consume it, and
NETIF_F_HW_HSR_DUP is advertised together with the TAG_RM and FWD
features. On RX, a new HostRecv BD bit selects which HSR duplicate goes
to the stack, and the driver strips the HSR tag while copying. The patch
depends on updated HSR/PRP PRU firmware that keeps the same file names.
> diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth.c b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
> index 0536b0965c222..3ee44fdffcb1f 100644
> --- a/drivers/net/ethernet/ti/icssm/icssm_prueth.c
> +++ b/drivers/net/ethernet/ti/icssm/icssm_prueth.c
[ ... ]
> @@ -608,6 +641,34 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
> free_blocks = buffer_desc_count;
> }
>
> + /* Fetch queue state for the second LRE port */
> + if (prueth_is_lre(prueth)) {
[ ... ]
> + if (write_block > read_block_other_port) {
> + free_blocks_other_port = buffer_desc_count -
> + write_block;
> + free_blocks_other_port += read_block_other_port;
> + } else if (write_block < read_block_other_port) {
> + free_blocks_other_port = read_block_other_port -
> + write_block;
> + } else {
> + free_blocks_other_port = buffer_desc_count;
> + }
> +
> + if (free_blocks_other_port < free_blocks)
> + free_blocks = free_blocks_other_port;
> + }
> pkt_block_size = DIV_ROUND_UP(pktlen, ICSS_BLOCK_SIZE);
> if (pkt_block_size > free_blocks) /* out of queue space */
> return -ENOBUFS;
[Severity: Medium]
Can this fill the shared ring exactly, so that a full ring looks the same
as an empty one?
When write_block equals the peer's read block, the whole ring counts as
free. A frame is rejected only when pkt_block_size > free_blocks. So a
frame whose block count equals min(free_blocks, free_blocks_other_port) is
accepted, and update_block then lands on the lagging consumer's read
block.
The same update_wr_ptr is then written to both descriptors further down:
writew(update_wr_ptr, &queue_desc->wr_ptr);
...
if (prueth_is_lre(prueth))
writew(update_wr_ptr, &queue_desc_other_port->wr_ptr);
The lagging PRU would then see wr_ptr == rd_ptr and treat its backlog as
empty. On the next xmit, free_blocks_other_port becomes buffer_desc_count
again. The driver could then overwrite OCMC blocks holding frames that PRU
has not sent yet, and those frames were already counted in tx_packets.
Would rejecting when pkt_block_size >= free_blocks, so one block always
stays unused, avoid this? The own-port check had the same off-by-one
before this patch. The second consumer on the shared ring is new here.
> @@ -657,6 +718,40 @@ static int icssm_prueth_tx_enqueue(struct prueth_emac *emac,
> if (PRUETH_IS_HSR(prueth))
> wr_buf_desc |= BIT(PRUETH_BD_HSR_FRAME_SHIFT);
>
> + if (prueth_is_lre(prueth)) {
> + /* LRE slave ports are driven only by the HSR/PRP master (no XDP
> + * TX), so the MAC header is always set.
> + */
> + ethhdr = (struct ethhdr *)skb_mac_header(skb);
> + proto = ethhdr->h_proto;
[Severity: Medium]
Is the MAC header always set here? A CAP_NET_RAW user can bind an AF_XDP
socket in copy mode to a slave port, which gives this path:
xsk_generic_xmit()
__dev_direct_xmit()
netdev_start_xmit()
icssm_emac_ndo_start_xmit()
icssm_prueth_tx_enqueue()
skb_mac_header(skb)
xsk_build_skb() never sets the MAC header. __dev_queue_xmit() calls
skb_reset_mac_header(), but __dev_direct_xmit() and validate_xmit_skb()
do not. skb->mac_header would still be (u16)~0, so skb_mac_header()
returns skb->head + 0xffff.
Would this read h_proto about 64KB past the skb head, and then use that
garbage value to choose the LAN A/LAN B/RED BD bits?
The rest of the function copies from skb->data, which has at least 60
linear bytes after eth_skb_pad(). Could the ethertype be read from
skb->data instead?
> +
> + if (proto == htons(ETH_P_8021Q)) {
> + vlan_hdr = (struct vlan_ethhdr *)ethhdr;
> + proto = vlan_hdr->h_vlan_encapsulated_proto;
> + }
> +
> + /* Check if the SKB has HSR tag */
> + if (PRUETH_IS_HSR(prueth) && proto == htons(ETH_P_HSR)) {
> + /* Enable flags to duplicate transmission in firmware */
> + wr_buf_desc |= PRUETH_BD_LAN_INFO_MASK;
> + wr_buf_desc |= PRUETH_BD_RED_PKT_MASK;
> + } else if (PRUETH_IS_PRP(prueth)) {
> + /* Check if the SKB has PRP tag */
> + prp_ethtype = get_unaligned_be16(skb_tail_pointer(skb) -
> + ETH_TLEN);
> +
> + if (prp_ethtype == ETH_P_PRP) {
> + wr_buf_desc |= PRUETH_BD_LAN_INFO_MASK;
> + wr_buf_desc |= PRUETH_BD_RED_PKT_MASK;
[Severity: Medium]
Is the 2-byte suffix check enough to identify an RCT? The HSR core checks
both the suffix and the LSDU size:
net/hsr/hsr_forward.c:prp_fill_frame_info() {
...
if (rct &&
prp_check_lsdu_size(skb, rct, frame->is_supervision)) {
...
}
The PRP master always adds a correct RCT through prp_fill_rct(). Untagged
frames on a slave therefore come from direct transmission on the slave
netdev, for example AF_PACKET. If such a frame's payload happens to end
in 0x88FB, it would be marked for both LANs and sent on the LAN it was not
meant for.
If the firmware rewrites the RCT LAN-ID for the second copy, could it
also change real payload bytes of such a frame?
[ ... ]
> @@ -681,8 +780,10 @@ void icssm_parse_packet_info(struct prueth *prueth, u32 buffer_descriptor,
> else
> pkt_info->start_offset = false;
>
> - pkt_info->port = (buffer_descriptor & PRUETH_BD_PORT_MASK) >>
> - PRUETH_BD_PORT_SHIFT;
> + /* Flag from BD to indicate packet is valid for HOST or not. */
> + pkt_info->host_recv_flag = !!(buffer_descriptor &
> + PRUETH_BD_HOST_RECV_MASK);
> +
[ ... ]
> @@ -742,11 +847,36 @@ int icssm_emac_rx_packet(struct prueth_emac *emac, u16 *bd_rd_ptr,
[ ... ]
> - /* Exclude the HSR tag bytes already stripped by firmware, if any. */
> - actual_pkt_len = pkt_info->length - start_offset;
> + if (PRUETH_IS_HSR(emac->prueth)) {
> + if (!pkt_info->host_recv_flag)
> + return 0;
[Severity: Medium]
What happens here when an older HSR/PRP firmware image is loaded? The
commit message says:
The firmware file names are unchanged.
With older HSR/PRP firmware images, HSR and PRP offload modes are not
functional.
icssm_prueth_sw_boot_prus() still loads fw_name[prueth->eth_type], for
example ti-pruss/am57xx-pru0-pruhsr-fw.elf. The only revision information
is the static per-SoC fw_rev in the match data, and nothing reads back
from the running PRU to confirm the new ABI.
The earlier patch in this series that added HSR/PRP offload was written
against the previous firmware ABI under these same file names. So images
with these names and the old ABI exist. That also makes the ABI change
inside the series a bisect hazard.
With such an image:
- this check silently drops every HSR frame whose BD bit 10 is clear,
and no counter is incremented
- the frame start is parsed as an unshifted MAC/VLAN/ethertype layout
- on TX, both ports' wr_ptrs advance over one shared buffer, with LAN
bits the old firmware does not understand
icssm_prueth_sw_emac_config() and icssm_prueth_lre_port_config() write
the new queue tables anyway. Could the driver detect the firmware
capability and refuse LRE offload, rather than running incorrectly with
no error?
[ ... ]
> @@ -1313,17 +1452,31 @@ static enum netdev_tx icssm_emac_ndo_start_xmit(struct sk_buff *skb,
> struct net_device *ndev)
> {
> struct prueth_emac *emac = netdev_priv(ndev);
> + /* Spinlock for Tx Queues */
> + spinlock_t *lock_queue;
> + unsigned long flags;
> int ret;
> u16 qid;
>
> qid = icssm_prueth_get_tx_queue_id(emac->prueth, skb);
> + /* Select the TX queue spin lock for this queue ID */
> + if (prueth_is_lre(emac->prueth))
> + lock_queue = &emac->prueth->lre_host_queue_lock[qid - 2];
> + else
> + lock_queue = &emac->host_queue_lock[qid - 2];
> +
> + if (eth_skb_pad(skb)) {
> + if (netif_msg_tx_err(emac) && net_ratelimit())
> + netdev_err(ndev, "packet pad failed\n");
> + atomic64_inc(&emac->stats.tx_dropped);
> + return NETDEV_TX_OK;
> + }
> +
> + spin_lock_irqsave(lock_queue, flags);
> ret = icssm_prueth_tx_enqueue(emac, skb, qid);
> - if (ret) {
> - if (ret != -ENOBUFS && netif_msg_tx_err(emac) &&
> - net_ratelimit())
> - netdev_err(ndev, "packet queue failed: %d\n", ret);
> + spin_unlock_irqrestore(lock_queue, flags);
> + if (ret)
> goto fail_tx;
> - }
[Severity: Low]
Does this need to disable interrupts? The lock is only taken here, and
icssm_prueth_tx_enqueue() has no other caller. Neither
icssm_emac_tx_timer_callback() nor any IRQ handler takes host_queue_lock
or lre_host_queue_lock.
With irqsave, hard IRQs stay off while up to about 1.5KB is copied into
uncached OCMC, plus several MMIO accesses. xmit already runs with BH
disabled, so would plain spin_lock() be enough?
In EMAC and switch modes, the per-emac host_queue_lock also looks
redundant. The driver is not LLTX and has a single TX queue, so
HARD_TX_LOCK already serializes xmit per netdev. The shared
lre_host_queue_lock does look necessary in LRE mode, since two netdevs
write the shared queue.
[ ... ]
> @@ -1338,10 +1491,6 @@ static enum netdev_tx icssm_emac_ndo_start_xmit(struct sk_buff *skb,
> us_to_ktime(HR_TIMER_TX_DELAY_US),
> HRTIMER_MODE_REL_PINNED);
> ret = NETDEV_TX_BUSY;
> - } else {
> - /* error */
> - atomic64_inc(&emac->stats.tx_dropped);
> - ret = NET_XMIT_DROP;
> }
>
> return ret;
[Severity: Low]
This isn't a bug, but the commit message only covers the dup offload, the
PRP RCT check, the feature coupling and HostRecv. These other changes in
the patch are not mentioned:
- the TX locking above also applies to dual-EMAC and switch modes
- eth_skb_pad() moves from icssm_prueth_tx_enqueue() into
icssm_emac_ndo_start_xmit()
- the "packet queue failed" print is removed
- the tx_dropped / NET_XMIT_DROP branch above is gone, so a future
non-ENOBUFS error from icssm_prueth_tx_enqueue() would be returned
from ndo_start_xmit as a raw errno
- icssm_prueth_common_emac_rx_packets() stops zeroing consumed BDs for
HSR and PRP
- port MAC programming moves from icssm_prueth_sw_port_config() to
icssm_prueth_sw_emac_config() and now runs on every call
- a new TX queue memory map is added (QUEUE_3/4_TXOPT_SIZE and the new
OCMC and shared RAM offsets)
Could these be described in the commit message, or split out into
separate patches?
[ ... ]
> diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
> index 992fadc1521f9..df749ca0a7b1b 100644
> --- a/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
> +++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_common.c
> @@ -177,16 +177,17 @@ static int icssm_prueth_common_emac_rx_packets(struct prueth_emac *emac,
> used++;
> }
>
> - /* Zero the BD after consuming it, a misaligned rd_ptr
> - * would otherwise mistake stale data for a valid incoming
> - * frame.
> + /* Leave the BD intact after reading. Firmware reuses it to
> + * forward the frame to the second LRE port.
> */
> if (port == 0) {
> - writel(0, shared_ram + bd_rd_ptr);
> + if (PRUETH_IS_SWITCH(prueth))
> + writel(0, shared_ram + bd_rd_ptr);
[Severity: Low]
Should this exception apply only to HSR? The new comment's reason,
firmware reusing the BD to forward the frame to the second LRE port,
describes HSR ring forwarding. PRP does no port-to-port forwarding:
prp_drop_frame() and prp_is_lan_dup() in the HSR core forbid LAN A <->
LAN B forwarding.
The old comment covered the misaligned rd_ptr case. A zeroed BD fails
the EMAC_MIN_PKTLEN check, and rd_ptr is resynced to wr_ptr. Without the
zeroing in PRP mode, a stale BD with a plausible length would be
processed and rd_ptr advanced by a stale amount.
Would skipping the writel(0) only when PRUETH_IS_HSR(prueth) keep that
safeguard for PRP?
[ ... ]
> diff --git a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
> index 9d67fc7c23acc..7bc65b2901cbb 100644
> --- a/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
> +++ b/drivers/net/ethernet/ti/icssm/icssm_prueth_switch.c
> @@ -199,6 +199,189 @@ static const struct prueth_queue_info rx_queue_infos[][NUM_QUEUES] = {
[ ... ]
> + [PRUETH_PORT_QUEUE_MII0] = {
> + [PRUETH_QUEUE1] = {
> + P0_Q3_BUFFER_OFFSET,
> + P0_Q3_BUFFER_OFFSET +
> + ((HOST_QUEUE_3_SIZE - 1) * ICSS_BLOCK_SIZE),
> + P0_Q3_BD_OFFSET,
> + P0_Q3_BD_OFFSET + ((HOST_QUEUE_1_SIZE - 1) * BD_SIZE),
> + },
> + [PRUETH_QUEUE2] = {
> + P0_Q4_BUFFER_OFFSET,
> + P0_Q4_BUFFER_OFFSET +
> + ((HOST_QUEUE_4_SIZE - 1) * ICSS_BLOCK_SIZE),
> + P0_Q4_BD_OFFSET,
> + P0_Q4_BD_OFFSET + ((HOST_QUEUE_2_SIZE - 1) * BD_SIZE),
> + },
[Severity: Low]
This isn't a bug today, but should buffer_desc_end here use
HOST_QUEUE_3_SIZE and HOST_QUEUE_4_SIZE? Each entry uses HOST_QUEUE_3/4
for the buffer end but HOST_QUEUE_1/2 for the BD end of the same ring.
lre_rx_queue_infos[PRUETH_PORT_QUEUE_MII0] describes the same BD ranges
with HOST_QUEUE_3/4_SIZE, and icssm_switch.h sizes P0_Q3/P0_Q4 with
HOST_QUEUE_3/4_SIZE.
icssm_prueth_lre_port_config() also writes HOST_QUEUE_1_SIZE and
HOST_QUEUE_2_SIZE into the firmware queue size table for Q1/Q2 of both
ports, whichever region backs them.
All HOST_QUEUE_*_SIZE values are currently 194, so nothing differs at
runtime. If one host queue size changes later, would the driver and
firmware disagree on the wrap point of these shared rings?
[ ... ]
> diff --git a/drivers/net/ethernet/ti/icssm/icssm_switch.h b/drivers/net/ethernet/ti/icssm/icssm_switch.h
> index b4d6b86605395..ec5925c812c36 100644
> --- a/drivers/net/ethernet/ti/icssm/icssm_switch.h
> +++ b/drivers/net/ethernet/ti/icssm/icssm_switch.h
[ ... ]
> @@ -51,18 +54,19 @@
> * address found in FDB). For switch only.
> * 8..12 Block_length number of valid bytes in this specific block.
> * Will be <=32 bytes on last block of packet
> + * 8 RED_INFO Set if the frame carries an HSR or PRP
> + * redundancy tag
> + * 10 HostRecv Set if the frame is destined for the host port.
> + * For HSR only.
[ ... ]
> @@ -86,14 +90,21 @@
> #define PRUETH_BD_SW_FLOOD_MASK BIT(7)
> #define PRUETH_BD_SW_FLOOD_SHIFT 7
>
> +#define PRUETH_BD_RED_PKT_MASK BIT(8)
> +#define PRUETH_BD_RED_PKT 8
> +
> +#define PRUETH_BD_HOST_RECV_MASK BIT(10)
> +#define PRUETH_BD_HOST_RECV_SHIFT 10
> +
> #define PRUETH_BD_SHADOW_MASK BIT(14)
> #define PRUETH_BD_SHADOW_SHIFT 14
>
> #define PRUETH_BD_TIMESTAMP_MASK BIT(15)
> #define PRUETH_BD_TIMESTAMP_SHIFT 15
>
> -#define PRUETH_BD_PORT_MASK GENMASK(17, 16)
> -#define PRUETH_BD_PORT_SHIFT 16
> +#define PRUETH_BD_LAN_INFO_MASK GENMASK(17, 16)
> +#define PRUETH_BD_LAN_A_SHIFT 16
> +#define PRUETH_BD_LAN_B_SHIFT 17
[Severity: Low]
This isn't a bug, but the BD layout table still lists bits 8..12 as
Block_length, and now also lists bit 8 as RED_INFO and bit 10 as
HostRecv. Could the Block_length entry be narrowed or removed so that
each bit has one meaning? The table already had similar overlaps before
this patch, with flag bits inside the Index 0..7 field.
The new macro names are also inconsistent with the other pairs.
PRUETH_BD_RED_PKT has no _SHIFT suffix, and PRUETH_BD_HOST_RECV_SHIFT and
PRUETH_BD_LAN_B_SHIFT are never used.
--
Sashiko AI review · https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20261113143127.282466-1-saivarun%40couthit.com
^ permalink raw reply [flat|nested] 7+ messages in thread