bnxt_en: Add support for ethtool get dump.

Add support to collect live firmware coredump via ethtool.

Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/broadcom/bnxt/bnxt_coredump.h

@@ -0,0 +1,66 @@
+/* Broadcom NetXtreme-C/E network driver.
+ *
+ * Copyright (c) 2018 Broadcom Inc
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation.
+ */
+
+#ifndef BNXT_COREDUMP_H
+#define BNXT_COREDUMP_H
+
+struct bnxt_coredump_segment_hdr {
+	__u8 signature[4];
+	__le32 component_id;
+	__le32 segment_id;
+	__le32 flags;
+	__u8 low_version;
+	__u8 high_version;
+	__le16 function_id;
+	__le32 offset;
+	__le32 length;
+	__le32 status;
+	__le32 duration;
+	__le32 data_offset;
+	__le32 instance;
+	__le32 rsvd[5];
+};
+
+struct bnxt_coredump_record {
+	__u8 signature[4];
+	__le32 flags;
+	__u8 low_version;
+	__u8 high_version;
+	__u8 asic_state;
+	__u8 rsvd0[5];
+	char system_name[32];
+	__le16 year;
+	__le16 month;
+	__le16 day;
+	__le16 hour;
+	__le16 minute;
+	__le16 second;
+	__le16 utc_bias;
+	__le16 rsvd1;
+	char commandline[256];
+	__le32 total_segments;
+	__le32 os_ver_major;
+	__le32 os_ver_minor;
+	__le32 rsvd2;
+	char os_name[32];
+	__le16 end_year;
+	__le16 end_month;
+	__le16 end_day;
+	__le16 end_hour;
+	__le16 end_minute;
+	__le16 end_second;
+	__le16 end_utc_bias;
+	__le32 asic_id1;
+	__le32 asic_id2;
+	__le32 coredump_status;
+	__u8 ioctl_low_version;
+	__u8 ioctl_high_version;
+	__le16 rsvd3[313];
+};
+#endif

drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c

@@ -16,12 +16,15 @@
 #include <linux/etherdevice.h>
 #include <linux/crc32.h>
 #include <linux/firmware.h>
+#include <linux/utsname.h>
+#include <linux/time.h>
 #include "bnxt_hsi.h"
 #include "bnxt.h"
 #include "bnxt_xdp.h"
 #include "bnxt_ethtool.h"
 #include "bnxt_nvm_defs.h"	/* NVRAM content constant and structure defs */
 #include "bnxt_fw_hdr.h"	/* Firmware hdr constant and structure defs */
+#include "bnxt_coredump.h"
 #define FLASH_NVRAM_TIMEOUT	((HWRM_CMD_TIMEOUT) * 100)
 #define FLASH_PACKAGE_TIMEOUT	((HWRM_CMD_TIMEOUT) * 200)
 #define INSTALL_PACKAGE_TIMEOUT	((HWRM_CMD_TIMEOUT) * 200)
@@ -2685,6 +2688,334 @@ static int bnxt_reset(struct net_device *dev, u32 *flags)
 	return rc;
 }
 
+static int bnxt_hwrm_dbg_dma_data(struct bnxt *bp, void *msg, int msg_len,
+				  struct bnxt_hwrm_dbg_dma_info *info)
+{
+	struct hwrm_dbg_cmn_output *cmn_resp = bp->hwrm_cmd_resp_addr;
+	struct hwrm_dbg_cmn_input *cmn_req = msg;
+	__le16 *seq_ptr = msg + info->seq_off;
+	u16 seq = 0, len, segs_off;
+	void *resp = cmn_resp;
+	dma_addr_t dma_handle;
+	int rc, off = 0;
+	void *dma_buf;
+
+	dma_buf = dma_alloc_coherent(&bp->pdev->dev, info->dma_len, &dma_handle,
+				     GFP_KERNEL);
+	if (!dma_buf)
+		return -ENOMEM;
+
+	segs_off = offsetof(struct hwrm_dbg_coredump_list_output,
+			    total_segments);
+	cmn_req->host_dest_addr = cpu_to_le64(dma_handle);
+	cmn_req->host_buf_len = cpu_to_le32(info->dma_len);
+	mutex_lock(&bp->hwrm_cmd_lock);
+	while (1) {
+		*seq_ptr = cpu_to_le16(seq);
+		rc = _hwrm_send_message(bp, msg, msg_len, HWRM_CMD_TIMEOUT);
+		if (rc)
+			break;
+
+		len = le16_to_cpu(*((__le16 *)(resp + info->data_len_off)));
+		if (!seq &&
+		    cmn_req->req_type == cpu_to_le16(HWRM_DBG_COREDUMP_LIST)) {
+			info->segs = le16_to_cpu(*((__le16 *)(resp +
+							      segs_off)));
+			if (!info->segs) {
+				rc = -EIO;
+				break;
+			}
+
+			info->dest_buf_size = info->segs *
+					sizeof(struct coredump_segment_record);
+			info->dest_buf = kmalloc(info->dest_buf_size,
+						 GFP_KERNEL);
+			if (!info->dest_buf) {
+				rc = -ENOMEM;
+				break;
+			}
+		}
+
+		if (info->dest_buf)
+			memcpy(info->dest_buf + off, dma_buf, len);
+
+		if (cmn_req->req_type ==
+				cpu_to_le16(HWRM_DBG_COREDUMP_RETRIEVE))
+			info->dest_buf_size += len;
+
+		if (!(cmn_resp->flags & HWRM_DBG_CMN_FLAGS_MORE))
+			break;
+
+		seq++;
+		off += len;
+	}
+	mutex_unlock(&bp->hwrm_cmd_lock);
+	dma_free_coherent(&bp->pdev->dev, info->dma_len, dma_buf, dma_handle);
+	return rc;
+}
+
+static int bnxt_hwrm_dbg_coredump_list(struct bnxt *bp,
+				       struct bnxt_coredump *coredump)
+{
+	struct hwrm_dbg_coredump_list_input req = {0};
+	struct bnxt_hwrm_dbg_dma_info info = {NULL};
+	int rc;
+
+	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_DBG_COREDUMP_LIST, -1, -1);
+
+	info.dma_len = COREDUMP_LIST_BUF_LEN;
+	info.seq_off = offsetof(struct hwrm_dbg_coredump_list_input, seq_no);
+	info.data_len_off = offsetof(struct hwrm_dbg_coredump_list_output,
+				     data_len);
+
+	rc = bnxt_hwrm_dbg_dma_data(bp, &req, sizeof(req), &info);
+	if (!rc) {
+		coredump->data = info.dest_buf;
+		coredump->data_size = info.dest_buf_size;
+		coredump->total_segs = info.segs;
+	}
+	return rc;
+}
+
+static int bnxt_hwrm_dbg_coredump_initiate(struct bnxt *bp, u16 component_id,
+					   u16 segment_id)
+{
+	struct hwrm_dbg_coredump_initiate_input req = {0};
+
+	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_DBG_COREDUMP_INITIATE, -1, -1);
+	req.component_id = cpu_to_le16(component_id);
+	req.segment_id = cpu_to_le16(segment_id);
+
+	return hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
+}
+
+static int bnxt_hwrm_dbg_coredump_retrieve(struct bnxt *bp, u16 component_id,
+					   u16 segment_id, u32 *seg_len,
+					   void *buf, u32 offset)
+{
+	struct hwrm_dbg_coredump_retrieve_input req = {0};
+	struct bnxt_hwrm_dbg_dma_info info = {NULL};
+	int rc;
+
+	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_DBG_COREDUMP_RETRIEVE, -1, -1);
+	req.component_id = cpu_to_le16(component_id);
+	req.segment_id = cpu_to_le16(segment_id);
+
+	info.dma_len = COREDUMP_RETRIEVE_BUF_LEN;
+	info.seq_off = offsetof(struct hwrm_dbg_coredump_retrieve_input,
+				seq_no);
+	info.data_len_off = offsetof(struct hwrm_dbg_coredump_retrieve_output,
+				     data_len);
+	if (buf)
+		info.dest_buf = buf + offset;
+
+	rc = bnxt_hwrm_dbg_dma_data(bp, &req, sizeof(req), &info);
+	if (!rc)
+		*seg_len = info.dest_buf_size;
+
+	return rc;
+}
+
+static void
+bnxt_fill_coredump_seg_hdr(struct bnxt *bp,
+			   struct bnxt_coredump_segment_hdr *seg_hdr,
+			   struct coredump_segment_record *seg_rec, u32 seg_len,
+			   int status, u32 duration, u32 instance)
+{
+	memset(seg_hdr, 0, sizeof(*seg_hdr));
+	strcpy(seg_hdr->signature, "sEgM");
+	if (seg_rec) {
+		seg_hdr->component_id = (__force __le32)seg_rec->component_id;
+		seg_hdr->segment_id = (__force __le32)seg_rec->segment_id;
+		seg_hdr->low_version = seg_rec->version_low;
+		seg_hdr->high_version = seg_rec->version_hi;
+	} else {
+		/* For hwrm_ver_get response Component id = 2
+		 * and Segment id = 0
+		 */
+		seg_hdr->component_id = cpu_to_le32(2);
+		seg_hdr->segment_id = 0;
+	}
+	seg_hdr->function_id = cpu_to_le16(bp->pdev->devfn);
+	seg_hdr->length = cpu_to_le32(seg_len);
+	seg_hdr->status = cpu_to_le32(status);
+	seg_hdr->duration = cpu_to_le32(duration);
+	seg_hdr->data_offset = cpu_to_le32(sizeof(*seg_hdr));
+	seg_hdr->instance = cpu_to_le32(instance);
+}
+
+static void
+bnxt_fill_coredump_record(struct bnxt *bp, struct bnxt_coredump_record *record,
+			  time64_t start, s16 start_utc, u16 total_segs,
+			  int status)
+{
+	time64_t end = ktime_get_real_seconds();
+	u32 os_ver_major = 0, os_ver_minor = 0;
+	struct tm tm;
+
+	time64_to_tm(start, 0, &tm);
+	memset(record, 0, sizeof(*record));
+	strcpy(record->signature, "cOrE");
+	record->flags = 0;
+	record->low_version = 0;
+	record->high_version = 1;
+	record->asic_state = 0;
+	strncpy(record->system_name, utsname()->nodename,
+		strlen(utsname()->nodename));
+	record->year = cpu_to_le16(tm.tm_year);
+	record->month = cpu_to_le16(tm.tm_mon);
+	record->day = cpu_to_le16(tm.tm_mday);
+	record->hour = cpu_to_le16(tm.tm_hour);
+	record->minute = cpu_to_le16(tm.tm_min);
+	record->second = cpu_to_le16(tm.tm_sec);
+	record->utc_bias = cpu_to_le16(start_utc);
+	strcpy(record->commandline, "ethtool -w");
+	record->total_segments = cpu_to_le32(total_segs);
+
+	sscanf(utsname()->release, "%u.%u", &os_ver_major, &os_ver_minor);
+	record->os_ver_major = cpu_to_le32(os_ver_major);
+	record->os_ver_minor = cpu_to_le32(os_ver_minor);
+
+	strcpy(record->os_name, utsname()->sysname);
+	time64_to_tm(end, 0, &tm);
+	record->end_year = cpu_to_le16(tm.tm_year + 1900);
+	record->end_month = cpu_to_le16(tm.tm_mon + 1);
+	record->end_day = cpu_to_le16(tm.tm_mday);
+	record->end_hour = cpu_to_le16(tm.tm_hour);
+	record->end_minute = cpu_to_le16(tm.tm_min);
+	record->end_second = cpu_to_le16(tm.tm_sec);
+	record->end_utc_bias = cpu_to_le16(sys_tz.tz_minuteswest * 60);
+	record->asic_id1 = cpu_to_le32(bp->chip_num << 16 |
+				       bp->ver_resp.chip_rev << 8 |
+				       bp->ver_resp.chip_metal);
+	record->asic_id2 = 0;
+	record->coredump_status = cpu_to_le32(status);
+	record->ioctl_low_version = 0;
+	record->ioctl_high_version = 0;
+}
+
+static int bnxt_get_coredump(struct bnxt *bp, void *buf, u32 *dump_len)
+{
+	u32 ver_get_resp_len = sizeof(struct hwrm_ver_get_output);
+	struct coredump_segment_record *seg_record = NULL;
+	u32 offset = 0, seg_hdr_len, seg_record_len;
+	struct bnxt_coredump_segment_hdr seg_hdr;
+	struct bnxt_coredump_record coredump_rec;
+	struct bnxt_coredump coredump = {NULL};
+	time64_t start_time;
+	u16 start_utc;
+	int rc = 0, i;
+
+	start_time = ktime_get_real_seconds();
+	start_utc = sys_tz.tz_minuteswest * 60;
+	seg_hdr_len = sizeof(seg_hdr);
+
+	/* First segment should be hwrm_ver_get response */
+	*dump_len = seg_hdr_len + ver_get_resp_len;
+	if (buf) {
+		bnxt_fill_coredump_seg_hdr(bp, &seg_hdr, NULL, ver_get_resp_len,
+					   0, 0, 0);
+		memcpy(buf + offset, &seg_hdr, seg_hdr_len);
+		offset += seg_hdr_len;
+		memcpy(buf + offset, &bp->ver_resp, ver_get_resp_len);
+		offset += ver_get_resp_len;
+	}
+
+	rc = bnxt_hwrm_dbg_coredump_list(bp, &coredump);
+	if (rc) {
+		netdev_err(bp->dev, "Failed to get coredump segment list\n");
+		goto err;
+	}
+
+	*dump_len += seg_hdr_len * coredump.total_segs;
+
+	seg_record = (struct coredump_segment_record *)coredump.data;
+	seg_record_len = sizeof(*seg_record);
+
+	for (i = 0; i < coredump.total_segs; i++) {
+		u16 comp_id = le16_to_cpu(seg_record->component_id);
+		u16 seg_id = le16_to_cpu(seg_record->segment_id);
+		u32 duration = 0, seg_len = 0;
+		unsigned long start, end;
+
+		start = jiffies;
+
+		rc = bnxt_hwrm_dbg_coredump_initiate(bp, comp_id, seg_id);
+		if (rc) {
+			netdev_err(bp->dev,
+				   "Failed to initiate coredump for seg = %d\n",
+				   seg_record->segment_id);
+			goto next_seg;
+		}
+
+		/* Write segment data into the buffer */
+		rc = bnxt_hwrm_dbg_coredump_retrieve(bp, comp_id, seg_id,
+						     &seg_len, buf,
+						     offset + seg_hdr_len);
+		if (rc)
+			netdev_err(bp->dev,
+				   "Failed to retrieve coredump for seg = %d\n",
+				   seg_record->segment_id);
+
+next_seg:
+		end = jiffies;
+		duration = jiffies_to_msecs(end - start);
+		bnxt_fill_coredump_seg_hdr(bp, &seg_hdr, seg_record, seg_len,
+					   rc, duration, 0);
+
+		if (buf) {
+			/* Write segment header into the buffer */
+			memcpy(buf + offset, &seg_hdr, seg_hdr_len);
+			offset += seg_hdr_len + seg_len;
+		}
+
+		*dump_len += seg_len;
+		seg_record =
+			(struct coredump_segment_record *)((u8 *)seg_record +
+							   seg_record_len);
+	}
+
+err:
+	if (buf) {
+		bnxt_fill_coredump_record(bp, &coredump_rec, start_time,
+					  start_utc, coredump.total_segs + 1,
+					  rc);
+		memcpy(buf + offset, &coredump_rec, sizeof(coredump_rec));
+	}
+	kfree(coredump.data);
+	*dump_len += sizeof(coredump_rec);
+
+	return rc;
+}
+
+static int bnxt_get_dump_flag(struct net_device *dev, struct ethtool_dump *dump)
+{
+	struct bnxt *bp = netdev_priv(dev);
+
+	if (bp->hwrm_spec_code < 0x10801)
+		return -EOPNOTSUPP;
+
+	dump->version = bp->ver_resp.hwrm_fw_maj_8b << 24 |
+			bp->ver_resp.hwrm_fw_min_8b << 16 |
+			bp->ver_resp.hwrm_fw_bld_8b << 8 |
+			bp->ver_resp.hwrm_fw_rsvd_8b;
+
+	return bnxt_get_coredump(bp, NULL, &dump->len);
+}
+
+static int bnxt_get_dump_data(struct net_device *dev, struct ethtool_dump *dump,
+			      void *buf)
+{
+	struct bnxt *bp = netdev_priv(dev);
+
+	if (bp->hwrm_spec_code < 0x10801)
+		return -EOPNOTSUPP;
+
+	memset(buf, 0, dump->len);
+
+	return bnxt_get_coredump(bp, buf, &dump->len);
+}
+
 void bnxt_ethtool_init(struct bnxt *bp)
 {
 	struct hwrm_selftest_qlist_output *resp = bp->hwrm_cmd_resp_addr;
@@ -2788,4 +3119,6 @@ const struct ethtool_ops bnxt_ethtool_ops = {
 	.set_phys_id		= bnxt_set_phys_id,
 	.self_test		= bnxt_self_test,
 	.reset			= bnxt_reset,
+	.get_dump_flag		= bnxt_get_dump_flag,
+	.get_dump_data		= bnxt_get_dump_data,
 };

drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h

@@ -22,6 +22,43 @@ struct bnxt_led_cfg {
 	u8 rsvd;
 };
 
+#define COREDUMP_LIST_BUF_LEN		2048
+#define COREDUMP_RETRIEVE_BUF_LEN	4096
+
+struct bnxt_coredump {
+	void		*data;
+	int		data_size;
+	u16		total_segs;
+};
+
+struct bnxt_hwrm_dbg_dma_info {
+	void *dest_buf;
+	int dest_buf_size;
+	u16 dma_len;
+	u16 seq_off;
+	u16 data_len_off;
+	u16 segs;
+};
+
+struct hwrm_dbg_cmn_input {
+	__le16 req_type;
+	__le16 cmpl_ring;
+	__le16 seq_id;
+	__le16 target_id;
+	__le64 resp_addr;
+	__le64 host_dest_addr;
+	__le32 host_buf_len;
+};
+
+struct hwrm_dbg_cmn_output {
+	__le16 error_code;
+	__le16 req_type;
+	__le16 seq_id;
+	__le16 resp_len;
+	u8 flags;
+	#define HWRM_DBG_CMN_FLAGS_MORE	1
+};
+
 #define BNXT_LED_DFLT_ENA				\
 	(PORT_LED_CFG_REQ_ENABLES_LED0_ID |		\
 	 PORT_LED_CFG_REQ_ENABLES_LED0_STATE |		\