bnxt_en: improve firmware timeout messaging

While it has always been possible to infer that an HWRM command was
abandoned due to an unhealthy firmware status by the shortened timeout
reported, this change improves the log messaging to account for this
case explicitly. In the interests of further clarity, the firmware
status is now also reported in these new messages.

v2: Remove inline keyword for hwrm_wait_must_abort() in .c file.

Reviewed-by: Andy Gospodarek <[email protected]>
Signed-off-by: Edwin Peer <[email protected]>
Signed-off-by: Michael Chan <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

Author: Edwin Peer

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

@@ -7694,19 +7694,6 @@ static void __bnxt_map_fw_health_reg(struct bnxt *bp, u32 reg)
 					 BNXT_FW_HEALTH_WIN_MAP_OFF);
 }
 
-bool bnxt_is_fw_healthy(struct bnxt *bp)
-{
-	if (bp->fw_health && bp->fw_health->status_reliable) {
-		u32 fw_status;
-
-		fw_status = bnxt_fw_health_readl(bp, BNXT_FW_HEALTH_REG);
-		if (fw_status && !BNXT_FW_IS_HEALTHY(fw_status))
-			return false;
-	}
-
-	return true;
-}
-
 static void bnxt_inv_fw_health_reg(struct bnxt *bp)
 {
 	struct bnxt_fw_health *fw_health = bp->fw_health;

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

@@ -2305,7 +2305,6 @@ int bnxt_cancel_reservations(struct bnxt *bp, bool fw_reset);
 int bnxt_hwrm_alloc_wol_fltr(struct bnxt *bp);
 int bnxt_hwrm_free_wol_fltr(struct bnxt *bp);
 int bnxt_hwrm_func_resc_qcaps(struct bnxt *bp, bool all);
-bool bnxt_is_fw_healthy(struct bnxt *bp);
 int bnxt_hwrm_fw_set_time(struct bnxt *);
 int bnxt_open_nic(struct bnxt *, bool, bool);
 int bnxt_half_open_nic(struct bnxt *bp);

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

@@ -444,6 +444,18 @@ static void hwrm_req_dbg(struct bnxt *bp, struct input *req)
 			netdev_err((bp)->dev, fmt, __VA_ARGS__);       \
 	} while (0)
 
+static bool hwrm_wait_must_abort(struct bnxt *bp, u32 req_type, u32 *fw_status)
+{
+	if (req_type == HWRM_VER_GET)
+		return false;
+
+	if (!bp->fw_health || !bp->fw_health->status_reliable)
+		return false;
+
+	*fw_status = bnxt_fw_health_readl(bp, BNXT_FW_HEALTH_REG);
+	return *fw_status && !BNXT_FW_IS_HEALTHY(*fw_status);
+}
+
 static int __hwrm_send(struct bnxt *bp, struct bnxt_hwrm_ctx *ctx)
 {
 	u32 doorbell_offset = BNXT_GRCPF_REG_CHIMP_COMM_TRIGGER;
@@ -455,8 +467,8 @@ static int __hwrm_send(struct bnxt *bp, struct bnxt_hwrm_ctx *ctx)
 	unsigned int i, timeout, tmo_count;
 	u32 *data = (u32 *)ctx->req;
 	u32 msg_len = ctx->req_len;
+	u32 req_type, sts;
 	int rc = -EBUSY;
-	u32 req_type;
 	u16 len = 0;
 	u8 *valid;
 
@@ -556,8 +568,11 @@ static int __hwrm_send(struct bnxt *bp, struct bnxt_hwrm_ctx *ctx)
 				usleep_range(HWRM_SHORT_MIN_TIMEOUT,
 					     HWRM_SHORT_MAX_TIMEOUT);
 			} else {
-				if (HWRM_WAIT_MUST_ABORT(bp, ctx))
-					break;
+				if (hwrm_wait_must_abort(bp, req_type, &sts)) {
+					hwrm_err(bp, ctx, "Resp cmpl intr abandoning msg: 0x%x due to firmware status: 0x%x\n",
+						 req_type, sts);
+					goto exit;
+				}
 				usleep_range(HWRM_MIN_TIMEOUT,
 					     HWRM_MAX_TIMEOUT);
 			}
@@ -608,15 +623,19 @@ static int __hwrm_send(struct bnxt *bp, struct bnxt_hwrm_ctx *ctx)
 				usleep_range(HWRM_SHORT_MIN_TIMEOUT,
 					     HWRM_SHORT_MAX_TIMEOUT);
 			} else {
-				if (HWRM_WAIT_MUST_ABORT(bp, ctx))
-					goto timeout_abort;
+				if (hwrm_wait_must_abort(bp, req_type, &sts)) {
+					hwrm_err(bp, ctx, "Abandoning msg {0x%x 0x%x} len: %d due to firmware status: 0x%x\n",
+						 req_type,
+						 le16_to_cpu(ctx->req->seq_id),
+						 len, sts);
+					goto exit;
+				}
 				usleep_range(HWRM_MIN_TIMEOUT,
 					     HWRM_MAX_TIMEOUT);
 			}
 		}
 
 		if (i >= tmo_count) {
-timeout_abort:
 			hwrm_err(bp, ctx, "Error (timeout: %u) msg {0x%x 0x%x} len:%d\n",
 				 hwrm_total_timeout(i), req_type,
 				 le16_to_cpu(ctx->req->seq_id), len);

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

@@ -82,10 +82,6 @@ void hwrm_update_token(struct bnxt *bp, u16 seq, enum bnxt_hwrm_wait_state s);
 #define HWRM_MIN_TIMEOUT		25
 #define HWRM_MAX_TIMEOUT		40
 
-#define HWRM_WAIT_MUST_ABORT(bp, ctx)					\
-	(le16_to_cpu((ctx)->req->req_type) != HWRM_VER_GET &&		\
-	 !bnxt_is_fw_healthy(bp))
-
 static inline unsigned int hwrm_total_timeout(unsigned int n)
 {
 	return n <= HWRM_SHORT_TIMEOUT_COUNTER ? n * HWRM_SHORT_MIN_TIMEOUT :