bnxt_en: Improve firmware fatal error shutdown sequence.
authorMichael Chan <michael.chan@broadcom.com>
Mon, 25 Jan 2021 07:08:19 +0000 (02:08 -0500)
committerJakub Kicinski <kuba@kernel.org>
Tue, 26 Jan 2021 03:20:04 +0000 (19:20 -0800)
In the event of a fatal firmware error, firmware will notify the host
and then it will proceed to do core reset when it sees that all functions
have disabled Bus Master.  To prevent Master Aborts and other hard
errors, we need to quiesce all activities in addition to disabling Bus
Master before the chip goes into core reset.

Reviewed-by: Edwin Peer <edwin.peer@broadcom.com>
Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
drivers/net/ethernet/broadcom/bnxt/bnxt.c

index 80dab4e..e7abb3b 100644 (file)
@@ -10905,11 +10905,18 @@ static void bnxt_rx_ring_reset(struct bnxt *bp)
 static void bnxt_fw_reset_close(struct bnxt *bp)
 {
        bnxt_ulp_stop(bp);
-       /* When firmware is fatal state, disable PCI device to prevent
-        * any potential bad DMAs before freeing kernel memory.
+       /* When firmware is in fatal state, quiesce device and disable
+        * bus master to prevent any potential bad DMAs before freeing
+        * kernel memory.
         */
-       if (test_bit(BNXT_STATE_FW_FATAL_COND, &bp->state))
+       if (test_bit(BNXT_STATE_FW_FATAL_COND, &bp->state)) {
+               bnxt_tx_disable(bp);
+               bnxt_disable_napi(bp);
+               bnxt_disable_int_sync(bp);
+               bnxt_free_irq(bp);
+               bnxt_clear_int_mode(bp);
                pci_disable_device(bp->pdev);
+       }
        __bnxt_close_nic(bp, true, false);
        bnxt_clear_int_mode(bp);
        bnxt_hwrm_func_drv_unrgtr(bp);