modules/swcore: one more bugfix preventing the transfer FSM to be stuck forever…

modules/swcore: one more bugfix preventing the transfer FSM to be stuck forever in S_DROP on high traffic load This time only one port was getting stuck (transfer FSM stuck in S_DROP) once in a while during the 18-port 100% load snake test. In situations when setting the usecnt was taking some longer time, it may happen that RCV_FSM was requesting force_free and force_free was done before TP_FSM went to S_DROP. In that case I register the mmu_force_free_done_i signal in S_SET_USECNT so that it can be used later in S_DROP (if needed).

modules/swcore: one more bugfix preventing the transfer FSM to be stuck forever…
modules/swcore: one more bugfix preventing the transfer FSM to be stuck forever in S_DROP on high traffic load This time only one port was getting stuck (transfer FSM stuck in S_DROP) once in a while during the 18-port 100% load snake test. In situations when setting the usecnt was taking some longer time, it may happen that RCV_FSM was requesting force_free and force_free was done before TP_FSM went to S_DROP. In that case I register the mmu_force_free_done_i signal in S_SET_USECNT so that it can be used later in S_DROP (if needed).
a68d4d2b · Grzegorz Daniluk · 9199cf32 · a68d4d2b
Commit a68d4d2b authored Jun 28, 2015 by Grzegorz Daniluk
Hide whitespace changes
Inline Side-by-side

Showing with 14 additions and 2 deletions

xswc_input_block.vhd modules/wrsw_swcore/xswc_input_block.vhd +14 -2

No files found.
--- a/modules/wrsw_swcore/xswc_input_block.vhd
+++ b/modules/wrsw_swcore/xswc_input_block.vhd
@@ -534,7 +534,7 @@ architecture syn of xswc_input_block is
  signal tp_drop                     : std_logic;
  signal tp_stuck                    : std_logic;
  signal tp_transfer_valid           : std_logic;
-
+  signal tp_ff_done_reg              :  std_logic;


  -- used by rcv_pck FSM to sync with ll_write. checked by rcv_pck on entering READY
@@ -1727,6 +1727,7 @@ begin
      pta_resource     <= (others => '0');
      pta_hp           <= '0';
      pta_prio         <= (others => '0');
+      tp_ff_done_reg   <= '0';
      --========================================
    else

@@ -1888,8 +1889,17 @@ begin
              end if;
            end if;
          end if;
+          
+          -- in case setting the usecnt takes some longer time, it may happen that
+          -- RCV_FSM requests force_free and force_free will be done before TP_FSM
+          -- goes to S_DROP. In that case we need to register the
+          -- mmu_force_free_done_i signal.
+          if(mmu_force_free_done_i = '1') then
+            tp_ff_done_reg <= '1';
+          end if;
          --===========================================================================================
        when S_TRANSFER =>
+          tp_ff_done_reg <= '0';
          --===========================================================================================
          -- TODO: think about enabling reception of new pck when still waiting for the transfer
          if(pta_transfer_ack_i = '1') then
@@ -1904,6 +1914,7 @@ begin
          --===========================================================================================
        when S_WAIT_WITH_TRANSFER =>  -- waits for ll_write to clear the first page
          --===========================================================================================            
+          tp_ff_done_reg  <= '0';
          if(lw_pckstart_pg_clred = '1') then
            s_transfer_pck   <= S_TRANSFER;
            pta_transfer_pck <= '1';
@@ -1939,7 +1950,8 @@ begin
          -- this is to prevent trashing of the drop process (it happened that force_free was
          -- re-done many times by rcv_pck FSM when transfer_pck FSM stayed in DROP state waiting
          -- for global sync
-          elsif(mmu_force_free_req = '1' and mmu_force_free_done_i = '1') then
+          elsif((mmu_force_free_req = '1' and mmu_force_free_done_i = '1') or tp_ff_done_reg='1') then
+            tp_ff_done_reg <= '0';
            s_transfer_pck <= S_IDLE;  
          -- we got here when RCV FSM did not use any page (memory was full), so we don't wait
          -- for  force_free_done, because it will never happen.