Updates due to updated debug spec #858

silabs-oysteink · 2023-05-22T11:44:07Z

Removed tdata3 and tcontrol
Added hit1, hit0, uncertain and uncertainen to mcontrol6
Updating all matching mcontrol6 hit fields when a debug trigger is taken

Signed-off-by: Oystein Knauserud <Oystein.Knauserud@silabs.com>

- Added hit1, hit0, uncerteain and uncertainen bits - Propagating trigger matches for all implemented triggers down the pipeline. - Updating all matcing mcontrol6 triggers' hit fields when a trigger debug entry happens. Signed-off-by: Oystein Knauserud <Oystein.Knauserud@silabs.com>

-Removed tdata3 and tcontrol -Updated tinfo to add version field Signed-off-by: Oystein Knauserud <Oystein.Knauserud@silabs.com>

Silabs-ArjanB · 2023-05-22T13:38:06Z

rtl/cv32e40x_debug_triggers.sv

-                                2'b00,                 // zero  26:25
-                                3'b000,                // zero, vs, vu, hit 24:22
+                                mcontrol6_uncertain_resolve(csr_wdata_i[MCONTROL_6_UNCERTAIN]), // uncertain  26
+                                mcontrol6_hit1_resolve(csr_wdata_i[MCONTROL_6_HIT1]), // hit1: 25


hit1 and hit0 must form a combined 2-bit WARL field (not two 1-bit WARL fields)

Fixed. Note that due to strange issues with bit-slicing a function return value a separate 2-bit logic was used as return value.

Silabs-ArjanB · 2023-05-22T13:49:36Z

rtl/cv32e40x_debug_triggers.sv

@@ -226,6 +223,18 @@ import cv32e40x_pkg::*;
              tdata1_n[idx] = {TTYPE_DISABLED, 1'b1, {27{1'b0}}};
            end
          end // tdata1_we_i
+
+          // Writes from the controller takes presedence over SW writes


This comment sounds suspicious. How can there ever be both a software write and a HW write? Can you add an assertion that this is impossible and change this comment?

Silabs-ArjanB · 2023-05-22T13:51:34Z

rtl/cv32e40x_ex_stage.sv

@@ -420,7 +421,7 @@ module cv32e40x_ex_stage import cv32e40x_pkg::*;

        // CSR illegal instruction detected in this stage, OR'ing in the status
        ex_wb_pipe_o.illegal_insn   <= id_ex_pipe_i.illegal_insn || csr_is_illegal;
-        ex_wb_pipe_o.trigger_match  <= id_ex_pipe_i.trigger_match;
+        ex_wb_pipe_o.trigger_match  <= {{(32-DBG_NUM_TRIGGERS){1'b0}}, id_ex_pipe_i.trigger_match[DBG_NUM_TRIGGERS-1:0]};


Why this expression?

Just add assertion on all those 32-bit signals that MSBs are always 0

It was a remainder after performing some synthesis checks, removed now.

Silabs-ArjanB · 2023-05-22T13:52:01Z

rtl/cv32e40x_id_stage.sv

@@ -639,7 +640,7 @@ module cv32e40x_id_stage import cv32e40x_pkg::*;
          id_ex_pipe_o.instr                <= if_id_pipe_i.instr;
        end

-        id_ex_pipe_o.trigger_match          <= if_id_pipe_i.trigger_match;
+        id_ex_pipe_o.trigger_match          <= {{(32-DBG_NUM_TRIGGERS){1'b0}}, if_id_pipe_i.trigger_match[DBG_NUM_TRIGGERS-1:0]};


Expression should not be needed (I think)

Same as above

Silabs-ArjanB · 2023-05-22T13:52:25Z

rtl/cv32e40x_if_stage.sv

@@ -431,7 +432,7 @@ module cv32e40x_if_stage import cv32e40x_pkg::*;
        if_id_pipe_o.illegal_c_insn   <= seq_valid ? 1'b0 : illegal_c_insn;

        if_id_pipe_o.priv_lvl         <= prefetch_priv_lvl;
-        if_id_pipe_o.trigger_match    <= trigger_match_i;
+        if_id_pipe_o.trigger_match    <= {{(32-DBG_NUM_TRIGGERS){1'b0}}, trigger_match_i[DBG_NUM_TRIGGERS-1:0]};


Why the expression?

Same as the other two

Silabs-ArjanB · 2023-05-22T13:57:32Z

rtl/cv32e40x_load_store_unit.sv

@@ -938,7 +938,7 @@ module cv32e40x_load_store_unit import cv32e40x_pkg::*;
      assign xif_mem_if.mem_resp.exccode = xif_mpu_err ? (
                                            trans.we ? EXC_CAUSE_STORE_FAULT : EXC_CAUSE_LOAD_FAULT
                                           ) : '0;
-      assign xif_mem_if.mem_resp.dbg     = xif_wpt_match;
+      assign xif_mem_if.mem_resp.dbg     = |xif_wpt_match; // TODO:XIF Is one bit enough?


I think so. This todo can be removed.

Silabs-ArjanB · 2023-05-22T14:00:23Z

rtl/include/cv32e40x_pkg.sv

+    return 1'b0;
+  endfunction
+
+  function automatic logic mcontrol6_hit1_resolve


hit1+hit0 should be combined WARL field

Silabs-ArjanB · 2023-05-22T14:03:24Z

sva/cv32e40x_debug_triggers_sva.sv

@@ -86,6 +89,32 @@ module cv32e40x_debug_triggers_sva
                  $stable(tdata2_q)) // Checking stability of ALL tdata2, not just the one selected
    else `uvm_error("debug_triggers", "tdata2_q changed after set/clear with rs1==0")

+// Check that tdata1.hit0 is set when a mcontrol6 trigger matched when debug is entered
+generate


Also check HIT1.
Also check that the triggers not hitting have their HIT* field unchanged.

rtl/cv32e40x_wpt.sv

Signed-off-by: Oystein Knauserud <Oystein.Knauserud@silabs.com>

silabs-oysteink added 3 commits May 19, 2023 14:59

Removed tdata3 and tcontrol.

15b88c8

Signed-off-by: Oystein Knauserud <Oystein.Knauserud@silabs.com>

Updated yaml file for debug:

cfb686f

-Removed tdata3 and tcontrol -Updated tinfo to add version field Signed-off-by: Oystein Knauserud <Oystein.Knauserud@silabs.com>

silabs-oysteink added the Component:RTL For issues in the RTL (e.g. for files in the rtl directory) label May 22, 2023

silabs-oysteink requested a review from Silabs-ArjanB May 22, 2023 11:44

Silabs-ArjanB requested changes May 22, 2023

View reviewed changes

Updates after review.

f3f3636

Signed-off-by: Oystein Knauserud <Oystein.Knauserud@silabs.com>

silabs-oysteink requested a review from Silabs-ArjanB May 22, 2023 15:55

Silabs-ArjanB approved these changes May 22, 2023

View reviewed changes

Silabs-ArjanB merged commit e675b64 into openhwgroup:master May 22, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Updates due to updated debug spec #858

Updates due to updated debug spec #858

silabs-oysteink commented May 22, 2023

Silabs-ArjanB May 22, 2023

silabs-oysteink May 22, 2023

Silabs-ArjanB May 22, 2023

silabs-oysteink May 22, 2023

Silabs-ArjanB May 22, 2023

Silabs-ArjanB May 22, 2023

silabs-oysteink May 22, 2023

Silabs-ArjanB May 22, 2023

silabs-oysteink May 22, 2023

Silabs-ArjanB May 22, 2023

silabs-oysteink May 22, 2023

Silabs-ArjanB May 22, 2023

silabs-oysteink May 22, 2023

Silabs-ArjanB May 22, 2023

silabs-oysteink May 22, 2023

Silabs-ArjanB May 22, 2023

silabs-oysteink May 22, 2023

Updates due to updated debug spec #858

Updates due to updated debug spec #858

Conversation

silabs-oysteink commented May 22, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment