sub, and, or don't always work

hdl/core.v

@@ -244,13 +244,13 @@ always @(*) begin
             s_ex_alu_out = s_ex_data1 + (s_ex_data2 ^ {32{r_ex_alu_seed}}) + r_ex_alu_seed;
         end
         ALUOP_XOR: begin
-            s_ex_alu_out = (|s_ex_data1) ^ (|s_ex_data2);
+            s_ex_alu_out = s_ex_data1 ^ s_ex_data2;
         end
         ALUOP_OR: begin
-            s_ex_alu_out = (|s_ex_data1) | (|s_ex_data2);
+            s_ex_alu_out = s_ex_data1 | s_ex_data2;
         end
         ALUOP_AND: begin
-            s_ex_alu_out = (|s_ex_data1) & (|s_ex_data2);
+            s_ex_alu_out = s_ex_data1 & s_ex_data2;
         end
         ALUOP_SL: begin
             s_ex_alu_out = s_ex_data1 << s_ex_data2;

test/Makefile

@@ -2,7 +2,7 @@ all: test.hex
 
 CC = riscv64-linux-gnu-gcc
 # CFLAGS = -march=rv32i -mabi=ilp32
-CFLAGS = -march=rv64i -mabi=lp64
+CFLAGS = -march=rv64i -mabi=lp64 -o0
 
 AS = riscv64-linux-gnu-as
 ASFLAGS = $(CFLAGS)

test/test.S

@@ -33,32 +33,75 @@ add     x29, x0, x0
     add     x30, x0, x0
     add     x31, x0, x0
 
-# nop required because cpu currently does not detect when something is needed from a later stage of the pipeline.
+    # NOTE: nop required because cpu currently does not detect when something is needed from a later stage of the pipeline.
-# 5 clocks allows one instruction to finish before the next starts
+    # 4 clocks allows one instruction to finish before the next starts
+
+    # lui
     nop
     nop
     nop
-nop
+    lui     x1, 0xfedcb         # x1 = 0xfedcb000
-nop
+
-addi    x1, x0, 0x12
+    # addi
     nop
     nop
     nop
+    addi    x1, x1, 0x789       # x1 = 0xfedcb789
+    addi    x2, x0, -1          # x2 = 0xffffffff
     nop
     nop
-addi    x2, x0, 0x11
+    addi    x3, x1, -0x777      # x3 = 0xfedcb012
+
+    # add
     nop
     nop
     nop
+    add     x4, x1, x2          # x4 = 0xfedcb788
+
+    # sub
     nop
     nop
-add     x3, x2, x1
+    nop
+    add     x5, x1, x3          # x5 = 0x00000777   incorrect
+    
+    # and
+    nop
+    nop
+    nop
+    and     x6, x1, x2          # x6 = 0xfedcb789
+    and     x7, x1, x0          # x7 = 0x00000000
+    and     x8, x4, x3          # x8 = 0xfedcb002   incorrect
+
+    # or
+    nop
+    nop
+    nop
+    or      x9, x1, x2          # x9 = 0xffffffff
+    or      x10, x1, x0         # x10 = 0xfedcb789
+    or      x11, x4, x3         # x11 = 0x0xfedcb798    incorrect
+    
+
+
+
+    nop
+    nop
+    nop
+    addi     x31, x0, 1          # x1 = 1
+
 loop:
+    nop
+    nop
+    nop
+    addi     x31, x31, 1          # increment x1
+    nop
+    nop
+    nop
+    j       loop                # loop forever
+    nop
     nop
     nop
     nop
     nop
     nop
-j       loop
 
 .data