add get_opcode and its test, fix a couple bugs + semantics things

3 files changed, 142 insertions, 5 deletions

diff --git a/twasm/README.md b/twasm/README.md
index ce6beec..d6d2ec5 100644
--- a/twasm/README.md
+++ b/twasm/README.md
@@ -144,6 +144,20 @@ type metadata for the different types is as follows:
 11b ; 64 bit
 ```
 
+#### `opcodes.by_id`
+
+entries are as follows:
+
+```
++----------+--------+----------+
+| 31    24 | 23  16 | 15     0 |
++----------+--------+----------+
+| reserved | opcode | token ID |
++----------+--------+----------+
+```
+
+note the lack of support for multiple-byte opcodes or multiple opcodes for one token ID; these features will likely be added at some point after the parser accumulates too much jank.
+
 ### token IDs
 
 supported tokens are listed below
diff --git a/twasm/asm/main.asm b/twasm/asm/main.asm
index 1dfad3b..12f54fd 100644
--- a/twasm/asm/main.asm
+++ b/twasm/asm/main.asm
@@ -18,6 +18,7 @@ STACK_ADDR equ 0x00060000 ; address to put the 64-bit stack at
 UNRECOGNISED_TOKEN_ID equ 0xFFFF  ; id of an unrecognised token
 UNRECOGNISED_ID_TYPE equ 0x0F     ; type of an unrecognised id
 UNRECOGNISED_ID_METADATA equ 0xFF ; metadata of an unrecognised id
+UNRECOGNISED_ID_OPCODE equ 0x90   ; opcode of an unrecognised id (NOP)
 
 TEST_LINE_LENGTH equ 80 ; right border of test suite results
 
@@ -81,24 +82,43 @@ assemble:
     je .operator
     jne .continue_operator
 
-  .operator
+  .operator:
     push rsi
     mov rsi, .msg_found_operator
     call print
     pop rsi
 
-  .continue_operator
+    push rdi
+
+    ; di = tte
+    call get_tte_typed_metadata
+    ; al = tte typed metadata
+
+    pop rdi
+
+    and al, 11b ; mask for # operands
+
+    cmp al, 0 ; 0 operands
+    je .operator_0
+    jne .continue_operator0
+
+  .operator_0:
+    jmp .continue_operator
+
+  .continue_operator0:
+
+  .continue_operator:
     cmp ax, 0x02 ; check if it's a register
     je .register
     jne .continue_register
 
-  .register
+  .register:
     push rsi
     mov rsi, .msg_found_register
     call print
     pop rsi
 
-  .continue_register
+  .continue_register:
     pop rax ; incrementer
     pop rdi ; total number of tokens
 
@@ -109,6 +129,8 @@ assemble:
     ret
   .msg_found_operator db "found operator", 0x0A, 0x00
   .msg_found_register db "found register", 0x0A, 0x00
+  .next_register dw UNRECOGNISED_TOKEN_ID
+  .next_next_register dw UNRECOGNISED_TOKEN_ID
 
 ; ------------------------------------------------------------------------------
 ; get_tte_type
@@ -185,6 +207,7 @@ get_tte_typed_metadata:
     ret
   .found:
     mov al, [3 + tokens.by_id + rax * 4]
+    and rax, 0xFF
     ret
 
 ; ------------------------------------------------------------------------------
@@ -199,7 +222,7 @@ get_tte_typed_metadata:
 ; si = token table entry `R/M`
 ;
 ; returned:
-; al = ModR/M byte
+; al = ModR/M byte; the rest of rax is zeroed
 ; ------------------------------------------------------------------------------
 
 get_direct_addressing_ModRM:
@@ -245,6 +268,43 @@ get_direct_addressing_ModRM:
   ret
 
 ; ------------------------------------------------------------------------------
+; get_opcode
+;
+; description:
+; given an operator token, returns its opcode
+;
+; parameters:
+; di = token table entry
+;
+; returned:
+; al = opcode; the rest of rax is zeroed
+; ------------------------------------------------------------------------------
+
+get_opcode:
+  and rdi, 0xFFFF
+  xor eax, eax
+
+  .loop:
+    cmp rax, (opcodes.by_id_end - opcodes.by_id) / 4 ; make sure it's still in range
+    jg .not_found
+
+    mov cx, [opcodes.by_id + rax * 4] ; next entry in opcodes.by_id
+
+    cmp cx, di
+    je .found
+
+    inc rax
+    jmp .loop
+  .not_found:
+    xor eax, eax
+    mov al, UNRECOGNISED_ID_OPCODE
+    ret
+  .found:
+    mov al, [2 + opcodes.by_id + rax * 4]
+    and rax, 0xFF ; mask
+    ret
+
+; ------------------------------------------------------------------------------
 ; tokenising
 ; ------------------------------------------------------------------------------
 
@@ -961,6 +1021,25 @@ tokens:
     db 0x00   ; # operands
   .by_id_end:
 
+opcodes:
+  .by_id:
+    dw 0x0053 ; xor
+    db 0x33
+    db 0x00   ; reserved
+
+    dw 0x0054 ; inc
+    db 0xFF
+    db 0x00   ; reserved
+
+    dw 0x0056 ; mov
+    db 0x8B
+    db 0x00   ; reserved
+
+    dw 0x004F ; hlt
+    db 0xF4
+    db 0x00   ; reserved
+  .by_id_end:
+
 msg_welcome db "Welcome to Twasm", 0x0A, 0x00
 msg_halt db "halted.", 0x0A, 0x00
 
diff --git a/twasm/asm/tests.asm b/twasm/asm/tests.asm
index 8480554..22d626a 100644
--- a/twasm/asm/tests.asm
+++ b/twasm/asm/tests.asm
@@ -37,6 +37,9 @@ run_tests:
   call clear_test_arena
   call test_get_direct_addressing_ModRM
 
+  call clear_test_arena
+  call test_get_opcode
+
   ret
   .msg db "running test suite...", 0x0A, 0x00
 
@@ -513,6 +516,47 @@ test_get_direct_addressing_ModRM:
     ret
   .msg db "test_get_direct_addressing_ModRM...", 0x00
 
+; ------------------------------------------------------------------------------
+; test_get_opcode
+;
+; description:
+; tests get_opcode described functionality
+; ------------------------------------------------------------------------------
+
+test_get_opcode:
+  mov rsi, .msg
+  call print
+
+  mov di, 0x0053 ; xor
+  call get_opcode
+  cmp al, 0x33
+  jne .fail
+
+  mov di, 0x0054  ; inc
+  call get_opcode
+  cmp al, 0xFF
+  jne .fail
+
+  mov di, 0x004F  ; hlt
+  call get_opcode
+  cmp al, 0xF4
+  jne .fail
+
+  mov di, 0x0003  ; rdx (not an operator)
+  call get_opcode
+  cmp al, UNRECOGNISED_ID_OPCODE
+  jne .fail
+
+  .pass:
+    mov rsi, msg_pass
+    call print
+    ret
+  .fail:
+    mov rsi, msg_fail
+    call print
+    ret
+  .msg db "test_get_opcode...", 0x00
+
 msg_pass:
   db 0x0A
   times (TEST_LINE_LENGTH + .start - .end) db " ", ; right align