From 5b67cb4e2565c5a38cbf18110a0bb4523d0a9dc9 Mon Sep 17 00:00:00 2001
From: Rokas Puzonas <rokas.puz@gmail.com>
Date: Sun, 9 Apr 2023 21:19:30 +0300
Subject: [PATCH] complete challenge homework 2

---
 src/asm.c | 107 ++++++++++++++++++++++++++++++++++++++++++++----------
 1 file changed, 87 insertions(+), 20 deletions(-)

diff --git a/src/asm.c b/src/asm.c
index 38bb483..b2c7387 100644
--- a/src/asm.c
+++ b/src/asm.c
@@ -6,6 +6,8 @@
 #define u32 uint32_t
 #define u16 uint16_t
 #define u8  uint8_t
+#define i16 int16_t
+#define i8  int8_t
 
 #define todo(...) fprintf(stderr, "TODO(%s:%d): ", __FILE__, __LINE__); fprintf(stderr, __VA_ARGS__); abort()
 #define dbg(...) printf("; "); printf(__VA_ARGS__); printf("\n")
@@ -13,7 +15,8 @@
 const char *wide_reg_names[8]     = { "ax", "cx", "dx", "bx", "sp", "bp", "si", "di" };
 const char *non_wide_reg_names[8] = { "al", "cl", "dl", "bl", "ah", "ch", "dh", "bh" };
 
-const char *mod_lookup_names[8] = {
+// eac - Effective address calculation
+const char *eac_names[8] = {
     "bx + si",
     "bx + di",
     "bp + si",
@@ -33,6 +36,37 @@ const char *lookup_reg_name(u8 idx, bool wide) {
     }
 }
 
+void snprintf_mod_with_disp(char *rm_name, size_t size, u8 rm, i16 displacement) {
+    if (displacement < 0) {
+        snprintf(rm_name, size, "[%s - %d]", eac_names[rm], -displacement);
+    } else {
+        snprintf(rm_name, size, "[%s + %d]", eac_names[rm], displacement);
+    }
+}
+
+void get_rm_name(char *rm_name, size_t size, FILE *src, u8 rm, u8 mod, bool wide) {
+    if (mod == 0b11000000) { // Mod = 0b11
+        strcpy(rm_name, lookup_reg_name(rm, wide));
+    } else if (mod == 0b10000000) { // Mod = 0b10
+        i16 displacement = fgetc(src) | (fgetc(src) << 8);
+        snprintf_mod_with_disp(rm_name, size, rm, displacement);
+    } else if (mod == 0b01000000) { // Mod = 0b01
+        i16 displacement = fgetc(src);
+        if (displacement & 0b10000000) {
+            displacement |= (0b11111111 << 8);
+        }
+        snprintf_mod_with_disp(rm_name, size, rm, displacement);
+    } else if (mod == 0b00000000) { // Mod = 0b00
+        if (rm == 0b110) { // Direct address
+            u16 displacement = fgetc(src) | (fgetc(src) << 8);
+            snprintf(rm_name, size, "[%d]", displacement);
+        } else {
+            snprintf(rm_name, size, "[%s]", eac_names[rm]);
+        }
+    }
+}
+
+// Handy reference: Table 4-12. 8086 Instruction Encoding
 void dissassemble(FILE *src, FILE *dst) {
     fprintf(dst, "bits 16\n\n");
     while (true) {
@@ -50,40 +84,73 @@ void dissassemble(FILE *src, FILE *dst) {
             u8 rm  =  byte2 & 0b00000111;
             const char *reg_name = lookup_reg_name(reg, wide);
             char rm_name[24] = { 0 };
-            if (mod == 0b11000000) { // Mod = 0b11
-                strcpy(rm_name, lookup_reg_name(rm, wide));
-            } else if (mod == 0b10000000) { // Mod = 0b10
-                u16 immidiate = fgetc(src) | (fgetc(src) << 8);
-                snprintf(rm_name, sizeof(rm_name), "[%s + %d]", mod_lookup_names[rm], immidiate);
-            } else if (mod == 0b01000000) { // Mod = 0b01
-                u8 immidiate = fgetc(src);
-                snprintf(rm_name, sizeof(rm_name), "[%s + %d]", mod_lookup_names[rm], immidiate);
-            } else if (mod == 0b00000000) { // Mod = 0b00
-                if (rm == 0b110) { // Direct address
-                    todo("direct address");
-                } else {
-                    snprintf(rm_name, sizeof(rm_name), "[%s]", mod_lookup_names[rm]);
-                }
-            }
+            get_rm_name(rm_name, sizeof(rm_name), src, rm, mod, wide);
 
             if (direction) {
                 fprintf(dst, "mov %s, %s\n", reg_name, rm_name);
             } else {
                 fprintf(dst, "mov %s, %s\n", rm_name, reg_name);
             }
+
         // Immediate to register
         } else if ((byte1 & 0b11110000) == 0b10110000) {
             bool wide = (byte1 & 0b1000) >> 3;
             u8 reg = byte1 & 0b111;
 
-            u16 immidiate;
+            u16 immediate;
             if (wide) {
-                immidiate = fgetc(src) | (fgetc(src) << 8);
+                immediate = fgetc(src) | (fgetc(src) << 8);
             } else {
-                immidiate = fgetc(src);
+                immediate = fgetc(src);
             }
 
-            fprintf(dst, "mov %s, %d\n", lookup_reg_name(reg, wide), immidiate);
+            fprintf(dst, "mov %s, %d\n", lookup_reg_name(reg, wide), immediate);
+
+        // Immediate to memory
+        } else if ((byte1 & 0b11111110) == 0b11000110) {
+            u8 byte2 = fgetc(src);
+
+            bool wide = byte1 & 0b1;
+            u8 mod = byte2 & 0b11000000;
+            u8 rm  = byte2 & 0b00000111;
+            char rm_name[24] = { 0 };
+            get_rm_name(rm_name, sizeof(rm_name), src, rm, mod, wide);
+
+            u16 immediate;
+            if (wide) {
+                immediate = fgetc(src) | (fgetc(src) << 8);
+                fprintf(dst, "mov %s, word %d\n", rm_name, immediate);
+            } else {
+                immediate = fgetc(src);
+                fprintf(dst, "mov %s, byte %d\n", rm_name, immediate);
+            }
+
+        // Memory to accumulator
+        } else if ((byte1 & 0b11111110) == 0b10100000) {
+            bool wide = byte1 & 0b1;
+
+            u16 immediate;
+            if (wide) {
+                immediate = fgetc(src) | (fgetc(src) << 8);
+            } else {
+                immediate = fgetc(src);
+            }
+
+            fprintf(dst, "mov ax, [%d]\n", immediate);
+
+        // Accumulator to memory
+        } else if ((byte1 & 0b11111110) == 0b10100010) {
+            bool wide = byte1 & 0b1;
+
+            u16 immediate;
+            if (wide) {
+                immediate = fgetc(src) | (fgetc(src) << 8);
+            } else {
+                immediate = fgetc(src);
+            }
+
+            fprintf(dst, "mov [%d], ax\n", immediate);
+
         } else {
             todo("unhandled byte %d", byte1);
         }