fix table management instructions' ({l,s}{g,i,l}dt) mem_size

these instructions, it turns out, have fixed operand size based on CPU
execution mode and regardless of prefixes. good to know!

8 files changed, 51 insertions, 27 deletions

diff --git a/CHANGELOG b/CHANGELOG
index 8d88ec7..540a5bf 100644
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -10,6 +10,8 @@
 * push-immediate, pushf, popf, enter, leave, and xlat now all report a correct memory
   access size, fixing the prior behavior of reporting no memory access size at
   all
+* table load/store instructions (lgdt, lidt, lldt, sgdt, sidt, sldt) have correct (mode-dependent)
+  memory access sizes, rather than incorrectly varying on operand-size overrides.
 * 64-bit mode: mov seg-to-reg uses 32-bit GPRs for the destination rather than 16-bit.
   * this is more accurate to the semantic of the instruction, which is why other disassemblers
     report it this way; for register destinations specifically the segment selector is
diff --git a/src/long_mode/mod.rs b/src/long_mode/mod.rs
index 2971313..4288eb9 100644
--- a/src/long_mode/mod.rs
+++ b/src/long_mode/mod.rs
@@ -8243,7 +8243,7 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::SGDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    instruction.mem_size = 10;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 1 {
@@ -8298,7 +8298,7 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::SIDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    instruction.mem_size = 10;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 2 {
@@ -8337,7 +8337,10 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::LGDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    // quoth SDM:
+                    // > In 64-bit mode, the operand size is fixed at 8+2 bytes. The instruction
+                    // > stores an 8-byte base and a 2-byte limit.
+                    instruction.mem_size = 10;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 3 {
@@ -8401,7 +8404,7 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::LIDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    instruction.mem_size = 10;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 4 {
diff --git a/src/protected_mode/mod.rs b/src/protected_mode/mod.rs
index 6701071..89c5dfd 100644
--- a/src/protected_mode/mod.rs
+++ b/src/protected_mode/mod.rs
@@ -8042,7 +8042,7 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::SGDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    instruction.mem_size = 6;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 1 {
@@ -8097,7 +8097,11 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::SIDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    // quoth SDM:
+                    // > In non-64-bit modes, the 16-bit limit field of the register is stored in
+                    // > the low 2 bytes of the memory location and the 32-bit base address is
+                    // > stored in the high 4 bytes.
+                    instruction.mem_size = 6;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 2 {
@@ -8136,7 +8140,10 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::LGDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    // quoth SDM:
+                    // > In legacy or compatibility mode, the destination operand is a 6-byte
+                    // > memory location.
+                    instruction.mem_size = 6;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 3 {
@@ -8200,7 +8207,7 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::LIDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    instruction.mem_size = 6;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 4 {
diff --git a/src/real_mode/mod.rs b/src/real_mode/mod.rs
index 7ac4c2e..559cac3 100644
--- a/src/real_mode/mod.rs
+++ b/src/real_mode/mod.rs
@@ -8087,7 +8087,7 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::SGDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    instruction.mem_size = 6;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 1 {
@@ -8142,7 +8142,11 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::SIDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    // quoth SDM:
+                    // > In non-64-bit modes, the 16-bit limit field of the register is stored in
+                    // > the low 2 bytes of the memory location and the 32-bit base address is
+                    // > stored in the high 4 bytes.
+                    instruction.mem_size = 6;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 2 {
@@ -8181,7 +8185,7 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::LGDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    instruction.mem_size = 6;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 3 {
@@ -8245,7 +8249,7 @@ fn read_operands<
                 } else {
                     instruction.opcode = Opcode::LIDT;
                     instruction.operand_count = 1;
-                    instruction.mem_size = 63;
+                    instruction.mem_size = 6;
                     instruction.operands[0] = read_E(words, instruction, modrm, bank, sink)?;
                 }
             } else if r == 4 {
diff --git a/test/long_mode/mod.rs b/test/long_mode/mod.rs
index dc82cd4..0b6dcc6 100644
--- a/test/long_mode/mod.rs
+++ b/test/long_mode/mod.rs
@@ -955,11 +955,11 @@ fn test_0f01() {
     }
     test_display(&[0x0f, 0x01, 0x38], "invlpg byte [rax]");
     test_display(&[0x0f, 0x01, 0x3f], "invlpg byte [rdi]");
-    test_display(&[0x0f, 0x01, 0x40, 0xff], "sgdt ptr [rax - 0x1]");
-    test_display(&[0x0f, 0x01, 0x41, 0xff], "sgdt ptr [rcx - 0x1]");
-    test_display(&[0x0f, 0x01, 0x49, 0xff], "sidt ptr [rcx - 0x1]");
-    test_display(&[0x0f, 0x01, 0x51, 0xff], "lgdt ptr [rcx - 0x1]");
-    test_display(&[0x0f, 0x01, 0x59, 0xff], "lidt ptr [rcx - 0x1]");
+    test_display(&[0x0f, 0x01, 0x40, 0xff], "sgdt mword [rax - 0x1]");
+    test_display(&[0x0f, 0x01, 0x41, 0xff], "sgdt mword [rcx - 0x1]");
+    test_display(&[0x0f, 0x01, 0x49, 0xff], "sidt mword [rcx - 0x1]");
+    test_display(&[0x0f, 0x01, 0x51, 0xff], "lgdt mword [rcx - 0x1]");
+    test_display(&[0x0f, 0x01, 0x59, 0xff], "lidt mword [rcx - 0x1]");
     test_display(&[0x0f, 0x01, 0x61, 0xff], "smsw word [rcx - 0x1]");
     test_invalid(&[0x0f, 0x01, 0x69, 0xff]);
     test_display(&[0x0f, 0x01, 0x71, 0xff], "lmsw word [rcx - 0x1]");
diff --git a/test/long_mode/operand.rs b/test/long_mode/operand.rs
index 09ac5a2..9f594b5 100644
--- a/test/long_mode/operand.rs
+++ b/test/long_mode/operand.rs
@@ -69,4 +69,12 @@ fn test_implied_memory_width() {
     assert_eq!(mem_size_of(&[0x0f, 0xa8]), Some(8));
     // pop gs
     assert_eq!(mem_size_of(&[0x0f, 0xa9]), Some(8));
+    // lidt
+    assert_eq!(mem_size_of(&[0x0f, 0x01, 0x18]), Some(10));
+    // sidt
+    assert_eq!(mem_size_of(&[0x0f, 0x01, 0x08]), Some(10));
+    // lgdt
+    assert_eq!(mem_size_of(&[0x0f, 0x01, 0x10]), Some(10));
+    // sgdt
+    assert_eq!(mem_size_of(&[0x0f, 0x01, 0x00]), Some(10));
 }
diff --git a/test/protected_mode/mod.rs b/test/protected_mode/mod.rs
index 78a061a..d5533cc 100644
--- a/test/protected_mode/mod.rs
+++ b/test/protected_mode/mod.rs
@@ -811,11 +811,11 @@ fn test_0f01() {
     // Number"
     test_display(&[0x0f, 0x01, 0x38], "invlpg byte [eax]");
     test_display(&[0x0f, 0x01, 0x3f], "invlpg byte [edi]");
-    test_display(&[0x0f, 0x01, 0x40, 0xff], "sgdt ptr [eax - 0x1]");
-    test_display(&[0x0f, 0x01, 0x41, 0xff], "sgdt ptr [ecx - 0x1]");
-    test_display(&[0x0f, 0x01, 0x49, 0xff], "sidt ptr [ecx - 0x1]");
-    test_display(&[0x0f, 0x01, 0x51, 0xff], "lgdt ptr [ecx - 0x1]");
-    test_display(&[0x0f, 0x01, 0x59, 0xff], "lidt ptr [ecx - 0x1]");
+    test_display(&[0x0f, 0x01, 0x40, 0xff], "sgdt far [eax - 0x1]");
+    test_display(&[0x0f, 0x01, 0x41, 0xff], "sgdt far [ecx - 0x1]");
+    test_display(&[0x0f, 0x01, 0x49, 0xff], "sidt far [ecx - 0x1]");
+    test_display(&[0x0f, 0x01, 0x51, 0xff], "lgdt far [ecx - 0x1]");
+    test_display(&[0x0f, 0x01, 0x59, 0xff], "lidt far [ecx - 0x1]");
     test_display(&[0x0f, 0x01, 0x61, 0xff], "smsw word [ecx - 0x1]");
     test_invalid(&[0x0f, 0x01, 0x69, 0xff]);
     test_display(&[0x0f, 0x01, 0x71, 0xff], "lmsw word [ecx - 0x1]");
diff --git a/test/real_mode/mod.rs b/test/real_mode/mod.rs
index 5fb109f..af8bfaf 100644
--- a/test/real_mode/mod.rs
+++ b/test/real_mode/mod.rs
@@ -104,11 +104,11 @@ fn test_real_mode() {
     test_display(&[0x0e], "push cs");
     test_display(&[0x0f, 0x01, 0x38], "invlpg byte [bx + si * 1]");
     test_display(&[0x0f, 0x01, 0x3f], "invlpg byte [bx]");
-    test_display(&[0x0f, 0x01, 0x40, 0xff], "sgdt ptr [bx + si * 1 - 0x1]");
-    test_display(&[0x0f, 0x01, 0x41, 0xff], "sgdt ptr [bx + di * 1 - 0x1]");
-    test_display(&[0x0f, 0x01, 0x49, 0xff], "sidt ptr [bx + di * 1 - 0x1]");
-    test_display(&[0x0f, 0x01, 0x51, 0xff], "lgdt ptr [bx + di * 1 - 0x1]");
-    test_display(&[0x0f, 0x01, 0x59, 0xff], "lidt ptr [bx + di * 1 - 0x1]");
+    test_display(&[0x0f, 0x01, 0x40, 0xff], "sgdt far [bx + si * 1 - 0x1]");
+    test_display(&[0x0f, 0x01, 0x41, 0xff], "sgdt far [bx + di * 1 - 0x1]");
+    test_display(&[0x0f, 0x01, 0x49, 0xff], "sidt far [bx + di * 1 - 0x1]");
+    test_display(&[0x0f, 0x01, 0x51, 0xff], "lgdt far [bx + di * 1 - 0x1]");
+    test_display(&[0x0f, 0x01, 0x59, 0xff], "lidt far [bx + di * 1 - 0x1]");
     test_display(&[0x0f, 0x01, 0x61, 0xff], "smsw word [bx + di * 1 - 0x1]");
     test_display(&[0x0f, 0x01, 0x71, 0xff], "lmsw word [bx + di * 1 - 0x1]");
     test_display(&[0x0f, 0x01, 0x79, 0xff], "invlpg byte [bx + di * 1 - 0x1]");