Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated: 1 00:00:00.05 --> 00:00:02.08 - [Narrator] In the early days of microcomputers, 2 00:00:02.08 --> 00:00:05.03 processors such as the Intel 8080 chip 3 00:00:05.03 --> 00:00:07.04 supported 8-bit computing. 4 00:00:07.04 --> 00:00:09.06 Computers have evolved a lot since then, 5 00:00:09.06 --> 00:00:11.02 progressing through 16-bit 6 00:00:11.02 --> 00:00:14.00 and then 32-bit processors has being used, 7 00:00:14.00 --> 00:00:16.09 through to the Intel 486 chip. 8 00:00:16.09 --> 00:00:19.02 The Microsoft Macro Assembler, MASM, 9 00:00:19.02 --> 00:00:21.09 was originally developed as a 32-bit assembler, 10 00:00:21.09 --> 00:00:25.02 and exists today in the Microsoft software development kits, 11 00:00:25.02 --> 00:00:27.02 as ml.exe. 12 00:00:27.02 --> 00:00:33.02 The MASM32 project supports the Microsoft 32-bit, ml.exe. 13 00:00:33.02 --> 00:00:36.08 Intel introduced 64-bit processing in the Pentium chip. 14 00:00:36.08 --> 00:00:40.01 And this is now the standard used in modern computers. 15 00:00:40.01 --> 00:00:42.05 A new version of MASM for 64-bits 16 00:00:42.05 --> 00:00:47.09 is now shipped with the SDKs and is called ml64.exe. 17 00:00:47.09 --> 00:00:51.01 However, much of the MASM code that we come across 18 00:00:51.01 --> 00:00:53.09 is legacy 32-bit. 19 00:00:53.09 --> 00:00:57.05 Registers are the primary data areas used by the processor, 20 00:00:57.05 --> 00:01:00.00 and they reflect the processing size. 21 00:01:00.00 --> 00:01:03.04 So earlier processes worked on 32-bit registers, 22 00:01:03.04 --> 00:01:07.04 but modern systems work on 64-bit registers. 23 00:01:07.04 --> 00:01:11.06 Fortunately, the X86 chips provide backwards compatibility, 24 00:01:11.06 --> 00:01:15.05 and can still use the 32-bit portions of the registers. 25 00:01:15.05 --> 00:01:17.08 The term Word Size is sometimes used to refer 26 00:01:17.08 --> 00:01:21.07 to the processor size, which is nowadays 64-bits. 27 00:01:21.07 --> 00:01:25.00 However, to retain consistency with early terminology, 28 00:01:25.00 --> 00:01:28.08 the Windows API continues to use the term "word" 29 00:01:28.08 --> 00:01:31.01 to be a 16-bit value. 30 00:01:31.01 --> 00:01:33.07 So 32-bits is a double word, 31 00:01:33.07 --> 00:01:37.04 and 64-bits is a quad word. 32 00:01:37.04 --> 00:01:40.00 There are also extended registers available 33 00:01:40.00 --> 00:01:42.09 in some of the more recent X64 chips. 34 00:01:42.09 --> 00:01:45.01 These are known as the XMM registers 35 00:01:45.01 --> 00:01:49.02 and allow operations using 128-bits. 36 00:01:49.02 --> 00:01:52.02 We'll take a look at the XMM registers later on. 37 00:01:52.02 --> 00:01:54.01 There's also YMM registers, 38 00:01:54.01 --> 00:01:58.04 which allow 256-bit operation and beyond. 39 00:01:58.04 --> 00:02:02.08 The 64-bit X86 architecture provides a set of 16 registers, 40 00:02:02.08 --> 00:02:07.01 which can be accessed to 64 bits, 32 bits, 16 bits, 41 00:02:07.01 --> 00:02:11.09 or in some cases 8-bits using the mnemonic shown. 42 00:02:11.09 --> 00:02:13.09 Racks can be used as a general purpose register, 43 00:02:13.09 --> 00:02:16.05 but is also typically used as the register 44 00:02:16.05 --> 00:02:19.01 for the return value from functions. 45 00:02:19.01 --> 00:02:21.09 RSP is used as the stack pointer. 46 00:02:21.09 --> 00:02:26.00 The processor also maintains an RIP or EIP register, 47 00:02:26.00 --> 00:02:29.00 which is the next instruction pointer. 48 00:02:29.00 --> 00:02:32.00 A typical sequence of events when coding an assembler, 49 00:02:32.00 --> 00:02:33.01 is to load a register with 50 00:02:33.01 --> 00:02:35.02 the contents of a memory location, 51 00:02:35.02 --> 00:02:38.05 process it and store it back into memory. 52 00:02:38.05 --> 00:02:39.08 Let's see how we code that 53 00:02:39.08 --> 00:02:42.05 in the 32-bit MASM assembly language. 54 00:02:42.05 --> 00:02:44.02 I've set up a new project 55 00:02:44.02 --> 00:02:47.04 and here we see the skeleton MASM console code. 56 00:02:47.04 --> 00:02:49.08 In the constant section at line two, 57 00:02:49.08 --> 00:02:54.02 I'll set a hexadecimal value of A-B-A-D-D-E-E-D, 58 00:02:54.02 --> 00:02:56.09 for the variable trouble. 59 00:02:56.09 --> 00:03:06.09 Trouble equals 0 A-B-A-D-D-E-E-D, 60 00:03:06.09 --> 00:03:10.08 and H to show that this is a hexadecimal constant. 61 00:03:10.08 --> 00:03:13.09 In the dissection at line eight, let's set up a string, 62 00:03:13.09 --> 00:03:16.06 which is stored as an array of bytes, 63 00:03:16.06 --> 00:03:21.09 a 16-bit word and a double word value. 64 00:03:21.09 --> 00:03:25.02 Buffer DB, 65 00:03:25.02 --> 00:03:37.01 A-B-C-D-E-F-G-H-I-J-K-L-M-N-O-P-Q-R-S-T-U-V-W-X-Y-Z, 66 00:03:37.01 --> 00:03:41.06 bword and the mnemonic for this is dw, 67 00:03:41.06 --> 00:03:46.00 and we'll set it to a value, 25. 68 00:03:46.00 --> 00:03:51.05 And bdouble, and the mnemonic for that is dd. 69 00:03:51.05 --> 00:04:03.06 And we'll set that to a hexadecimal string DEADBEEF. 70 00:04:03.06 --> 00:04:05.07 Now let's add some code starting at line 20 71 00:04:05.07 --> 00:04:08.09 to move these values into the racks register 72 00:04:08.09 --> 00:04:12.00 being careful to ensure we match the register mnemonic 73 00:04:12.00 --> 00:04:15.05 with the data size 74 00:04:15.05 --> 00:04:18.01 move, al 75 00:04:18.01 --> 00:04:19.09 and we put the contents of the first character 76 00:04:19.09 --> 00:04:22.01 in the buffer, 77 00:04:22.01 --> 00:04:22.09 and then we'll follow 78 00:04:22.09 --> 00:04:25.02 that by moving again into al 79 00:04:25.02 --> 00:04:30.05 and replacing the value with the second entry 80 00:04:30.05 --> 00:04:32.04 in the string buffer. 81 00:04:32.04 --> 00:04:40.08 We'll move the word bword into AX 82 00:04:40.08 --> 00:04:46.00 and we'll move into EAX the double word, 83 00:04:46.00 --> 00:04:54.01 bdouble we'll, then move into EAX trouble 84 00:04:54.01 --> 00:04:58.01 and we'll use the move instruction again 85 00:04:58.01 --> 00:05:03.00 to store back into bdouble the value in EAX. 86 00:05:03.00 --> 00:05:08.01 Okay we'll build, build project one dot EXE, 87 00:05:08.01 --> 00:05:10.02 and we can see we have no errors. 88 00:05:10.02 --> 00:05:11.07 Let's see what this looks like. 89 00:05:11.07 --> 00:05:14.04 As we step through it by selecting, 90 00:05:14.04 --> 00:05:17.07 build debug project 1, 91 00:05:17.07 --> 00:05:22.03 we're in the debugger and we'll press run to use a code 92 00:05:22.03 --> 00:05:28.00 and then we'll press F7 to get down to our code. 93 00:05:28.00 --> 00:05:30.05 Okay we can see the code that we wrote, 94 00:05:30.05 --> 00:05:33.09 I'll go to the memory map tab, 95 00:05:33.09 --> 00:05:38.02 and I'll double click the data section 96 00:05:38.02 --> 00:05:42.05 at the bottom left we can see the data values that we set. 97 00:05:42.05 --> 00:05:48.03 Let's step to the first move using F8, 98 00:05:48.03 --> 00:05:55.04 the two instructions at 40101C and 401021 99 00:05:55.04 --> 00:06:02.04 I move AL by pointer data segment 404.004 100 00:06:02.04 --> 00:06:07.05 and move al by pointer data segment 404005, 101 00:06:07.05 --> 00:06:18.00 and the machine code is 0044400-00540400 102 00:06:18.00 --> 00:06:19.03 we can say the bottom left. 103 00:06:19.03 --> 00:06:21.02 The address of the first bit in the buffer 104 00:06:21.02 --> 00:06:24.05 is indeed 404 power 4 105 00:06:24.05 --> 00:06:25.03 and we can see that 106 00:06:25.03 --> 00:06:27.06 the four bits of the machine code address 107 00:06:27.06 --> 00:06:29.04 are in little Indian form 108 00:06:29.04 --> 00:06:33.05 which means they're held in reverse order bit wise. 109 00:06:33.05 --> 00:06:35.06 Okay let's start with first instruction 110 00:06:35.06 --> 00:06:40.04 and watch the value of EAX in the top right panel. 111 00:06:40.04 --> 00:06:44.03 We can see the bottom bit of EAX has changed to 41. 112 00:06:44.03 --> 00:06:46.02 The value from our buffer 113 00:06:46.02 --> 00:06:48.01 let's step again, 114 00:06:48.01 --> 00:06:48.09 And we can see 115 00:06:48.09 --> 00:06:50.09 that the bottom bit is now 42, 116 00:06:50.09 --> 00:06:53.00 the hexadecimal value of B, 117 00:06:53.00 --> 00:06:55.08 the next bit from our character string. 118 00:06:55.08 --> 00:06:59.06 The next instruction has an address 40401E 119 00:06:59.06 --> 00:07:03.03 which we can see at the bottom left contains 19 120 00:07:03.03 --> 00:07:05.07 followed by zero zero. 121 00:07:05.07 --> 00:07:07.02 As we step over this, 122 00:07:07.02 --> 00:07:09.04 we can see the last two bits of EAX 123 00:07:09.04 --> 00:07:12.01 become 0019. 124 00:07:12.01 --> 00:07:15.06 The hexadecimal value for 25, 125 00:07:15.06 --> 00:07:19.00 the next instruction points to 404020 126 00:07:19.00 --> 00:07:25.00 and we can see the four bytes, E-F-B-E-A-D-D-E. 127 00:07:25.00 --> 00:07:31.06 As we step over it, we see EAX turn into DEADBEEF. 128 00:07:31.06 --> 00:07:36.09 We can now load EAX with the constant trouble, 129 00:07:36.09 --> 00:07:41.00 and we can see EX has become a bad deed, 130 00:07:41.00 --> 00:07:44.01 and when we, again, step, we can see the data 131 00:07:44.01 --> 00:07:48.03 at the bottom left at 404020, 132 00:07:48.03 --> 00:07:53.02 has become E-D-D-E-8D-8B. 133 00:07:53.02 --> 00:07:54.05 So we've now load the data 134 00:07:54.05 --> 00:07:57.00 from memory into a 32 bit register, 135 00:07:57.00 --> 00:07:59.02 and also at 16 and eight big forms 136 00:07:59.02 --> 00:08:02.03 loaded a constant value into a register 137 00:08:02.03 --> 00:08:05.00 and stored a register back into memory. 10853