From e0eb2d32257ff7848b91bf10f5f150e934b74725 Mon Sep 17 00:00:00 2001 From: Malfurious Date: Mon, 13 Feb 2023 02:29:00 -0500 Subject: lactf 2023 results Signed-off-by: Malfurious --- scores.txt | 2 ++ 1 file changed, 2 insertions(+) diff --git a/scores.txt b/scores.txt index 8eb89b4..835a726 100644 --- a/scores.txt +++ b/scores.txt @@ -21,3 +21,5 @@ Metasploit Community CTF 2021 1300 22 /265 (727) picoCTF 2022 13100 140 /7794 angstromCTF 2022 2111 80 /1179 (1319) X-MAS CTF 2022 1123 66 /816 (1451) + +lactf 2023 5683 98 /980 -- cgit v1.2.3 From 660e93e1f372ab867d8165bf8caab24fc6e4978d Mon Sep 17 00:00:00 2001 From: Malfurious Date: Thu, 16 Feb 2023 14:05:17 -0500 Subject: Writeup LACTF 2023 / One more time pad Signed-off-by: Malfurious --- .../2023/lactf/crypto/one-more-time-pad.txt | 55 ++++++++++++++++++++++ 1 file changed, 55 insertions(+) create mode 100644 docs/writeups/2023/lactf/crypto/one-more-time-pad.txt diff --git a/docs/writeups/2023/lactf/crypto/one-more-time-pad.txt b/docs/writeups/2023/lactf/crypto/one-more-time-pad.txt new file mode 100644 index 0000000..5c190af --- /dev/null +++ b/docs/writeups/2023/lactf/crypto/one-more-time-pad.txt @@ -0,0 +1,55 @@ +I heard the onetime pad is perfectly secure so I used it to send an important +message to a friend, but now a UCLA competition is asking for the key? I threw +that out a long time ago! Can you help me recover it? + + + + +The problem description implies a weakness through key reuse, however we can +easily recover the key because we are given both a plaintext and corresponding +ciphertext for a simple XOR cipher. + +The key is made up of the flag data, which is shorter than the actual message, +so it is repeated using Python itertools.cycle to pad it out. + +``` +from itertools import cycle +pt = b"Long ago, the four nations lived together in harmony ..." + +key = cycle(b"lactf{??????????????}") + +ct = "" + +for i in range(len(pt)): + b = (pt[i] ^ next(key)) + ct += f'{b:02x}' +print("ct =", ct) + +#ct = 200e0d13461a055b4e592b0054543902462d1000042b045f1c407f18581b56194c150c13030f0a5110593606111c3e1f5e305e174571431e +``` + +To get the flag, we ran this algorithm in reverse: + +``` +#!/usr/bin/env python3 + +ct = ( + b"\x20\x0e\x0d\x13\x46\x1a\x05\x5b\x4e\x59\x2b\x00\x54\x54\x39\x02" + b"\x46\x2d\x10\x00\x04\x2b\x04\x5f\x1c\x40\x7f\x18\x58\x1b\x56\x19" + b"\x4c\x15\x0c\x13\x03\x0f\x0a\x51\x10\x59\x36\x06\x11\x1c\x3e\x1f" + b"\x5e\x30\x5e\x17\x45\x71\x43\x1e" ) + +pt = b"Long ago, the four nations lived together in harmony ..." +key = "" + +for i in range(len(pt)): + b = (pt[i] ^ ct[i]) + key += chr(b) + +print(key) +``` + +Because the key was cycled, we see repeated characters in the output, but the +full flag is there. + +lactf{b4by_h1t_m3_0ne_m0r3_t1m3}lactf{b4by_h1t_m3_0ne_m0 -- cgit v1.2.3 From 7458cf4933ccbb545ed7eef01671cd0e7ba0207f Mon Sep 17 00:00:00 2001 From: Malfurious Date: Thu, 16 Feb 2023 14:09:31 -0500 Subject: Writeup LACTF 2023 / A hacker's notes Signed-off-by: Malfurious --- docs/writeups/2023/lactf/misc/a-hackers-notes.txt | 104 ++++++++++++++++++++++ 1 file changed, 104 insertions(+) create mode 100644 docs/writeups/2023/lactf/misc/a-hackers-notes.txt diff --git a/docs/writeups/2023/lactf/misc/a-hackers-notes.txt b/docs/writeups/2023/lactf/misc/a-hackers-notes.txt new file mode 100644 index 0000000..37719da --- /dev/null +++ b/docs/writeups/2023/lactf/misc/a-hackers-notes.txt @@ -0,0 +1,104 @@ +We managed to get ahold of a flash drive which we think contains the decryption +keys for the ransomware that a hacker group tried to deploy on our computer +network! However, it seems like the hacker encrypted the flash drive. We know +that the organization uses passwords in the format hacker### (hacker + 3 digits) +for their disks, but a much stronger encryption password once you login. Can +you try to get access to their notes? + + + + +LUKS decryption +--------------- +We are given a zip file containing an image file of the flash drive. As stated +in the problem description, it is a LUKS encrypted volume. Fortunately, the +passphrase is of a known format and we only have a small search space to guess +the three unknown digits at the end. We wrote a short shell script to +brute force the password guessing. + +``` +#!/bin/bash + +for num in {0..1000}; do + echo "hacker${num}" | sudo cryptsetup open /dev/loop0 hackerdrive + if [ $? -eq 0 ]; then + echo "hacker${num}" + break + fi +done +``` + +/dev/loop0 is a loop device backed by the unzipped image. When run, this script +will eventually print the password 'hacker765' and a mapped device will be +unlocked. + + + + +Recon +----- +With the unlocked volume mounted, we performed a manual search for interesting +files. Among these were: + +/note_to_self.txt +``` +Note to self: delete notes and notes_normalized tables in +.config/joplin/database.sqlite when not in use; allow encrypted sync to restore +notes after +``` + +/.sqlite_history +``` +[...] +pragma secure_delete; +select * from notes_normalized; +delete from notes_normalized; +select * from notes_normalized; +vacuum; +.exit +``` + +/.config/joplin/log.txt +``` +[...] +2023-01-16 01:06:52: "Initializing tables..." +2023-01-16 01:06:52: "KeychainService: checking if keychain supported" +2023-01-16 01:06:52: "KeychainService: could not set test password - keychain support will be disabled" +2023-01-16 01:06:52: e2ee/utils: "Master password is not set - trying to get it from the active master key..." +2023-01-16 01:06:52: handleSyncStartupOperation: "Processing operation:", "0" +2023-01-16 01:06:52: App: "Client ID: 5250b22a001e444bbfc4b332e840dea3" +2023-01-16 01:06:52: "First start: detected locale as en_GB" +2023-01-16 01:06:52: models/Setting: "Skipping all default migrations..." +2023-01-16 01:06:52: e2ee/utils: "Trying to load 0 master keys..." +2023-01-16 01:06:52: e2ee/utils: "Loaded master keys: 0" +[...] +``` + +/.config/joplin/settings.json +``` +{ + "$schema": "https://joplinapp.org/schema/settings.json", + "locale": "en_GB", + "api.token": "5c3c596604f44ea76007d85c35e97d3a3e7307079a3f9a68e91b62a4ab66b8a8ce0da3693d1e0226709e80887b9428f8a79d281fd468c81b0385000fc6f31052", + "markdown.plugin.softbreaks": false, + "markdown.plugin.typographer": false, + "editor": "emacs", + "sync.target": 2, + "sync.2.path": "/home/h4ck3r/encrypted-notes" +} +``` + + + + +Joplin +------ +Joplin's website says: "Joplin is an open source note-taking app. Capture your +thoughts and securely access them from any device." After a while of studying +the files shown above, we conculded that the encrypted notes store was also +included on the hacker's flash drive, at /encrypted-notes/. Futrhermore, that +the Joplin app was configured to be capable of performing a sync automatically. + +We setup a Joplin instance and manually imported the hacker's config prompting +the program to autonomously update with the content of the encrypted notes +database tables. The flag was contained within them. -- cgit v1.2.3 From f3ded28abdff053cf9ac1de8219da4526ff90266 Mon Sep 17 00:00:00 2001 From: Malfurious Date: Fri, 17 Feb 2023 14:14:49 -0500 Subject: Writeup LACTF 2023 / CTFd plus Signed-off-by: Malfurious --- docs/writeups/2023/lactf/rev/ctfd-plus.txt | 125 +++++++++++++++++++++++++++++ 1 file changed, 125 insertions(+) create mode 100644 docs/writeups/2023/lactf/rev/ctfd-plus.txt diff --git a/docs/writeups/2023/lactf/rev/ctfd-plus.txt b/docs/writeups/2023/lactf/rev/ctfd-plus.txt new file mode 100644 index 0000000..cf8df62 --- /dev/null +++ b/docs/writeups/2023/lactf/rev/ctfd-plus.txt @@ -0,0 +1,125 @@ +CTFd is too insufferably slow. You know why? Because they use an SQL database +that's bogged down by JOINs instead of a web scale database like MongoDB. +MongoDB is web scale. You turn it on and it scales right up. You know what's +more web scale though? Nothing. That's right, the throughput of /dev/null is +off the charts. Behold, CTFd+, the first databaseless CTF platform. Can you +get the flag for the only challenge? + + + + +RE +-- +We are given a binary that presents a welcome message along with a dummy PWN +challenge, then asks for the flag for said challenge. As this is a mock CTF +platform, it is a flag checker RE challenge. The content of the embedded PWN +is not relevant: the program just calls `puts` then exits. + +Ghidra shows that the flag we supply is validated character by character +against the result of some computation performed on a static array of data. +(Some markup provided by me). + +``` +undefined8 main(void) +{ + int flag_check_char; + size_t flag_in_len; + long idx; + int *static_data; + char flag_in [256]; + + puts("Welcome to CTFd+!"); + puts( + "So far, we only have one challenge, which is one more than the number of databases we have.\n " + ); + puts("Very Doable Pwn - 500 points, 0 solves"); + puts("Can you help me pwn this program?"); + puts("#include \nint main(void) {\n puts(\"Bye!\");\n return 0;\n}\n"); + puts("Enter the flag:"); + fgets(flag_in,0x100,stdin); + flag_in_len = strcspn(flag_in,"\n"); + idx = 0; + static_data = INT_ARRAY_00104060; + flag_in[flag_in_len] = '\0'; + do { + flag_check_char = flag_gen(static_data[idx]); + if ((char)flag_check_char != flag_in[idx]) { + puts("Incorrect flag."); + return 0; + } + idx = idx + 1; + } while (idx != 0x2f); + puts("You got the flag! Unfortunately we don\'t exactly have a database to store the solve in...") + ; + return 0; +} +``` + +If at any point in the validation a character fails, the whole flag is +considered incorrect. + + + + +Solution +-------- +To extract the expected flag, I ported the static data and the `flag_gen` +computation logic to a standalone C program and simply print out the value of +each character. We know the expected size of the flag string from the bounds +of `main`'s while loop (`while (idx != 0x2f)`). + +``` +#include +#include +#include + +typedef unsigned char byte; + +static const unsigned char DATA[] = { + 0x74, 0x92, 0x7f, 0x9c, 0xe9, 0x0c, 0xbb, 0xc2, 0x29, 0x19, 0x9b, 0x40, + 0xd7, 0x6e, 0x3d, 0xbe, 0x04, 0xe1, 0x83, 0x4f, 0x83, 0xd4, 0x85, 0x91, + 0xaf, 0x70, 0x5d, 0xfd, 0xf1, 0x7f, 0xc4, 0x88, 0x71, 0xfa, 0x78, 0xe6, + 0x0d, 0xdf, 0xcb, 0x72, 0xa7, 0x3d, 0xb6, 0xf4, 0x3d, 0x9e, 0x29, 0x54, + 0xf4, 0x7b, 0x05, 0xaa, 0xa3, 0x4d, 0x14, 0x14, 0x3c, 0x02, 0xc6, 0xe1, + 0x39, 0xb5, 0xb9, 0x74, 0x0f, 0xd8, 0x5f, 0x54, 0x29, 0x73, 0x7a, 0x04, + 0x3f, 0xd9, 0x41, 0xad, 0xd0, 0xbc, 0x16, 0x96, 0x50, 0x62, 0x59, 0x76, + 0x0f, 0xec, 0xa7, 0xaa, 0x2f, 0xf2, 0xb1, 0x21, 0x7e, 0xb3, 0x80, 0x87, + 0x15, 0x14, 0x8d, 0x76, 0x60, 0xad, 0xf3, 0x56, 0x4d, 0x6f, 0x84, 0x2c, + 0x3e, 0x57, 0x38, 0x15, 0x9e, 0x7b, 0x95, 0x6a, 0x70, 0x08, 0x03, 0xaa, + 0xbc, 0xbf, 0xc7, 0x27, 0x4d, 0x88, 0x2e, 0x47, 0x71, 0x09, 0x34, 0xbc, + 0x94, 0xc0, 0x70, 0x95, 0xea, 0x21, 0x55, 0xd6, 0xbe, 0x14, 0x84, 0x86, + 0x8d, 0xec, 0xf7, 0xff, 0xff, 0x65, 0x14, 0xaa, 0xa7, 0x6a, 0xd1, 0x21, + 0x0c, 0xc1, 0x97, 0x84, 0xf7, 0xd2, 0x3a, 0x51, 0xca, 0xbb, 0x11, 0x62, + 0xe5, 0xc8, 0x99, 0x87, 0xbd, 0xfc, 0x37, 0xb5, 0xed, 0x29, 0xcc, 0x44, + 0x5d, 0xd9, 0x8a, 0x40, 0xb1, 0x02, 0x09, 0x2d }; + +int flag_gen(uint param_1) +{ + byte bVar1; + uint uVar2; + int iVar3; + + uVar2 = 0; + iVar3 = 0; + do { + bVar1 = (byte)iVar3 & 0x1f; + iVar3 = iVar3 + 1; + param_1 = (param_1 * param_1 >> bVar1 | param_1 * param_1 << 0x20 - bVar1) * 0x1337 + 0x4201337 + ^ uVar2; + uVar2 = uVar2 + 0x13371337; + } while (iVar3 != 0x20); + return (param_1 >> 8) + (param_1 >> 0x10) + param_1 + (param_1 >> 0x18); +} + +int main(void) { + const int *arr = (void *)DATA; + for (int i = 0; i < 47; i++) { + int c = flag_gen(arr[i]); + printf("%c", (char)c); + } + printf("\n"); + return 0; +} +``` + +lactf{m4yb3_th3r3_1s_s0m3_m3r1t_t0_us1ng_4_db} -- cgit v1.2.3 From 44e95dc49dc6b8afa4bb49c6dd1cb2002866b0ca Mon Sep 17 00:00:00 2001 From: Malfurious Date: Fri, 17 Feb 2023 14:54:13 -0500 Subject: Writeup LACTF 2023 / Switcheroo Signed-off-by: Malfurious --- docs/writeups/2023/lactf/rev/switcheroo.txt | 296 ++++++++++++++++++++++++++++ 1 file changed, 296 insertions(+) create mode 100644 docs/writeups/2023/lactf/rev/switcheroo.txt diff --git a/docs/writeups/2023/lactf/rev/switcheroo.txt b/docs/writeups/2023/lactf/rev/switcheroo.txt new file mode 100644 index 0000000..c7cd4e5 --- /dev/null +++ b/docs/writeups/2023/lactf/rev/switcheroo.txt @@ -0,0 +1,296 @@ +Never fear, the flag shall not appear (in memory)! I've switched it out for a +more secure system! + +Note: The actual flag consists purely of non-whitespace printable characters +(aka ascii values 0x21 through 0x7e). + + + + +RE +-- +The switcheroo binary is compiled from assembly language, rather than C or +anything else higher-level. However, it is still very useful to use Ghidra to +analyze strings and add additional comments and markup in the code listing view. + +Given the above, the file is actually quite small and simple. We have just a +couple boilerplate sections in the ELF, .data contains 3 strings and an array of +64-bit integers, and all of the logic lies sequentially under .text. + +The program is a flag checker. When run, it first asks you to input the flag, +then asserts that the size of input == 64, so we know the length of the expected +flag. To validate the content of the proposed flag, we enter a loop to iterate +over each input character (this is denoted by the `loop_start` label below). + +For each input character, we start by using the value of that character (ascii +as an int) as an index into the 64-bit integer array found in the program. The +value we lookup will encode what position(s) the character index may correctly +occupy in the real flag. We enter an inner loop to decode the 64-bit value. + +At each step, an 8-bit left bit rotation is performed on the value, we then look +at the resulting lowest-order byte in the 64-bit value. The first value we +extract is a count value. + +The code asserts that the value of the current flag character position (as a +byte) exists in the looked-up value within the next `count` byte rotations. +This area is denoted by the `inner_loop_start` label below. If we hit a +matching position value, the outer loop advances. However, if we exhaust +`count` many rotations before that occurs, we increment another register that +records how many failed characters were given, and the outer loop continues +regardless. + +If the outer loop finishes without incrementing the failure register, the +correct flag was given. See the full .text logic here, with some markup from +me: + +``` + undefined __stdcall entry(void) + undefined AL:1 + undefined1 Stack[0x0]:1 local_res0 XREF[1]: 00401040(*) + entry XREF[4]: Entry Point(*), 00400018(*), + 00400088(*), + _elfSectionHeaders::00000090(*) +00401000 48 8d 34 LEA RSI,[s_Give_me_the_flag:_00402000] = "Give me the flag: " + 25 00 20 + 40 00 +00401008 bf 01 00 MOV EDI,0x1 + 00 00 +0040100d ba 12 00 MOV EDX,0x12 + 00 00 +00401012 b8 01 00 MOV EAX,0x1 + 00 00 + write(stdout, "Give me the flag", size) +00401017 0f 05 SYSCALL +00401019 31 ff XOR EDI,EDI +0040101b 48 89 e6 MOV RSI,RSP +0040101e ba 64 00 MOV EDX,0x64 + 00 00 +00401023 31 c0 XOR EAX,EAX + read(stdin, &stack, 100) +00401025 0f 05 SYSCALL +00401027 48 83 f8 40 CMP RAX,0x40 + if read() != 64: fail +0040102b 75 5e JNZ fail +0040102d 4d 31 e4 XOR R12,R12 +00401030 4c 8d 1c LEA R11,[static_data] + 25 3c 20 + 40 00 +00401038 4d 31 d2 XOR R10,R10 + loop_start XREF[1]: 00401071(j) +0040103b 49 39 c2 CMP R10,RAX +0040103e 7d 33 JGE loop_break + R9 = stack[R10] // get current char +00401040 4e 0f b6 MOVZX R9,byte ptr [RSP + R10*0x1]=>local_res0 + 0c 14 + R8 = static_data[R9] // lookup long by current char +00401045 4e 8b 04 MOV R8,qword ptr [static_data + R9*0x8] + cd 3c 20 + 40 00 + rotate left by 8-bits (1-byte) +0040104d 49 c1 c0 08 ROL R8,0x8 + R13 = (char)(static_data[stack[i]] >> 56) +00401051 4d 0f b6 e8 MOVZX R13,R8B +00401055 4d 31 f6 XOR R14,R14 + while R14 < R13 + inner_loop_start XREF[1]: 00401069(j) +00401058 4d 39 ee CMP R14,R13 +0040105b 7d 0e JGE inner_loop_break_bad +0040105d 49 c1 c0 08 ROL R8,0x8 +00401061 45 38 d0 CMP R8B,R10B +00401064 74 08 JZ inner_loop_break_good +00401066 49 ff c6 INC R14 +00401069 eb ed JMP inner_loop_start + inner_loop_break_bad XREF[1]: 0040105b(j) +0040106b 49 ff c4 INC R12 + inner_loop_break_good XREF[1]: 00401064(j) +0040106e 49 ff c2 INC R10 +00401071 eb c8 JMP loop_start + loop_break XREF[1]: 0040103e(j) +00401073 4d 85 e4 TEST R12,R12 +00401076 75 13 JNZ fail +00401078 eb 02 JMP success +0040107a eb ?? EBh +0040107b 00 ?? 00h + success XREF[1]: 00401078(j) +0040107c 48 8d 34 LEA RSI,[s_That_was_the_flag!_0040202a] = "That was the flag!" + 25 2a 20 + 40 00 +00401084 ba 12 00 MOV EDX,0x12 + 00 00 +00401089 eb 0f JMP exit + fail XREF[2]: 0040102b(j), 00401076(j) +0040108b 48 8d 34 LEA RSI,[s_That_was_not_the_flag_:(_00402012] = "That was not the flag :(" + 25 12 20 + 40 00 +00401093 ba 18 00 MOV EDX,0x18 + 00 00 +00401098 eb 00 JMP exit + exit XREF[2]: 00401089(j), 00401098(j) +0040109a bf 01 00 MOV EDI,0x1 + 00 00 +0040109f b8 01 00 MOV EAX,0x1 + 00 00 + write(stdout, "That was [not] the flag", size) +004010a4 0f 05 SYSCALL +004010a6 6a 0a PUSH 0xa +004010a8 48 89 e6 MOV RSI,RSP +004010ab ba 01 00 MOV EDX,0x1 + 00 00 +004010b0 b8 01 00 MOV EAX,0x1 + 00 00 + write(stdout, "\n", 1) +004010b5 0f 05 SYSCALL +004010b7 31 ff XOR EDI,EDI +004010b9 48 c7 c0 MOV RAX,0x3c + 3c 00 00 00 + exit(0) +004010c0 0f 05 SYSCALL +``` + + + + +Solution +-------- +To extract the flag, I extracted the 64-bit integer array to attempt to fully +decode it. With the rules we learned during RE, we know that each entry can be +used to validate whether or not it resides at any given position in the flag +string. + +I wrote a C program to iterate over a flag string (aka position from 0 -> 64) +and print any character value (aka array entry at an ascii index) that _would_ +pass validation. This initially gave some ambiguous results, however when my +search was restricted per the problem description (0x21 -> 0x7e), a unique flag +appears. + +``` +#include + +static const unsigned char A[] = { + 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00, 0x07, 0x7e, 0xfc, 0x3c, 0x78, + 0xfb, 0xf4, 0x04, 0x3e, 0x03, 0x87, 0x30, 0x04, 0x3d, 0x81, 0x70, 0x9d, + 0x4d, 0xa7, 0x3d, 0x09, 0xf8, 0xa1, 0x51, 0x4a, 0xfc, 0x02, 0x8b, 0xa8, + 0xb2, 0x50, 0x49, 0x0c, 0x4b, 0xfd, 0xf0, 0x62, 0x92, 0x1d, 0xd2, 0x02, + 0xa9, 0x46, 0xbe, 0xaa, 0xbe, 0x81, 0xe7, 0x88, 0x94, 0x19, 0xbb, 0x85, + 0x0e, 0x0f, 0x09, 0x77, 0xfb, 0x49, 0x3f, 0x35, 0x24, 0xb8, 0x4e, 0x56, + 0x47, 0x9b, 0x62, 0xeb, 0x9d, 0x7a, 0x0b, 0x7f, 0xd1, 0x18, 0x6e, 0x14, + 0x30, 0xb1, 0x3f, 0x01, 0xed, 0x65, 0xd4, 0x4d, 0x7e, 0xe1, 0x50, 0xe6, + 0xe7, 0xab, 0x3d, 0x80, 0x04, 0x67, 0xd3, 0x81, 0x7a, 0x8c, 0x71, 0x73, + 0xa3, 0x80, 0xce, 0x02, 0x73, 0x48, 0x25, 0x30, 0x09, 0x51, 0x2c, 0x82, + 0x4e, 0x68, 0x70, 0x50, 0x47, 0x37, 0x08, 0x57, 0xb5, 0x14, 0x10, 0x32, + 0xec, 0x34, 0xd7, 0x3b, 0xb3, 0x98, 0xdd, 0x7e, 0xe3, 0x1e, 0x5e, 0x97, + 0xfd, 0x4c, 0x0d, 0xb6, 0xfe, 0x44, 0x81, 0x15, 0x91, 0x42, 0x15, 0xac, + 0x6c, 0x7c, 0x40, 0x88, 0x1b, 0xfc, 0x0d, 0x92, 0x2b, 0x19, 0xae, 0x0f, + 0x80, 0x26, 0x76, 0x7c, 0x6e, 0x67, 0x0c, 0xfc, 0x27, 0xef, 0x94, 0x6d, + 0xd3, 0xb8, 0x50, 0x48, 0xa2, 0x4d, 0x88, 0x0d, 0x72, 0x66, 0x64, 0x08, + 0xdc, 0x48, 0x4f, 0xc1, 0x53, 0x04, 0x7e, 0xac, 0xd8, 0xdf, 0x61, 0x24, + 0xad, 0x68, 0x25, 0x7c, 0x08, 0x1f, 0x91, 0xde, 0x5d, 0x06, 0x3b, 0x47, + 0x71, 0x58, 0xe6, 0x42, 0xd7, 0x1f, 0xed, 0x1e, 0x48, 0x25, 0xb8, 0xc1, + 0xc3, 0xed, 0x81, 0x6c, 0xff, 0x02, 0x18, 0xb6, 0x71, 0x9f, 0x16, 0x03, + 0xe0, 0x3f, 0xae, 0x92, 0x67, 0x39, 0xa7, 0xe4, 0xca, 0xb7, 0x10, 0x6b, + 0xac, 0x08, 0xcf, 0x90, 0xa0, 0x81, 0xf0, 0x31, 0x09, 0xa1, 0x4c, 0x00, + 0x3b, 0x68, 0x89, 0x8f, 0xf1, 0x3d, 0xf5, 0x00, 0x49, 0x8f, 0xde, 0xb1, + 0xbc, 0x42, 0x42, 0x00, 0xca, 0x84, 0xac, 0xce, 0x00, 0x59, 0x7e, 0x00, + 0xa4, 0x0e, 0x11, 0xe8, 0x0f, 0x6f, 0x70, 0x00, 0x4e, 0x50, 0x62, 0x20, + 0xb4, 0x78, 0x22, 0x00, 0x9c, 0x2b, 0xa5, 0x5d, 0xa9, 0x05, 0x2a, 0x00, + 0x49, 0x0e, 0xdf, 0xb9, 0xd0, 0xc5, 0x4d, 0x00, 0xc4, 0xf0, 0x6c, 0xe0, + 0x41, 0x36, 0x6c, 0x00, 0x6e, 0x4a, 0x37, 0x8e, 0xa6, 0x47, 0x79, 0x00, + 0x02, 0x68, 0x33, 0x2b, 0x63, 0xd2, 0xd8, 0x00, 0x2e, 0x02, 0xb5, 0x2f, + 0xc4, 0x56, 0x38, 0x01, 0x18, 0x39, 0x08, 0x1a, 0xdb, 0x14, 0x3b, 0x00, + 0xbb, 0xe1, 0x2f, 0xb1, 0x3a, 0x67, 0xf8, 0x00, 0x24, 0x12, 0xed, 0x14, + 0xb0, 0x08, 0xa1, 0x00, 0x04, 0x3b, 0xef, 0x6a, 0x2e, 0x89, 0xb2, 0x00, + 0xd6, 0x24, 0x00, 0x18, 0xd3, 0x3c, 0x2c, 0x02, 0x34, 0x5f, 0xfc, 0xd0, + 0xaa, 0x11, 0x39, 0x02, 0xe6, 0x4e, 0x4b, 0x71, 0xc9, 0x07, 0x08, 0x02, + 0x27, 0x23, 0x30, 0x09, 0x1c, 0x33, 0x1a, 0x07, 0x45, 0xd2, 0xa9, 0x21, + 0x29, 0x06, 0x18, 0x04, 0x53, 0x92, 0x0f, 0x2b, 0x34, 0x16, 0x1f, 0x05, + 0x7c, 0xe9, 0x0a, 0xda, 0xbf, 0x19, 0x3b, 0x01, 0x1e, 0x2f, 0x19, 0x35, + 0x25, 0x17, 0x12, 0x07, 0x95, 0xd2, 0x10, 0xc9, 0x0b, 0x2a, 0x32, 0x03, + 0x94, 0x63, 0x19, 0xfa, 0xff, 0x70, 0xc0, 0x00, 0xdd, 0x99, 0x49, 0xa8, + 0x72, 0xb0, 0x9a, 0x00, 0xac, 0xaa, 0xa1, 0x74, 0xe1, 0x82, 0x4e, 0x00, + 0xb9, 0x84, 0x95, 0x41, 0x05, 0x52, 0xaa, 0x00, 0xd5, 0xe3, 0x1e, 0x8c, + 0x2b, 0x8a, 0x67, 0x00, 0xd0, 0xe7, 0xd8, 0xc9, 0xac, 0xbc, 0x45, 0x00, + 0xf9, 0xd4, 0xb0, 0x84, 0xc3, 0x19, 0x3a, 0x00, 0x1b, 0xc5, 0x82, 0xa5, + 0x60, 0x19, 0x8b, 0x00, 0xcd, 0x2a, 0x2a, 0x6c, 0x50, 0xc4, 0x4e, 0x00, + 0xe1, 0x86, 0x57, 0x60, 0xc8, 0x73, 0x74, 0x00, 0x7f, 0xb3, 0xba, 0x6b, + 0x53, 0x16, 0x13, 0x01, 0xef, 0xde, 0xdb, 0x7f, 0x24, 0x0d, 0x0f, 0x00, + 0x76, 0x1e, 0xde, 0xa9, 0x08, 0xf6, 0xad, 0x00, 0xe3, 0x82, 0xcf, 0x68, + 0x08, 0x54, 0x9f, 0x00, 0x1d, 0x80, 0x2b, 0x31, 0xfc, 0xf5, 0x14, 0x00, + 0xfe, 0x24, 0x32, 0xc4, 0x14, 0x6f, 0x10, 0x00, 0xc5, 0x37, 0x8b, 0x5d, + 0xad, 0x5b, 0xb8, 0x00, 0x0d, 0x99, 0x8c, 0x84, 0x17, 0xce, 0xed, 0x00, + 0xc4, 0xbc, 0x11, 0xdf, 0x13, 0xdc, 0xcc, 0x00, 0x9b, 0x55, 0xa0, 0x76, + 0x67, 0x2d, 0x0c, 0x02, 0xf7, 0x21, 0x75, 0xea, 0x64, 0x1b, 0x0a, 0x02, + 0x82, 0xbe, 0x90, 0xd9, 0x24, 0x1d, 0x39, 0x00, 0xa8, 0x3c, 0xd7, 0x9d, + 0x71, 0x00, 0x75, 0x00, 0x37, 0x2a, 0xa8, 0xb2, 0xad, 0x2e, 0xef, 0x00, + 0x60, 0x14, 0x96, 0x40, 0x0a, 0xf1, 0x8e, 0x00, 0xa5, 0x67, 0x3f, 0xfe, + 0x2c, 0x9a, 0x33, 0x00, 0x9d, 0x87, 0x65, 0x63, 0x30, 0x71, 0xe0, 0x00, + 0xf0, 0xdc, 0x75, 0x2a, 0x3d, 0x6e, 0x6e, 0x00, 0xf0, 0x73, 0x82, 0xc1, + 0x54, 0x72, 0x2e, 0x01, 0x7f, 0x66, 0x9b, 0xcd, 0xad, 0x08, 0x48, 0x00, + 0x95, 0x63, 0xea, 0x30, 0x12, 0x80, 0x10, 0x01, 0x38, 0xfc, 0xf1, 0xf1, + 0xc7, 0x98, 0xfa, 0x00, 0xd9, 0x7b, 0xfe, 0xe6, 0x0f, 0x40, 0x0d, 0x01, + 0x4d, 0xb9, 0x08, 0x00, 0x15, 0x9e, 0x47, 0x00, 0x95, 0x48, 0x13, 0x38, + 0xcf, 0x80, 0xf9, 0x00, 0xac, 0x5d, 0xcb, 0x7b, 0xfe, 0x0f, 0x2a, 0x00, + 0x99, 0x9b, 0xce, 0x0f, 0x4b, 0x0b, 0x75, 0x00, 0x58, 0x77, 0xb3, 0x79, + 0xfe, 0x6e, 0xbd, 0x00, 0x28, 0x36, 0x15, 0x31, 0x24, 0x0e, 0x20, 0x07, + 0x02, 0xbc, 0x32, 0x40, 0xf1, 0xaa, 0x7c, 0x00, 0xbb, 0x95, 0x83, 0x61, + 0x1f, 0x01, 0x3d, 0x02, 0x9e, 0x6f, 0xfe, 0x34, 0x78, 0x80, 0x53, 0x00, + 0xc8, 0x70, 0x86, 0xe8, 0x3d, 0x30, 0x02, 0x01, 0xf8, 0xd8, 0x08, 0xc7, + 0x3a, 0x0c, 0xf5, 0x00, 0xbc, 0x75, 0x9d, 0xa9, 0xf8, 0x5e, 0xff, 0x00, + 0x7e, 0x51, 0x0e, 0x10, 0x12, 0x3a, 0x04, 0x02, 0x06, 0xd7, 0x49, 0xd0, + 0x19, 0x2d, 0xe8, 0x00, 0x14, 0x42, 0x62, 0xdb, 0x31, 0x26, 0x14, 0x02, + 0x0b, 0xa6, 0xff, 0xb0, 0x63, 0x13, 0xa9, 0x00, 0xf9, 0xb2, 0x7d, 0x13, + 0xa1, 0x8e, 0x85, 0x00, 0xec, 0x18, 0xb0, 0x06, 0xfa, 0x4f, 0x6f, 0x00, + 0x34, 0xf0, 0x80, 0x76, 0xe7, 0x57, 0x00, 0x01, 0x72, 0x72, 0x8c, 0x8b, + 0xb8, 0x5a, 0x5a, 0x00, 0x82, 0x1d, 0x08, 0x34, 0x68, 0x19, 0x1d, 0x01, + 0xac, 0xea, 0xfa, 0x56, 0x07, 0xee, 0x6d, 0x00, 0x78, 0x00, 0x16, 0x0c, + 0xe8, 0x61, 0x16, 0x00, 0x13, 0xc3, 0x58, 0x1a, 0x7a, 0xa6, 0xcc, 0x00, + 0xe8, 0x13, 0x1b, 0xb8, 0x83, 0x36, 0x22, 0x01, 0x8d, 0xd6, 0x45, 0x03, + 0xc2, 0xca, 0x65, 0x00, 0xdc, 0xec, 0x50, 0x60, 0x1a, 0x1c, 0x03, 0x01, + 0x05, 0x19, 0x2c, 0x97, 0xec, 0x04, 0x37, 0x01, 0x25, 0x0b, 0x6c, 0x8b, + 0xe4, 0xc9, 0x4e, 0x00, 0xdd, 0xe1, 0xa5, 0x12, 0x0d, 0xc4, 0x55, 0x00, + 0x0e, 0xe4, 0x20, 0xc4, 0xfc, 0xd4, 0xf0, 0x00, 0xf3, 0x27, 0x8b, 0x99, + 0xc1, 0xb8, 0x9b, 0x00, 0xf6, 0x1a, 0x04, 0x04, 0x5d, 0xc8, 0xca, 0x00, + 0x05, 0xba, 0x8e, 0x89, 0xaf, 0x41, 0x05, 0x02, 0x95, 0x10, 0x67, 0x3d, + 0x26, 0xb0, 0x2c, 0x00, 0x43, 0x36, 0xac, 0xe7, 0xff, 0x42, 0x3e, 0x02, + 0x87, 0x60, 0x09, 0x4e, 0x62, 0xf3, 0x19, 0x00, 0xec, 0x88, 0xa7, 0x5a, + 0x99, 0xd3, 0xf8, 0x26 }; + +static const unsigned long *B = (unsigned long *)A; + +static void rol(long *v) { + long tmp = *v << 8; + tmp |= *v >> 56; + *v = tmp; +} + +int check_char_in_pos(int c, int pos) { + long value = B[c]; + rol(&value); + + for (int count = value & 0xff; count > 0; count--) { + rol(&value); + + if ((value & 0xff) == pos) + return 1; + } + + return 0; +} + +int main(void) { + for (int pos = 0; pos < 0x40; pos++) { + for (int c = 0x21; c <= 0x7e; c++) { + if (check_char_in_pos(c, pos)) { + printf("%c", (char)c); + break; + } + } + } + + printf("\n"); + return 0; +} +``` + +lactf{4223M8LY_5W17Ch_57473M3n75_4r3_7h3_4850LU73_8357_u+1f60a} + +I still don't know where the switch statement was... -- cgit v1.2.3