update : 05.09.2024 : This "scratchpad" repo is theoretically closed. I won't be posting anything more here. These are just a few topics that interest me. Probably end of 2025 I'll be back to posting if there's something interesting finally worth sharing and developing. TIME TO GO TO NEXT LEVEL ...in programming skills also.
update : 07.09.2024 : About "fix typo on README.md and inside code comments" - I apologize for my low quality and many of a typo mistakes. After some time I feel like I could have put more effort into it and spell checked it before posting the code, but I won't be correcting every comment anymore because it's unnecessary. It's about the code, and the demos. I think the comment captures the general gist or context of what is going on here. It is sufficiently readable and understandable. Besides, it is a "scratchpad" for me. Not a public repo. But like I said, I feel like I could have put more care and effort into it.
# IMPORTANT # update : 11.09.2024 : I do not intend to copy the work of others. Steal someone's work. I DO NOT DO THIS HERE. Links to repositories and materials which I put here is for build memory map for my learning path about programming. For me this is very useful stuff. This is for me only to build knowledge and make progress.
update : 14.09.2024 : still open, but I stop adding new topics here. This is only for educational reason.
update : 18.09.2024 : added last few notes to repos for myself in README. At the final I added example from ammo.js webgl_demo_test_ray to machine learning directory. This is base for further learning.
update : 24.09.2024 : TODO in 2025 to get more serious tasks with ML. (1) language translator similar to translate.google (2) minimalist project similar to whisper to encode speech from wideo/music to text - more advanced is translator i.e. en -> pl, en -> ua, en -> ru, ru -> pl, ua -> pl. I need some mix between deutch, english, russian, ukrainian (3) run GPT-2 124M (karpathy/llm.c#481), 774M (karpathy/llm.c#580) and 1.6B (karpathy/llm.c#677) model on 8 x A100 and get some expiriences from that. First I need run 124M model. This is a minimum target. For this purpose I don't need 8xA100, but, the case is to learn MPI and stuff like that to learn more advanced parallel computing. But this is not important right now. I need start somewhere. Model with 124M parameters is the first target to achieve. A lot of ready stuff is here https://www.tensorflow.org/tutorials/audio/simple_audio?hl=pl and https://pytorch.org/tutorials/ but... you have to take it in your hands and process it... I don't know if it will be possible to achieve it, till the end of 2025, but keep in mind. I need learn some basics, because I want. Because after so many years of doing what others want from me, I want to finally start doing what gives me pleasure and satisfaction. That's it.
update : 25.09.2024 : note in Embedding_RPI4_ . Keep in mind only. Maybe one day I'll come back to this, I'd like to learn more about it.
update : 28.09.2024 : When I added Win32/web_renderer/ repo I think, At the end of 2025 this all things it need to be DONE! Everything needs to be done step by step and probably thrown into a small system emulator.
Some time ago, looking at these types of projects, I thought it was a waste of time, but I need to put it together, because it's been cluttering my head for ~10 years. I just need exercises like this to get it over with. Then I can do something else with my head free from all this... So, probably in 2025 I'll work it out and at the end I'll upload a new repo where everything will be accumulated. Maybe some small documentation will be created for the future. (...) Or I'll just implement it for Linux and Windows. I'll think about it, but I have to put it together. Do these exercises.
update : 4.11.2024 : I uploaded 3 another demos to win32 repo... But when I look at it, it doesn't make sense for me to upload it any more. This scratchpad was supposed to be used to start doing something towards learning certain topics that interest me. But when I look at it now 1) it has terribly low quality and many errors and language typos 2) It wastes a lot of space on github in such a form as those .gif animations that I uploaded and a lot of junk that I treated as a notebook - better for this case create own website or just a folder on a disk on my computer and when I refine certain things to the stage that it makes sense and you don't have to compile it as a project in MSVC but via cl and you can easily explain the compilation and operation 3) "Scratchpad" was supposed to be used to simply collect a few things in one place to get back to learning programming and topics around IT. And this task was fulfilled! (...) I'm looking at this scratchpad/Win32/Multi Windows Manager/ repo now and there are a lot of similar things like hooking window procedures, MouseProc, KeyboardProc and using DefWindowProc, SetWindowsHookEx or like the part from main.cpp where it's SetWindowLongPtr(hChild, GWLP_WNDPROC, (LONG_PTR)ProcessDialogProc2); - it just needs more explanation of how it works from the low level than just copy and paste to compile and run simple demo. Because if you start a new topic, there are a lot of questions, new problems, coding rules, common mistakes, etc. and that is just as important in educational path. And recording this and the conclusions "what did I learn from this".
So. Ok, I stop here! Today I understood that it is better to create a folder on the disk on a local computer, finish demos and then possibly upload everything when it will have educational value for me that I will be able to come back to after years and it will make me actually make progress thanks to this "work".
Update 11-12-2024 - I added WinPE, tianocore EDK2 and ASL/AML Intel quick intro. At first I threw in some pieces of code to do something as a warm-up. At the end of 2024 I think it's time to change the approach to uploading ready, finished work. It was mainly planning what I have to do now, to remind myself of the basics. There has to be some purpose to it. It can't be learning for the sake of learning, but I have to push myself because this is not trivial. Typical, of what I was used to so far.. Such things are not done in public, in fact I despise such challenges if someone does it in public, but in a way it can help. But only up to a point. Because some things are better done quietly. After a while, this type of "challenge", which was originally intended to serve as a journal, begins to take time, and instead of spending time studying and improving your skills in what you do, you focus more on what you publish. For example. Will learning to play the guitar and posting it on YT make you better guitar player like Jimi Hendrix, Slash, etc.? Nope. No, it doesn't. It's just practice and you either have the talent for it or you don't. This is just an example to have a point of reference.(...) But since I already have this repo on github, I need to "summarize it somehow". Because, willingly or unwillingly, I started publishing certain things and thoughts.
Update 23-12-2024 - OK, last word in '24. Something inside tells me that I would rather focus on it now, as long as there is an opportunity for free learning and development myself. I just prefer to focus on it for a year and in the end I'll find out whether I know something and can do it and it makes sense or not. And when I look at, for example, where I can then train Machine Learning, I chose well to create a "playground" on what I have already done (3d board folder). I will grasp the basics faster knowing what I want from the application. It's just about training and pushing myself, and building a Virtual Machine seems like a good challenge for this because it touches on general issues as a whole when it comes to computers (from the software development side, not from the hardware and chip building side). And in a year we will find out whether it is good or nonsense. It is that simple...
That's it for 2024 !!!
NtQueryInformationProcess_run_example - dump.png
64 bit - dla 32 bitowego procesu offsety są inne i nie wczyta adresu PEB a co z tym idzie reszty odczytywanej przez ReadProcessMemory.
Dodaj ten kod po lini 85, czyli po printf("%zu \n", sizeof(PVOID));
Te struktury i pola bitowe jak IMAGE_DOS_SIGNATURE są w winnt.h
IMAGE_DOS_HEADER *idh = (IMAGE_DOS_HEADER*)hdr;
printf("e_magic: %p %p %d %d %p\n", idh, image, IMAGE_DOS_SIGNATURE, idh->e_magic, idh->e_lfanew);
IMAGE_NT_HEADERS32 *inh32 = (IMAGE_NT_HEADERS32*)((BYTE*)hdr + idh->e_lfanew);
printf("%d %d \n", inh32->Signature, IMAGE_NT_SIGNATURE);
Zwróci coś takiego:
e_magic: 000000B6B3A0E0F0 00007FF6B9400000 23117 23117 00000000000000F8
17744 17744
Czyli pobrane pola są takie same jak IMAGE_DOS_SIGNATURE i IMAGE_NT_SIGNATURE. W ten sposób można dostawać się do pól tych struktur.
http://pinvoke.net/default.aspx/Structures.IMAGE_DOS_HEADER
Ten offset 0xf8 z idh->e_lfanew jest ten sam który obliczyłem "ręcznie" tylko mi wyszło 247.
https://learn.microsoft.com/en-us/windows/win32/api/winnt/ns-winnt-image_optional_header32
IMAGE_NT_HEADERS32 *inh32 = (IMAGE_NT_HEADERS32*)((BYTE*)hdr + idh->e_lfanew);
printf("%d %d \n", inh32->Signature, IMAGE_NT_SIGNATURE);
printf("%p %d %p %d\n", inh32->OptionalHeader.AddressOfEntryPoint,
inh32->OptionalHeader.SizeOfImage, inh32->OptionalHeader.ImageBase, inh32->OptionalHeader.SizeOfCode);
0000000000011276 159744 0000000000007FF7 37376
update 23-07-2023 -
NtQueryInformationProcess_run_example
Linie 145-147 - odczytanie linii poleceń. To jest przydatne gdy aplikacja jest wywoływana z argumentami, a tak jest np przy otwarciu pliku w notatniku (nie pustego notatnika, tylko pliku .txt). Wtedy w lini poleceń jest ścieżka do pliku. Tak samo proces msedge.exe, czyli przeglądarka internetowa, w lini poleceń jest sporo informacji do wyciągnięcia.
ReadProcessMemory(h, rupp.CommandLine.Buffer, cmd, rupp.CommandLine.Length - 2, &rb4);
printf("%d %ws\n", rb4, cmd);
Jeśli chciałbym znaleźć aktualnie otwartą kartę w przeglądarce, aktualnie aktywne okno i to co widnieje w "title" to jak to zrobić? Okazuje się że EnumWindows w tej formie wyciągnie te informacje, ale potrzeba moim zdaniem jest jeszcze kolejna funkcja PrintProcessNameAndID która zwraca ściżkę do procesu (ścieżkę uruchomionego programu np. msedge.exe) oraz PID. Dzięki temu można w dalszym kroku powiązać ze sobą te dwa wyniki zwracane przez funkcje i filtrować (wyszukiwać) konkretnego PID, w tym przypadku dla procesu "msedge.exe", jeśi chcę znać aktualną nawzę otwartego okna. To jest tylko mały przykład od czego zacząć.
Gdyby użyć polecenia CMD tasklist /v to nie dostaniemy wystarczajaco informacji o otwartym oknie. Niektóre otwarte strony nawet nie zwrócą tutaj żadnyuch informacji tylko zobaczymy N/A przy wywołaniu tego polecenia. Dlatego trzeba użyć funkcji z winapi żeby dostać się chociaż do szczątkowych informacji o tym jaka jest aktualna nazwa okna / wyswietlana strona w przeglądarce.
Może to jest najlepsza opcja żeby śledzić otwarte okna w systemie Windows. Jest najbardziej oczywista. Oprócz nazwy okna, PID procesu w drugiej lini jest nazwa procesu, żeby miec pewność który dokładnie proces był w danym momencie był zapisywany w logach / które okno było aktualnie na pierwszym planie wyświetlane. Czyli dostajemy w ten sposób nazwy otwartych stron (nie adresy, ale nazwy z title) oraz nazwę i PID procesu.
#define _CRT_SECURE_NO_WARNINGS
#include <string>
#include <iostream>
#include <windows.h>
#include <stdio.h>
#include <psapi.h>
#include <tchar.h>
#define BUF_SIZ 500
FILE* stream_log;
void save_to_file(SYSTEMTIME lt, char *retbuf, int counter)
{
// logs
char log_buf[140] = { 0 };
char log_tmp_buf[100] = { 0 };
strcat(log_buf, "c:\\__bin\\detect_ext_logs\\v2-");
memset(log_tmp_buf, 0, 100);
_itoa(lt.wYear, log_tmp_buf, 10);
strcat(log_buf, log_tmp_buf);
strcat(log_buf, "-");
memset(log_tmp_buf, 0, 100);
_itoa(lt.wMonth, log_tmp_buf, 10);
strcat(log_buf, log_tmp_buf);
strcat(log_buf, "-");
memset(log_tmp_buf, 0, 100);
_itoa(lt.wDay, log_tmp_buf, 10);
strcat(log_buf, log_tmp_buf);
memset(log_tmp_buf, 0, 100);
strcat(log_buf, ".txt");
//fopen_s(&stream, log_buf, "a");
printf("%s \n", log_buf);
fopen_s(&stream_log, log_buf, "a");
fprintf(stream_log, "%d - %02d | %02d:%02d:%02d ", counter, lt.wDayOfWeek, lt.wDay, lt.wMonth, lt.wYear);
fprintf(stream_log, "%02d:%02d:%02d", lt.wHour, lt.wMinute, lt.wSecond);
//fprintf(stream_log, "%s\n", " --logs only.");
fprintf(stream_log, " %s\n", retbuf);
fclose(stream_log);
}
int main()
{
//HANDLE hProcess = OpenProcess(PROCESS_QUERY_LIMITED_INFORMATION, FALSE, 9200);
volatile int i = 0;
// temporary buffer for first get window name
char* buffer = new char[BUF_SIZ];
// the same size - 500
char* retbuf = new char[BUF_SIZ];
char buf_name[BUF_SIZ];
SYSTEMTIME st;
while (1) {
HWND hWnd = GetForegroundWindow();
GetWindowTextW(hWnd, (LPWSTR)buffer, BUF_SIZ);
int len = GetWindowTextLengthA(hWnd);
//printf("size of buffer %d %d\n", sizeof(buffer), len);
int j;
int jj = 0;
for (j = 0; j < len*2; j++) {
if (buffer[j] == 0) continue;
//printf("%c", buffer[j]);
retbuf[jj] = buffer[j];
jj++;
}
retbuf[jj++] = '\0';
GetLocalTime(&st);
DWORD pid = 0;
GetWindowThreadProcessId(hWnd, &pid);
printf("%d [%d] - %d:%d:%d - %s\n", i, pid, st.wHour, st.wMinute, st.wSecond, retbuf);
HANDLE hProcess = OpenProcess(PROCESS_QUERY_INFORMATION, FALSE, pid);
DWORD dump_pid_name = GetProcessImageFileNameA(hProcess, (LPSTR)buf_name, BUF_SIZ);
printf("%d %s \n", dump_pid_name, buf_name);
//save_to_file(st, retbuf, i);
//printf("%d %ws \n", i, buffer);
// inc logs counter
i++;
Sleep(1000);
}
return 0;
}