Pinwhell

Results 17 comments of Pinwhell

Same Happend for me, analizign an arm32 lib, thumb mode

![image](https://user-images.githubusercontent.com/60289470/203178555-e3d39e87-2e60-4bbb-b1b1-8c88733cd4ce.png)

this wasnt happending in old versions

![image](https://user-images.githubusercontent.com/60289470/218264036-2bc1fbdc-e899-4ea7-b351-ed3b4372ed38.png) Bug Manages to disappear when this code is commented, also the address space base do changes without leaving any second effects ( at least visually ), also, notice, that...

100% Understood those challenges and limitations we are currently facing

i see good, suggestion, i will add a case for that, also, the reason why it consume so much cycles, its because the type of overhead from data structures potentially,...

Regarding your comment > You are using SIMD to try and compare multiple bytes at a given position (acting like memcmp), but 99% of the time, even just the first...

[Duration Bench](https://github.com/pinwhell/findpattern-bench) in the duration aspect of TBS searching is insanely fast, locally it outperforms all the other libraries as shown next: ``` FindPattern benchmark Page size: 4096, allocating 64...

> Instead, you could check if any of those 16 bytes match a specific byte of the pattern. That's like memchr ```c++ inline bool CompareWithMask(const UByte* chunk1, const UByte* chunk2,...

> I think you need to clear the cache in runOne, since it does multiple iterations of each pattern you are correct in pointing that out. it looks like it...