Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Programs packaged with vmp do not result. #17

Closed
maskelihileci opened this issue Nov 20, 2021 · 13 comments
Closed

Programs packaged with vmp do not result. #17

maskelihileci opened this issue Nov 20, 2021 · 13 comments

Comments

@maskelihileci
Copy link

Good day, I was using tiny racer v 1.4. I was getting results without any issues

I updated to version 2.0 today Can analyze unpackaged programs nicely but software packaged with vmp keeps working forever.
I waited for 20 minutes, the pin is working in the background. The program runs in the background, but it never completes. So it never reaches the oep point.

779a61;section: [.vmp1]
82404d;kernel32.LocalAlloc
7972b0;kernel32.GetModuleHandleA
7972b0;kernel32.GetModuleHandleA
8b56e0;kernel32.GetModuleFileNameW
7df6d1;kernel32.LocalAlloc

Windows 10 1607
pin-3.20
i7 6700K
Visual Studio 2019 (v142)

@maskelihileci
Copy link
Author

Current status ; I extended the waiting time a little longer and wanted it to work until the result came. The result has arrived
But its performance is very slow . The source of this slowness is probably one of the newly added features.
So my advice to you is to set a config that will disable this feature and improve performance.

@hasherezade
Copy link
Owner

hi! thank you for reporting.
I suspect what may be slowing it down, but I need to make some more tests. Could you please share the sample on which you was testing? I checked on some VMProtect-protected sample but didn't notice such a drastic performance loss (possibly my sample was packed with different settings), so I would prefer to use the same what you was using, in order to be precise. You can drop it here (packed with a password) or send to my e-mail: hasherezade-at-pm.me.
Also, please share what are the settings of TinyTracer that you use (the INI file).

@hasherezade
Copy link
Owner

hasherezade commented Nov 25, 2021

I reviewed the changes in the TinyTracer, and the feature that I suspected to be causing the slow down turns out to be added before 1.4. So, could you please just make a test for me, to make sure that the slow down is not caused by the changes in the Pin itself?

I added a project TinyTracer.old.vcxproj that allows to build the latest TinyTracer with Pin < 3.19. Could build the version 2.0 of TinyTracer with the old Pin, which you used with 1.4, and check if the same problem occurs?

@hasherezade
Copy link
Owner

@maskelihileci please let me know the results ^

@maskelihileci
Copy link
Author

Sorry for the late reply I was sick
I will now review

@maskelihileci
Copy link
Author

Pin-3.17 used on 2.0 unfortunately still slow
CPU usage is the same 12%
The only difference I've found now is the pincrt.dll module.

Tiny 1.4 = pincrt.dll not loading
Tiny 2.0 = pincrt.dll loading

Maybe it has nothing to do with it, but this is what I see right now.

@maskelihileci
Copy link
Author

If there is a discord server we can contact, we can solve the problem more easily.

@hasherezade
Copy link
Owner

I am not on Discord, but I am on Wire (https://wire.com), my login is the same as here.
If you can just send me the sample that is causing you the problems I will handle it. Without reproducing it it is very hard.

@hasherezade
Copy link
Owner

@maskelihileci - ok, now you can find me on Discord too. I am hasherezade#9405

@maskelihileci
Copy link
Author

I added you as a friend and sent a message

@hasherezade
Copy link
Owner

hasherezade commented Jan 19, 2022

Thank you, I reproduced this problem, and working on the fix.
Those are my benchmarks:

demo.zip ; pass: demo

version 1.9
bench_1 9

version 1.9.5
benchmark_1 9 5

version 2.0
benchmark_2 0

The commit that caused the slowdown to the unacceptable values was this: 8493e3c

benchmark_commit_8493

I reverted those changes and applied on 2.0. You can find it in the branch: faster

However, I am still trying to improve it, so it is not the final fix yet.

hasherezade added a commit that referenced this issue Jan 19, 2022
@hasherezade
Copy link
Owner

I merged it, and run the benchmark, this is my result:

benchmark_2 1

I think this is acceptable, what do you think?

@maskelihileci
Copy link
Author

Problem solved, thanks for your attention.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants