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

flush instruction cache after performing relocations #59

Open
wants to merge 5 commits into
base: master
Choose a base branch
from
Open

flush instruction cache after performing relocations #59

wants to merge 5 commits into from

Conversation

hMihaiDavid
Copy link

Flush instruction cache to avoid executing stale code after performing relocations.
According to MSDN: "Applications should call FlushInstructionCache if they generate or modify code in memory. The CPU cannot detect the change, and may execute the old code it cached.".
After performing relocation we have modified executable code, so if we don't flush the cache maybe the old code without relocation is executed instead. So far the code has woked without flushing instruction cache but it's better to be safe.

https://msdn.microsoft.com/en-us/library/windows/desktop/ms679350(v=vs.85).aspx

Flush instruction cache to avoid executing stale code after performing relocations.
According to MSDN: "Applications should call FlushInstructionCache if they generate or modify code in memory. The CPU cannot detect the change, and may execute the old code it cached.".
After performing relocation we have modified executable code, so if we don't flush the cache maybe the old code without relocation is executed instead. So far the code has woked without flushing instruction cache but it's better to be safe.

https://msdn.microsoft.com/en-us/library/windows/desktop/ms679350(v=vs.85).aspx
flush instruction cache after performing relocations
@hMihaiDavid
Copy link
Author

Now the instruction cache is flushed only for each page-sized block of image where relocations have been performed. The FlushInstructionCache is now called inside PerformBaseRelocations, and it uses GetCurrentProcess instead of hard-coding (HANDLE)-1

@Elmue
Copy link

Elmue commented Jul 2, 2020

I cannot see ANY sense in calling FlushInstructionCache().
You want to execute this function after relocating the code section.
But this happens long before the code is executed by calling the entry point.
You want to call FlushInstructionCache() when the section has not even been marked as executable with VirtualProtect()

@akasandra
Copy link

akasandra commented Jul 3, 2020

The correct usage of this function may be called by user code, not MemoryModule itself - after multiple loads in a row, for example.

This also may be an undesirable behavior to call additional APIs with unclear functions.

@Elmue
Copy link

Elmue commented Jul 4, 2020

I don't know what you want to say with "additional APIs with unclear functions" ????
However I cannot see the sense of calling this function.
You did not help to clarify this.

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

Successfully merging this pull request may close these issues.

4 participants