Mark Mankins

When you take the recording, if you add -complete-rr-snapshot to the panda command line - do you get the same replay errors?

The existing documentation was very helpful in getting me up to speed with PyPANDA. It's super easy to look at the docs and see how panda callbacks such as before_block_exec...

Maybe it's just the tests that require protobuf. Without protobuf some of the tests fail... Traceback (most recent call last): File "file_hook.py", line 5, in from pandare import * AttributeError:...

It would be nice if taint2 would allow one to label a virtual memory address instead of a "RamOffset". More of a panda problem than a PyPanda problem, but this...

PandaPhysicalAddressToRamOffset(). It's defined in common.h: https://github.com/panda-re/panda/blob/dev/panda/include/panda/common.h. See tstringsearch.cpp for an example of how it is used.

This change combined with #1445 introduces an inconsistency in the output of callstack_instr. Running the same recording multiple times can result in different output from callstack_instr...here's an example of what...

> Oh no, I was expecting a difference with output from callstack_instr before/after these changes, but you're saying it's just producing varying output just running repeatedly with these changes? Correct...

I tested with the latest code and I'm still seeing issues where callstack_instr outputs different results for the same replay.

Yes, no issues on dev.

Oh - sorry - didn't realize the recent push was just a rebase - I thought you made some changes - ignore me. :)