You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a problem with this code [0], which seems to be designed to make the debugger "continue" if we're not yet in a correct source file. On windows under dbgeng, step_info.current_location.path is
Which are two very different strings. The symptom is that dbgeng doesn't implement "go"; but even if it did, this would still be broken.
Ultimately I don't think we should be trying to resolve the paths that the debugger coughs up into paths on the actual filesystem -- or if we do, we should make that the debugger drivers problem, not generic code.
Aaaannnddd, this turns out to be necessary for things like lldb to not be infinitely stepping through the implementation of printf and the like. So this portion of code does need to be sticking around.
We should probably call os.path.normpath in that code that you highlighted to make sure that when comparing paths we're actually comparing like for like.
I have a problem with this code [0], which seems to be designed to make the debugger "continue" if we're not yet in a correct source file. On windows under dbgeng,
step_info.current_location.path
isAnd
self.context.options.source_files
isWhich are two very different strings. The symptom is that dbgeng doesn't implement "go"; but even if it did, this would still be broken.
Ultimately I don't think we should be trying to resolve the paths that the debugger coughs up into paths on the actual filesystem -- or if we do, we should make that the debugger drivers problem, not generic code.
[0]
dexter/dex/debugger/DebuggerBase.py
Line 158 in 861102e
The text was updated successfully, but these errors were encountered: