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 accidentally pasted directory path instead of full file path. This resulted in analysis getting stuck.
From a quick inspection seems like failure to open file resulted in rizin/cutter assuming that the valid memory range goes got 0x7fffffffffffff and thus analyses was taking forever to process endless range of nothing.
Trying to do similar thing in plain rizin resulted in ERROR: [d] Cannot open directory. I assume that similar thing happens within rizin, but Cutter doesn't check the error reported by rizin and proceeds with further steps after opening attempt which then acts badly .
To Reproduce
Steps to reproduce the behavior:
Open "open file" window
Type or paste directory path in the path input
Proceed to open and analyze
Observe that analysis gets stuck
Expected behavior
Cutter shows an error.
Additional potential edge cases:
macos .app
non file IO modes, some of which expect special filesystem objects which are not files, and some of them non fiilesytem path strings
It might be better to improve the checking of errors returned by rizin, instead of adding early file/folder check within Cutter. That would potentially allow dealing with other reasons causing file opening to fail, it would also reduce the need to deal with the various edgecases of non file IO modes.
Screenshots
Additional context
The text was updated successfully, but these errors were encountered:
After looking at it more the error checking on rizin side for this might not be as good as I thought. Searching for the error string I saw, it seems like the folder check, that prevented opening folder using plain rizin, happens within rz_main_rizin which cutter of course doesn't use .
Environment information
Describe the bug
I accidentally pasted directory path instead of full file path. This resulted in analysis getting stuck.
From a quick inspection seems like failure to open file resulted in rizin/cutter assuming that the valid memory range goes got
0x7fffffffffffff
and thus analyses was taking forever to process endless range of nothing.Trying to do similar thing in plain rizin resulted in
ERROR: [d] Cannot open directory
. I assume that similar thing happens within rizin, but Cutter doesn't check the error reported by rizin and proceeds with further steps after opening attempt which then acts badly .To Reproduce
Steps to reproduce the behavior:
Expected behavior
Cutter shows an error.
Additional potential edge cases:
It might be better to improve the checking of errors returned by rizin, instead of adding early file/folder check within Cutter. That would potentially allow dealing with other reasons causing file opening to fail, it would also reduce the need to deal with the various edgecases of non file IO modes.
Screenshots
Additional context
The text was updated successfully, but these errors were encountered: