Post #2425117
2026-05-08 05:33 UTC
I couldn't sleep, so I got up at 5 am this morning.
To kill some time before work, I wrote a new Python-based proof of concept for Copy Fail based on the original, but one that is easier to read (I don't care about the byte size bragging rights) and which does not push a payload into the page cache of a setuid exec (which is bad). Instead it just overwrites text on a temp text file. It works great!
I got the idea from the copy-fail-c implementation on GitHub.
#CopyFail #Linux #Python
Replies (2)
-
@veronica@mastodon.online 2026-05-08 05:35
I mostly wrote it so I can test all my Linux boxes and VMs for when the fix is in. My main systems have had a proper kernel patch, but some only have the affected module disabled (mainly Ubuntu VMs). My script detects the difference between the two in a readable way.
-
@bortzmeyer@mastodon.gougere.fr 2026-05-08 06:07
@veronica@mastodon.online And now a port for DirtyFrag?