Gack. Now the question is, how much breaks if you implement the spec only (QUERY_STRING has the query params)?
And indeed it has nothing to do with threads. It just means the internal state of stdio (or anything not AS-safe) will be inconsistent after longjmp out of it from a signal handler.
-
-
I don't understand the argument here. If we're providing a safe gets(), we're necessarily free to write down our own refinement of POSIX (which we're implementing). That can include more permissive AS-safety rules. Clients targeting the general POSIX gets() were already hosed.
-
Nah, my point was you can aftermarket set up a choke collar for gets. I didn't say it was a good idea, just not impossible. :)
-
Exactly.
@landley was talking about mitigation of gets() badness on the application side, not the implementation side. -
On the implementation side, the only mitigation is to have a memory-safe C implementation (highly nontrivial) so that it necessarily traps when gets() overflows.
-
Ah yes, all agreed. A memory-safe C implementation is what I'm working on. :-) (It promises to suck less than the others, too.)https://github.com/stephenrkell/libcrunch …
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.