Any idea why the fuzzer didn't catch the new one? Is it triggerable by the functions the fuzzer tests? (I.e. lack of input mutation or lack of testing the right functions?)
-
-
-
No, I still don't understand how it could trigger the memory leak but not the memory corruption. I promised
@kayseesee I'd investigate and improve the fuzzer in oss-fuzz when I get a chance
- 4 more replies
New conversation -
-
-
No good deed goes unpunished.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
When I was an engineering manager, in a codebase in C with non-standard memory semantics, my rule of thumb is that 1 in every 10 bug fixes would introduce a bug. The lesson is that C is hard. It's hard because it's a portable macro-assembler, and assembler is hard.
-
Sometimes I have a feeling that fixing bug doesn't really *fix* it - it just converts known (high probability) bug into unknown bugs with lower probability to occur. And the final purpose is just to make probability of bugs low enough so that they won't be triggered in production
- 1 more reply
New conversation -
-
-
Making big bugs while fixing little ones. Robots, they're just like us.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
its bugs all the way down!
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Maybe the lesson is C++. Use after free is a lot harder when the object has fallen out of scope
-
At gnutls we now added automatic NULLifying after free(). No more double frees or dangerous use after free, even if the dev makes a mistake.
- 2 more replies
New conversation -
-
-
Thx for working on it !
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.