Potential memory corruption after many concurrent requests? #21
This issue is really hard to reproduce, but every once in a while, I can get Telodendria to segfault when I spam it with many requests. I don't have any more details than that, because I was never able to figure out why it does this or under what specific circumstances. Clearly there's a bug somewhere, I just haven't had the time to locate it yet. And maybe it's fixed by now, I'm not sure. I'm just throwing this issue out there to see if anyone can reproduce it or wants to try to tackle it.
What worked the best for me in reproducing the issue was opening the browser and going to
/_matrix/client/v3/auth/m.login.registration_token/fallback/web?session=x, entering a bogus registration token, and then holding down the enter key, forcing the browser to send many, many requests to Telodnedria. Since there's no rate limiting, it tries to handle them as quickly as they come in. Eventually, after a few minutes (or maybe less if you get lucky), it will crash, but never in the same place.
Maybe rate limiting, as recommended by the spec, will fix this, but we should still eventually figure out the root cause.
No due date set.
No dependencies set.
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?