Milan Miladinovic

Results 37 comments of Milan Miladinovic

https://github.com/cloudflare/workerd/compare/035450664ec894cc6d26540e096d7c4f612652fa..a529139f3b54dacedd1b6b77b0b3bebe2babc63d ^ tries to make it a bit easier to track when `Entry`s are added to the lists and when we're transitioning `EntrySyncStatus` states.

https://github.com/cloudflare/workerd/compare/a529139f3b54dacedd1b6b77b0b3bebe2babc63d..4feb0f11e12a52065c9f38936ff165d4af9c622e was a rebase to fix conflicts. I accidentally pulled 1 extra commit so the second force push just gets us back to this feature branch on top of `main`.

> Just a partial review of the first handful of commits for now, sorry. I'll continue reviewing the rest, although have a couple questions that could use a response as...

@DylanRJohnston Can you provide a repro? When do you see the segfault? Also are you on M1 mac as well?

@thecatanon I thought it might be which is why I tried closing out that bug, but I suspect it's unrelated. I don't see how that would have caused a segfault.

> still seeing it in the latest release @thecatanon are you referring to the [latest release of Workerd](https://github.com/cloudflare/workerd/releases/tag/v1.20240117.0) or Miniflare (which I think is still pointing to a prev version...

@codefrau do you have a repro we could take a look at? Are you also on mac?

@nvie thanks for the repro! I see a segfault on linux too (very surprising tbh). It looks like it's segfaulting in the [`ActorContainerRef` destructor](https://github.com/cloudflare/workerd/blob/main/src/workerd/server/server.c%2B%2B#L1668-L1673). Not sure why and will need...

Let me know if this is still a problem after the next release goes out. It looks like we aren't putting one out this week, so I suspect it should...

@nvie I'm running your repro with `[email protected]` and `[email protected]` and I'm not seeing a crash or segfault, can you confirm it's fixed for you too?