AaronBallman wrote: > For publicly-available code, it's not clear to me how much of the burden > should fall on people that identify the problem. I want to do as much of this > work as I can, it's difficult to balance the urgency of providing some > reproducer (it gets hard to push for a fix if we wait a week for a good > reproducers), the quality of reduction, and other work/deadlines. (As > mentioned, the timing was difficult this time as this landed just before a > holiday).
(Not talking about the revert policy, but just about reduced reproducers.) I don't think we have (or want) hard-and-fast rules for how much of the burden falls on folks who find the issue -- we want to encourage people to report issues even if they don't have the time or ability to reduce the problem, but we also don't want to overload the community with issues that need significant reduction work. My suggestion is: initially, provide *anything* that demonstrates the issue so that folks are aware a problem exists, but with an understanding that the harder it is to reproduce (or the larger the reproducer is), the less likely it is the issue will be addressed in a timely manner. So if something is of critical importance to you, then the bar is higher for you to provide as much help as possible in getting to the point we can fix the issue on a timeline that works for you. But it's definitely a balancing act and the community should be willing to pitch in to help with reduction and mitigate fallout as much as possible (as what seems to have happened in this case). https://github.com/llvm/llvm-project/pull/89807 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
