Hacker News Reader

Comment by naasking | Hacker News Reader

> There are zero GC pauses. Unless you claim that a C++ alloc/feee call is “garbage collection”.

Alloc/free can introduce arbitrary pauses last I checked, so yes, there are pauses. Any time doing book keeping for resources rather than running your code counts as GC time.

View on HN

> Any time doing book keeping for resources rather than running your code counts as GC time.

Perhaps a nitpick: memory management time, yes, but not GC time.

alloc/free is manual memory management, not garbage collection.

If you're using alloc/free in your GC, which is what was being implied, then that counts as GC time.

On any OS which is not hard realtime, there could be arbitrary pauses with any syscall. This is just nitpicking.

Nitpicking: arbitrary pauses can occur even without syscalls, when the OS preempts the program.

More nitpicking: on x86-64, SMI interrupts can cause arbitrary pauses even without any software control involved. Hard realtime on x86-64 is not possible.

More nitpicking: Your computer might turn off, cosmic rays might blow fuck up your RAM/CPU, Capital G God could reach down and pause the system, there's a universal quantum pause every 5.391247 × 10^-44 so that the universe can reboot, etc etc etc

Orrrrr, GC pause just means pauses caused by the GC as part of its implementation's work to manage memory.

This is not nitpicking. The OP was implying they were using alloc/free in their GC. Those calls can be doing all sorts of book keeping behind the scenes, including cascading destructor calls which could free a chain of objects the size of live memory. Suffice it to say, that would pause.

Reply on news.ycombinator.com