Threadwell

Recent Notes

DemetriSpanos
Oct 06, 2026
NOTimothyLottes
Oct 06, 2026
bonzajplc
Oct 06, 2026

bmcnett

16 items with this tag.

Aug 26, 2026
bmcnett
Dec 30, 2024
The simplest multi-producer-multi-consumer ring buffer data structure would seem to need one "trit" (ternary bit) of state per buffer entry, for the states "awaiting write", "awaiting read", "busy"
Dec 23, 2023
even among performance minded folks, there's a persistent myth that malloc is "pretty fast" when in fact it's so performance hostile that big game productions ban it at link time, to eliminate the damage it causes
Dec 30, 2024
I just do combinations of separate SPSC ring buffers with a known amount of input and output threads.
Dec 30, 2024
@bmcnett There is another way, one can group items and just do counting per group, release the whole group only when all items in group are ready
Dec 30, 2024
@bmcnett Qubits:
Dec 30, 2023
@bmcnett You need it in tools, it's impossible to produce good enough tools on time without it (or new/delete).
Dec 25, 2023
@bmcnett @forked_franz malloc() actually is pretty fast on most implementation.
Dec 24, 2023
@bmcnett BTW, it's usually not malloc that kills you.
Dec 24, 2023
@bmcnett writing a reasonably performant memory allocator is incredibly eye opening, the amount of work that can happen for 1 allocation is certainly not something you want in hot paths
Dec 23, 2023
Yes, you should use malloc sparingly.
Dec 23, 2023
@bmcnett Any performance minded person should know that any memory allocation is expensive.
Dec 23, 2023
@bmcnett @SheriefFYI Perf enlightment happens when you no longer use dynamic memory allocation, and thus spend your time complaining about page table related bottlenecks at load time.
Dec 23, 2023
@bmcnett How can you ban that at link time?
Dec 23, 2023
@bmcnett Consoles.
Dec 23, 2023
@bmcnett Right at the start of a project, add a few million clocks of time slug to your malloc() and free() calls and never tell anybody else what you did.

Created with Quartz v5.0.0 © 2026

GitHub