On Mon, Feb 10, 2025 at 03:07:24AM -0800, Breno Leitao wrote: > Hello Bagas, > > On Fri, Feb 07, 2025 at 08:50:22AM +0700, Bagas Sanjaya wrote: > > On Thu, Feb 06, 2025 at 03:05:59AM -0800, Breno Leitao wrote: > > > +.. note:: > > > + > > > + If the user has set a conflicting `cpu` key in the userdata > > > dictionary, > > > + both keys will be reported, with the kernel-populated entry appearing > > > after > > > + the user one. For example:: > > > > In that case, shouldn't the kernel autopopulates numbers of the rest of > > CPUs? > > Do you mean listing all the CPUs that are *not* sending the current > message?
Nope. > > Let me come up with an example to try to understand this better. Let's > suppopse I have a machine with 64 cores, and cpu=42 is sending that > current message, then I would see the following on the dictionary: > > cpu=42 > > You are suggesting we send all the other cpus, except 42 in a "key"? Sort of. I mean, on the dictionary, we would see user-defined cpu number on one cpu, and kernel-generated numbers on the rest. Thanks. -- An old man doll... just what I always wanted! - Clara
signature.asc
Description: PGP signature
