On Wed, May 07, 2025 at 03:35:50PM +0100, Jonathan Cameron wrote: > On Wed, 7 May 2025 12:55:00 +0100 > Sudeep Holla <sudeep.holla@xxxxxxx> wrote: > [...] > > > > Indeed and also we should have private resources like L1 cache described > > after the initial 20 bytes of the node. So I am bit worried if this will > > just hide other problems while it may solve this problem by looks of it. > > This example doesn't look like a proper PPTT matching real systems. > > > > Assuming I'm understanding the bug correctly... > > SMT systems will hit this. There will typically be no private resources > for a thread as the L1I/D shared by multiple threads (which are processor > nodes IIRC). Note we are trying to improve the cache description in QEMU > at the moment as it would definitely be better to present caches in PPTT, > but that isn't the main issue here. > Indeed, I just replied in the other thread that I clearly missed SMT. -- Regards, Sudeep
