plappermaul
plappermaul
1 VPE = 1 THREAD. so we have: - 8380: 1 core, 1 thread: 1 timer + 1 clock - 8390: 1 core, 2 threads: 2 timer + 1 clock...
@oliv3r If there is anything that is really needed (major) in this PR please tell me. Regarding the DT naming conventions please elaborate with @svanheule before I change them multiple...
> Sander and I are contemplating we basically have 2 9300 in the 9310, e.g. 2 CPU with 2 vpe each, so 5 timers? That would be. But maybe we...
> In my [proposed binding](https://lore.kernel.org/linux-devicetree/7c53821386b8f4c1c0ac440f1cd186e09f4a0456.1642369117.git.sander@svanheule.net/), I went for a single reg per node More than discouraged to go on here. So once again I must be honest: - adding this...
New version posted. - DT is now hardware-describing and no longer driver-describing. - NR_CPUS removed
After the SMP fix we should concentrate this. At least we some steps to go to avoid Ollivers out of tree builds. - Integrate this - use this for RTL930x...
v3 pushed - update documentation (6th IRQ on RTL930x unknown and left out for now) - update DTSI - disable new config for 93xx targets
trying to clean up others leftovers, always complaining, never happy, compiling for hours ...
v4 pushed - init timer with 0 instead of 0x0fffffff (seems to work) - change clocksource initialization - update documentation - change DTSI
> I've implemented @oliv3r's suggestion to rename the driver symbol, swapped the order of the compatibles, and dropped some unused MIPS dependencies. Merged to my staging tree, should land in...