Ryzen CPU HyperThreads break if 100% busy and interrupted to top of Memory

nly wrote: post #3

copx wrote: post #2

I wonder if the Ryzen bugs can really be fixed with a microcode update..

They will probably just issue an update to turn off hyperthreading.

Even with SMT disabled it causes the faults

Re: Ryzen CPU HyperThreads break if 100% busy and interrupted to top of Memory

#5

Post by cma » Fri, Aug 04, 2017, 9:56 PM UTC

> When a cpu-thread stalls in this manner it appears to stall INSIDE the microcode for IRETQ. It doesn't make it to the return pc, and the cpu thread cannot take any IPIs or other hardware interrupts while in this state.

So maybe fixable with a microcode update?

Re: Ryzen CPU HyperThreads break if 100% busy and interrupted to top of Memory

#6

Post by strmpnk » Fri, Aug 04, 2017, 10:09 PM UTC

kartD wrote: post #4

nly wrote: post #3

Earlier quoted context omitted.

They will probably just issue an update to turn off hyperthreading.

Even with SMT disabled it causes the faults

I don't think that is the case for this particular issue. There may be other reported SMT bugs (still some instability) but here it has to be a pair of HTs:

    if one hyperthread is in a cpu-bound loop of any kind
    (can be in user mode), and the other hyperthread is 
    returning from an interrupt via IRETQ ...

Re: Ryzen CPU HyperThreads break if 100% busy and interrupted to top of Memory

#7

Post by lewisj489 » Fri, Aug 04, 2017, 10:09 PM UTC

nly wrote: post #3

copx wrote: post #2

I wonder if the Ryzen bugs can really be fixed with a microcode update..

They will probably just issue an update to turn off hyperthreading.

Ryzen CPU's aren't made by Intel.

Re: Ryzen CPU HyperThreads break if 100% busy and interrupted to top of Memory

#8

Post by loeg » Fri, Aug 04, 2017, 10:17 PM UTC

Note that we still don't really know what's going on; just that this change seems to alleviate the symptoms.

Re: Ryzen CPU HyperThreads break if 100% busy and interrupted to top of Memory

#9

Post by sqeaky » Fri, Aug 04, 2017, 10:31 PM UTC

Does this affect only BSDs then?

I have a Ryzen machine I bought and put into a Jenkins cluster. It does builds on a bunch of VMs occasionally and pegs the CPUs. I have had no issue so far.

Re: Ryzen CPU HyperThreads break if 100% busy and interrupted to top of Memory

#10

Post by 0x0 » Fri, Aug 04, 2017, 10:31 PM UTC

Not the first bug reported for Ryzen. Wasn't there a couple of others too, one with linux locking up and another triggered by the ocaml compiler emitting opcodes refering AH/BH/CH/DH registers in a tight loop?

Edit: Sorry, the ocaml bug was Intel Skylake. It's interesting how so many new CPUs have breaking bugs. Feels like it's been quiet since the original pentium F00F bug and then all of a sudden everyone's new CPUs break.