Re: [PATCH 1/2] KVM: PPC: Book3S HV: trace_tlbie must not be called in realmode

Michael Ellerman wrote:

Nicholas Piggin [off-list ref] writes:
=20

quoted

On Sun, 8 Apr 2018 20:17:47 +1000
Balbir Singh [off-list ref] wrote:

quoted

On Fri, Apr 6, 2018 at 3:56 AM, Nicholas Piggin [off-list ref] wro=

te:

quoted

quoted

quoted

This crashes with a "Bad real address for load" attempting to load
from the vmalloc region in realmode (faulting address is in DAR).

  Oops: Bad interrupt in KVM entry/exit code, sig: 6 [#1]
  LE SMP NR_CPUS=3D2048 NUMA PowerNV
  CPU: 53 PID: 6582 Comm: qemu-system-ppc Not tainted 4.16.0-01530-g4=

3d1859f0994

quoted

quoted

quoted

  NIP:  c0000000000155ac LR: c0000000000c2430 CTR: c000000000015580
  REGS: c000000fff76dd80 TRAP: 0200   Not tainted  (4.16.0-01530-g43d=

1859f0994)

quoted

quoted

quoted

  MSR:  9000000000201003 <SF,HV,ME,RI,LE>  CR: 48082222  XER: 0000000=

0

quoted

quoted

quoted

  CFAR: 0000000102900ef0 DAR: d00017fffd941a28 DSISR: 00000040 SOFTE:=

 3

quoted

quoted

quoted

  NIP [c0000000000155ac] perf_trace_tlbie+0x2c/0x1a0
  LR [c0000000000c2430] do_tlbies+0x230/0x2f0

I suspect the reason is the per-cpu data is not in the linear chunk.
This could be restored if that was able to be fixed, but for now,
just remove the tracepoints. =20

=20
Could you share the stack trace as well? I've not observed this in my t=

esting.

quoted

I can't seem to find it, I can try reproduce tomorrow. It was coming
from h_remove hcall from the guest. It's 176 logical CPUs.

quoted

May be I don't have as many cpus. I presume your talking about the per =

cpu

quoted

quoted

data offsets for per cpu trace data?

It looked like it was dereferencing virtually mapped per-cpu data, yes.
Probably the perf_events deref.

=20
Naveen has posted a series to (hopefully) fix this, which just missed
the merge window:
=20
  https://patchwork.ozlabs.org/patch/894757/

I'm afraid that won't actually help here :(
That series is specific to the function tracer, while this is using=20
static tracepoints.

We could convert trace_tlbie() to a TRACE_EVENT_CONDITION() and guard it=20
within a check for paca->ftrace_enabled, but that would only be useful=20
if the below callsites can ever be hit outside of KVM guest mode.

- Naveen

=