[9fans] SMC SYS-5018A-FTN4 lapic weirdness (9front)

[9fans] SMC SYS-5018A-FTN4 lapic weirdness (9front)

[Lyndon Nerenberg]

I have a stack of Supermicro SYS-5018A-FTN4 servers upon which  I'm trying to
spin up 9front.

For the most part they work, but one annoyance is the *endless* stream of

  cpu0: lapicerror: 0x00000080

messages the kernel prints out.  Sometimes these originate from cpu1 as well.
The hardware has eight CPU cores.  I don't think I've seen anything from
cpu>1, but in the blizzard of messages, who knows.

I poked a wee bit inside the kernel source, but I don't have time right now
to chase this.  The hardware runs fine, other than refusing to reboot, which
I put down to the usual BIOS ACPI table stupidity.

For the time being I'm going to put a filter on the kprints, but I'm curious
if this sounds familiar to anyone.

Note this happens with both the 32- and 64-bit kernels.

I don't have a quick way to attach sysinfo(1) to this message, but if somebody
can use that I'll figure something out.

--lyndon

Re: [9fans] SMC SYS-5018A-FTN4 lapic weirdness (9front)

[Rodrigo G. López]

[-- Attachment #1: Type: text/plain, Size: 1174 bytes --]

Hi.

I get the exact same error every time I (in|de)crease the backlight
brightness in the x220.

On Thu, Apr 18, 2019, 3:49 AM Lyndon Nerenberg <lyndon@orthanc.ca> wrote:

> I have a stack of Supermicro SYS-5018A-FTN4 servers upon which  I'm trying
> to
> spin up 9front.
>
> For the most part they work, but one annoyance is the *endless* stream of
>
>   cpu0: lapicerror: 0x00000080
>
> messages the kernel prints out.  Sometimes these originate from cpu1 as
> well.
> The hardware has eight CPU cores.  I don't think I've seen anything from
> cpu>1, but in the blizzard of messages, who knows.
>
> I poked a wee bit inside the kernel source, but I don't have time right now
> to chase this.  The hardware runs fine, other than refusing to reboot,
> which
> I put down to the usual BIOS ACPI table stupidity.
>
> For the time being I'm going to put a filter on the kprints, but I'm
> curious
> if this sounds familiar to anyone.
>
> Note this happens with both the 32- and 64-bit kernels.
>
> I don't have a quick way to attach sysinfo(1) to this message, but if
> somebody
> can use that I'll figure something out.
>
> --lyndon
>
>

[-- Attachment #2: Type: text/html, Size: 1597 bytes --]

Re: [9fans] SMC SYS-5018A-FTN4 lapic weirdness (9front)

[Lyndon Nerenberg]

cinap_lenrek@felloff.net writes:
> bit 7 is "illegal register access". try to print the pc from the ureg
> passed to the first argument in lapicerror() in /sys/src/9/pc/apic.c.

A quick printf hack says ureg->pc = 0xffffffff801103b3 mostly
(>99.9%), but a few other oddballs are:

  cpu0  0x2400c8
  cpu0  0x226df6
  cpu0  0xffffffff80208b1c
  cpu0  0xffffffff802083f9

etc.  During boot I see complaints from cpu1, but once the machine is
fully booted only cpu0 spits messages.

This is from a generic Dick Tracy amd64 kernel, with a U->pc printf
added to lapicerror().

--lyndon

Re: [9fans] SMC SYS-5018A-FTN4 lapic weirdness (9front)

[Lyndon Nerenberg]

cinap_lenrek@felloff.net writes:
> bit 7 is "illegal register access". try to print the pc from the ureg
> passed to the first argument in lapicerror() in /sys/src/9/pc/apic.c.

Okay.  Likely not 'til the weekend though ...

Re: [9fans] SMC SYS-5018A-FTN4 lapic weirdness (9front)

[cinap_lenrek]

typo. illegal register *ADDRESS*.

--
cinap

Re: [9fans] SMC SYS-5018A-FTN4 lapic weirdness (9front)

[cinap_lenrek]

bit 7 is "illegal register access". try to print the pc from the ureg
passed to the first argument in lapicerror() in /sys/src/9/pc/apic.c.

--
cinap