[Bug 219399] System panics after several hours of 14-threads-compilation orgies using poudriere on AMD Ryzen...

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=219399

--- Comment #165 from Mark Millard <markmi at dsl-only.net> ---
(In reply to Don Lewis from comment #163)

Note: I can not claim to know for

In:

Physical Memory Array
. . .
        Error Correction Type: Multi-bit ECC

I wonder if it says that even with non-ECC RAM in
one or more (or all) DIMMs. That section is not
DIMM specific at all from what I can tell. But
may be some of it is a summary over all DIMMs?
(My guess is: no.)

In:

Memory Device
. . .
        Error Information Handle: 0x002D
        Total Width: 128 bits
        Data Width: 64 bits

vs. the older, narrower context's

Memory Device
. . .
        Error Information Handle: Not Provided
        Total Width: 72 bits
        Data Width: 64 bits

The 128 bits instead of 128+8 or 128+2*8 bits
suggests lack of ECC use at the DIMM in
question.

But I'm unable to test a Ryzen context for now
so take my notes in this area as speculation
(or as a suggestion for contrasting tests that
I might have performed if the Ryzen system still
worked, presuming some non-ECC RAM is available
to substitute temporarily).

-- 
You are receiving this mail because:
You are the assignee for the bug.