interesting arm issue/bug
Andrew Turner
andrew at fubar.geek.nz
Tue Sep 13 10:21:09 UTC 2016
On Mon, 12 Sep 2016 14:36:43 -0700
John-Mark Gurney <jmg at funkthat.com> wrote:
> I'm not sure if we run on this hardware or not, but I thought it'd be
> good to bring this to the attention of the list:
> http://www.mono-project.com/news/2016/09/12/arm64-icache/
>
We don't boot on this hardware. It is also my understanding having a
mismatch in the CTR_EL0 register is a violation of the architecture,
and therefore a hardware bug.
If someone was to port FreeBSD to this SoC they would need to do
something similar to the proposed Linux fix and trap access to the
CTR_EL0 register to return a stable minimum value.
I note there is a similar issue with an earlier Samsung ARMv7 SoC,
however we don't support multi-cluster on 32-bit ARM so this isn't a
problem for us.
Andrew
More information about the freebsd-arm
mailing list