Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 13 Jan 2022 10:12:31 +0200
From:      Andriy Gapon <avg@FreeBSD.org>
To:        FreeBSD Hackers <freebsd-hackers@FreeBSD.org>
Subject:   Re: IO-APIC interrupt delivery mode on Ryzen
Message-ID:  <54e5f61c-64fd-0ebf-c3cd-d4a7a4cbb7e7@FreeBSD.org>
In-Reply-To: <c0a41342-4480-76e7-8aad-98a007fb7b88@FreeBSD.org>
References:  <c0a41342-4480-76e7-8aad-98a007fb7b88@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On 14/02/2018 09:16, Andriy Gapon wrote:
> 
> A while ago I discovered that all AMD IO-APICs (in separate southbridges and in
> FCHs) had a bug with respect to how the interrupt delivery mode got interpreted.
> I am curious if the problem has been fixed in Ryzen or if it is being carried on.
> 
> I would appreciate any help with testing that.
> 
> The discussion of the problem and the tests I used can be found in this thread:
> https://mail.coreboot.org/pipermail/coreboot/2016-October/082148.html
> https://mail.coreboot.org/pipermail/coreboot/2016-October/082156.html
> 
> If you are interested and it's not really clear how to conduct a test, please
> write me.

FWIW, Ryzen IO-APICs do not have the problem.
They interpret the delivery mode according to the APIC specification, not the 
HyperTransport one.


-- 
Andriy Gapon



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?54e5f61c-64fd-0ebf-c3cd-d4a7a4cbb7e7>