Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 30 May 2018 08:47:48 -0500
From:      Justin Hibbits <jhibbits@freebsd.org>
To:        Peter Grehan <grehan@freebsd.org>
Cc:        src-committers <src-committers@freebsd.org>, svn-src-all@freebsd.org,  svn-src-head@freebsd.org
Subject:   Re: svn commit: r334365 - head/sys/dev/pci
Message-ID:  <CAHSQbTDk2C%2B_YMqegM69LgWiNGrf3iyAJzD1TkbGbpH9hmfsAQ@mail.gmail.com>
In-Reply-To: <25bd4319-6d7f-e668-88e4-4c3adbc1e8e9@freebsd.org>
References:  <201805300241.w4U2fmvB033942@repo.freebsd.org> <25bd4319-6d7f-e668-88e4-4c3adbc1e8e9@freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, May 30, 2018 at 4:42 AM, Peter Grehan <grehan@freebsd.org> wrote:
>>    PCIe only permits 1 device on an endpoint, so some devices ignore the
>> device
>>    part of B:D:F probing.  Although ARI likely fixes this, not all
>> platforms
>>    support ARI completely or correctly, so some devices end up showing up
>> 32
>>    times on the bus.
>
>
>  I think this might have broken bhyve - a fake PCIe capability is put on the
> root port so that guests will use MSI/MSI-x, but otherwise it looks like
> parallel PCI. Not exactly spec-compliant, but then neither is most of the
> world of PCI/PCIe.
>
>  It may be worth #ifdef'ing this with powerpc.
>
> later,
>
> Peter.

I can special case i386 and amd64, but this change should (eventually)
remove the majority of the current special case overrides for
alternative architectures.

I think it'd be better to have a bhyve root bridge driver that itself
overrides pci_maxslots(), and have the 'compliant' path be the common.

Thoughts?

- Justin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAHSQbTDk2C%2B_YMqegM69LgWiNGrf3iyAJzD1TkbGbpH9hmfsAQ>