Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 30 Nov 2019 06:36:21 +0000
From:      bugzilla-noreply@freebsd.org
To:        virtualization@FreeBSD.org
Subject:   [Bug 203874] [patch] MSI/MSI-X interrupts don't work in VMware virtual machines
Message-ID:  <bug-203874-27103-POd8z47nK8@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-203874-27103@https.bugs.freebsd.org/bugzilla/>
References:  <bug-203874-27103@https.bugs.freebsd.org/bugzilla/>

next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D203874

Mahmoud Al-Qudsi <mqudsi@neosmart.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mqudsi@neosmart.net

--- Comment #28 from Mahmoud Al-Qudsi <mqudsi@neosmart.net> ---
Hopefully this isn't too late but I know there were some issues caused with
pass-through PCI-e devices and MSI-X support under ESX 5.5+, but I cannot f=
or
the life of me find the issue that I know exists somewhere in bugzilla (eit=
her
this one or on the FreeNAS one).

This is not the issue I was searching for (which is one that I've commented
on), but here is a report that was originally filed against a FreeBSD system
running under ESX 5.1 (not pre-5.0 as claimed in some of the comments above=
):
https://redmine.ixsystems.com/issues/2056

This must have been before the hack was added to pci.c because the fix back
then was to globally set hw.pci.enable_msix=3D0 (or else individually black=
list
the LSI device).

I personally ran into interrupt storms with MSI-X enabled under 6.0 and a L=
SI
9211-8i a few years back, which is how I first came across this back in the
day. Googling for some variation of `esxi, freebsd, mps, interrupt storm, l=
si,
pass-through, vmware` brings up a lot of relevant results, almost all of wh=
ich
are post ESX 5.0

I am reminded heavily of Chesterton's fence here and would recommend procee=
ding
only with extreme caution (and after looking into LSI HBA passthrough
configurations and scenarios).

--=20
You are receiving this mail because:
You are the assignee for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-203874-27103-POd8z47nK8>