Date: Wed, 1 Nov 2017 11:54:21 -0700 From: John Baldwin <jhb@FreeBSD.org> To: Stephen Hurd <shurd@FreeBSD.org>, src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: Re: svn commit: r325166 - head/sys/net Message-ID: <ed2c3fb2-1b68-06b9-a9df-e9d9e883fac2@FreeBSD.org> In-Reply-To: <201710302108.v9UL8CMP071651@repo.freebsd.org> References: <201710302108.v9UL8CMP071651@repo.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 10/30/17 9:08 PM, Stephen Hurd wrote: > Author: shurd > Date: Mon Oct 30 21:08:12 2017 > New Revision: 325166 > URL: https://svnweb.freebsd.org/changeset/base/325166 > > Log: > Avoid enabling MSI-X if MSI-X is disabled globally > > It was reported on the community call that with > hw.pci.enable_msix=0, iflib would enable MSI-X on the device and attempt > to use it, which caused issues. Test the sysctl explicitly and do not > enable MSI-X if it's disabled globally. There are also various blacklists you should probably be honoring. If you just use the existing APIs (pci_*_msix()) that should all happen transparently however. Also, the 'enable_msi=0' tunable disables both MSI and MSI-X (again, if you used the existing API that is already handled) -- John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?ed2c3fb2-1b68-06b9-a9df-e9d9e883fac2>