Content-Type: multipart/mixed; boundary="===============9018754689968596997==" MIME-Version: 1.0 From: Joerg Roedel To: lkp@lists.01.org Subject: Re: [iommu/vt] 571dbbd4d04: DMAR-IR: Failed to copy IR table for dmar1 from previous kernel Date: Thu, 02 Jul 2015 08:39:25 +0200 Message-ID: <20150702063925.GC16233@suse.de> In-Reply-To: <1435814058.5489.105.camel@intel.com> List-Id: --===============9018754689968596997== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable On Thu, Jul 02, 2015 at 01:14:18PM +0800, Huang Ying wrote: > FYI, we noticed the below changes on > = > git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master > commit 571dbbd4d044e11c78bc077acb3ccef4c77b096e ("iommu/vt-d: Don't disab= le IR when it was previously enabled") > = > The following new message in kernel log may make end user confusing. > = > [ 0.503764] DMAR-IR: IRQ remapping was enabled on dmar0 but we are not= in kdump mode > [ 0.512418] DMAR-IR: Failed to copy IR table for dmar0 from previous k= ernel > [ 0.520501] DMAR-IR: IRQ remapping was enabled on dmar1 but we are not= in kdump mode > [ 0.529154] DMAR-IR: Failed to copy IR table for dmar1 from previous k= ernel > [ 0.537077] DMAR-IR: Enabled IRQ remapping in x2apic mode This is not a problem, the kernel detected that the IOMMU was already enabled, but it is not booting into a kdump kernel so it is not trying to copy over translation tables. The messages for this case are indeed a bit misleading, it looks like the kernel tried to copy the translation tables, but it didn't. I'll fix this. What were you doing to trigger this, it is quite uncommon to find the IOMMU enabled on a normal kernel boot. Did you boot with kexec into a new kernel? Joerg --===============9018754689968596997==--