From mboxrd@z Thu Jan 1 00:00:00 1970 From: bugzilla-daemon@bugzilla.kernel.org Subject: [Bug 116611] New: scsi-block passthrough breaks with external pci-express SATA controller Date: Mon, 18 Apr 2016 05:26:59 +0000 Message-ID: Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit To: kvm@vger.kernel.org Return-path: Received: from mail.kernel.org ([198.145.29.136]:39371 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751185AbcDRF1E (ORCPT ); Mon, 18 Apr 2016 01:27:04 -0400 Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id EF2DF20272 for ; Mon, 18 Apr 2016 05:27:02 +0000 (UTC) Received: from bugzilla1.web.kernel.org (bugzilla1.web.kernel.org [172.20.200.51]) by mail.kernel.org (Postfix) with ESMTP id E563B20221 for ; Mon, 18 Apr 2016 05:26:59 +0000 (UTC) Sender: kvm-owner@vger.kernel.org List-ID: https://bugzilla.kernel.org/show_bug.cgi?id=116611 Bug ID: 116611 Summary: scsi-block passthrough breaks with external pci-express SATA controller Product: Virtualization Version: unspecified Kernel Version: 4.4 Hardware: x86-64 OS: Linux Tree: Mainline Status: NEW Severity: normal Priority: P1 Component: kvm Assignee: virtualization_kvm@kernel-bugs.osdl.org Reporter: mutedbytes@gmail.com Regression: No My hardware provides 6 SATA ports directly from Intel Z97 chipset, and 2 additional SATA ports from a Marvell 88SE9172 attached to the Z97 PCI-e bus. A physical block device attached directly to the chipset ports can successfully be passed through using virtio-scsi-pci controller and scsi-block device to Windows 8/8.1/10 guest using qemu 2.5. However, attaching the same physical drive to the Marvell controller results in missing drives or weird/corrupted behavior in the guest when attempting to perform the same passthrough (for example, no hard drives are found during guest installation procedure after loading virtio drivers). This seems to have broken somewhere in 4.2 or later, as it works fine as expected in 4.1. The broken behavior appears in 4.4 and 4.5, though I have not been able to test 4.2 and 4.3. -- You are receiving this mail because: You are watching the assignee of the bug.