This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


Re: [Xen-devel] IOMMU issue on HP DL385g7 servers - workaround_bogus_bio

To: Wei Wang2 <wei.wang2@xxxxxxx>
Subject: Re: [Xen-devel] IOMMU issue on HP DL385g7 servers - workaround_bogus_bios / workaround_bug_bios doesn't help
From: Mark Schneider <ms@xxxxxxxxxxxxxxxxxxxxxx>
Date: Fri, 15 Jul 2011 17:40:37 +0200
Cc: xen-devel@xxxxxxxxxxxxxxxxxxx, ms@xxxxxxxxxxxxxxxxxxxxxx, Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>
Delivery-date: Fri, 15 Jul 2011 10:33:32 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
In-reply-to: <201107141827.41736.wei.wang2@xxxxxxx>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
References: <4E1DA642.9080500@xxxxxxxxxxxxxxxxxxxxxx> <201107141827.41736.wei.wang2@xxxxxxx>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv: Gecko/20110307 Icedove/3.0.11
Am 14.07.2011 18:27, schrieb Wei Wang2:
Wednesday 13 July 2011 16:05:54 Mark Schneider wrote:

I have cretaed wheezy xen 4.1 debian live image (s. the link below to
it) but even I tried different boot options like "iommu=1 |
workaround_bios_bug | workaround_bogus_bios" I am getting such entry
"(XEN) AMD-Vi: IOMMU not found!" in the log (s. for more details in
L385g7_NOK.txt )

link to xen 4.0 (squeeze) and xen 4.1 (wheezy) debian live images -

do you have any idea how to solve this issue? - thank you in advance for
any hints (different one than killing the hardware vendor (HP) ;) ...
for the missing IOMMU BIOS option

Linux xen41dom0 2.6.39-2-amd64 #1 SMP Wed Jun 8 11:01:04 UTC 2011 x86_64

Thank you in advance for any hints.

Regards, Mark
workaround_bios_bug  and workaround_bogus_bios only works for vtd not amd.
You should have IVRS table build by BIOS to enable amd iommu. It could look
like this:
(XEN) ACPI: IVRS AFE7C1ED, 00C8 (r1  AMD     RD890S  6040000 AMD         0)

Thanks a lot for your hints Wei.

All ACPI entries I see in "xm dmesg" are only like below. I didn't find any IVRS entry
# --
(XEN) Xen-e820 RAM map:
(XEN)  0000000000000000 - 000000000009e800 (usable)
(XEN)  000000000009e800 - 00000000000a0000 (reserved)
(XEN)  00000000000f0000 - 0000000000100000 (reserved)
(XEN)  0000000000100000 - 00000000bfc1c000 (usable)
(XEN)  00000000bfc1c000 - 00000000bfc4c000 (ACPI data)
(XEN)  00000000bfc4c000 - 00000000bfc4d000 (usable)
(XEN)  00000000bfc4d000 - 00000000d0000000 (reserved)
(XEN)  00000000fec00000 - 00000000fee10000 (reserved)
(XEN)  00000000ff800000 - 0000000100000000 (reserved)
(XEN)  0000000100000000 - 000000043ffff000 (usable)
(XEN) ACPI: RSDP 000F4EE0, 0024 (r2 HP    )
(XEN) ACPI: XSDT BFC1D3C0, 00A4 (r1 HP     ProLiant        2   Ò^D     162E)
(XEN) ACPI: FACP BFC1D4C0, 00F4 (r3 HP     ProLiant        2   Ò^D     162E)
(XEN) ACPI: DSDT BFC1D5C0, CCBB (r1 HP         DSDT        1 INTL 20061109)
(XEN) ACPI: FACS BFC1C140, 0040
(XEN) ACPI: SPCR BFC1C180, 0050 (r1 HP     SPCRRBSU        1   Ò^D     162E)
(XEN) ACPI: MCFG BFC1C200, 003C (r1 HP     ProLiant        1             0)
(XEN) ACPI: HPET BFC1C240, 0038 (r1 HP     ProLiant        2   Ò^D     162E)
(XEN) ACPI: SPMI BFC1C280, 0040 (r5 HP     ProLiant        1   Ò^D     162E)
(XEN) ACPI: ERST BFC1C2C0, 01D0 (r1 HP     ProLiant        1   Ò^D     162E)
(XEN) ACPI: APIC BFC1C4C0, 011E (r1 HP     ProLiant        2             0)
(XEN) ACPI: SRAT BFC1C7C0, 02A0 (r2 AMD    F10             1 AMD         1)
(XEN) ACPI: FFFF BFC1CFC0, 0176 (r1 HP     ProLiant        1   Ò^D     162E)
(XEN) ACPI: BERT BFC1D140, 0030 (r1 HP     ProLiant        1   Ò^D     162E)
(XEN) ACPI: HEST BFC1D180, 018C (r1 HP     ProLiant        1   Ò^D     162E)
(XEN) ACPI: FFFF BFC1D340, 0064 (r2 HP     ProLiant        2   Ò^D     162E)
(XEN) ACPI: SSDT BFC2A280, 0126 (r3     HP  PCIE1X8        2 INTL 20090625)
(XEN) ACPI: SSDT BFC2A3C0, 024A (r3     HP PCIE2N3X        2 INTL 20090625)
(XEN) ACPI: SSDT BFC2A640, 0125 (r3     HP  CRSPCI0        2   HP        1)
(XEN) ACPI: SSDT BFC2A780, 0377 (r1     HP     pmab        1 INTL 20090625)
(XEN) Initing memory sharing.
(XEN) AMD-Vi: IOMMU not found!
(XEN) I/O virtualisation disabled
(XEN)  -> Using new ACK method
(XEN) Platform timer is 14.318MHz HPET
(XEN) Allocated console ring of 64 KiB.
(XEN) HVM: ASIDs enabled.
(XEN) SVM: Supported advanced features:
(XEN)  - Nested Page Tables (NPT)
(XEN)  - Last Branch Record (LBR) Virtualisation
(XEN)  - Next-RIP Saved on #VMEXIT
(XEN)  - Pause-Intercept Filter
(XEN) HVM: SVM enabled
(XEN) HVM: Hardware Assisted Paging detected.
(XEN) Brought up 24 CPUs
(XEN) Xenoprofile: AMD IBS detected (0x0000001f)
# ---

It looks like HP DL385g7 BIOS doesn't offer such IVRS table. Maybe there is a kind of workaround to use the IOMMU support (890FX chipset / 6176SE CPUs support it) as it is required for "passing through" PCI devices to domu domains.

I have tested xen enabled kernels 2.6.32-5, 2.6.39-2 and 3.0.0-rc7 (with aufs support)
# ---
root@xen41dom0:~# uname -a
Linux xen41dom0 3.0.0-rc7 #2 SMP Fri Jul 15 02:10:42 CEST 2011 x86_64 GNU/Linux

I will upload today the last test xen live iso image (s. below) for downloads.

Please take a look for details at: http://www.it-infrastrukturen.com/fileadmin/linux/debian-live-xen/README.xen

In the (selfmade) used kernel 3.0.0-rc7 packages I have activated IOMMU-debugging however I have to look for a debug howto.

Thank you in advance for your time.



Attachment: xm_dmesg.boot
Description: Binary data

Attachment: dmesg.boot
Description: Text document

Xen-devel mailing list