qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Tobias Geiger <tobias.geiger@vido.info>
To: Anthony PERARD <anthony.perard@citrix.com>
Cc: xen-devel@lists.xensource.com, QEMU-devel <qemu-devel@nongnu.org>,
	Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Subject: Re: [Qemu-devel] [Xen-devel] [PATCH V7 00/11] Xen PCI Passthrough
Date: Mon, 20 Feb 2012 20:58:57 +0100	[thread overview]
Message-ID: <4F42A601.7000701@vido.info> (raw)
In-Reply-To: <CAJJyHj+5NJsHRESxn=socWA_bum3O19trhYUZE9+qdAji8hxwg@mail.gmail.com>

Hi

With the patch you mentioned, xl create looks more familiar:

libxl: error: libxl_pci.c:756:libxl__device_pci_reset: write to 
/sys/bus/pci/devices/0000:03:00.0/reset returned -1: Invalid argument
libxl: error: libxl_pci.c:756:libxl__device_pci_reset: write to 
/sys/bus/pci/devices/0000:03:00.1/reset returned -1: Invalid argument


This is actually "normal", at least compared to traditional qemu-dm - i 
guess due to missing FLR (Bios is capable, but Card not i assume).

Anyhow - the downside of the patch is, qemu doesnt even start :)

This is qemu-dm.log (with debugging enabled):

[00:06.0] pci_intx: intx=1
[00:06.0] pt_initfn: Real physical device 03:00.0 registered successfuly!
[00:07.0] pt_initfn: Assigning real physical device 03:00.1 to devfn 0x38
[00:07.0] pt_register_regions: IO region 0 registered (size=0x00004000 
base_addr=0xc3120000 type: 0)
[00:07.0] pci_intx: intx=2
[00:07.0] pt_initfn: Real physical device 03:00.1 registered successfuly!
[00:08.0] pt_initfn: Assigning real physical device 00:1d.0 to devfn 0x40
[00:08.0] pt_register_regions: IO region 4 registered (size=0x00000020 
base_addr=0x00005080 type: 0x1)
[00:08.0] pci_intx: intx=1
[00:08.0] pt_initfn: Real physical device 00:1d.0 registered successfuly!
[00:09.0] pt_initfn: Assigning real physical device 00:1d.1 to devfn 0x48
[00:09.0] pt_register_regions: IO region 4 registered (size=0x00000020 
base_addr=0x00005060 type: 0x1)
[00:09.0] pci_intx: intx=2
[00:09.0] pt_initfn: Real physical device 00:1d.1 registered successfuly!
[00:0a.0] pt_initfn: Assigning real physical device 00:1d.2 to devfn 0x50
[00:0a.0] pt_register_regions: IO region 4 registered (size=0x00000020 
base_addr=0x00005040 type: 0x1)
[00:0a.0] pci_intx: intx=3
[00:0a.0] pt_initfn: Real physical device 00:1d.2 registered successfuly!
[00:0b.0] pt_initfn: Assigning real physical device 00:1d.7 to devfn 0x58
[00:0b.0] pt_register_regions: IO region 0 registered (size=0x00000400 
base_addr=0xc3321000 type: 0)
[00:0b.0] pci_intx: intx=1
[00:0b.0] pt_initfn: Real physical device 00:1d.7 registered successfuly!
[00:06.0] pt_pci_read_config: address=0x000a val=0x00000300 len=2
[00:06.0] pt_pci_read_config: address=0x0000 val=0x00001002 len=2
[00:06.0] pt_pci_read_config: address=0x0002 val=0x00006718 len=2
[00:06.0] pt_pci_read_config: address=0x0010 val=0x0000000c len=4
[00:06.0] pt_pci_write_config: address=0x0010 val=0xffffffff len=4
[00:06.0] pt_iomem_map: BAR 0, e_phys=0xffffffffffffffff 
maddr=0xe0000000 len=0x10000000 first_map=1
[00:06.0] pt_pci_read_config: address=0x0010 val=0xf000000c len=4
[00:06.0] pt_pci_write_config: address=0x0010 val=0x0000000c len=4
[00:06.0] pt_iomem_map: BAR 0, e_phys=0xffffffffffffffff 
maddr=0xe0000000 len=0x10000000 first_map=0
[00:06.0] pt_pci_read_config: address=0x0018 val=0x00000004 len=4
[00:06.0] pt_pci_write_config: address=0x0018 val=0xffffffff len=4
[00:06.0] pt_iomem_map: BAR 2, e_phys=0xffffffffffffffff 
maddr=0xc3100000 len=0x20000 first_map=1
[00:06.0] pt_pci_read_config: address=0x0018 val=0xfffe0004 len=4
[00:06.0] pt_pci_write_config: address=0x0018 val=0x00000004 len=4
[00:06.0] pt_iomem_map: BAR 2, e_phys=0xffffffffffffffff 
maddr=0xc3100000 len=0x20000 first_map=0
[00:06.0] pt_pci_read_config: address=0x0020 val=0x00000001 len=4
[00:06.0] pt_pci_write_config: address=0x0020 val=0xffffffff len=4
[00:06.0] pt_ioport_map: BAR 4, e_phys=0xffffffffffffffff 
pio_base=0x2000 len=256 first_map=1
[00:06.0] pt_pci_read_config: address=0x0020 val=0xffffff01 len=4
[00:06.0] pt_pci_write_config: address=0x0020 val=0x00000001 len=4
[00:06.0] pt_ioport_map: BAR 4, e_phys=0xffffffffffffffff 
pio_base=0x2000 len=256 first_map=0
[00:06.0] pt_pci_read_config: address=0x0024 val=0x00000000 len=4
[00:06.0] pt_pci_write_config: address=0x0024 val=0xffffffff len=4
[00:06.0] pt_pci_read_config: address=0x0024 val=0x00000000 len=4
[00:06.0] pt_pci_write_config: address=0x0024 val=0x00000000 len=4
[00:06.0] pt_pci_read_config: address=0x0030 val=0x00000000 len=4
[00:06.0] pt_pci_write_config: address=0x0030 val=0xffffffff len=4
[00:06.0] pt_iomem_map: BAR 6, e_phys=0xffffffffffffffff 
maddr=0xc3140000 len=0x20000 first_map=1
[00:06.0] pt_pci_read_config: address=0x0030 val=0xfffe0001 len=4
[00:06.0] pt_pci_write_config: address=0x0030 val=0x00000000 len=4
[00:06.0] pt_iomem_map: BAR 6, e_phys=0xffffffffffffffff 
maddr=0xc3140000 len=0x20000 first_map=0
[00:06.0] pt_pci_read_config: address=0x003d val=0x00000001 len=1
[00:06.0] pt_pci_write_config: address=0x003c val=0x0000000b len=1
[00:06.0] pt_pci_read_config: address=0x0004 val=0x00000000 len=2
[00:06.0] pt_pci_write_config: address=0x0004 val=0x00000004 len=2
qemu-system-x86_64: 
/usr/src/xen-with-upstream-qemu-patch/upstream-qemu/qemu/memory.c:1378: 
memory_region_del_subregion: Assertion `subregion->parent == mr' failed.
Waiting for data...



Am 20.02.2012 18:08, schrieb Anthony PERARD:
> On Mon, Feb 20, 2012 at 16:11, Tobias Geiger<tobias.geiger@vido.info>  wrote:
>> My Domu (Win7_64) boots fine, but 2 of my 6 passed-through pcidevices dont get
>> passed through at all:
>>
>> libxl: error: libxl_pci.c:756:libxl__device_pci_reset: write to
>> /sys/bus/pci/devices/0000:03:00.0/reset returned -1: Invalid argument
>> libxl: error: libxl_qmp.c:239:qmp_handle_error_response: received an error
>> message from QMP server: Device 'xen-pci-passthrough' could not be initialized
>> libxl: error: libxl_pci.c:756:libxl__device_pci_reset: write to
>> /sys/bus/pci/devices/0000:03:00.1/reset returned -1: Invalid argument
>> libxl: error: libxl_qmp.c:239:qmp_handle_error_response: received an error
>> message from QMP server: Device 'xen-pci-passthrough' could not be initialized
> These two errors are probably because my patch series depend on one
> other patch sent earlier:
> http://lists.xen.org/archives/html/xen-devel/2012-02/msg01027.html
>
>> The other 4 PCI-Ids (USB Controller) get passed through, but the USB-Devices
>> attached to them are "Not Working" according to Windows Device-Manager.
>>
>> All Devices worked with the old qemu-dm (traditional).
>>
>> Anything i can do to debug this further?
> You can send the log of qemu ( /var/log/xen/qemu-dm-$vm_name.log ).
> There is also a way to have more from QEMU by decommenting this two lines:
> /* #define PT_LOGGING_ENABLED */
> /* #define PT_DEBUG_PCI_CONFIG_ACCESS */
> in hw/xen_pci_passthrough.h
>
> Thanks for the testing,
>

  reply	other threads:[~2012-02-20 19:59 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-02-17 17:08 [Qemu-devel] [PATCH V7 00/11] Xen PCI Passthrough Anthony PERARD
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 01/11] pci_ids: Add INTEL_82599_VF id Anthony PERARD
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 02/11] pci_regs: Fix value of PCI_EXP_TYPE_RC_EC Anthony PERARD
2012-02-20 20:27   ` Michael S. Tsirkin
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 03/11] pci_regs: Add PCI_EXP_TYPE_PCIE_BRIDGE Anthony PERARD
2012-02-20 20:30   ` Michael S. Tsirkin
2012-02-21 16:07     ` Anthony PERARD
2012-02-23 20:20   ` Michael S. Tsirkin
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 04/11] configure: Introduce --enable-xen-pci-passthrough Anthony PERARD
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 05/11] Introduce HostPCIDevice to access a pci device on the host Anthony PERARD
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 06/11] pci.c: Add pci_check_bar_overlap Anthony PERARD
2012-02-19 13:35   ` Michael S. Tsirkin
2012-02-28 11:19     ` Anthony PERARD
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 07/11] Introduce Xen PCI Passthrough, qdevice (1/3) Anthony PERARD
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 08/11] Introduce Xen PCI Passthrough, PCI config space helpers (2/3) Anthony PERARD
2012-02-19 13:32   ` Michael S. Tsirkin
2012-02-20 20:35   ` Michael S. Tsirkin
2012-02-21 10:18     ` Stefano Stabellini
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 09/11] Introduce apic-msidef.h Anthony PERARD
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 10/11] Introduce Xen PCI Passthrough, MSI (3/3) Anthony PERARD
2012-02-17 17:08 ` [Qemu-devel] [PATCH V7 11/11] xen passthrough: clean up MSI-X table handling Anthony PERARD
2012-02-21  9:34   ` Jan Beulich
2012-02-21 16:15     ` Anthony PERARD
2012-02-20 10:51 ` [Qemu-devel] [Xen-devel] [PATCH V7 00/11] Xen PCI Passthrough Tobias Geiger
2012-02-20 11:31   ` Anthony PERARD
2012-02-20 16:11     ` Tobias Geiger
2012-02-20 17:08       ` Anthony PERARD
2012-02-20 19:58         ` Tobias Geiger [this message]
2012-02-20 23:31           ` Anthony PERARD
2012-02-22 14:56             ` Anthony PERARD

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4F42A601.7000701@vido.info \
    --to=tobias.geiger@vido.info \
    --cc=anthony.perard@citrix.com \
    --cc=qemu-devel@nongnu.org \
    --cc=stefano.stabellini@eu.citrix.com \
    --cc=xen-devel@lists.xensource.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).