* nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB @ 2009-08-25 15:47 Mr. Teo En Ming (Zhang Enming) 2009-08-25 15:56 ` Mr. Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Mr. Teo En Ming (Zhang Enming) @ 2009-08-25 15:47 UTC (permalink / raw) To: xen-devel Dear All, I have managed to do PCI-e VGA passthrough with the open source Xen but the work is still in progress because although Windows XP guest can see the REAL PCI-e x16 graphics card instead of an emulated graphics driver, it cannot be initialized yet. Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris Derzhavets, Marc, Caz Yokoyama, and others who have helped me and shared their knowledge with me along the way. System Configuration: Intel Desktop Board DQ45CB with BIOS upgraded to 0093 Onboard Intel GMA 4500 Graphics (IGD) nVidia Geforce 8400 GS PCI Express x16 Graphics Card Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 Primary Video Adapter in BIOS: IGD Please see the screenshots and my blog at the link here: http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400-gs-pci-express-x16.html -- Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com Alma Maters: Singapore Polytechnic, National University of Singapore ^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-25 15:47 nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Mr. Teo En Ming (Zhang Enming) @ 2009-08-25 15:56 ` Mr. Teo En Ming (Zhang Enming) 2009-08-25 17:25 ` nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB djmagee 0 siblings, 1 reply; 25+ messages in thread From: Mr. Teo En Ming (Zhang Enming) @ 2009-08-25 15:56 UTC (permalink / raw) To: enming.teo; +Cc: xen-devel I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as suggested by Weidong. -- Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com Alma Maters: Singapore Polytechnic, National University of Singapore On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: > Dear All, > > I have managed to do PCI-e VGA passthrough with the open source Xen > but the work is still in progress because although Windows XP guest > can see the REAL PCI-e x16 graphics card instead of an emulated > graphics driver, it cannot be initialized yet. > > Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris > Derzhavets, Marc, Caz Yokoyama, and others who have helped me and > shared their knowledge with me along the way. > > System Configuration: > > Intel Desktop Board DQ45CB with BIOS upgraded to 0093 > Onboard Intel GMA 4500 Graphics (IGD) > nVidia Geforce 8400 GS PCI Express x16 Graphics Card > > Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host Operating > System > Xen 3.5 Unstable/Development Type 1 Hypervisor > Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 > Primary Video Adapter in BIOS: IGD > > Please see the screenshots and my blog at the link here: > > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400-gs-pci-express-x16.html > > ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB 2009-08-25 15:56 ` Mr. Teo En Ming (Zhang Enming) @ 2009-08-25 17:25 ` djmagee [not found] ` <f4527be0908251056s295eaf0fle31922e7d90faa1c@mail.gmail.com> 2009-08-26 8:26 ` Han, Weidong 0 siblings, 2 replies; 25+ messages in thread From: djmagee @ 2009-08-25 17:25 UTC (permalink / raw) To: enming.teo; +Cc: xen-devel As I've pointed out on this list before, there are not enough PCIe lanes on the DQ45CB to drive both the internal graphics adapter and the add-on adapter at the same time. I believe the onboard one may be able to operate in some sort of VGA only mode when there is a card installed and used as the primary adapter. I have a similar setup, using the same motherboard, and an ATI 4770. I used the VGA passthrough patches from an earlier posting (I saw you followed the same set of instructions), with limited success. I've been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made modifications to the dom0 portion of the patches so they would apply to my xenified 2.6.29 kernel. These patches include code that will copy the VGA bios to the guest bios instead of using the emulated vga bios. I've had very little success, however. I have only tried passing through the ATI adapter. In all instances, the guest bios messages appear on my monitor, so this much works. In some cases, the guest essentially stops there; xm list show's about 2sec CPU usage and nothing ever happens after that point. In other cases, the guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot all the way, but in very low res/color mode, and cannot properly initialize the video device. Once or twice, it actually did recognize the device and had a reasonable default color/resolution combination. In all cases where the guest actually boots, the system eventually freezes. In some cases, I get endless streams of iommu page faults. I have 8GB ram installed. In all cases I've limited dom0 memory to 2GB. In all cases, my guest has been assigned 2GB of memory. I have a dual core e6600. I've tried allowing dom0 to use both cores, offlining one core (using xend dom0_vcpu setting) after boot, and restricting dom0 to only one core using the dom0_max_vcpus xen hypervisor parameter. In all of these cases I've tried both one and two vcpus for the guest. My success with VGA passthrough seems somewhat random and no combination of cpu assignment seems to have any effect. I have not tried with the 2.6.18-xen kernel as I haven't gotten it to boot on my hardware; it can never find my volume group, even if I create a initrd with all of the required modules, or build those drivers into the kernel. I have not spent more than maybe a half an hour on this problem; I suspect it may have something to do with the version of mkinitrd I'm using (from Fedora 9 x64). If anyone else has any insight or similar experience I'd also love to hear it. Doug Magee djmagee@mageenet.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM To: enming.teo@asiasoftsea.net Cc: xen-devel@lists.xensource.com Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as suggested by Weidong. -- Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com Alma Maters: Singapore Polytechnic, National University of Singapore On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: > Dear All, > > I have managed to do PCI-e VGA passthrough with the open source Xen > but the work is still in progress because although Windows XP guest > can see the REAL PCI-e x16 graphics card instead of an emulated > graphics driver, it cannot be initialized yet. > > Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris > Derzhavets, Marc, Caz Yokoyama, and others who have helped me and > shared their knowledge with me along the way. > > System Configuration: > > Intel Desktop Board DQ45CB with BIOS upgraded to 0093 > Onboard Intel GMA 4500 Graphics (IGD) > nVidia Geforce 8400 GS PCI Express x16 Graphics Card > > Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host Operating > System > Xen 3.5 Unstable/Development Type 1 Hypervisor > Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 > Primary Video Adapter in BIOS: IGD > > Please see the screenshots and my blog at the link here: > > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400-gs-pci-express-x16.html > > _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel ^ permalink raw reply [flat|nested] 25+ messages in thread
[parent not found: <f4527be0908251056s295eaf0fle31922e7d90faa1c@mail.gmail.com>]
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB [not found] ` <f4527be0908251056s295eaf0fle31922e7d90faa1c@mail.gmail.com> @ 2009-08-25 19:57 ` djmagee 0 siblings, 0 replies; 25+ messages in thread From: djmagee @ 2009-08-25 19:57 UTC (permalink / raw) To: Andrew Lyon; +Cc: xen-devel [-- Attachment #1: Type: text/plain, Size: 7170 bytes --] Here are the patches I used. The file ending with a timestamp is the set of patches that was posted on xen-devel a couple of months ago by someone with a similar H/W set up and goals. I applied this to my xen-3.4-testing tree as is. As I mentioned I'm using a xenified 2.6.29.6. Of course, I had no idea what parts of the kernel code in that patch I needed to apply to this much newer kernel, so I just tried as is. It failed patching include/linux/pci_regs.h. I saw that most of the stuff in the vgapt patch was already in the 2.6.29 kernel, but some of the PCI_AF_* defines had different values. Vga-pt-patch-2.6.29.6-pci_regs.patch is my patch to that file, and vga-pt-patch-kernel-only contains the kernel patches from the original file but after a global replace to make sure they used the slightly different PCI_AF_* names. I simply made sure that the values would end up the same as they would have been in the 2.6.18-xen kernel. I wouldn't know if this was the right approach. I did not apply the FLR patch that is in xen-unstable, and I was under the impression the same capability is in the patch I've attached. For what it's worth, when I start a guest with the video card assigned and vga_passthrough=1, I can hear the fan on the video card spin up for a few seconds just as it does when the real machine (re)boots. If I (or my test environment) can be of any assistance please let me know. I saw that intel plans on releasing vga passthrough patches soon so I'll be looking out for those. Doug Magee -----Original Message----- From: Andrew Lyon [mailto:andrew.lyon@gmail.com] Sent: Tuesday, August 25, 2009 1:56 PM To: djmagee@mageenet.net Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB On Tue, Aug 25, 2009 at 6:25 PM, <djmagee@mageenet.net> wrote: > As I've pointed out on this list before, there are not enough PCIe lanes on the DQ45CB to drive both the internal graphics adapter and the add-on adapter at the same time. I believe the onboard one may be able to operate in some sort of VGA only mode when there is a card installed and used as the primary adapter. > > I have a similar setup, using the same motherboard, and an ATI 4770. I used the VGA passthrough patches from an earlier posting (I saw you followed the same set of instructions), with limited success. I've been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made modifications to the dom0 portion of the patches so they would apply to my xenified 2.6.29 kernel. These patches include code that will copy the VGA bios to the guest bios instead of using the emulated vga bios. I've had very little success, however. I have only tried passing through the ATI adapter. In all instances, the guest bios messages appear on my monitor, so this much works. In some cases, the guest essentially stops there; xm list show's about 2sec CPU usage and nothing ever happens after that point. In other cases, the guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot all the way, but in very low res/color mode, and cannot properly initialize the video device. Once or twice, it actually did recognize the device and had a reasonable default color/resolution combination. In all cases where the guest actually boots, the system eventually freezes. In some cases, I get endless streams of iommu page faults. Can you post the patches you used? specifically the parts for copying the vga bios. did you also use the patch for flr? I've ported it to 2.6.29 but not had time to try vga passthru yet. > > I have 8GB ram installed. In all cases I've limited dom0 memory to 2GB. In all cases, my guest has been assigned 2GB of memory. > > I have a dual core e6600. I've tried allowing dom0 to use both cores, offlining one core (using xend dom0_vcpu setting) after boot, and restricting dom0 to only one core using the dom0_max_vcpus xen hypervisor parameter. In all of these cases I've tried both one and two vcpus for the guest. My success with VGA passthrough seems somewhat random and no combination of cpu assignment seems to have any effect. > > I have not tried with the 2.6.18-xen kernel as I haven't gotten it to boot on my hardware; it can never find my volume group, even if I create a initrd with all of the required modules, or build those drivers into the kernel. I have not spent more than maybe a half an hour on this problem; I suspect it may have something to do with the version of mkinitrd I'm using (from Fedora 9 x64). > > If anyone else has any insight or similar experience I'd also love to hear it. > > Doug Magee > djmagee@mageenet.com > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo En Ming (Zhang Enming) > Sent: Tuesday, August 25, 2009 11:57 AM > To: enming.teo@asiasoftsea.net > Cc: xen-devel@lists.xensource.com > Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB > > I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as suggested > by Weidong. > > -- > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > Alma Maters: Singapore Polytechnic, National University of Singapore > > > > On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >> Dear All, >> >> I have managed to do PCI-e VGA passthrough with the open source Xen >> but the work is still in progress because although Windows XP guest >> can see the REAL PCI-e x16 graphics card instead of an emulated >> graphics driver, it cannot be initialized yet. >> >> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >> shared their knowledge with me along the way. >> >> System Configuration: >> >> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >> Onboard Intel GMA 4500 Graphics (IGD) >> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >> >> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host Operating >> System >> Xen 3.5 Unstable/Development Type 1 Hypervisor >> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >> Primary Video Adapter in BIOS: IGD >> >> Please see the screenshots and my blog at the link here: >> >> http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400-gs-pci-express-x16.html >> >> > > > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > [-- Attachment #2: vga-pt-patch-2.6.29.6-pci_regs.patch --] [-- Type: application/octet-stream, Size: 951 bytes --] --- linux-2.6.29.6/include/linux/pci_regs.h 2009-07-02 19:41:20.000000000 -0400 +++ linux-2.6.29.6-vgapt/include/linux/pci_regs.h 2009-07-15 06:04:09.000000000 -0400 @@ -240,6 +240,11 @@ #define PCI_PM_CTRL_DATA_SEL_MASK 0x1e00 /* Data select (??) */ #define PCI_PM_CTRL_DATA_SCALE_MASK 0x6000 /* Data scale (??) */ #define PCI_PM_CTRL_PME_STATUS 0x8000 /* PME pin status */ +#define PCI_PM_CTRL_DATA_DSTATE_MASK 0x3 /* D0 - D3 */ +#define PCI_PM_CTRL_D0 0x0 +#define PCI_PM_CTRL_D1 0x1 +#define PCI_PM_CTRL_D2 0x2 +#define PCI_PM_CTRL_D3HOT 0x3 #define PCI_PM_PPB_EXTENSIONS 6 /* PPB support extensions (??) */ #define PCI_PM_PPB_B2_B3 0x40 /* Stop clock when in D3hot (??) */ #define PCI_PM_BPCC_ENABLE 0x80 /* Bus power/clock control enable (??) */ @@ -321,6 +326,7 @@ #define PCI_AF_LENGTH 2 #define PCI_AF_CAP 3 +#define PCI_AF_WHAT 0x06 #define PCI_AF_CAP_TP 0x01 #define PCI_AF_CAP_FLR 0x02 #define PCI_AF_CTRL 4 [-- Attachment #3: vga-pt-patch-kernel-only.patch --] [-- Type: application/octet-stream, Size: 42450 bytes --] diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/controller.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/controller.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/controller.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/controller.c 2009-06-07 19:04:11.000000000 -0400 @@ -208,7 +208,7 @@ void pciback_release_pci_dev(struct pcib } spin_unlock_irqrestore(&dev_data->lock, flags); - pcistub_put_pci_dev(found_dev); + pcistub_put_pci_dev(found_dev, 0); } int pciback_init_devices(struct pciback_device *pdev) @@ -396,7 +396,7 @@ void pciback_release_devices(struct pcib list_for_each_entry_safe(dev_entry, d, &cntrl_entry->dev_list, list) { list_del(&dev_entry->list); - pcistub_put_pci_dev(dev_entry->dev); + pcistub_put_pci_dev(dev_entry->dev, 0); kfree(dev_entry); } list_del(&cntrl_entry->list); diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/passthrough.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/passthrough.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/passthrough.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/passthrough.c 2009-06-07 19:04:11.000000000 -0400 @@ -88,7 +88,7 @@ void pciback_release_pci_dev(struct pcib spin_unlock_irqrestore(&dev_data->lock, flags); if (found_dev) - pcistub_put_pci_dev(found_dev); + pcistub_put_pci_dev(found_dev, 1); } int pciback_init_devices(struct pciback_device *pdev) @@ -157,7 +157,7 @@ void pciback_release_devices(struct pcib list_for_each_entry_safe(dev_entry, t, &dev_data->dev_list, list) { list_del(&dev_entry->list); - pcistub_put_pci_dev(dev_entry->dev); + pcistub_put_pci_dev(dev_entry->dev, 1); kfree(dev_entry); } diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback.h b/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback.h --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback.h 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback.h 2009-06-07 19:04:11.000000000 -0400 @@ -25,6 +25,14 @@ struct pci_dev_entry { #define _PCIB_op_pending (1) #define PCIB_op_pending (1<<(_PCIB_op_pending)) +#define PCIBACK_TYPE_UNKNOWN 0 +#define PCIBACK_TYPE_PCIe_ENDPOINT 1 +#define PCIBACK_TYPE_PCIe_BRIDGE 2 +#define PCIBACK_TYPE_PCI_BRIDGE 3 +#define PCIBACK_TYPE_PCI 4 + +#define DEV_CLASS_PCI_PCI_BRIDGE 0x0604 + struct pciback_device { void *pci_dev_data; spinlock_t dev_lock; @@ -48,6 +56,13 @@ struct pciback_dev_data { struct list_head config_fields; int permissive; int warned_on_write; + u32 dev_type; + int no_flr; + int exp_flr_offset; + int af_flr_offset; + int use_sbr; + int use_d3r; + u8 *cfg_space; /* saved config space for device */ }; /* Get/Put PCI Devices that are hidden from the PCI Backend Domain */ @@ -56,11 +71,25 @@ struct pci_dev *pcistub_get_pci_dev_by_s int slot, int func); struct pci_dev *pcistub_get_pci_dev(struct pciback_device *pdev, struct pci_dev *dev); -void pcistub_put_pci_dev(struct pci_dev *dev); +void pcistub_put_pci_dev(struct pci_dev *dev, int do_flr); + +/* Reference/unreference PCI Devices and stubs without changing the state */ +struct pci_dev *pcistub_ref_pci_dev(struct pci_dev *dev); +void pcistub_unref_pci_dev(struct pci_dev *dev); + +/* Store/reload config space for devices */ +void pciback_store_config_space(struct pci_dev *dev); +void pciback_reload_config_space(struct pci_dev *dev); /* Ensure a device is turned off or reset */ void pciback_reset_device(struct pci_dev *pdev); +/* Do a function level reset (or approximage functionality) for device */ +void pciback_flr_device(struct pci_dev *dev); + +/* Helper to classify the device type */ +void pciback_classify_device(struct pci_dev *dev); + /* Access a virtual configuration space for a PCI device */ int pciback_config_init(void); int pciback_config_init_dev(struct pci_dev *dev); @@ -102,6 +131,10 @@ void pciback_release_devices(struct pcib irqreturn_t pciback_handle_event(int irq, void *dev_id, struct pt_regs *regs); void pciback_do_op(void *data); +/* Parse and load device specific module parameters */ +int pciback_parse_device_params(const char *device_args, int type, + int (*add_func) (int domain, int bus, int slot, int func, int type)); + int pciback_xenbus_register(void); void pciback_xenbus_unregister(void); diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback_ops.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback_ops.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback_ops.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback_ops.c 2009-06-07 19:04:11.000000000 -0400 @@ -5,20 +5,189 @@ */ #include <linux/module.h> #include <linux/wait.h> +#include <linux/delay.h> /* For mdelay function */ #include <asm/bitops.h> #include <xen/evtchn.h> #include "pciback.h" +#define PCIBACK_VENDOR_INTEL 0x8086 +#define PCIBACK_CLASS_ID_USB 0x0c03 +#define PCIBACK_CLASS_ID_VGA 0x0300 +#define PCIBACK_USB_FLRCTRL 0x4 + +#define PCIBACK_IGFX_CAP09_OFFSET 0xa4 +#define PCIBACK_IGFX_CAP13_OFFSET 0xa4 + +#define PCIBACK_IGFX_MEDIARST 0x0d +#define PCIBACK_IGFX_MEDIARST_OFFSET 0xc0 + int verbose_request = 0; module_param(verbose_request, int, 0644); +struct pcistub_sbr_entry { + struct list_head dev_list; + struct pci_dev *dev; +}; + +struct pcistub_sbr_list { + struct list_head dev_list; + struct pci_dev *bridge; + struct pci_dev *dev; + int find_all; + int err; +}; + +/* Used to store the config state so it can be restored after + * resets. + */ +void pciback_store_config_space(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data = pci_get_drvdata(dev); + u32 *ptr = (u32*)dev_data->cfg_space; + int i, count = dev->cfg_size/sizeof(u32); + + for (i = 0; i < count; i += sizeof(u32), ptr++) + pci_read_config_dword(dev, i, ptr); +} + +/* Used to reload the config state after resets. + */ +void pciback_reload_config_space(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data = pci_get_drvdata(dev); + u32 *ptr = (u32*)dev_data->cfg_space; + int i, val, count = dev->cfg_size/sizeof(u32); + + for (i = 0; i < count; i += sizeof(u32), ptr++) { + pci_read_config_dword(dev, i, &val); + if (val != *ptr) + pci_write_config_dword(dev, i, *ptr); + } +} + +static void pciback_walk_bus_cb(struct pci_dev *dev, void *userdata) +{ + struct pcistub_sbr_list *list = (struct pcistub_sbr_list*)userdata; + struct pcistub_sbr_entry *entry; + struct pci_dev *dev_tmp; + + if (list->err != 0) + return; + + /* For PCIe endpoints we are only looking for co-assigned functions */ + if (!list->find_all && + (dev->bus->number != list->dev->bus->number || + PCI_SLOT(dev->devfn) != PCI_SLOT(list->dev->devfn))) + return; + + dev_tmp = pcistub_ref_pci_dev(dev); + if (dev_tmp == NULL) { + /* not controlled by pciback, fail */ + list->err = ENXIO; + return; + } + + entry = kzalloc(sizeof(*entry), GFP_ATOMIC); + if (entry == NULL) { + pcistub_unref_pci_dev(dev_tmp); + list->err = ENOMEM; + return; + } + + entry->dev = dev_tmp; + list_add_tail(&entry->dev_list, &list->dev_list); +} + +static void pciback_cleanup_sbr_list(struct pcistub_sbr_list *list) +{ + struct pcistub_sbr_entry *entry; + + list_for_each_entry(entry, &list->dev_list, dev_list) { + pcistub_unref_pci_dev(entry->dev); + kfree(entry); + } +} + +/* Routine to find all devices and bridges that need to be reset + * during a secondary bus reset. For PCIe this is simply all the + * functions on the particular device. For PCI this is all devices + * and bridges below the topmost PCI/PCI-X bridge. Note for PCI, + * there is at least one something->PCI/PCI-X bridge to find since + * the device is not on the host bus 0 and is on a PCI bus. + */ +static int pciback_get_sbr_list(struct pci_dev *dev, + struct pcistub_sbr_list *list, int pcie_endpoint) +{ + struct pci_dev *bridge = dev->bus->self; + struct pci_dev *last = NULL; + int exp_pos; + u16 exp_caps = 0; + + list->err = 0; + list->dev = dev; + INIT_LIST_HEAD(&list->dev_list); + + if (!pcie_endpoint) { + while (bridge) { + /* Looking for the uppermost PCI/PCI-X bridge. If it is not PCIe then + * this is a PCI/PCI-X bridge. If it is PCIe then except the PCIe to + * PCI/PCI-X type 7, the rest of the bridge types are PCIe so the last + * bridge encountered was the topmost PCI/PCI-X bridge. + */ + exp_pos = pci_find_capability(bridge, PCI_CAP_ID_EXP); + if (exp_pos != 0) { + pci_read_config_word(bridge, exp_pos + PCI_EXP_FLAGS, &exp_caps); + if (((exp_caps & PCI_EXP_FLAGS_TYPE) >> 4) != PCI_EXP_TYPE_PCI_BRIDGE) + break; /* don't want it in the list if it is a PCIe bridge */ + } + last = bridge; + bridge = last->bus->self; + } + list->bridge = last; + list->find_all = 1; /* find all devices/bridges below the topmost */ + } + else { + if (bridge) { + /* For PCIe, SBR logic is limited to PCIe endpoints behind a root/switch + * port. + */ + exp_pos = pci_find_capability(bridge, PCI_CAP_ID_EXP); + if (likely(exp_pos != 0)) { + pci_read_config_word(bridge, exp_pos + PCI_EXP_FLAGS, &exp_caps); + exp_caps = ((exp_caps & PCI_EXP_FLAGS_TYPE) >> 4); + if (exp_caps == PCI_EXP_TYPE_ROOT_PORT || + exp_caps == PCI_EXP_TYPE_UPSTREAM || + exp_caps == PCI_EXP_TYPE_DOWNSTREAM) + last = bridge; + } + } + list->bridge = last; + list->find_all = 0; /* find just functions on this slot */ + } + + /* Sanity check, there may not be any appropriate bridge to reset */ + if (!list->bridge) { + dev_dbg(&dev->dev, "No appropriate bridge to reset\n"); + return ENXIO; + } + + pci_walk_bus(list->bridge->subordinate, pciback_walk_bus_cb, list); + + if (list->err) { + pciback_cleanup_sbr_list(list); + return list->err; + } + + return 0; +} + /* Ensure a device is "turned off" and ready to be exported. * (Also see pciback_config_reset to ensure virtual configuration space is * ready to be re-exported) */ void pciback_reset_device(struct pci_dev *dev) { - u16 cmd; + u16 cmd = 0; /* Disable devices (but not bridges) */ if (dev->hdr_type == PCI_HEADER_TYPE_NORMAL) { @@ -38,6 +207,425 @@ void pciback_reset_device(struct pci_dev } } } + +/* Do a PCIe type function level reset for a single function on this + * device. + */ +static void pciback_do_pcie_flr(struct pci_dev *dev, int exp_pos) +{ + u16 status = 0; + + dev_dbg(&dev->dev, "doing PCIe FLR\n"); + + pci_block_user_cfg_access(dev); + + /* Wait for Transaction Pending bit clean */ + msleep(100); + pci_read_config_word(dev, exp_pos + PCI_EXP_DEVSTA, &status); + if (status & PCI_EXP_DEVSTA_TRPND) { + dev_dbg(&dev->dev, "Busy after 100ms while trying to reset; sleeping for 1 second\n"); + ssleep(1); + pci_read_config_word(dev, exp_pos + PCI_EXP_DEVSTA, &status); + if (status & PCI_EXP_DEVSTA_TRPND) + dev_warn(&dev->dev, "Still busy after 1s; proceeding with reset anyway\n"); + } + + pci_write_config_word(dev, exp_pos + PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_BCR_FLR); + mdelay(200); + + pciback_reload_config_space(dev); + + pci_unblock_user_cfg_access(dev); +} + +/* Do a PCI type function level reset for a single function on this + * device. This uses the Advanced Features Capability extensions to + * the PCI spec. + */ +static void pciback_do_pci_flr(struct pci_dev *dev, int af_pos, int clear_cmd) +{ + u8 status = 0; + + dev_dbg(&dev->dev, "doing PCI FLR\n"); + + pci_block_user_cfg_access(dev); + + /* Clear the command register to prevent new transactions */ + if (clear_cmd) + pci_write_config_word(dev, PCI_COMMAND, 0); + + /* Wait for Transaction Pending bit clean */ + msleep(100); + pci_read_config_byte(dev, af_pos + PCI_AF_STATUS, &status); + if (status & PCI_AF_STATUS_TP) { + dev_dbg(&dev->dev, "Busy after 100ms while trying to reset; sleeping for 1 second\n"); + ssleep(1); + pci_read_config_byte(dev, af_pos + PCI_AF_STATUS, &status); + if (status & PCI_AF_STATUS_TP) + dev_warn(&dev->dev, "Still busy after 1s; proceeding with reset anyway\n"); + } + + pci_write_config_byte(dev, af_pos + PCI_AF_CTRL, PCI_AF_CTRL_FLR); + mdelay(200); + + pciback_reload_config_space(dev); + + pci_unblock_user_cfg_access(dev); +} + +/* Vendor specific resets. These can be set in the vendor specific + * capabilities structures. Currently only the Intel USB and iGFX + * reset is supported. + */ +static int pciback_do_vendor_specific_reset(struct pci_dev *dev) +{ + struct pci_dev *gmch; + int vendor_pos, i; + u32 reg32 = 0; + u16 device_id, cmd; + u8 reg8 = 0; + + dev_dbg(&dev->dev, "doing vendor specific resets\n"); + + if (dev->vendor != PCIBACK_VENDOR_INTEL) + return -ENXIO; + + if ((dev->class >> 8) == PCIBACK_CLASS_ID_VGA) { + if (dev->bus->number != 0 || dev->devfn != PCI_DEVFN(2,0)) + return -ENXIO; + + /* Locate the GMCH (north bridge) and test for specific Intel devices */ + gmch = pci_get_bus_and_slot(0, PCI_DEVFN(0,0)); + if (!gmch) + return -ENXIO; + + device_id = gmch->device; + pci_dev_put(gmch); + + if (device_id != PCI_DEVICE_ID_INTEL_GMCHGM45) + return -ENXIO; + + /* Correct device and platform, assume vendor specific offset */ + pci_read_config_dword(dev, PCIBACK_IGFX_CAP09_OFFSET, ®32); + if ((reg32 & 0x000000FF) != PCI_CAP_ID_VNDR || + ((reg32 >> 16) & 0x000000FF) != 0x06 || + ((reg32 >> 24) & 0x000000F0) != 0x20) + return -ENXIO; + + vendor_pos = PCIBACK_IGFX_CAP09_OFFSET; + } else if ((dev->class >> 8) == PCIBACK_CLASS_ID_USB) { + vendor_pos = pci_find_capability(dev, PCI_CAP_ID_VNDR); + if (vendor_pos == 0) + return -ENXIO; + } + else + return -ENXIO; + + if ((dev->class >> 8) == PCIBACK_CLASS_ID_VGA) { + pci_write_config_byte(dev, PCIBACK_IGFX_MEDIARST_OFFSET, PCIBACK_IGFX_MEDIARST); + for (i = 0; i <= 10; i++) { + msleep(100); + pci_read_config_byte(dev, PCIBACK_IGFX_MEDIARST_OFFSET, ®8); + if ((reg8 & 0x01) == 0) + break; + if (i == 10) { + dev_warn(&dev->dev, "media not reset after 1s; skipping FLR\n"); + goto out; + } + } + + /* This specific reset will hang if the command register does not have + * memory space access enabled */ + pci_read_config_word(dev, PCI_COMMAND, &cmd); + pci_write_config_word(dev, PCI_COMMAND, (cmd | PCI_COMMAND_MEMORY)); + /* The rest is the same as a PCI AF FLR - use the same routine */ + pciback_do_pci_flr(dev, vendor_pos, 0); + pci_write_config_word(dev, PCI_COMMAND, cmd); + } else { + pci_block_user_cfg_access(dev); + + pci_write_config_byte(dev, vendor_pos + PCIBACK_USB_FLRCTRL, 1); + mdelay(200); + + pciback_reload_config_space(dev); + + pci_unblock_user_cfg_access(dev); + } + +out: + return 0; +} + +/* Use a D0-D3-D0 device state transition to reset the device. This + * is a good enough reset for some devices (like NICs). + */ +static int pciback_do_dstate_transition_reset(struct pci_dev *dev) +{ + int pm_pos; + u32 pm_ctl = 0; + + pm_pos = pci_find_capability(dev, PCI_CAP_ID_PM); + if (pm_pos == 0) + return -ENXIO; + + dev_dbg(&dev->dev, "doing Dstate transition reset\n"); + + /* No_Soft_Reset - When set 1, this bit indicates that devices + * transitioning from D3hot to D0 because of PowerState commands + * do not perform an internal reset. + */ + pci_read_config_dword(dev, pm_pos + PCI_PM_CTRL, &pm_ctl); + if (pm_ctl & PCI_PM_CTRL_NO_SOFT_RESET) + return -ENXIO; + + pci_block_user_cfg_access(dev); + + pm_ctl &= ~PCI_PM_CTRL_STATE_MASK; + pm_ctl |= PCI_PM_CTRL_D3HOT; + pci_write_config_word(dev, pm_pos + PCI_PM_CTRL, pm_ctl); + mdelay(10); + + pm_ctl &= ~PCI_PM_CTRL_STATE_MASK; + pm_ctl |= PCI_PM_CTRL_D0; + pci_write_config_word(dev, pm_pos + PCI_PM_CTRL, pm_ctl); + mdelay(10); + + pciback_reload_config_space(dev); + + pci_unblock_user_cfg_access(dev); + + return 0; +} + +/* Do a secondary bus reset on a bridge. This is only done if all + * co-assignment rules are satisfied and if it was explicitly + * requested via pciback parameters. + */ +static int pciback_do_secondary_bus_reset(struct pci_dev *dev, u32 dev_type) +{ + struct pcistub_sbr_list sbr_list; + struct pcistub_sbr_entry *entry; + u16 pci_bctl = 0; + int err = 0; + + /* Call helper to get the device list needed for the device type. */ + err = pciback_get_sbr_list(dev, &sbr_list, + (dev_type == PCIBACK_TYPE_PCIe_ENDPOINT ? 1 : 0)); + if (err) { + dev_warn(&dev->dev, + "secondary bus reset failed for device - all functions need to be co-assigned - err: %d\n", err); + return err; + } + + pci_block_user_cfg_access(dev); + + /* Reset the secondary bus and restore the PCI space for all the devfn found above. + */ + pci_read_config_word(sbr_list.bridge, PCI_BRIDGE_CONTROL, &pci_bctl); + pci_write_config_word(sbr_list.bridge, PCI_BRIDGE_CONTROL, pci_bctl | PCI_BRIDGE_CTL_BUS_RESET); + msleep(200); + pci_write_config_word(sbr_list.bridge, PCI_BRIDGE_CONTROL, pci_bctl); + msleep(200); + + list_for_each_entry(entry, &sbr_list.dev_list, dev_list) { + pciback_reload_config_space(entry->dev); + } + + pci_unblock_user_cfg_access(dev); + + pciback_cleanup_sbr_list(&sbr_list); + + return 0; +} + +/* This function is used to do a function level reset on a singe + * device/function. FLRs must be done on devices before they are + * unassigned from one domain and passed through to another. The + * preferred method is to do an actual FLR on the device but the + * functionality may not be present or exposed. In the later case + * we attempt to locate the capability even though it is not + * chained into the capabilities list. + * + * In some cases, there is no way to perform the actual FLR so we + * fall back to some alternate methods (which are not as effective + * or useful). + */ +void pciback_flr_device(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data = pci_get_drvdata(dev); + int err = 0; + + if (dev_data->no_flr) { + dev_dbg(&dev->dev, "FLR disabled for device\n"); + return; + } + dev_dbg(&dev->dev, "FLR invoked for device\n"); + + do { + /* First, always try to do an FLR */ + if (dev_data->dev_type == PCIBACK_TYPE_PCIe_ENDPOINT && + dev_data->exp_flr_offset != 0) { + pciback_do_pcie_flr(dev, dev_data->exp_flr_offset); + break; + } + if (dev_data->dev_type == PCIBACK_TYPE_PCI && + dev_data->af_flr_offset != 0) { + pciback_do_pci_flr(dev, dev_data->af_flr_offset, 1); + break; + } + + /* Next for integrated devices on the host bus 0, try some other methods */ + if (dev->bus->number == 0) { + err = pciback_do_vendor_specific_reset(dev); + if (err && dev_data->use_d3r) + err = pciback_do_dstate_transition_reset(dev); + if (err) + dev_warn(&dev->dev, "FLR functionality not supported; " + "attempts to use vendor FLR or D-states unsuccessful\n"); + break; + } + + /* Else attempt a secondary bus reset if all conditions are met */ + if (dev_data->use_sbr) { + err = pciback_do_secondary_bus_reset(dev, dev_data->dev_type); + if (err) + dev_warn(&dev->dev, "FLR functionality not supported; " + "attempts to use secondary bus reset unsuccessful;\n"); + break; + } + + err = -ENODEV; + } while (0); + + if (err) + dev_warn(&dev->dev, "FLR not performed for device\n"); +} + +/* Helper used to location the FLR capabilities for a PCIe device. + * When the capability cannot be found in the chain but is present, + * special logic is used to attempt to locate functionality. + * + * returns: the offset to the capability, zero if not found. + */ +static int pciback_find_pcie_flr_caps(struct pci_dev *dev) +{ + int exp_pos; + u32 cap = 0; + + /* First look for the PCIe FLR capabilities using the capabilities list */ + exp_pos = pci_find_capability(dev, PCI_CAP_ID_EXP); + if (exp_pos) { + pci_read_config_dword(dev, exp_pos + PCI_EXP_DEVCAP, &cap); + if (cap & PCI_EXP_DEVCAP_FLR) { + return exp_pos; + } + } + + return 0; +} + +/* Helper used to location the AF FLR capabilities for a PCI device. + * When the capability cannot be found in the chain but is present, + * special logic is used to attempt to locate functionality. + * + * returns: the offset to the capability, zero if not found. + */ +static int pciback_find_pci_flr_caps(struct pci_dev *dev) +{ + struct pci_dev *gmch; + int af_pos; + u16 device_id; + u8 cap = 0, reg8 = 0; + + /* First look for the PCI AF capabilities for FLR using the capabilities list. This + * is only used on the devices on the root/host bus (integrated devices). + */ + if (dev->bus->number != 0) + return 0; + + af_pos = pci_find_capability(dev, PCI_CAP_ID_AF); + if (af_pos) { + pci_read_config_byte(dev, af_pos + PCI_AF_CAP, &cap); + if (cap & PCI_AF_CAP_FLR) { + return af_pos; + } + } + + /* Next look for the unchained AF capabilities for FLR using specific + * logic. Currently only the graphics device on the Intel Q45 etc + * systems has special logic for locating the hidden FLR caps. + */ + do { + if (dev->bus->number != 0 || dev->devfn != PCI_DEVFN(2,0) || + dev->vendor != PCIBACK_VENDOR_INTEL || (dev->class >> 8) != PCIBACK_CLASS_ID_VGA) + break; + + /* Locate the GMCH (north bridge) and test for specific Intel devices */ + gmch = pci_get_bus_and_slot(0, PCI_DEVFN(0,0)); + if (!gmch) + break; + + device_id = gmch->device; + pci_dev_put(gmch); + + if (device_id != PCI_DEVICE_ID_INTEL_GMCHQ45 && + device_id != PCI_DEVICE_ID_INTEL_GMCHG45 && + device_id != PCI_DEVICE_ID_INTEL_GMCHG41) + break; + + /* Correct device and platform, assume AF offset */ + af_pos = PCIBACK_IGFX_CAP13_OFFSET; + pci_read_config_byte(dev, af_pos + PCI_AF_LENGTH, ®8); + if (reg8 == PCI_AF_WHAT) { + pci_read_config_byte(dev, af_pos + PCI_AF_CAP, &cap); + if (cap & PCI_AF_CAP_FLR) { + return af_pos; + } + } + } while (0); + + /* Else not found */ + return 0; +} + +/* Classify the device, specifically determine if it is PCIe/PCI + * and whether it is a PCIe endpoint, bridge, or other PCI device. + */ +void pciback_classify_device(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data; + int exp_pos; + u16 exp_caps = 0; + + dev_data = pci_get_drvdata(dev); + dev_data->dev_type = PCIBACK_TYPE_UNKNOWN; + + exp_pos = pci_find_capability(dev, PCI_CAP_ID_EXP); + + if ((dev->class >> 8) != DEV_CLASS_PCI_PCI_BRIDGE) { + if (exp_pos != 0) { + dev_data->dev_type = PCIBACK_TYPE_PCIe_ENDPOINT; + dev_data->exp_flr_offset = pciback_find_pcie_flr_caps(dev); + } else { + dev_data->dev_type = PCIBACK_TYPE_PCI; + dev_data->af_flr_offset = pciback_find_pci_flr_caps(dev); + } + goto classify_done; + } + + if (exp_pos == 0) { + dev_data->dev_type = PCIBACK_TYPE_PCI_BRIDGE; + goto classify_done; + } + + pci_read_config_word(dev, exp_pos + PCI_EXP_FLAGS, &exp_caps); + dev_data->dev_type = (((exp_caps & PCI_EXP_FLAGS_TYPE) >> 4) == PCI_EXP_TYPE_PCI_BRIDGE) ? PCIBACK_TYPE_PCI_BRIDGE : PCIBACK_TYPE_PCIe_BRIDGE; + +classify_done: + + return; +} + extern wait_queue_head_t aer_wait_queue; extern struct workqueue_struct *pciback_wq; /* @@ -132,3 +720,51 @@ irqreturn_t pciback_handle_event(int irq return IRQ_HANDLED; } + +/* Helper routine used to parse command line parameters passed to the + * pciback module from the boot loader. These params all have the form + * of a list of one or more devices, e.g.: + * (XXXX:XX:XX.X)(XXXX:XX:XX.X) + * Which is: (domain/segment:bus:dev.func) + */ +int pciback_parse_device_params(const char *device_args, int type, + int (*add_func) (int domain, int bus, int slot, int func, int type)) +{ + int pos = 0; + int err = 0; + int domain, bus, slot, func; + int parsed; + + if (device_args && *device_args) { + do { + parsed = 0; + + err = sscanf(device_args + pos, + " (%x:%x:%x.%x) %n", + &domain, &bus, &slot, &func, &parsed); + if (err != 4) { + domain = 0; + err = sscanf(device_args + pos, + " (%x:%x.%x) %n", + &bus, &slot, &func, &parsed); + if (err != 3) + goto parse_error; + } + + err = add_func(domain, bus, slot, func, type); + if (err) + goto out; + + /* if parsed<=0, we've reached the end of the string */ + pos += parsed; + } while (parsed > 0 && device_args[pos]); + } + +out: + return err; + +parse_error: + printk(KERN_ERR "pciback: Error parsing device parameters \"%s\" at \"%s\"\n", + device_args, device_args + pos); + return -EINVAL; +} diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/pci_stub.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/pci_stub.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/pci_stub.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/pci_stub.c 2009-06-07 19:04:11.000000000 -0400 @@ -24,10 +24,28 @@ wait_queue_head_t aer_wait_queue; * We want to avoid in middle of AER ops, pciback devices is being removed */ static DECLARE_RWSEM(pcistub_sem); -module_param_named(hide, pci_devs_to_hide, charp, 0444); +module_param_named(hide, pci_devs_to_hide, charp, S_IRUGO); + +static char *pci_devs_use_sbr = NULL; +module_param_named(sbr, pci_devs_use_sbr, charp, S_IRUGO); + +static char *pci_devs_use_d3r = NULL; +module_param_named(d3r, pci_devs_use_d3r, charp, S_IRUGO); + +static char *pci_devs_no_flr = NULL; +module_param_named(noflr, pci_devs_no_flr, charp, S_IRUGO); + +/* Device id list holding different device type listings + * for hiding devices and reset logic. + */ +#define PCIBACK_ID_TYPE_HIDE 1 +#define PCIBACK_ID_TYPE_SBR 2 +#define PCIBACK_ID_TYPE_D3R 3 +#define PCIBACK_ID_TYPE_NOFLR 4 struct pcistub_device_id { struct list_head slot_list; + int type; int domain; unsigned char bus; unsigned int devfn; @@ -56,6 +74,8 @@ static LIST_HEAD(pcistub_devices); static int initialize_devices = 0; static LIST_HEAD(seized_devices); +static int disable_all_flr = 0; + static struct pcistub_device *pcistub_device_alloc(struct pci_dev *dev) { struct pcistub_device *psdev; @@ -78,6 +98,23 @@ static struct pcistub_device *pcistub_de return psdev; } +static struct pciback_dev_data *pcistub_dev_data_alloc(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data; + + dev_dbg(&dev->dev, "pcistub_dev_data_alloc\n"); + + dev_data = kzalloc(sizeof(*dev_data) + dev->cfg_size, GFP_ATOMIC); + if (!dev_data) + return NULL; + + pci_set_drvdata(dev, dev_data); + + dev_data->cfg_space = (u8*)(dev_data) + sizeof(*dev_data); + + return dev_data; +} + /* Don't call this directly as it's called by pcistub_device_put */ static void pcistub_device_release(struct kref *kref) { @@ -200,7 +237,7 @@ struct pci_dev *pcistub_get_pci_dev(stru return found_dev; } -void pcistub_put_pci_dev(struct pci_dev *dev) +void pcistub_put_pci_dev(struct pci_dev *dev, int do_flr) { struct pcistub_device *psdev, *found_psdev = NULL; unsigned long flags; @@ -220,6 +257,13 @@ void pcistub_put_pci_dev(struct pci_dev * pcistub and pciback when AER is in processing */ down_write(&pcistub_sem); + + /* For pass-through devices, do an FLR (or approximate) for the device + * before it is put back and ready for the next domain + */ + if (!disable_all_flr && do_flr) + pciback_flr_device(dev); + /* Cleanup our device * (so it's ready for the next domain) */ @@ -235,6 +279,43 @@ void pcistub_put_pci_dev(struct pci_dev up_write(&pcistub_sem); } +struct pci_dev *pcistub_ref_pci_dev(struct pci_dev *dev) +{ + struct pcistub_device *psdev; + struct pci_dev *found_dev = NULL; + unsigned long flags; + + spin_lock_irqsave(&pcistub_devices_lock, flags); + + list_for_each_entry(psdev, &pcistub_devices, dev_list) { + if (psdev->dev == dev) { + pcistub_device_get(psdev); /* just a ref count */ + found_dev = psdev->dev; + break; + } + } + + spin_unlock_irqrestore(&pcistub_devices_lock, flags); + return found_dev; +} + +void pcistub_unref_pci_dev(struct pci_dev *dev) +{ + struct pcistub_device *psdev; + unsigned long flags; + + spin_lock_irqsave(&pcistub_devices_lock, flags); + + list_for_each_entry(psdev, &pcistub_devices, dev_list) { + if (psdev->dev == dev) { + pcistub_device_get(psdev); /* just an unref count */ + break; + } + } + + spin_unlock_irqrestore(&pcistub_devices_lock, flags); +} + static int __devinit pcistub_match_one(struct pci_dev *dev, struct pcistub_device_id *pdev_id) { @@ -255,7 +336,7 @@ static int __devinit pcistub_match_one(s return 0; } -static int __devinit pcistub_match(struct pci_dev *dev) +static int __devinit pcistub_match(struct pci_dev *dev, int type) { struct pcistub_device_id *pdev_id; unsigned long flags; @@ -263,6 +344,8 @@ static int __devinit pcistub_match(struc spin_lock_irqsave(&device_ids_lock, flags); list_for_each_entry(pdev_id, &pcistub_device_ids, slot_list) { + if (pdev_id->type != type) + continue; if (pcistub_match_one(dev, pdev_id)) { found = 1; break; @@ -285,12 +368,11 @@ static int __devinit pcistub_init_device * would need to be called somewhere to free the memory allocated * here and then to call kfree(pci_get_drvdata(psdev->dev)). */ - dev_data = kzalloc(sizeof(*dev_data), GFP_ATOMIC); + dev_data = pcistub_dev_data_alloc(dev); if (!dev_data) { err = -ENOMEM; goto out; } - pci_set_drvdata(dev, dev_data); dev_dbg(&dev->dev, "initializing config\n"); @@ -317,6 +399,22 @@ static int __devinit pcistub_init_device dev_dbg(&dev->dev, "reset device\n"); pciback_reset_device(dev); + /* Classify the device so we know if it is PCI/PCIe and if it is + * a bridge - this information is used for FLR logic. Also store + * values if SBR/D3R reset logic was requested. + */ + pciback_classify_device(dev); + dev_data->no_flr = pcistub_match(dev, PCIBACK_ID_TYPE_NOFLR); + if (!dev_data->no_flr) { + dev_data->use_sbr = pcistub_match(dev, PCIBACK_ID_TYPE_SBR); + dev_data->use_d3r = pcistub_match(dev, PCIBACK_ID_TYPE_D3R); + } + + /* Store the config space here where the device is off and ready to be + * exported before any FLRs or other resets are done + */ + pciback_store_config_space(dev); + return 0; config_release: @@ -414,7 +512,7 @@ static int __devinit pcistub_probe(struc dev_dbg(&dev->dev, "probing...\n"); - if (pcistub_match(dev)) { + if (pcistub_match(dev, PCIBACK_ID_TYPE_HIDE)) { if (dev->hdr_type != PCI_HEADER_TYPE_NORMAL && dev->hdr_type != PCI_HEADER_TYPE_BRIDGE) { @@ -851,7 +949,7 @@ static inline int str_to_quirk(const cha return -EINVAL; } -static int pcistub_device_id_add(int domain, int bus, int slot, int func) +static int pcistub_device_id_add(int domain, int bus, int slot, int func, int type) { struct pcistub_device_id *pci_dev_id; unsigned long flags; @@ -860,12 +958,13 @@ static int pcistub_device_id_add(int dom if (!pci_dev_id) return -ENOMEM; + pci_dev_id->type = type; pci_dev_id->domain = domain; pci_dev_id->bus = bus; pci_dev_id->devfn = PCI_DEVFN(slot, func); - pr_debug("pciback: wants to seize %04x:%02x:%02x.%01x\n", - domain, bus, slot, func); + pr_debug("pciback: adding device ID type: %d for %04x:%02x:%02x.%01x\n", + type, domain, bus, slot, func); spin_lock_irqsave(&device_ids_lock, flags); list_add_tail(&pci_dev_id->slot_list, &pcistub_device_ids); @@ -874,7 +973,7 @@ static int pcistub_device_id_add(int dom return 0; } -static int pcistub_device_id_remove(int domain, int bus, int slot, int func) +static int pcistub_device_id_remove(int domain, int bus, int slot, int func, int type) { struct pcistub_device_id *pci_dev_id, *t; int devfn = PCI_DEVFN(slot, func); @@ -884,7 +983,7 @@ static int pcistub_device_id_remove(int spin_lock_irqsave(&device_ids_lock, flags); list_for_each_entry_safe(pci_dev_id, t, &pcistub_device_ids, slot_list) { - if (pci_dev_id->domain == domain + if (pci_dev_id->type == type && pci_dev_id->domain == domain && pci_dev_id->bus == bus && pci_dev_id->devfn == devfn) { /* Don't break; here because it's possible the same * slot could be in the list more than once @@ -939,6 +1038,32 @@ static int pcistub_reg_add(int domain, i return err; } +static int pcistub_device_do_flr(int domain, int bus, int slot, int func) +{ + int err = 0; + struct pcistub_device *psdev; + struct pci_dev *dev; + + psdev = pcistub_device_find(domain, bus, slot, func); + if (!psdev || !psdev->dev) { + err = -ENODEV; + goto out; + } + dev = psdev->dev; + + /* Do an FLR (or approximate) for the device on demand and + * reload config + */ + if (!disable_all_flr) { + pciback_flr_device(dev); + } + else + dev_dbg(&dev->dev, "FLR disabled for all devices\n"); + +out: + return err; +} + static ssize_t pcistub_slot_add(struct device_driver *drv, const char *buf, size_t count) { @@ -949,7 +1074,7 @@ static ssize_t pcistub_slot_add(struct d if (err) goto out; - err = pcistub_device_id_add(domain, bus, slot, func); + err = pcistub_device_id_add(domain, bus, slot, func, PCIBACK_ID_TYPE_HIDE); out: if (!err) @@ -969,7 +1094,7 @@ static ssize_t pcistub_slot_remove(struc if (err) goto out; - err = pcistub_device_id_remove(domain, bus, slot, func); + err = pcistub_device_id_remove(domain, bus, slot, func, PCIBACK_ID_TYPE_HIDE); out: if (!err) @@ -987,6 +1112,10 @@ static ssize_t pcistub_slot_show(struct spin_lock_irqsave(&device_ids_lock, flags); list_for_each_entry(pci_dev_id, &pcistub_device_ids, slot_list) { + /* only want devices set for hide, not reset entries */ + if (pci_dev_id->type != PCIBACK_ID_TYPE_HIDE) + continue; + if (count >= PAGE_SIZE) break; @@ -1068,7 +1197,7 @@ static ssize_t pcistub_quirk_show(struct DRIVER_ATTR(quirks, S_IRUSR | S_IWUSR, pcistub_quirk_show, pcistub_quirk_add); -static ssize_t permissive_add(struct device_driver *drv, const char *buf, +static ssize_t pcistub_permissive_add(struct device_driver *drv, const char *buf, size_t count) { int domain, bus, slot, func; @@ -1109,7 +1238,7 @@ static ssize_t permissive_add(struct dev return err; } -static ssize_t permissive_show(struct device_driver *drv, char *buf) +static ssize_t pcistub_permissive_show(struct device_driver *drv, char *buf) { struct pcistub_device *psdev; struct pciback_dev_data *dev_data; @@ -1132,7 +1261,68 @@ static ssize_t permissive_show(struct de return count; } -DRIVER_ATTR(permissive, S_IRUSR | S_IWUSR, permissive_show, permissive_add); +DRIVER_ATTR(permissive, S_IRUSR | S_IWUSR, pcistub_permissive_show, pcistub_permissive_add); + +static ssize_t pcistub_do_flr(struct device_driver *drv, const char *buf, + size_t count) +{ + int domain, bus, slot, func; + int err; + + err = str_to_slot(buf, &domain, &bus, &slot, &func); + if (err) + goto out; + + err = pcistub_device_do_flr(domain, bus, slot, func); + +out: + if (!err) + err = count; + return err; +} + +DRIVER_ATTR(do_flr, S_IWUSR, NULL, pcistub_do_flr); + +static ssize_t pcistub_resets(struct device_driver *drv, const char *buf, + size_t count) +{ + int domain, bus, slot, func; + int type, err = 0; + + /* string begins with reset type specifier sbr=|dr3=|noflr= */ + if (!strncmp(buf, "sbr=", 4)) { + type = PCIBACK_ID_TYPE_SBR; + buf += 4; + } else if (!strncmp(buf, "d3r=", 4)) { + type = PCIBACK_ID_TYPE_D3R; + buf += 4; + } else if (!strncmp(buf, "noflr=", 6)) { + type = PCIBACK_ID_TYPE_NOFLR; + buf += 6; + } else { + err = -EINVAL; + goto out; + } + + /* check special wildcard noflr */ + if (type == PCIBACK_ID_TYPE_NOFLR && !strncmp(buf, "(*)", 3)) { + disable_all_flr = 1; + goto out; + } + + err = str_to_slot(buf, &domain, &bus, &slot, &func); + if (err) + goto out; + + err = pcistub_device_id_add(domain, bus, slot, func, type); + +out: + if (!err) + err = count; + return err; +} + +DRIVER_ATTR(resets, S_IWUSR, NULL, pcistub_resets); #ifdef CONFIG_PCI_MSI @@ -1158,6 +1348,8 @@ static void pcistub_exit(void) driver_remove_file(&pciback_pci_driver.driver, &driver_attr_slots); driver_remove_file(&pciback_pci_driver.driver, &driver_attr_quirks); driver_remove_file(&pciback_pci_driver.driver, &driver_attr_permissive); + driver_remove_file(&pciback_pci_driver.driver, &driver_attr_do_flr); + driver_remove_file(&pciback_pci_driver.driver, &driver_attr_resets); pci_unregister_driver(&pciback_pci_driver); WARN_ON(unregister_msi_get_owner(pciback_get_owner)); @@ -1165,35 +1357,27 @@ static void pcistub_exit(void) static int __init pcistub_init(void) { - int pos = 0; int err = 0; - int domain, bus, slot, func; - int parsed; - if (pci_devs_to_hide && *pci_devs_to_hide) { - do { - parsed = 0; - - err = sscanf(pci_devs_to_hide + pos, - " (%x:%x:%x.%x) %n", - &domain, &bus, &slot, &func, &parsed); - if (err != 4) { - domain = 0; - err = sscanf(pci_devs_to_hide + pos, - " (%x:%x.%x) %n", - &bus, &slot, &func, &parsed); - if (err != 3) - goto parse_error; - } + /* Parse device lists for hide, sbr, and d3r */ + err = pciback_parse_device_params(pci_devs_to_hide, PCIBACK_ID_TYPE_HIDE, pcistub_device_id_add); + if (err) + goto out; - err = pcistub_device_id_add(domain, bus, slot, func); - if (err) - goto out; + err = pciback_parse_device_params(pci_devs_use_sbr, PCIBACK_ID_TYPE_SBR, pcistub_device_id_add); + if (err) + goto out; - /* if parsed<=0, we've reached the end of the string */ - pos += parsed; - } while (parsed > 0 && pci_devs_to_hide[pos]); - } + err = pciback_parse_device_params(pci_devs_use_d3r, PCIBACK_ID_TYPE_D3R, pcistub_device_id_add); + if (err) + goto out; + + if (pci_devs_no_flr && *pci_devs_no_flr && !strncmp(pci_devs_no_flr, "(*)", 3)) + disable_all_flr = 1; /* check special wildcard noflr */ + else + err = pciback_parse_device_params(pci_devs_no_flr, PCIBACK_ID_TYPE_NOFLR, pcistub_device_id_add); + if (err) + goto out; /* If we're the first PCI Device Driver to register, we're the * first one to get offered PCI devices as they become @@ -1217,6 +1401,12 @@ static int __init pcistub_init(void) if (!err) err = driver_create_file(&pciback_pci_driver.driver, &driver_attr_permissive); + if (!err) + err = driver_create_file(&pciback_pci_driver.driver, + &driver_attr_do_flr); + if (!err) + err = driver_create_file(&pciback_pci_driver.driver, + &driver_attr_resets); if (!err) err = register_msi_get_owner(pciback_get_owner); @@ -1225,11 +1415,6 @@ static int __init pcistub_init(void) out: return err; - - parse_error: - printk(KERN_ERR "pciback: Error parsing pci_devs_to_hide at \"%s\"\n", - pci_devs_to_hide + pos); - return -EINVAL; } #ifndef MODULE diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/slot.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/slot.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/slot.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/slot.c 2009-06-07 19:04:11.000000000 -0400 @@ -109,7 +109,7 @@ void pciback_release_pci_dev(struct pcib spin_unlock_irqrestore(&slot_dev->lock, flags); if (found_dev) - pcistub_put_pci_dev(found_dev); + pcistub_put_pci_dev(found_dev, 0); } int pciback_init_devices(struct pciback_device *pdev) @@ -149,7 +149,7 @@ void pciback_release_devices(struct pcib for (slot = 0; slot < PCI_SLOT_MAX; slot++) { dev = slot_dev->slots[bus][slot]; if (dev != NULL) - pcistub_put_pci_dev(dev); + pcistub_put_pci_dev(dev, 0); } kfree(slot_dev); diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/vpci.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/vpci.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/vpci.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/vpci.c 2009-06-07 19:04:11.000000000 -0400 @@ -162,7 +162,7 @@ void pciback_release_pci_dev(struct pcib spin_unlock_irqrestore(&vpci_dev->lock, flags); if (found_dev) - pcistub_put_pci_dev(found_dev); + pcistub_put_pci_dev(found_dev, 0); } int pciback_init_devices(struct pciback_device *pdev) @@ -202,7 +202,7 @@ void pciback_release_devices(struct pcib list_for_each_entry_safe(e, tmp, &vpci_dev->dev_list[slot], list) { list_del(&e->list); - pcistub_put_pci_dev(e->dev); + pcistub_put_pci_dev(e->dev, 0); kfree(e); } } diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/include/linux/pci_ids.h b/linux-2.6.18-xen.hg/include/linux/pci_ids.h --- a/linux-2.6.18-xen.hg/include/linux/pci_ids.h 2009-06-07 20:48:23.000000000 -0400 +++ b/linux-2.6.18-xen.hg/include/linux/pci_ids.h 2009-06-07 19:04:11.000000000 -0400 @@ -2259,6 +2259,14 @@ #define PCI_DEVICE_ID_INTEL_IXP2800 0x9004 #define PCI_DEVICE_ID_INTEL_S21152BB 0xb152 +#define PCI_DEVICE_ID_INTEL_GMCHQ45 0x2e10 +#define PCI_DEVICE_ID_INTEL_GMCHG45 0x2e20 +#define PCI_DEVICE_ID_INTEL_MCHP45 0x2e20 +#define PCI_DEVICE_ID_INTEL_GMCHG41 0x2e30 +#define PCI_DEVICE_ID_INTEL_GMCHGM45 0x2a40 + +#define PCI_DEVICE_ID_INTEL_GMCHG41 0x2e30 + #define PCI_VENDOR_ID_SCALEMP 0x8686 #define PCI_DEVICE_ID_SCALEMP_VSMP_CTL 0x1010 [-- Attachment #4: vga-pt-patch.20090607212059 --] [-- Type: application/octet-stream, Size: 86445 bytes --] diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/controller.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/controller.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/controller.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/controller.c 2009-06-07 19:04:11.000000000 -0400 @@ -208,7 +208,7 @@ void pciback_release_pci_dev(struct pcib } spin_unlock_irqrestore(&dev_data->lock, flags); - pcistub_put_pci_dev(found_dev); + pcistub_put_pci_dev(found_dev, 0); } int pciback_init_devices(struct pciback_device *pdev) @@ -396,7 +396,7 @@ void pciback_release_devices(struct pcib list_for_each_entry_safe(dev_entry, d, &cntrl_entry->dev_list, list) { list_del(&dev_entry->list); - pcistub_put_pci_dev(dev_entry->dev); + pcistub_put_pci_dev(dev_entry->dev, 0); kfree(dev_entry); } list_del(&cntrl_entry->list); diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/passthrough.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/passthrough.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/passthrough.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/passthrough.c 2009-06-07 19:04:11.000000000 -0400 @@ -88,7 +88,7 @@ void pciback_release_pci_dev(struct pcib spin_unlock_irqrestore(&dev_data->lock, flags); if (found_dev) - pcistub_put_pci_dev(found_dev); + pcistub_put_pci_dev(found_dev, 1); } int pciback_init_devices(struct pciback_device *pdev) @@ -157,7 +157,7 @@ void pciback_release_devices(struct pcib list_for_each_entry_safe(dev_entry, t, &dev_data->dev_list, list) { list_del(&dev_entry->list); - pcistub_put_pci_dev(dev_entry->dev); + pcistub_put_pci_dev(dev_entry->dev, 1); kfree(dev_entry); } diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback.h b/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback.h --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback.h 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback.h 2009-06-07 19:04:11.000000000 -0400 @@ -25,6 +25,14 @@ struct pci_dev_entry { #define _PCIB_op_pending (1) #define PCIB_op_pending (1<<(_PCIB_op_pending)) +#define PCIBACK_TYPE_UNKNOWN 0 +#define PCIBACK_TYPE_PCIe_ENDPOINT 1 +#define PCIBACK_TYPE_PCIe_BRIDGE 2 +#define PCIBACK_TYPE_PCI_BRIDGE 3 +#define PCIBACK_TYPE_PCI 4 + +#define DEV_CLASS_PCI_PCI_BRIDGE 0x0604 + struct pciback_device { void *pci_dev_data; spinlock_t dev_lock; @@ -48,6 +56,13 @@ struct pciback_dev_data { struct list_head config_fields; int permissive; int warned_on_write; + u32 dev_type; + int no_flr; + int exp_flr_offset; + int af_flr_offset; + int use_sbr; + int use_d3r; + u8 *cfg_space; /* saved config space for device */ }; /* Get/Put PCI Devices that are hidden from the PCI Backend Domain */ @@ -56,11 +71,25 @@ struct pci_dev *pcistub_get_pci_dev_by_s int slot, int func); struct pci_dev *pcistub_get_pci_dev(struct pciback_device *pdev, struct pci_dev *dev); -void pcistub_put_pci_dev(struct pci_dev *dev); +void pcistub_put_pci_dev(struct pci_dev *dev, int do_flr); + +/* Reference/unreference PCI Devices and stubs without changing the state */ +struct pci_dev *pcistub_ref_pci_dev(struct pci_dev *dev); +void pcistub_unref_pci_dev(struct pci_dev *dev); + +/* Store/reload config space for devices */ +void pciback_store_config_space(struct pci_dev *dev); +void pciback_reload_config_space(struct pci_dev *dev); /* Ensure a device is turned off or reset */ void pciback_reset_device(struct pci_dev *pdev); +/* Do a function level reset (or approximage functionality) for device */ +void pciback_flr_device(struct pci_dev *dev); + +/* Helper to classify the device type */ +void pciback_classify_device(struct pci_dev *dev); + /* Access a virtual configuration space for a PCI device */ int pciback_config_init(void); int pciback_config_init_dev(struct pci_dev *dev); @@ -102,6 +131,10 @@ void pciback_release_devices(struct pcib irqreturn_t pciback_handle_event(int irq, void *dev_id, struct pt_regs *regs); void pciback_do_op(void *data); +/* Parse and load device specific module parameters */ +int pciback_parse_device_params(const char *device_args, int type, + int (*add_func) (int domain, int bus, int slot, int func, int type)); + int pciback_xenbus_register(void); void pciback_xenbus_unregister(void); diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback_ops.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback_ops.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback_ops.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/pciback_ops.c 2009-06-07 19:04:11.000000000 -0400 @@ -5,20 +5,189 @@ */ #include <linux/module.h> #include <linux/wait.h> +#include <linux/delay.h> /* For mdelay function */ #include <asm/bitops.h> #include <xen/evtchn.h> #include "pciback.h" +#define PCIBACK_VENDOR_INTEL 0x8086 +#define PCIBACK_CLASS_ID_USB 0x0c03 +#define PCIBACK_CLASS_ID_VGA 0x0300 +#define PCIBACK_USB_FLRCTRL 0x4 + +#define PCIBACK_IGFX_CAP09_OFFSET 0xa4 +#define PCIBACK_IGFX_CAP13_OFFSET 0xa4 + +#define PCIBACK_IGFX_MEDIARST 0x0d +#define PCIBACK_IGFX_MEDIARST_OFFSET 0xc0 + int verbose_request = 0; module_param(verbose_request, int, 0644); +struct pcistub_sbr_entry { + struct list_head dev_list; + struct pci_dev *dev; +}; + +struct pcistub_sbr_list { + struct list_head dev_list; + struct pci_dev *bridge; + struct pci_dev *dev; + int find_all; + int err; +}; + +/* Used to store the config state so it can be restored after + * resets. + */ +void pciback_store_config_space(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data = pci_get_drvdata(dev); + u32 *ptr = (u32*)dev_data->cfg_space; + int i, count = dev->cfg_size/sizeof(u32); + + for (i = 0; i < count; i += sizeof(u32), ptr++) + pci_read_config_dword(dev, i, ptr); +} + +/* Used to reload the config state after resets. + */ +void pciback_reload_config_space(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data = pci_get_drvdata(dev); + u32 *ptr = (u32*)dev_data->cfg_space; + int i, val, count = dev->cfg_size/sizeof(u32); + + for (i = 0; i < count; i += sizeof(u32), ptr++) { + pci_read_config_dword(dev, i, &val); + if (val != *ptr) + pci_write_config_dword(dev, i, *ptr); + } +} + +static void pciback_walk_bus_cb(struct pci_dev *dev, void *userdata) +{ + struct pcistub_sbr_list *list = (struct pcistub_sbr_list*)userdata; + struct pcistub_sbr_entry *entry; + struct pci_dev *dev_tmp; + + if (list->err != 0) + return; + + /* For PCIe endpoints we are only looking for co-assigned functions */ + if (!list->find_all && + (dev->bus->number != list->dev->bus->number || + PCI_SLOT(dev->devfn) != PCI_SLOT(list->dev->devfn))) + return; + + dev_tmp = pcistub_ref_pci_dev(dev); + if (dev_tmp == NULL) { + /* not controlled by pciback, fail */ + list->err = ENXIO; + return; + } + + entry = kzalloc(sizeof(*entry), GFP_ATOMIC); + if (entry == NULL) { + pcistub_unref_pci_dev(dev_tmp); + list->err = ENOMEM; + return; + } + + entry->dev = dev_tmp; + list_add_tail(&entry->dev_list, &list->dev_list); +} + +static void pciback_cleanup_sbr_list(struct pcistub_sbr_list *list) +{ + struct pcistub_sbr_entry *entry; + + list_for_each_entry(entry, &list->dev_list, dev_list) { + pcistub_unref_pci_dev(entry->dev); + kfree(entry); + } +} + +/* Routine to find all devices and bridges that need to be reset + * during a secondary bus reset. For PCIe this is simply all the + * functions on the particular device. For PCI this is all devices + * and bridges below the topmost PCI/PCI-X bridge. Note for PCI, + * there is at least one something->PCI/PCI-X bridge to find since + * the device is not on the host bus 0 and is on a PCI bus. + */ +static int pciback_get_sbr_list(struct pci_dev *dev, + struct pcistub_sbr_list *list, int pcie_endpoint) +{ + struct pci_dev *bridge = dev->bus->self; + struct pci_dev *last = NULL; + int exp_pos; + u16 exp_caps = 0; + + list->err = 0; + list->dev = dev; + INIT_LIST_HEAD(&list->dev_list); + + if (!pcie_endpoint) { + while (bridge) { + /* Looking for the uppermost PCI/PCI-X bridge. If it is not PCIe then + * this is a PCI/PCI-X bridge. If it is PCIe then except the PCIe to + * PCI/PCI-X type 7, the rest of the bridge types are PCIe so the last + * bridge encountered was the topmost PCI/PCI-X bridge. + */ + exp_pos = pci_find_capability(bridge, PCI_CAP_ID_EXP); + if (exp_pos != 0) { + pci_read_config_word(bridge, exp_pos + PCI_EXP_FLAGS, &exp_caps); + if (((exp_caps & PCI_EXP_FLAGS_TYPE) >> 4) != PCI_EXP_TYPE_PCI_BRIDGE) + break; /* don't want it in the list if it is a PCIe bridge */ + } + last = bridge; + bridge = last->bus->self; + } + list->bridge = last; + list->find_all = 1; /* find all devices/bridges below the topmost */ + } + else { + if (bridge) { + /* For PCIe, SBR logic is limited to PCIe endpoints behind a root/switch + * port. + */ + exp_pos = pci_find_capability(bridge, PCI_CAP_ID_EXP); + if (likely(exp_pos != 0)) { + pci_read_config_word(bridge, exp_pos + PCI_EXP_FLAGS, &exp_caps); + exp_caps = ((exp_caps & PCI_EXP_FLAGS_TYPE) >> 4); + if (exp_caps == PCI_EXP_TYPE_ROOT_PORT || + exp_caps == PCI_EXP_TYPE_UPSTREAM || + exp_caps == PCI_EXP_TYPE_DOWNSTREAM) + last = bridge; + } + } + list->bridge = last; + list->find_all = 0; /* find just functions on this slot */ + } + + /* Sanity check, there may not be any appropriate bridge to reset */ + if (!list->bridge) { + dev_dbg(&dev->dev, "No appropriate bridge to reset\n"); + return ENXIO; + } + + pci_walk_bus(list->bridge->subordinate, pciback_walk_bus_cb, list); + + if (list->err) { + pciback_cleanup_sbr_list(list); + return list->err; + } + + return 0; +} + /* Ensure a device is "turned off" and ready to be exported. * (Also see pciback_config_reset to ensure virtual configuration space is * ready to be re-exported) */ void pciback_reset_device(struct pci_dev *dev) { - u16 cmd; + u16 cmd = 0; /* Disable devices (but not bridges) */ if (dev->hdr_type == PCI_HEADER_TYPE_NORMAL) { @@ -38,6 +207,425 @@ void pciback_reset_device(struct pci_dev } } } + +/* Do a PCIe type function level reset for a single function on this + * device. + */ +static void pciback_do_pcie_flr(struct pci_dev *dev, int exp_pos) +{ + u16 status = 0; + + dev_dbg(&dev->dev, "doing PCIe FLR\n"); + + pci_block_user_cfg_access(dev); + + /* Wait for Transaction Pending bit clean */ + msleep(100); + pci_read_config_word(dev, exp_pos + PCI_EXP_DEVSTA, &status); + if (status & PCI_EXP_DEVSTA_TRPND) { + dev_dbg(&dev->dev, "Busy after 100ms while trying to reset; sleeping for 1 second\n"); + ssleep(1); + pci_read_config_word(dev, exp_pos + PCI_EXP_DEVSTA, &status); + if (status & PCI_EXP_DEVSTA_TRPND) + dev_warn(&dev->dev, "Still busy after 1s; proceeding with reset anyway\n"); + } + + pci_write_config_word(dev, exp_pos + PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_BCR_FLR); + mdelay(200); + + pciback_reload_config_space(dev); + + pci_unblock_user_cfg_access(dev); +} + +/* Do a PCI type function level reset for a single function on this + * device. This uses the Advanced Features Capability extensions to + * the PCI spec. + */ +static void pciback_do_pci_flr(struct pci_dev *dev, int af_pos, int clear_cmd) +{ + u8 status = 0; + + dev_dbg(&dev->dev, "doing PCI FLR\n"); + + pci_block_user_cfg_access(dev); + + /* Clear the command register to prevent new transactions */ + if (clear_cmd) + pci_write_config_word(dev, PCI_COMMAND, 0); + + /* Wait for Transaction Pending bit clean */ + msleep(100); + pci_read_config_byte(dev, af_pos + PCI_AF_STA, &status); + if (status & PCI_AF_STA_TP) { + dev_dbg(&dev->dev, "Busy after 100ms while trying to reset; sleeping for 1 second\n"); + ssleep(1); + pci_read_config_byte(dev, af_pos + PCI_AF_STA, &status); + if (status & PCI_AF_STA_TP) + dev_warn(&dev->dev, "Still busy after 1s; proceeding with reset anyway\n"); + } + + pci_write_config_byte(dev, af_pos + PCI_AF_CTRL, PCI_AF_CTRL_FLR); + mdelay(200); + + pciback_reload_config_space(dev); + + pci_unblock_user_cfg_access(dev); +} + +/* Vendor specific resets. These can be set in the vendor specific + * capabilities structures. Currently only the Intel USB and iGFX + * reset is supported. + */ +static int pciback_do_vendor_specific_reset(struct pci_dev *dev) +{ + struct pci_dev *gmch; + int vendor_pos, i; + u32 reg32 = 0; + u16 device_id, cmd; + u8 reg8 = 0; + + dev_dbg(&dev->dev, "doing vendor specific resets\n"); + + if (dev->vendor != PCIBACK_VENDOR_INTEL) + return -ENXIO; + + if ((dev->class >> 8) == PCIBACK_CLASS_ID_VGA) { + if (dev->bus->number != 0 || dev->devfn != PCI_DEVFN(2,0)) + return -ENXIO; + + /* Locate the GMCH (north bridge) and test for specific Intel devices */ + gmch = pci_get_bus_and_slot(0, PCI_DEVFN(0,0)); + if (!gmch) + return -ENXIO; + + device_id = gmch->device; + pci_dev_put(gmch); + + if (device_id != PCI_DEVICE_ID_INTEL_GMCHGM45) + return -ENXIO; + + /* Correct device and platform, assume vendor specific offset */ + pci_read_config_dword(dev, PCIBACK_IGFX_CAP09_OFFSET, ®32); + if ((reg32 & 0x000000FF) != PCI_CAP_ID_VNDR || + ((reg32 >> 16) & 0x000000FF) != 0x06 || + ((reg32 >> 24) & 0x000000F0) != 0x20) + return -ENXIO; + + vendor_pos = PCIBACK_IGFX_CAP09_OFFSET; + } else if ((dev->class >> 8) == PCIBACK_CLASS_ID_USB) { + vendor_pos = pci_find_capability(dev, PCI_CAP_ID_VNDR); + if (vendor_pos == 0) + return -ENXIO; + } + else + return -ENXIO; + + if ((dev->class >> 8) == PCIBACK_CLASS_ID_VGA) { + pci_write_config_byte(dev, PCIBACK_IGFX_MEDIARST_OFFSET, PCIBACK_IGFX_MEDIARST); + for (i = 0; i <= 10; i++) { + msleep(100); + pci_read_config_byte(dev, PCIBACK_IGFX_MEDIARST_OFFSET, ®8); + if ((reg8 & 0x01) == 0) + break; + if (i == 10) { + dev_warn(&dev->dev, "media not reset after 1s; skipping FLR\n"); + goto out; + } + } + + /* This specific reset will hang if the command register does not have + * memory space access enabled */ + pci_read_config_word(dev, PCI_COMMAND, &cmd); + pci_write_config_word(dev, PCI_COMMAND, (cmd | PCI_COMMAND_MEMORY)); + /* The rest is the same as a PCI AF FLR - use the same routine */ + pciback_do_pci_flr(dev, vendor_pos, 0); + pci_write_config_word(dev, PCI_COMMAND, cmd); + } else { + pci_block_user_cfg_access(dev); + + pci_write_config_byte(dev, vendor_pos + PCIBACK_USB_FLRCTRL, 1); + mdelay(200); + + pciback_reload_config_space(dev); + + pci_unblock_user_cfg_access(dev); + } + +out: + return 0; +} + +/* Use a D0-D3-D0 device state transition to reset the device. This + * is a good enough reset for some devices (like NICs). + */ +static int pciback_do_dstate_transition_reset(struct pci_dev *dev) +{ + int pm_pos; + u32 pm_ctl = 0; + + pm_pos = pci_find_capability(dev, PCI_CAP_ID_PM); + if (pm_pos == 0) + return -ENXIO; + + dev_dbg(&dev->dev, "doing Dstate transition reset\n"); + + /* No_Soft_Reset - When set 1, this bit indicates that devices + * transitioning from D3hot to D0 because of PowerState commands + * do not perform an internal reset. + */ + pci_read_config_dword(dev, pm_pos + PCI_PM_CTRL, &pm_ctl); + if (pm_ctl & PCI_PM_CTRL_NO_SOFT_RESET) + return -ENXIO; + + pci_block_user_cfg_access(dev); + + pm_ctl &= ~PCI_PM_CTRL_STATE_MASK; + pm_ctl |= PCI_PM_CTRL_D3HOT; + pci_write_config_word(dev, pm_pos + PCI_PM_CTRL, pm_ctl); + mdelay(10); + + pm_ctl &= ~PCI_PM_CTRL_STATE_MASK; + pm_ctl |= PCI_PM_CTRL_D0; + pci_write_config_word(dev, pm_pos + PCI_PM_CTRL, pm_ctl); + mdelay(10); + + pciback_reload_config_space(dev); + + pci_unblock_user_cfg_access(dev); + + return 0; +} + +/* Do a secondary bus reset on a bridge. This is only done if all + * co-assignment rules are satisfied and if it was explicitly + * requested via pciback parameters. + */ +static int pciback_do_secondary_bus_reset(struct pci_dev *dev, u32 dev_type) +{ + struct pcistub_sbr_list sbr_list; + struct pcistub_sbr_entry *entry; + u16 pci_bctl = 0; + int err = 0; + + /* Call helper to get the device list needed for the device type. */ + err = pciback_get_sbr_list(dev, &sbr_list, + (dev_type == PCIBACK_TYPE_PCIe_ENDPOINT ? 1 : 0)); + if (err) { + dev_warn(&dev->dev, + "secondary bus reset failed for device - all functions need to be co-assigned - err: %d\n", err); + return err; + } + + pci_block_user_cfg_access(dev); + + /* Reset the secondary bus and restore the PCI space for all the devfn found above. + */ + pci_read_config_word(sbr_list.bridge, PCI_BRIDGE_CONTROL, &pci_bctl); + pci_write_config_word(sbr_list.bridge, PCI_BRIDGE_CONTROL, pci_bctl | PCI_BRIDGE_CTL_BUS_RESET); + msleep(200); + pci_write_config_word(sbr_list.bridge, PCI_BRIDGE_CONTROL, pci_bctl); + msleep(200); + + list_for_each_entry(entry, &sbr_list.dev_list, dev_list) { + pciback_reload_config_space(entry->dev); + } + + pci_unblock_user_cfg_access(dev); + + pciback_cleanup_sbr_list(&sbr_list); + + return 0; +} + +/* This function is used to do a function level reset on a singe + * device/function. FLRs must be done on devices before they are + * unassigned from one domain and passed through to another. The + * preferred method is to do an actual FLR on the device but the + * functionality may not be present or exposed. In the later case + * we attempt to locate the capability even though it is not + * chained into the capabilities list. + * + * In some cases, there is no way to perform the actual FLR so we + * fall back to some alternate methods (which are not as effective + * or useful). + */ +void pciback_flr_device(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data = pci_get_drvdata(dev); + int err = 0; + + if (dev_data->no_flr) { + dev_dbg(&dev->dev, "FLR disabled for device\n"); + return; + } + dev_dbg(&dev->dev, "FLR invoked for device\n"); + + do { + /* First, always try to do an FLR */ + if (dev_data->dev_type == PCIBACK_TYPE_PCIe_ENDPOINT && + dev_data->exp_flr_offset != 0) { + pciback_do_pcie_flr(dev, dev_data->exp_flr_offset); + break; + } + if (dev_data->dev_type == PCIBACK_TYPE_PCI && + dev_data->af_flr_offset != 0) { + pciback_do_pci_flr(dev, dev_data->af_flr_offset, 1); + break; + } + + /* Next for integrated devices on the host bus 0, try some other methods */ + if (dev->bus->number == 0) { + err = pciback_do_vendor_specific_reset(dev); + if (err && dev_data->use_d3r) + err = pciback_do_dstate_transition_reset(dev); + if (err) + dev_warn(&dev->dev, "FLR functionality not supported; " + "attempts to use vendor FLR or D-states unsuccessful\n"); + break; + } + + /* Else attempt a secondary bus reset if all conditions are met */ + if (dev_data->use_sbr) { + err = pciback_do_secondary_bus_reset(dev, dev_data->dev_type); + if (err) + dev_warn(&dev->dev, "FLR functionality not supported; " + "attempts to use secondary bus reset unsuccessful;\n"); + break; + } + + err = -ENODEV; + } while (0); + + if (err) + dev_warn(&dev->dev, "FLR not performed for device\n"); +} + +/* Helper used to location the FLR capabilities for a PCIe device. + * When the capability cannot be found in the chain but is present, + * special logic is used to attempt to locate functionality. + * + * returns: the offset to the capability, zero if not found. + */ +static int pciback_find_pcie_flr_caps(struct pci_dev *dev) +{ + int exp_pos; + u32 cap = 0; + + /* First look for the PCIe FLR capabilities using the capabilities list */ + exp_pos = pci_find_capability(dev, PCI_CAP_ID_EXP); + if (exp_pos) { + pci_read_config_dword(dev, exp_pos + PCI_EXP_DEVCAP, &cap); + if (cap & PCI_EXP_DEVCAP_FLR) { + return exp_pos; + } + } + + return 0; +} + +/* Helper used to location the AF FLR capabilities for a PCI device. + * When the capability cannot be found in the chain but is present, + * special logic is used to attempt to locate functionality. + * + * returns: the offset to the capability, zero if not found. + */ +static int pciback_find_pci_flr_caps(struct pci_dev *dev) +{ + struct pci_dev *gmch; + int af_pos; + u16 device_id; + u8 cap = 0, reg8 = 0; + + /* First look for the PCI AF capabilities for FLR using the capabilities list. This + * is only used on the devices on the root/host bus (integrated devices). + */ + if (dev->bus->number != 0) + return 0; + + af_pos = pci_find_capability(dev, PCI_CAP_ID_AF); + if (af_pos) { + pci_read_config_byte(dev, af_pos + PCI_AF_DEVCAP, &cap); + if (cap & PCI_AF_CAP_FLR) { + return af_pos; + } + } + + /* Next look for the unchained AF capabilities for FLR using specific + * logic. Currently only the graphics device on the Intel Q45 etc + * systems has special logic for locating the hidden FLR caps. + */ + do { + if (dev->bus->number != 0 || dev->devfn != PCI_DEVFN(2,0) || + dev->vendor != PCIBACK_VENDOR_INTEL || (dev->class >> 8) != PCIBACK_CLASS_ID_VGA) + break; + + /* Locate the GMCH (north bridge) and test for specific Intel devices */ + gmch = pci_get_bus_and_slot(0, PCI_DEVFN(0,0)); + if (!gmch) + break; + + device_id = gmch->device; + pci_dev_put(gmch); + + if (device_id != PCI_DEVICE_ID_INTEL_GMCHQ45 && + device_id != PCI_DEVICE_ID_INTEL_GMCHG45 && + device_id != PCI_DEVICE_ID_INTEL_GMCHG41) + break; + + /* Correct device and platform, assume AF offset */ + af_pos = PCIBACK_IGFX_CAP13_OFFSET; + pci_read_config_byte(dev, af_pos + PCI_AF_LENFLD, ®8); + if (reg8 == PCI_AF_LENGTH) { + pci_read_config_byte(dev, af_pos + PCI_AF_DEVCAP, &cap); + if (cap & PCI_AF_CAP_FLR) { + return af_pos; + } + } + } while (0); + + /* Else not found */ + return 0; +} + +/* Classify the device, specifically determine if it is PCIe/PCI + * and whether it is a PCIe endpoint, bridge, or other PCI device. + */ +void pciback_classify_device(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data; + int exp_pos; + u16 exp_caps = 0; + + dev_data = pci_get_drvdata(dev); + dev_data->dev_type = PCIBACK_TYPE_UNKNOWN; + + exp_pos = pci_find_capability(dev, PCI_CAP_ID_EXP); + + if ((dev->class >> 8) != DEV_CLASS_PCI_PCI_BRIDGE) { + if (exp_pos != 0) { + dev_data->dev_type = PCIBACK_TYPE_PCIe_ENDPOINT; + dev_data->exp_flr_offset = pciback_find_pcie_flr_caps(dev); + } else { + dev_data->dev_type = PCIBACK_TYPE_PCI; + dev_data->af_flr_offset = pciback_find_pci_flr_caps(dev); + } + goto classify_done; + } + + if (exp_pos == 0) { + dev_data->dev_type = PCIBACK_TYPE_PCI_BRIDGE; + goto classify_done; + } + + pci_read_config_word(dev, exp_pos + PCI_EXP_FLAGS, &exp_caps); + dev_data->dev_type = (((exp_caps & PCI_EXP_FLAGS_TYPE) >> 4) == PCI_EXP_TYPE_PCI_BRIDGE) ? PCIBACK_TYPE_PCI_BRIDGE : PCIBACK_TYPE_PCIe_BRIDGE; + +classify_done: + + return; +} + extern wait_queue_head_t aer_wait_queue; extern struct workqueue_struct *pciback_wq; /* @@ -132,3 +720,51 @@ irqreturn_t pciback_handle_event(int irq return IRQ_HANDLED; } + +/* Helper routine used to parse command line parameters passed to the + * pciback module from the boot loader. These params all have the form + * of a list of one or more devices, e.g.: + * (XXXX:XX:XX.X)(XXXX:XX:XX.X) + * Which is: (domain/segment:bus:dev.func) + */ +int pciback_parse_device_params(const char *device_args, int type, + int (*add_func) (int domain, int bus, int slot, int func, int type)) +{ + int pos = 0; + int err = 0; + int domain, bus, slot, func; + int parsed; + + if (device_args && *device_args) { + do { + parsed = 0; + + err = sscanf(device_args + pos, + " (%x:%x:%x.%x) %n", + &domain, &bus, &slot, &func, &parsed); + if (err != 4) { + domain = 0; + err = sscanf(device_args + pos, + " (%x:%x.%x) %n", + &bus, &slot, &func, &parsed); + if (err != 3) + goto parse_error; + } + + err = add_func(domain, bus, slot, func, type); + if (err) + goto out; + + /* if parsed<=0, we've reached the end of the string */ + pos += parsed; + } while (parsed > 0 && device_args[pos]); + } + +out: + return err; + +parse_error: + printk(KERN_ERR "pciback: Error parsing device parameters \"%s\" at \"%s\"\n", + device_args, device_args + pos); + return -EINVAL; +} diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/pci_stub.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/pci_stub.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/pci_stub.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/pci_stub.c 2009-06-07 19:04:11.000000000 -0400 @@ -24,10 +24,28 @@ wait_queue_head_t aer_wait_queue; * We want to avoid in middle of AER ops, pciback devices is being removed */ static DECLARE_RWSEM(pcistub_sem); -module_param_named(hide, pci_devs_to_hide, charp, 0444); +module_param_named(hide, pci_devs_to_hide, charp, S_IRUGO); + +static char *pci_devs_use_sbr = NULL; +module_param_named(sbr, pci_devs_use_sbr, charp, S_IRUGO); + +static char *pci_devs_use_d3r = NULL; +module_param_named(d3r, pci_devs_use_d3r, charp, S_IRUGO); + +static char *pci_devs_no_flr = NULL; +module_param_named(noflr, pci_devs_no_flr, charp, S_IRUGO); + +/* Device id list holding different device type listings + * for hiding devices and reset logic. + */ +#define PCIBACK_ID_TYPE_HIDE 1 +#define PCIBACK_ID_TYPE_SBR 2 +#define PCIBACK_ID_TYPE_D3R 3 +#define PCIBACK_ID_TYPE_NOFLR 4 struct pcistub_device_id { struct list_head slot_list; + int type; int domain; unsigned char bus; unsigned int devfn; @@ -56,6 +74,8 @@ static LIST_HEAD(pcistub_devices); static int initialize_devices = 0; static LIST_HEAD(seized_devices); +static int disable_all_flr = 0; + static struct pcistub_device *pcistub_device_alloc(struct pci_dev *dev) { struct pcistub_device *psdev; @@ -78,6 +98,23 @@ static struct pcistub_device *pcistub_de return psdev; } +static struct pciback_dev_data *pcistub_dev_data_alloc(struct pci_dev *dev) +{ + struct pciback_dev_data *dev_data; + + dev_dbg(&dev->dev, "pcistub_dev_data_alloc\n"); + + dev_data = kzalloc(sizeof(*dev_data) + dev->cfg_size, GFP_ATOMIC); + if (!dev_data) + return NULL; + + pci_set_drvdata(dev, dev_data); + + dev_data->cfg_space = (u8*)(dev_data) + sizeof(*dev_data); + + return dev_data; +} + /* Don't call this directly as it's called by pcistub_device_put */ static void pcistub_device_release(struct kref *kref) { @@ -200,7 +237,7 @@ struct pci_dev *pcistub_get_pci_dev(stru return found_dev; } -void pcistub_put_pci_dev(struct pci_dev *dev) +void pcistub_put_pci_dev(struct pci_dev *dev, int do_flr) { struct pcistub_device *psdev, *found_psdev = NULL; unsigned long flags; @@ -220,6 +257,13 @@ void pcistub_put_pci_dev(struct pci_dev * pcistub and pciback when AER is in processing */ down_write(&pcistub_sem); + + /* For pass-through devices, do an FLR (or approximate) for the device + * before it is put back and ready for the next domain + */ + if (!disable_all_flr && do_flr) + pciback_flr_device(dev); + /* Cleanup our device * (so it's ready for the next domain) */ @@ -235,6 +279,43 @@ void pcistub_put_pci_dev(struct pci_dev up_write(&pcistub_sem); } +struct pci_dev *pcistub_ref_pci_dev(struct pci_dev *dev) +{ + struct pcistub_device *psdev; + struct pci_dev *found_dev = NULL; + unsigned long flags; + + spin_lock_irqsave(&pcistub_devices_lock, flags); + + list_for_each_entry(psdev, &pcistub_devices, dev_list) { + if (psdev->dev == dev) { + pcistub_device_get(psdev); /* just a ref count */ + found_dev = psdev->dev; + break; + } + } + + spin_unlock_irqrestore(&pcistub_devices_lock, flags); + return found_dev; +} + +void pcistub_unref_pci_dev(struct pci_dev *dev) +{ + struct pcistub_device *psdev; + unsigned long flags; + + spin_lock_irqsave(&pcistub_devices_lock, flags); + + list_for_each_entry(psdev, &pcistub_devices, dev_list) { + if (psdev->dev == dev) { + pcistub_device_get(psdev); /* just an unref count */ + break; + } + } + + spin_unlock_irqrestore(&pcistub_devices_lock, flags); +} + static int __devinit pcistub_match_one(struct pci_dev *dev, struct pcistub_device_id *pdev_id) { @@ -255,7 +336,7 @@ static int __devinit pcistub_match_one(s return 0; } -static int __devinit pcistub_match(struct pci_dev *dev) +static int __devinit pcistub_match(struct pci_dev *dev, int type) { struct pcistub_device_id *pdev_id; unsigned long flags; @@ -263,6 +344,8 @@ static int __devinit pcistub_match(struc spin_lock_irqsave(&device_ids_lock, flags); list_for_each_entry(pdev_id, &pcistub_device_ids, slot_list) { + if (pdev_id->type != type) + continue; if (pcistub_match_one(dev, pdev_id)) { found = 1; break; @@ -285,12 +368,11 @@ static int __devinit pcistub_init_device * would need to be called somewhere to free the memory allocated * here and then to call kfree(pci_get_drvdata(psdev->dev)). */ - dev_data = kzalloc(sizeof(*dev_data), GFP_ATOMIC); + dev_data = pcistub_dev_data_alloc(dev); if (!dev_data) { err = -ENOMEM; goto out; } - pci_set_drvdata(dev, dev_data); dev_dbg(&dev->dev, "initializing config\n"); @@ -317,6 +399,22 @@ static int __devinit pcistub_init_device dev_dbg(&dev->dev, "reset device\n"); pciback_reset_device(dev); + /* Classify the device so we know if it is PCI/PCIe and if it is + * a bridge - this information is used for FLR logic. Also store + * values if SBR/D3R reset logic was requested. + */ + pciback_classify_device(dev); + dev_data->no_flr = pcistub_match(dev, PCIBACK_ID_TYPE_NOFLR); + if (!dev_data->no_flr) { + dev_data->use_sbr = pcistub_match(dev, PCIBACK_ID_TYPE_SBR); + dev_data->use_d3r = pcistub_match(dev, PCIBACK_ID_TYPE_D3R); + } + + /* Store the config space here where the device is off and ready to be + * exported before any FLRs or other resets are done + */ + pciback_store_config_space(dev); + return 0; config_release: @@ -414,7 +512,7 @@ static int __devinit pcistub_probe(struc dev_dbg(&dev->dev, "probing...\n"); - if (pcistub_match(dev)) { + if (pcistub_match(dev, PCIBACK_ID_TYPE_HIDE)) { if (dev->hdr_type != PCI_HEADER_TYPE_NORMAL && dev->hdr_type != PCI_HEADER_TYPE_BRIDGE) { @@ -851,7 +949,7 @@ static inline int str_to_quirk(const cha return -EINVAL; } -static int pcistub_device_id_add(int domain, int bus, int slot, int func) +static int pcistub_device_id_add(int domain, int bus, int slot, int func, int type) { struct pcistub_device_id *pci_dev_id; unsigned long flags; @@ -860,12 +958,13 @@ static int pcistub_device_id_add(int dom if (!pci_dev_id) return -ENOMEM; + pci_dev_id->type = type; pci_dev_id->domain = domain; pci_dev_id->bus = bus; pci_dev_id->devfn = PCI_DEVFN(slot, func); - pr_debug("pciback: wants to seize %04x:%02x:%02x.%01x\n", - domain, bus, slot, func); + pr_debug("pciback: adding device ID type: %d for %04x:%02x:%02x.%01x\n", + type, domain, bus, slot, func); spin_lock_irqsave(&device_ids_lock, flags); list_add_tail(&pci_dev_id->slot_list, &pcistub_device_ids); @@ -874,7 +973,7 @@ static int pcistub_device_id_add(int dom return 0; } -static int pcistub_device_id_remove(int domain, int bus, int slot, int func) +static int pcistub_device_id_remove(int domain, int bus, int slot, int func, int type) { struct pcistub_device_id *pci_dev_id, *t; int devfn = PCI_DEVFN(slot, func); @@ -884,7 +983,7 @@ static int pcistub_device_id_remove(int spin_lock_irqsave(&device_ids_lock, flags); list_for_each_entry_safe(pci_dev_id, t, &pcistub_device_ids, slot_list) { - if (pci_dev_id->domain == domain + if (pci_dev_id->type == type && pci_dev_id->domain == domain && pci_dev_id->bus == bus && pci_dev_id->devfn == devfn) { /* Don't break; here because it's possible the same * slot could be in the list more than once @@ -939,6 +1038,32 @@ static int pcistub_reg_add(int domain, i return err; } +static int pcistub_device_do_flr(int domain, int bus, int slot, int func) +{ + int err = 0; + struct pcistub_device *psdev; + struct pci_dev *dev; + + psdev = pcistub_device_find(domain, bus, slot, func); + if (!psdev || !psdev->dev) { + err = -ENODEV; + goto out; + } + dev = psdev->dev; + + /* Do an FLR (or approximate) for the device on demand and + * reload config + */ + if (!disable_all_flr) { + pciback_flr_device(dev); + } + else + dev_dbg(&dev->dev, "FLR disabled for all devices\n"); + +out: + return err; +} + static ssize_t pcistub_slot_add(struct device_driver *drv, const char *buf, size_t count) { @@ -949,7 +1074,7 @@ static ssize_t pcistub_slot_add(struct d if (err) goto out; - err = pcistub_device_id_add(domain, bus, slot, func); + err = pcistub_device_id_add(domain, bus, slot, func, PCIBACK_ID_TYPE_HIDE); out: if (!err) @@ -969,7 +1094,7 @@ static ssize_t pcistub_slot_remove(struc if (err) goto out; - err = pcistub_device_id_remove(domain, bus, slot, func); + err = pcistub_device_id_remove(domain, bus, slot, func, PCIBACK_ID_TYPE_HIDE); out: if (!err) @@ -987,6 +1112,10 @@ static ssize_t pcistub_slot_show(struct spin_lock_irqsave(&device_ids_lock, flags); list_for_each_entry(pci_dev_id, &pcistub_device_ids, slot_list) { + /* only want devices set for hide, not reset entries */ + if (pci_dev_id->type != PCIBACK_ID_TYPE_HIDE) + continue; + if (count >= PAGE_SIZE) break; @@ -1068,7 +1197,7 @@ static ssize_t pcistub_quirk_show(struct DRIVER_ATTR(quirks, S_IRUSR | S_IWUSR, pcistub_quirk_show, pcistub_quirk_add); -static ssize_t permissive_add(struct device_driver *drv, const char *buf, +static ssize_t pcistub_permissive_add(struct device_driver *drv, const char *buf, size_t count) { int domain, bus, slot, func; @@ -1109,7 +1238,7 @@ static ssize_t permissive_add(struct dev return err; } -static ssize_t permissive_show(struct device_driver *drv, char *buf) +static ssize_t pcistub_permissive_show(struct device_driver *drv, char *buf) { struct pcistub_device *psdev; struct pciback_dev_data *dev_data; @@ -1132,7 +1261,68 @@ static ssize_t permissive_show(struct de return count; } -DRIVER_ATTR(permissive, S_IRUSR | S_IWUSR, permissive_show, permissive_add); +DRIVER_ATTR(permissive, S_IRUSR | S_IWUSR, pcistub_permissive_show, pcistub_permissive_add); + +static ssize_t pcistub_do_flr(struct device_driver *drv, const char *buf, + size_t count) +{ + int domain, bus, slot, func; + int err; + + err = str_to_slot(buf, &domain, &bus, &slot, &func); + if (err) + goto out; + + err = pcistub_device_do_flr(domain, bus, slot, func); + +out: + if (!err) + err = count; + return err; +} + +DRIVER_ATTR(do_flr, S_IWUSR, NULL, pcistub_do_flr); + +static ssize_t pcistub_resets(struct device_driver *drv, const char *buf, + size_t count) +{ + int domain, bus, slot, func; + int type, err = 0; + + /* string begins with reset type specifier sbr=|dr3=|noflr= */ + if (!strncmp(buf, "sbr=", 4)) { + type = PCIBACK_ID_TYPE_SBR; + buf += 4; + } else if (!strncmp(buf, "d3r=", 4)) { + type = PCIBACK_ID_TYPE_D3R; + buf += 4; + } else if (!strncmp(buf, "noflr=", 6)) { + type = PCIBACK_ID_TYPE_NOFLR; + buf += 6; + } else { + err = -EINVAL; + goto out; + } + + /* check special wildcard noflr */ + if (type == PCIBACK_ID_TYPE_NOFLR && !strncmp(buf, "(*)", 3)) { + disable_all_flr = 1; + goto out; + } + + err = str_to_slot(buf, &domain, &bus, &slot, &func); + if (err) + goto out; + + err = pcistub_device_id_add(domain, bus, slot, func, type); + +out: + if (!err) + err = count; + return err; +} + +DRIVER_ATTR(resets, S_IWUSR, NULL, pcistub_resets); #ifdef CONFIG_PCI_MSI @@ -1158,6 +1348,8 @@ static void pcistub_exit(void) driver_remove_file(&pciback_pci_driver.driver, &driver_attr_slots); driver_remove_file(&pciback_pci_driver.driver, &driver_attr_quirks); driver_remove_file(&pciback_pci_driver.driver, &driver_attr_permissive); + driver_remove_file(&pciback_pci_driver.driver, &driver_attr_do_flr); + driver_remove_file(&pciback_pci_driver.driver, &driver_attr_resets); pci_unregister_driver(&pciback_pci_driver); WARN_ON(unregister_msi_get_owner(pciback_get_owner)); @@ -1165,35 +1357,27 @@ static void pcistub_exit(void) static int __init pcistub_init(void) { - int pos = 0; int err = 0; - int domain, bus, slot, func; - int parsed; - if (pci_devs_to_hide && *pci_devs_to_hide) { - do { - parsed = 0; - - err = sscanf(pci_devs_to_hide + pos, - " (%x:%x:%x.%x) %n", - &domain, &bus, &slot, &func, &parsed); - if (err != 4) { - domain = 0; - err = sscanf(pci_devs_to_hide + pos, - " (%x:%x.%x) %n", - &bus, &slot, &func, &parsed); - if (err != 3) - goto parse_error; - } + /* Parse device lists for hide, sbr, and d3r */ + err = pciback_parse_device_params(pci_devs_to_hide, PCIBACK_ID_TYPE_HIDE, pcistub_device_id_add); + if (err) + goto out; - err = pcistub_device_id_add(domain, bus, slot, func); - if (err) - goto out; + err = pciback_parse_device_params(pci_devs_use_sbr, PCIBACK_ID_TYPE_SBR, pcistub_device_id_add); + if (err) + goto out; - /* if parsed<=0, we've reached the end of the string */ - pos += parsed; - } while (parsed > 0 && pci_devs_to_hide[pos]); - } + err = pciback_parse_device_params(pci_devs_use_d3r, PCIBACK_ID_TYPE_D3R, pcistub_device_id_add); + if (err) + goto out; + + if (pci_devs_no_flr && *pci_devs_no_flr && !strncmp(pci_devs_no_flr, "(*)", 3)) + disable_all_flr = 1; /* check special wildcard noflr */ + else + err = pciback_parse_device_params(pci_devs_no_flr, PCIBACK_ID_TYPE_NOFLR, pcistub_device_id_add); + if (err) + goto out; /* If we're the first PCI Device Driver to register, we're the * first one to get offered PCI devices as they become @@ -1217,6 +1401,12 @@ static int __init pcistub_init(void) if (!err) err = driver_create_file(&pciback_pci_driver.driver, &driver_attr_permissive); + if (!err) + err = driver_create_file(&pciback_pci_driver.driver, + &driver_attr_do_flr); + if (!err) + err = driver_create_file(&pciback_pci_driver.driver, + &driver_attr_resets); if (!err) err = register_msi_get_owner(pciback_get_owner); @@ -1225,11 +1415,6 @@ static int __init pcistub_init(void) out: return err; - - parse_error: - printk(KERN_ERR "pciback: Error parsing pci_devs_to_hide at \"%s\"\n", - pci_devs_to_hide + pos); - return -EINVAL; } #ifndef MODULE diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/slot.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/slot.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/slot.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/slot.c 2009-06-07 19:04:11.000000000 -0400 @@ -109,7 +109,7 @@ void pciback_release_pci_dev(struct pcib spin_unlock_irqrestore(&slot_dev->lock, flags); if (found_dev) - pcistub_put_pci_dev(found_dev); + pcistub_put_pci_dev(found_dev, 0); } int pciback_init_devices(struct pciback_device *pdev) @@ -149,7 +149,7 @@ void pciback_release_devices(struct pcib for (slot = 0; slot < PCI_SLOT_MAX; slot++) { dev = slot_dev->slots[bus][slot]; if (dev != NULL) - pcistub_put_pci_dev(dev); + pcistub_put_pci_dev(dev, 0); } kfree(slot_dev); diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/drivers/xen/pciback/vpci.c b/linux-2.6.18-xen.hg/drivers/xen/pciback/vpci.c --- a/linux-2.6.18-xen.hg/drivers/xen/pciback/vpci.c 2009-06-07 20:48:21.000000000 -0400 +++ b/linux-2.6.18-xen.hg/drivers/xen/pciback/vpci.c 2009-06-07 19:04:11.000000000 -0400 @@ -162,7 +162,7 @@ void pciback_release_pci_dev(struct pcib spin_unlock_irqrestore(&vpci_dev->lock, flags); if (found_dev) - pcistub_put_pci_dev(found_dev); + pcistub_put_pci_dev(found_dev, 0); } int pciback_init_devices(struct pciback_device *pdev) @@ -202,7 +202,7 @@ void pciback_release_devices(struct pcib list_for_each_entry_safe(e, tmp, &vpci_dev->dev_list[slot], list) { list_del(&e->list); - pcistub_put_pci_dev(e->dev); + pcistub_put_pci_dev(e->dev, 0); kfree(e); } } diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/include/linux/pci_ids.h b/linux-2.6.18-xen.hg/include/linux/pci_ids.h --- a/linux-2.6.18-xen.hg/include/linux/pci_ids.h 2009-06-07 20:48:23.000000000 -0400 +++ b/linux-2.6.18-xen.hg/include/linux/pci_ids.h 2009-06-07 19:04:11.000000000 -0400 @@ -2259,6 +2259,14 @@ #define PCI_DEVICE_ID_INTEL_IXP2800 0x9004 #define PCI_DEVICE_ID_INTEL_S21152BB 0xb152 +#define PCI_DEVICE_ID_INTEL_GMCHQ45 0x2e10 +#define PCI_DEVICE_ID_INTEL_GMCHG45 0x2e20 +#define PCI_DEVICE_ID_INTEL_MCHP45 0x2e20 +#define PCI_DEVICE_ID_INTEL_GMCHG41 0x2e30 +#define PCI_DEVICE_ID_INTEL_GMCHGM45 0x2a40 + +#define PCI_DEVICE_ID_INTEL_GMCHG41 0x2e30 + #define PCI_VENDOR_ID_SCALEMP 0x8686 #define PCI_DEVICE_ID_SCALEMP_VSMP_CTL 0x1010 diff -rupN --ignore-blank-lines-X diffignore a/linux-2.6.18-xen.hg/include/linux/pci_regs.h b/linux-2.6.18-xen.hg/include/linux/pci_regs.h --- a/linux-2.6.18-xen.hg/include/linux/pci_regs.h 2009-06-07 20:48:23.000000000 -0400 +++ b/linux-2.6.18-xen.hg/include/linux/pci_regs.h 2009-06-07 19:04:11.000000000 -0400 @@ -201,6 +201,7 @@ #define PCI_CAP_ID_SHPC 0x0C /* PCI Standard Hot-Plug Controller */ #define PCI_CAP_ID_EXP 0x10 /* PCI Express */ #define PCI_CAP_ID_MSIX 0x11 /* MSI-X */ +#define PCI_CAP_ID_AF 0x13 /* Advanced Features Capability */ #define PCI_CAP_LIST_NEXT 1 /* Next capability in the list */ #define PCI_CAP_FLAGS 2 /* Capability defined flags (16 bits) */ #define PCI_CAP_SIZEOF 4 @@ -229,6 +230,11 @@ #define PCI_PM_CTRL_DATA_SEL_MASK 0x1e00 /* Data select (??) */ #define PCI_PM_CTRL_DATA_SCALE_MASK 0x6000 /* Data scale (??) */ #define PCI_PM_CTRL_PME_STATUS 0x8000 /* PME pin status */ +#define PCI_PM_CTRL_DATA_DSTATE_MASK 0x3 /* D0 - D3 */ +#define PCI_PM_CTRL_D0 0x0 +#define PCI_PM_CTRL_D1 0x1 +#define PCI_PM_CTRL_D2 0x2 +#define PCI_PM_CTRL_D3HOT 0x3 #define PCI_PM_PPB_EXTENSIONS 6 /* PPB support extensions (??) */ #define PCI_PM_PPB_B2_B3 0x40 /* Stop clock when in D3hot (??) */ #define PCI_PM_BPCC_ENABLE 0x80 /* Bus power/clock control enable (??) */ @@ -346,6 +352,7 @@ #define PCI_EXP_DEVCAP_PWR_IND 0x4000 /* Power Indicator Present */ #define PCI_EXP_DEVCAP_PWR_VAL 0x3fc0000 /* Slot Power Limit Value */ #define PCI_EXP_DEVCAP_PWR_SCL 0xc000000 /* Slot Power Limit Scale */ +#define PCI_EXP_DEVCAP_FLR 0x10000000 /* Function Level Reset */ #define PCI_EXP_DEVCTL 8 /* Device Control */ #define PCI_EXP_DEVCTL_CERE 0x0001 /* Correctable Error Reporting En. */ #define PCI_EXP_DEVCTL_NFERE 0x0002 /* Non-Fatal Error Reporting Enable */ @@ -358,6 +365,7 @@ #define PCI_EXP_DEVCTL_AUX_PME 0x0400 /* Auxiliary Power PM Enable */ #define PCI_EXP_DEVCTL_NOSNOOP_EN 0x0800 /* Enable No Snoop */ #define PCI_EXP_DEVCTL_READRQ 0x7000 /* Max_Read_Request_Size */ +#define PCI_EXP_DEVCTL_BCR_FLR 0x8000 /* Bridge Configuration Retry / FLR */ #define PCI_EXP_DEVSTA 10 /* Device Status */ #define PCI_EXP_DEVSTA_CED 0x01 /* Correctable Error Detected */ #define PCI_EXP_DEVSTA_NFED 0x02 /* Non-Fatal Error Detected */ @@ -482,6 +490,17 @@ #define PCI_ARI_CTRL_ACS 0x0002 /* ACS Function Groups Enable */ #define PCI_ARI_CTRL_FG(x) (((x) >> 4) & 7) /* Function Group */ +/* Advanced Features Capability */ +#define PCI_AF_LENFLD 0x02 /* Device length offset */ +#define PCI_AF_LENGTH 0x06 +#define PCI_AF_DEVCAP 0x03 /* Device capabilities offset */ +#define PCI_AF_CAP_TP 0x01 +#define PCI_AF_CAP_FLR 0x02 +#define PCI_AF_CTRL 0x04 /* Device CTRL offset */ +#define PCI_AF_CTRL_FLR 0x01 +#define PCI_AF_STA 0x05 /* Device STATUS offset */ +#define PCI_AF_STA_TP 0x01 + /* Single Root I/O Virtualization */ #define PCI_SRIOV_CAP 0x04 /* SR-IOV Capabilities */ #define PCI_SRIOV_CAP_VFM 0x01 /* VF Migration Capable */ diff -rupN --ignore-blank-lines-X diffignore a/tools/firmware/hvmloader/hvmloader.c b/tools/firmware/hvmloader/hvmloader.c --- a/tools/firmware/hvmloader/hvmloader.c 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/firmware/hvmloader/hvmloader.c 2009-06-07 19:04:11.000000000 -0400 @@ -673,6 +673,7 @@ int main(void) break; default: printf("No emulated VGA adaptor ...\n"); + vgabios_sz = round_option_rom((*(uint8_t *)(VGABIOS_PHYSICAL_ADDRESS+2)) * 512); break; } diff -rupN --ignore-blank-lines-X diffignore a/tools/ioemu-remote/console.h b/tools/ioemu-remote/console.h --- a/tools/ioemu-remote/console.h 2009-06-07 21:14:47.000000000 -0400 +++ b/tools/ioemu-remote/console.h 2009-06-07 19:04:11.000000000 -0400 @@ -290,6 +290,9 @@ void vga_hw_update(void); void vga_hw_invalidate(void); void vga_hw_screen_dump(const char *filename); +void unset_vga_acc(void); +void set_vga_acc(void); + int is_graphic_console(void); int is_fixedsize_console(void); CharDriverState *text_console_init(const char *p); @@ -341,4 +344,9 @@ const char *readline_get_history(unsigne void readline_start(const char *prompt, int is_password, ReadLineFunc *readline_func, void *opaque); +/* intel.c */ +int intel_enter(void); +int intel_leave(void); +void intel_display_init(DisplayState *ds); + #endif diff -rupN --ignore-blank-lines-X diffignore a/tools/ioemu-remote/hw/pass-through.c b/tools/ioemu-remote/hw/pass-through.c --- a/tools/ioemu-remote/hw/pass-through.c 2009-06-07 21:14:48.000000000 -0400 +++ b/tools/ioemu-remote/hw/pass-through.c 2009-06-07 19:04:11.000000000 -0400 @@ -90,6 +90,8 @@ #include "qemu-xen.h" #include <unistd.h> +extern int vga_passthrough; + struct php_dev { struct pt_dev *pt_dev; uint8_t valid; @@ -1573,10 +1575,11 @@ static int pt_dev_is_virtfn(struct pci_d static int pt_register_regions(struct pt_dev *assigned_device) { - int i = 0; + int i = 0, ret = 0; uint32_t bar_data = 0; struct pci_dev *pci_dev = assigned_device->pci_dev; PCIDevice *d = &assigned_device->dev; + uint16_t class, vendor_id; /* Register PIO/MMIO BARs */ for ( i = 0; i < PCI_BAR_ENTRIES; i++ ) @@ -1632,6 +1635,25 @@ static int pt_register_regions(struct pt (uint32_t)(pci_dev->rom_size), (uint32_t)(pci_dev->rom_base_addr)); } + /* Map legacy ioport and iomem, for specific devices */ + vendor_id = pci_read_word(pci_dev, 0x00); + class = pci_read_word(pci_dev, 0x0a); + + PT_LOG("Real device vendor_id=0x%x class=0x%x\n", vendor_id, class); + if ( vga_passthrough && class == 0x0300 ) + { + PT_LOG("add an intel graphic card\n"); + + ret = xc_domain_ioport_mapping(xc_handle, domid, 0x3B0, 0x3B0, 0xb, DPCI_ADD_MAPPING); + ret = xc_domain_ioport_mapping(xc_handle, domid, 0x3C0, 0x3C0, 32, DPCI_ADD_MAPPING); + ret |= xc_domain_memory_mapping(xc_handle, domid, 0xa0, 0xa0, 32, DPCI_ADD_MAPPING); + if ( ret != 0 ) + { + PT_LOG("legacy mapping failed!\n"); + return ret; + } + } + return 0; } @@ -1640,6 +1662,7 @@ static void pt_unregister_regions(struct int i, type, ret; uint32_t e_size; PCIDevice *d = (PCIDevice*)assigned_device; + uint16_t class, vendor_id; for ( i = 0; i < PCI_NUM_REGIONS; i++ ) { @@ -1681,6 +1704,24 @@ static void pt_unregister_regions(struct } + /* unmap legacy ioport and iomem, for specific devices */ + vendor_id = pci_read_word(assigned_device->pci_dev, 0x00); + class = pci_read_word(assigned_device->pci_dev, 0x0a); + + PT_LOG("Real device vendor_id=0x%x class=0x%x\n", vendor_id, class); + if ( vga_passthrough && class == 0x0300 ) + { + PT_LOG("remove an intel graphic card\n"); + + ret = xc_domain_ioport_mapping(xc_handle, domid, 0x3B0, 0x3B0, 0xb, DPCI_REMOVE_MAPPING); + ret = xc_domain_ioport_mapping(xc_handle, domid, 0x3C0, 0x3C0, 32, DPCI_REMOVE_MAPPING); + ret |= xc_domain_memory_mapping(xc_handle, domid, 0xa0, 0xa0, 32, DPCI_REMOVE_MAPPING); + if ( ret != 0 ) + { + PT_LOG("legacy unmapping failed !\n"); + } + } + } static uint8_t find_cap_offset(struct pci_dev *pci_dev, uint8_t cap) @@ -3759,7 +3800,7 @@ static struct pt_dev * register_real_dev struct pci_config_cf8 machine_bdf; char *key, *val; int msi_translate, power_mgmt; - + PT_LOG("Assigning real physical device %02x:%02x.%x ...\n", r_bus, r_dev, r_func); @@ -3931,7 +3972,7 @@ static struct pt_dev * register_real_dev } out: - PT_LOG("Real physical device %02x:%02x.%x registered successfuly!\n" + PT_LOG("Real physical device %02x:%02x.%x registered successfully!\n" "IRQ type = %s\n", r_bus, r_dev, r_func, assigned_device->msi_trans_en? "MSI-INTx":"INTx"); @@ -4113,3 +4154,47 @@ err: return status; } +u8 pt_pci_host_read_byte(int bus, int dev, int fn, u32 addr) +{ + struct pci_dev *pci_dev; + u8 val; + + pci_dev = pci_get_dev(dpci_infos.pci_access, 0, bus, dev, fn); + if (!pci_dev) + return 0; + + val = pci_read_byte(pci_dev, addr); + pci_free_dev(pci_dev); + + return val; +} + +u16 pt_pci_host_read_word(int bus, int dev, int fn, u32 addr) +{ + struct pci_dev *pci_dev; + u16 val; + + pci_dev = pci_get_dev(dpci_infos.pci_access, 0, bus, dev, fn); + if (!pci_dev) + return 0; + + val = pci_read_word(pci_dev, addr); + pci_free_dev(pci_dev); + + return val; +} + +u32 pt_pci_host_read_long(int bus, int dev, int fn, u32 addr) +{ + struct pci_dev *pci_dev; + u32 val; + + pci_dev = pci_get_dev(dpci_infos.pci_access, 0, bus, dev, fn); + if (!pci_dev) + return 0; + + val = pci_read_long(pci_dev, addr); + pci_free_dev(pci_dev); + + return val; +} diff -rupN --ignore-blank-lines-X diffignore a/tools/ioemu-remote/hw/pc.c b/tools/ioemu-remote/hw/pc.c --- a/tools/ioemu-remote/hw/pc.c 2009-06-07 21:14:47.000000000 -0400 +++ b/tools/ioemu-remote/hw/pc.c 2009-06-07 19:04:11.000000000 -0400 @@ -65,6 +65,8 @@ void tpm_tis_init(SetIRQFunc *set_irq, v extern uint8_t *acpi_tables; extern size_t acpi_tables_len; +extern int vga_passthrough; + static fdctrl_t *floppy_controller; static RTCState *rtc_state; static PITState *pit; @@ -984,6 +986,7 @@ vga_bios_error: register_ioport_write(0xf0, 1, 1, ioportF0_write, NULL); + if (!vga_passthrough) { if (cirrus_vga_enabled) { if (pci_enabled) { pci_cirrus_vga_init(pci_bus, @@ -1010,6 +1013,7 @@ vga_bios_error: vga_ram_addr, vga_ram_size); } } + } #ifdef CONFIG_PASSTHROUGH /* Pass-through Initialization diff -rupN --ignore-blank-lines-X diffignore a/tools/ioemu-remote/hw/pci.c b/tools/ioemu-remote/hw/pci.c --- a/tools/ioemu-remote/hw/pci.c 2009-06-07 21:14:47.000000000 -0400 +++ b/tools/ioemu-remote/hw/pci.c 2009-06-07 19:04:11.000000000 -0400 @@ -28,11 +28,14 @@ #include "virtio-net.h" #include "sysemu.h" +#include "pass-through.h" #include "exec-all.h" #include "qemu-xen.h" //#define DEBUG_PCI +extern int vga_passthrough; + struct PCIBus { int bus_num; int devfn_min; @@ -611,7 +614,24 @@ uint32_t pci_data_read(void *opaque, uin goto the_end; } config_addr = addr & 0xff; - val = pci_dev->config_read(pci_dev, config_addr, len); + if (vga_passthrough && pci_dev->devfn == 0x00) //Host Bridge + { + val = pci_dev->config_read(pci_dev, config_addr, len); + + if (config_addr == 0x52) // GMCH + val = pt_pci_host_read_word(0, 0, 0, 0x52); + if (config_addr == 0x02) // Device ID + { + if (len == 2) + val = pt_pci_host_read_word(0, 0, 0, 0x00); + else if (len == 4) + val = pt_pci_host_read_long(0, 0, 0, 0x00); + } + } else if (vga_passthrough && pci_dev->devfn == 0x10 && // intel graphic card + config_addr == 0xfc) // OpRegion address + val = 0; // force to fall back to SMI mode + else + val = pci_dev->config_read(pci_dev, config_addr, len); #if defined(DEBUG_PCI) printf("pci_config_read: %s: addr=%02x val=%08x len=%d\n", pci_dev->name, config_addr, val, len); diff -rupN --ignore-blank-lines-X diffignore a/tools/ioemu-remote/hw/vga.c b/tools/ioemu-remote/hw/vga.c --- a/tools/ioemu-remote/hw/vga.c 2009-06-07 21:14:47.000000000 -0400 +++ b/tools/ioemu-remote/hw/vga.c 2009-06-07 19:04:11.000000000 -0400 @@ -34,9 +34,9 @@ #include "qemu-timer.h" -//#define DEBUG_VGA -//#define DEBUG_VGA_MEM -//#define DEBUG_VGA_REG +#define DEBUG_VGA +#define DEBUG_VGA_MEM +#define DEBUG_VGA_REG //#define DEBUG_BOCHS_VBE @@ -161,6 +161,18 @@ static uint8_t expand4to8[16]; static void vga_bios_init(VGAState *s); static void vga_screen_dump(void *opaque, const char *filename); +static VGAState *xen_vga_state; + +void set_vga_acc(void) +{ + set_vram_mapping(xen_vga_state, xen_vga_state->lfb_addr, xen_vga_state->lfb_end); +} + +void unset_vga_acc(void) +{ + unset_vram_mapping(xen_vga_state); +} + static void vga_dumb_update_retrace_info(VGAState *s) { (void) s; @@ -2473,8 +2485,6 @@ static void vga_bios_init(VGAState *s) } -static VGAState *xen_vga_state; - /* Allocate video memory in the GPFN space */ void xen_vga_populate_vram(uint64_t vram_addr, uint32_t vga_ram_size) { diff -rupN --ignore-blank-lines-X diffignore a/tools/ioemu-remote/intel.c b/tools/ioemu-remote/intel.c --- a/tools/ioemu-remote/intel.c 1969-12-31 19:00:00.000000000 -0500 +++ b/tools/ioemu-remote/intel.c 2009-06-07 19:04:11.000000000 -0400 @@ -0,0 +1,494 @@ +#include <stdio.h> +#include <stdlib.h> +#include <stdint.h> +#include <sys/mman.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <assert.h> +#include <signal.h> +#include <pci/pci.h> + +#include "qemu-common.h" +#include "console.h" +#include "sysemu.h" + +#define INTEL_DEBUG(format, args...) \ + fprintf (stderr, "intel.c:%d:%s " format , __LINE__, __func__, ## args) + +#define TileW 128 +#define TileH 8 + +#define REG_DR_DSPASURF 0x7019C +#define REG_DR_DSPACNTR 0x70180 +#define REG_DR_DSPASTRIDE 0x70188 +#define REG_DR_PIPEACONF 0x70008 + +#define REG_DR_DSPBSURF 0x7119C +#define REG_DR_DSPBCNTR 0x71180 +#define REG_DR_DSPBSTRIDE 0x71188 +#define REG_DR_PIPEBCONF 0x71008 + +#define REG_DE_PIPEASRC 0x6001c + +extern int vga_passthrough; +uint32_t guest_framebuffer; + +static int display = 0; + +static int mmio_fd = -1; +static int mem_fd = -1; +static uint8_t *intel_mem = NULL; +static uint8_t *intel_mmio = NULL; +static int intel_force_full_update = 0; +static int intel_have_focus; +static int IntelPitch = 16; +static int IntelX = 1280; +static int IntelY = 1024; +static DisplayState *lds = NULL; +static uint8_t *old_data = NULL; +static uint32_t intel_fb_base, intel_mmio_base; +static uint32_t map_s, map_d, map_size; +static int refresh; + +static void set_data_mappings(void); +static void unset_data_mappings(int mapping); +static void set_data_pointer(void); +static void intel_resize(DisplayState *ds); + +static inline unsigned int intel_get_reg(unsigned int reg) +{ + return *(unsigned int*)(intel_mmio + reg); +} + +static inline int is_linear(void) +{ + unsigned int *dspacntr = (unsigned int *)(intel_mmio + REG_DR_DSPACNTR); + if (((*dspacntr) & (1 << 10)) == 0) + return 1; + else + return 0; +} + +static inline unsigned int intel_get_pitch(void) +{ + unsigned int *dspastride = (unsigned int *)(intel_mmio + REG_DR_DSPASTRIDE); + return *dspastride; +} + +static inline unsigned int intel_get_offset(DisplaySurface *ds, int x, int y) +{ + return (y * ds->width + x) * 4; +} + +static void intel_update_linear(DisplaySurface *ds, int x, int y, int w, int h) +{ + int i, bpp = ds->pf.depth / 8; + unsigned char *s, *d; + s = ds->data; + d = (unsigned char *)(intel_mem + intel_get_reg(REG_DR_DSPASURF)); + s += (ds->linesize * y) + bpp * x; + d += (ds->linesize * y) + bpp * x; + for (i = 0; i < h; i++) { + memcpy(d, s, w * bpp); + s += ds->linesize; + d += ds->linesize; + } +} + +static void intel_force_linear(int linesize) +{ + unsigned int *dspacntr = (unsigned int *)(intel_mmio + REG_DR_DSPACNTR); + unsigned int *pipeaconf = (unsigned int *)(intel_mmio + REG_DR_PIPEACONF); + unsigned int *dspasurf = (unsigned int *)(intel_mmio + REG_DR_DSPASURF); + unsigned int *dspastride = (unsigned int *)(intel_mmio + REG_DR_DSPASTRIDE); + + unsigned int *dspbcntr = (unsigned int *)(intel_mmio + REG_DR_DSPBCNTR); + unsigned int *pipebconf = (unsigned int *)(intel_mmio + REG_DR_PIPEBCONF); + unsigned int *dspbsurf = (unsigned int *)(intel_mmio + REG_DR_DSPBSURF); + unsigned int *dspbstride = (unsigned int *)(intel_mmio + REG_DR_DSPBSTRIDE); + + unsigned int surfa = 0, surfb = 0, pipea = 0, pipeb = 0; + char pipebenabled = !!(*pipebconf & (1 << 30)); + + + INTEL_DEBUG("DSPASURF CTRL: 0x%x\n", intel_get_reg(REG_DR_DSPACNTR)); + + /* Disable surface */ + pipea = *pipeaconf & (0x3 << 18); + *pipeaconf &= ~(0x3 << 18); + *dspacntr |= (1 << 31); + /* Address of the surface to map to */ + surfa = *dspasurf; + *dspasurf = 0x00000000; + *dspacntr &= ~(1 << 31); + *dspasurf = 0x00000000; + *pipeaconf |= pipea; + + if (pipebenabled) { + INTEL_DEBUG("PIPEBCONF enabled.\n"); + + /* Disable surface */ + pipeb = *pipebconf & (0x3 << 18); + *pipebconf &= ~(0x3 << 18); + *dspbcntr |= (1 << 31); + /* Address of the surface to map to */ + surfb = *dspbsurf; + *dspbsurf = 0x00000000; + *dspbcntr &= ~(1 << 31); + *dspbsurf = 0x00000000; + *pipebconf |= pipeb; + } + + usleep(20000); + + *pipeaconf &= ~(0x3 << 18); + /* Enable surface linear mode */ + *dspacntr &= ~(1 << 10); + if (linesize) *dspastride = linesize; + *dspasurf = surfa; + *dspacntr |= (1 << 31); + *pipeaconf |= pipea; + + if (pipebenabled) { + *pipebconf &= ~(0x3 << 18); + /* Enable surface linear mode */ + *dspbcntr &= ~(1 << 10); + if (linesize) *dspbstride = linesize; + *dspbsurf = surfb; + *dspbcntr |= (1 << 31); + *pipebconf |= pipeb; + } + + usleep(20000); +} + +static void intel_update(DisplayState *ds, int x, int y, int w, int h) +{ + if (intel_have_focus && !old_data && !map_size) + intel_update_linear(ds->surface, x, y, w, h); +} + +static void set_fb_mapping(void) +{ + DisplaySurface *surf = lds->surface; + int rc; + unsigned long nr_pfn; + + unset_vga_acc(); + fprintf(stderr, "set_fb_mapping: %x %x\n", (intel_fb_base + intel_get_reg(REG_DR_DSPASURF)), guest_framebuffer); + nr_pfn = (surf->linesize * surf->height) >> TARGET_PAGE_BITS; + + rc = xc_domain_memory_mapping(xc_handle, + domid, + (guest_framebuffer >> TARGET_PAGE_BITS), + ((intel_fb_base + intel_get_reg(REG_DR_DSPASURF)) >> TARGET_PAGE_BITS), + nr_pfn, + DPCI_ADD_MAPPING); + if (rc) { + fprintf(stderr, "xc_domain_memory_mapping failed %d\n", rc); + return; + } + map_s = ((intel_fb_base + intel_get_reg(REG_DR_DSPASURF)) >> TARGET_PAGE_BITS); + map_d = (guest_framebuffer >> TARGET_PAGE_BITS); + map_size = nr_pfn; +} + +static void unset_fb_mapping(void) +{ + int rc; + + fprintf(stderr, "unset_fb_mapping: %x %x\n", map_d, map_s); + + rc = xc_domain_memory_mapping(xc_handle, + domid, + map_d, + map_s, + map_size, + DPCI_REMOVE_MAPPING); + if (rc) { + fprintf(stderr, "xc_domain_memory_mapping failed %d\n", rc); + return; + } + + set_vga_acc(); + map_s = 0; + map_d = 0; + map_size = 0; +} + +static void intel_setdata(DisplayState *ds) +{ + if (map_size) + unset_fb_mapping(); + set_fb_mapping(); +} + +static void intel_resize_shared(DisplayState *ds, int w, int h, int depth, int linesize, void *pixels) +{ + DisplaySurface *surf = ds->surface; + + if (!intel_have_focus) { + surf->width = w; + surf->height = h; + intel_resize(ds); + return; + } + if (depth == 32 && w == IntelX && h == IntelY) + surf->flags = QEMU_ALLOCATED_FLAG; + else + surf->flags &= ~QEMU_ALLOCATED_FLAG; + if (surf->flags & QEMU_ALLOCATED_FLAG) { + surf->width = w; + surf->height = h; + surf->pf.depth = 32; + surf->linesize = linesize; + /* adjust linesize */ + intel_force_linear(linesize); + set_data_mappings(); + if (refresh) { + memcpy(surf->data, pixels, surf->linesize * surf->height); + refresh = 0; + } + surf->data = pixels; + intel_setdata(ds); + } else { + surf->width = w; + surf->height = h; + intel_resize(ds); + } +} + +static void intel_resize(DisplayState *ds) +{ + DisplaySurface *surf = ds->surface; + int old_linesize = surf->linesize; + + if (surf->pf.depth == 32 && surf->width == IntelX && surf->height == IntelY) + surf->flags = QEMU_ALLOCATED_FLAG; + else + surf->flags &= ~QEMU_ALLOCATED_FLAG; + + if (is_buffer_shared(surf)) + { + INTEL_DEBUG("intel_resize_shared: enable shared buffer, linesize %d\n", + surf->linesize); + intel_force_linear(surf->linesize); + set_data_mappings(); + if (refresh) + { + // Pixels doesn't exist anymore ?? + //memcpy(surf->data, pixels, surf->linesize * surf->height); + refresh = 0; + } + intel_setdata(ds); + return; + } + + INTEL_DEBUG("intel_resize: no shared buffer, linesize=%d\n", surf->linesize); + surf->linesize = intel_get_pitch(); + if (map_size) { + unset_fb_mapping(); + unset_data_mappings(1); + } + if (intel_have_focus && !is_linear()) { + intel_force_linear(0); + } + surf->flags &= ~QEMU_ALLOCATED_FLAG; + if (intel_have_focus && !old_data && + surf->width * surf->height <= IntelX * IntelY) + set_data_mappings(); + else if (intel_have_focus && old_data && + surf->width * surf->height > IntelX * IntelY) + unset_data_mappings(0); + if (!old_data) { + qemu_free(surf->data); + surf->data = qemu_mallocz(surf->height * surf->linesize); + } else { + INTEL_DEBUG("intel_resize: set_data_pointer\n"); + set_data_pointer(); + } + if (intel_have_focus) + memset((unsigned char *)(intel_mem + intel_get_reg(REG_DR_DSPASURF)), 0x0, IntelX * IntelY); + if (refresh) { + if (old_data) { + unsigned char *s, *d; + int i; + s = old_data; + d = surf->data; + for (i = 0; i < surf->height; i++) { + memcpy(d, s, surf->width * 4); + s += old_linesize; + d += surf->linesize; + } + } + refresh = 0; + } +} + +static void intel_refresh(DisplayState *ds) +{ + vga_hw_update(); +} + +static void intel_init_mapping(void) +{ + struct pci_access *pci_bus; + struct pci_dev *pci_dev; + + mmio_fd = open("/dev/mem", O_RDWR); + if (mmio_fd == -1) + { + perror("open"); + exit(1); + } + mem_fd = open("/dev/mem", O_RDWR); + if (mem_fd == -1) + { + perror("open"); + exit(1); + } + + pci_bus = pci_alloc(); + pci_init(pci_bus); + pci_dev = pci_get_dev(pci_bus, 0, 0, 2, 0); + pci_fill_info(pci_dev, PCI_FILL_BASES); + intel_fb_base = pci_dev->base_addr[2] & 0xfffff000; + intel_mmio_base = pci_dev->base_addr[0] & 0xfffff000; + pci_free_dev(pci_dev); + pci_cleanup(pci_bus); + + INTEL_DEBUG("Map intel main mem 0x%x\n", intel_fb_base); + intel_mem = mmap(NULL, 0x10000000, PROT_READ | PROT_WRITE, MAP_SHARED, + mem_fd, intel_fb_base); + if (intel_mem == MAP_FAILED) + { + perror("mmap"); + exit(1); + } + + INTEL_DEBUG("Map intel mmio 0x%x\n", intel_mmio_base); + intel_mmio = mmap(NULL, 4 * 1024 * 1024, PROT_READ | PROT_WRITE, MAP_SHARED, + mmio_fd, intel_mmio_base); + if (intel_mem == MAP_FAILED) + { + perror("mmap"); + exit(1); + } +} + +static void set_data_pointer(void) +{ + DisplaySurface *surf = lds->surface; + + surf->data = (unsigned char *)(intel_mem + intel_get_reg(REG_DR_DSPASURF)); + surf->data = surf->data + + surf->linesize * ((IntelY - surf->height) / 2) + + 4 * ((IntelX - surf->width) / 2); +} + +static void set_data_mappings(void) +{ + INTEL_DEBUG("set_data_mappings\n"); + if (!old_data) + old_data = lds->surface->data; + set_data_pointer(); +} + +static void unset_data_mappings(int mapping) +{ + DisplaySurface *surf = lds->surface; + if (!old_data) + return; + if (mapping) { + uint8_t * buffer_pointer = surf->data; + surf->data = old_data; + old_data = NULL; + surf->data = realloc(surf->data, surf->linesize * surf->height); + memcpy(surf->data, + (unsigned char *)(intel_mem + intel_get_reg(REG_DR_DSPASURF)), + surf->linesize * surf->height); + memcpy(buffer_pointer, + surf->data, + surf->linesize * surf->height); + } else { + uint8_t * buffer_pointer = surf->data; + surf->data = old_data; + old_data = NULL; + surf->data = realloc(surf->data, surf->linesize * surf->height); + memcpy(surf->data, + buffer_pointer, + surf->linesize * surf->height); + } + INTEL_DEBUG("unset_data_mappings %d: success\n", mapping); +} + +static int intel_getfocus(void) +{ + return intel_have_focus; +} + +static void intel_focus(int focus) +{ + if (intel_have_focus == focus) + return; + + INTEL_DEBUG("intel_focus %d\n", focus); + intel_have_focus = focus; + if (focus) { + if (!is_linear()) { + IntelPitch = intel_get_reg(REG_DR_DSPASTRIDE); + IntelX = ((intel_get_reg(REG_DE_PIPEASRC) >> 16) & 0xfff) + 1; + IntelY = (intel_get_reg(REG_DE_PIPEASRC) & 0xfff) + 1; + INTEL_DEBUG("Resolution is %dx%d\n", IntelX, IntelY); + } + refresh = 1; + lds->listeners->dpy_resize = intel_resize; + lds->listeners->dpy_setdata = intel_setdata; + vga_hw_invalidate(); + } else { + if (map_size) { + unset_fb_mapping(); + unset_data_mappings(1); + } else if (old_data) { + unset_data_mappings(0); + } + lds->listeners->dpy_resize = NULL; + lds->listeners->dpy_setdata = NULL; + lds->surface->flags &= ~QEMU_ALLOCATED_FLAG; + } +} + +int intel_enter(void) +{ + intel_focus(1); + return 1; +} + +int intel_leave(void) +{ + intel_focus(0); + return 1; +} + +void intel_display_init(DisplayState *ds) +{ + DisplaySurface *surf = ds->surface; + + INTEL_DEBUG("\n"); + + intel_init_mapping(); + + INTEL_DEBUG("Frambuffer is at 0x%x\n", intel_get_reg(REG_DR_DSPASURF)); + + surf->flags = 0; + surf->width = 640; + surf->height = 480; + surf->pf.depth = 32; + intel_resize(ds); + lds = ds; + + ds->listeners->dpy_update = intel_update; + ds->listeners->dpy_resize = intel_resize; + ds->listeners->dpy_refresh = intel_refresh; +} \ No newline at end of file diff -rupN --ignore-blank-lines-X diffignore a/tools/ioemu-remote/vl.c b/tools/ioemu-remote/vl.c --- a/tools/ioemu-remote/vl.c 2009-06-07 21:14:47.000000000 -0400 +++ b/tools/ioemu-remote/vl.c 2009-06-07 19:29:12.000000000 -0400 @@ -233,6 +233,9 @@ CharDriverState *virtcon_hds[MAX_VIRTIO_ #ifdef TARGET_I386 int win2k_install_hack = 0; int rtc_td_hack = 0; +int vga_passthrough = 0; +const char *dom0_input = NULL; +int intel = 0; #endif int usb_enabled = 0; int smp_cpus = 1; @@ -4039,6 +4042,9 @@ static void help(int exitcode) "-disable-opengl disable OpenGL rendering, using SDL" #endif #endif + "-vga_passthrough enable graphics card passthrough\n" + "-dom0-input enable dom0 controlling qemu\n" + "-intel use intel gfx\n" "-portrait rotate graphical output 90 deg left (only PXA LCD)\n" "-vga [std|cirrus|vmware|none]\n" " select video card type\n" @@ -4275,6 +4281,9 @@ enum { QEMU_OPTION_domainname, QEMU_OPTION_acpi, QEMU_OPTION_vcpus, + QEMU_OPTION_vga_passthrough, + QEMU_OPTION_dom0_input, + QEMU_OPTION_intel, /* Debug/Expert options: */ QEMU_OPTION_serial, @@ -4448,6 +4457,9 @@ static const QEMUOption qemu_options[] = { "pciemulation", HAS_ARG, QEMU_OPTION_pci_emulation }, { "vncunused", 0, QEMU_OPTION_vncunused }, { "vcpus", HAS_ARG, QEMU_OPTION_vcpus }, + { "vga_passthrough", 0, QEMU_OPTION_vga_passthrough }, + { "dom0-input", HAS_ARG, QEMU_OPTION_dom0_input }, + { "intel", 0, QEMU_OPTION_intel }, #if defined(CONFIG_XEN) && !defined(CONFIG_DM) { "xen-domid", HAS_ARG, QEMU_OPTION_xen_domid }, { "xen-create", 0, QEMU_OPTION_xen_create }, @@ -5281,6 +5293,15 @@ int main(int argc, char **argv, char **e case QEMU_OPTION_disable_opengl: opengl_enabled = 0; break; + case QEMU_OPTION_vga_passthrough: + vga_passthrough = 1; + break; + case QEMU_OPTION_dom0_input: + dom0_input = optarg; + break; + case QEMU_OPTION_intel: + intel = 1; + break; case QEMU_OPTION_direct_pci: direct_pci = optarg; break; @@ -5876,6 +5897,14 @@ int main(int argc, char **argv, char **e fprintf(stderr, "fatal: -nographic can't be used with -curses\n"); exit(1); } + +#ifdef CONFIG_DM + if(vga_passthrough) + fprintf(stderr, "Replace if initializing dom0_driver\n" ); +#else + dumb_display_init(ds); +#endif + } else { #if defined(CONFIG_CURSES) if (curses) { diff -rupN --ignore-blank-lines-X diffignore a/tools/ioemu-remote/xen-hooks.mak b/tools/ioemu-remote/xen-hooks.mak --- a/tools/ioemu-remote/xen-hooks.mak 2009-06-07 21:14:47.000000000 -0400 +++ b/tools/ioemu-remote/xen-hooks.mak 2009-06-07 19:04:12.000000000 -0400 @@ -35,6 +35,7 @@ OBJS += exec-dm.o OBJS += pci_emulation.o OBJS += helper2.o OBJS += battery_mgmt.o +OBJS += intel.o ifdef CONFIG_STUBDOM CPPFLAGS += $(TARGET_CPPFLAGS) -DNEED_CPU_H \ diff -rupN --ignore-blank-lines-X diffignore a/tools/libxc/xc_hvm_build.c b/tools/libxc/xc_hvm_build.c --- a/tools/libxc/xc_hvm_build.c 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/libxc/xc_hvm_build.c 2009-06-07 19:04:12.000000000 -0400 @@ -1,7 +1,3 @@ -/****************************************************************************** - * xc_hvm_build.c - */ - #include <stddef.h> #include <inttypes.h> #include <stdlib.h> @@ -66,6 +62,82 @@ static void build_hvm_info(void *hvm_inf hvm_info->checksum = -sum; } +static int init_vgabios(int xc_handle, + uint32_t dom, + unsigned char *buffer, + uint32_t bios_size) +{ + char *va_bios = NULL; + uint32_t va_size = 0; + + va_size = bios_size + bios_size % XC_PAGE_SIZE; + va_bios = xc_map_foreign_range(xc_handle, dom, va_size, + PROT_READ | PROT_WRITE, 0xC0); + if (!va_bios) + { + IPRINTF("Unable to map vga bios!\n"); + return -1; + } + + if ( buffer != NULL) + memcpy(va_bios, buffer, bios_size); + else + memset(va_bios, 0, bios_size); + + munmap(va_bios, va_size); + return 0; +} + +static int setup_vga_pt(int xc_handle, + uint32_t dom) +{ + int rc = 0; + unsigned char *bios = NULL; + int bios_size = 0; + char *c = NULL; + char checksum = 0; + + IPRINTF("Setting up vga passthrough.\n"); + + /* Allocated 64K for the vga bios */ + if (!(bios = malloc(64 * 1024))) { + IPRINTF("Error allocating memory for vga bios.\n"); + return -1; + } + +#ifdef __linux__ + bios_size = xc_get_vgabios(bios, 64 * 1024); +#else + bios_size = 0; +#endif /* __linux__ */ + + if (bios_size == 0) + { + IPRINTF("vga bios size is 0!\n"); + rc = -1; + goto error; + } + + /* Adjust the bios checksum */ + for ( c = (char*)bios; c < ((char*)bios + bios_size); c++ ) + checksum += *c; + if (checksum) + bios[bios_size - 1] -= checksum; + init_vgabios(xc_handle, dom, bios, bios_size); + +error: + free(bios); + + if( rc == -1 ) { + IPRINTF("Error setting up vga passthrough.\n"); + } + else { + IPRINTF("Success setting up vga passthrough.\n"); + } + + return rc; +} + static int loadelfimage( struct elf_binary *elf, int xch, uint32_t dom, unsigned long *parray) { @@ -381,7 +453,8 @@ int xc_hvm_build_target_mem(int xc_handl uint32_t domid, int memsize, int target, - const char *image_name) + const char *image_name, + int vga_pt_enabled) { char *image; int sts; @@ -392,6 +465,12 @@ int xc_hvm_build_target_mem(int xc_handl return -1; sts = xc_hvm_build_internal(xc_handle, domid, memsize, target, image, image_size); + if ( vga_pt_enabled ) { + sts |= setup_vga_pt(xc_handle, domid); + } else { + sts |= init_vgabios(xc_handle, domid, NULL, 0x800); + } + free(image); diff -rupN --ignore-blank-lines-X diffignore a/tools/libxc/xc_linux.c b/tools/libxc/xc_linux.c --- a/tools/libxc/xc_linux.c 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/libxc/xc_linux.c 2009-06-07 19:04:12.000000000 -0400 @@ -562,6 +562,57 @@ int xc_gnttab_set_max_grants(int xcg_han return 0; } +int xc_get_vgabios(unsigned char *buf, + int len) +{ + int mem; + uint32_t start, size = 0; + uint16_t magic = 0; + + start = 0xC0000; + if (len < size) + return 0; + if ((mem = open("/dev/mem", O_RDONLY)) < 0) + return 0; + + /* + ** Check if it a real bios extension. + ** The magic number is 0xAA55. + */ + if (start != lseek(mem, start, SEEK_SET)) + goto out; + if (read(mem, &magic, 2) != 2) + goto out; + if (magic != 0xAA55) + goto out; + + /* Find the size of the rom extension */ + if (start != lseek(mem, start, SEEK_SET)) + goto out; + if (lseek(mem, 2, SEEK_CUR) != (start + 2)) + goto out; + if (read(mem, &size, 1) != 1) + goto out; + /* This size is in 512K */ + size *= 512; + + /* + ** Set the file to the begining of the rombios, + ** to start the copy. + */ + if (start != lseek(mem, start, SEEK_SET)) + { + size = 0; + goto out; + } + if (size != read(mem, buf, size)) + size = 0; + +out: + close(mem); + return size; +} + /* * Local variables: * mode: C diff -rupN --ignore-blank-lines-X diffignore a/tools/libxc/xenctrl.h b/tools/libxc/xenctrl.h --- a/tools/libxc/xenctrl.h 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/libxc/xenctrl.h 2009-06-07 19:04:12.000000000 -0400 @@ -145,6 +145,10 @@ int xc_waitdomain( int *status, int options); +int xc_get_vgabios( + unsigned char *bios, + int len); + #endif /* __linux__ */ /* diff -rupN --ignore-blank-lines-X diffignore a/tools/libxc/xenguest.h b/tools/libxc/xenguest.h --- a/tools/libxc/xenguest.h 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/libxc/xenguest.h 2009-06-07 19:04:12.000000000 -0400 @@ -134,7 +134,8 @@ int xc_hvm_build_target_mem(int xc_handl uint32_t domid, int memsize, int target, - const char *image_name); + const char *image_name, + int vga_pt_enabled); int xc_hvm_build_mem(int xc_handle, uint32_t domid, diff -rupN --ignore-blank-lines-X diffignore a/tools/python/xen/lowlevel/xc/xc.c b/tools/python/xen/lowlevel/xc/xc.c --- a/tools/python/xen/lowlevel/xc/xc.c 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/python/xen/lowlevel/xc/xc.c 2009-06-07 19:04:12.000000000 -0400 @@ -890,21 +890,21 @@ static PyObject *pyxc_hvm_build(XcObject int i; #endif char *image; - int memsize, target=-1, vcpus = 1, acpi = 0, apic = 1; + int memsize, target=-1, vcpus = 1, acpi = 0, apic = 1, vga_pt = 0; static char *kwd_list[] = { "domid", "memsize", "image", "target", "vcpus", "acpi", - "apic", NULL }; - if ( !PyArg_ParseTupleAndKeywords(args, kwds, "iis|iiii", kwd_list, + "apic", "vga_pt", NULL }; + if ( !PyArg_ParseTupleAndKeywords(args, kwds, "iis|iiiii", kwd_list, &dom, &memsize, &image, &target, &vcpus, - &acpi, &apic) ) + &acpi, &apic, &vga_pt) ) return NULL; if ( target == -1 ) target = memsize; if ( xc_hvm_build_target_mem(self->xc_handle, dom, memsize, - target, image) != 0 ) + target, image, vga_pt) != 0 ) return pyxc_error_to_exception(); #if !defined(__ia64__) diff -rupN --ignore-blank-lines-X diffignore a/tools/python/xen/xend/image.py b/tools/python/xen/xend/image.py --- a/tools/python/xen/xend/image.py 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/python/xen/xend/image.py 2009-06-07 19:04:12.000000000 -0400 @@ -279,6 +279,9 @@ class ImageHandler: vnc_config = {} has_vnc = int(vmConfig['platform'].get('vnc', 0)) != 0 has_sdl = int(vmConfig['platform'].get('sdl', 0)) != 0 + has_vga_passthrough = int(vmConfig['platform'].get('vga_passthrough', 0)) != 0 + has_intel = int(vmConfig['platform'].get('intel', 0)) != 0 + opengl = 1 keymap = vmConfig['platform'].get("keymap") for dev_uuid in vmConfig['console_refs']: @@ -302,7 +305,7 @@ class ImageHandler: ret.append("-k") ret.append(keymap) - if has_vnc: + if has_vnc and not has_vga_passthrough: if not vnc_config: for key in ('vncunused', 'vnclisten', 'vncdisplay', 'vncpasswd'): @@ -353,6 +356,15 @@ class ImageHandler: if int(vmConfig['platform'].get('monitor', 0)) != 0: ret = ret + ['-monitor', 'vc'] + + if has_vga_passthrough: + ret.append('-vga_passthrough') + dom0_input = str(vmConfig['platform'].get('dom0_input')) + ret = ret + ['-dom0-input', dom0_input] + + if has_intel: + ret.append('-intel') + return ret def getDeviceModelArgs(self, restore = False): @@ -754,7 +766,9 @@ class HVMImageHandler(ImageHandler): self.apic = int(vmConfig['platform'].get('apic', 0)) self.acpi = int(vmConfig['platform'].get('acpi', 0)) self.guest_os_type = vmConfig['platform'].get('guest_os_type') - + self.vga_pt = int(vmConfig['platform'].get('vga_passthrough', 0)) + self.dom0_input = str(vmConfig['platform'].get('dom0_input')) + self.intel = int(vmConfig['platform'].get('intel', 0)) # Return a list of cmd line args to the device models based on the # xm config file @@ -869,6 +883,7 @@ class HVMImageHandler(ImageHandler): log.debug("vcpus = %d", self.vm.getVCpuCount()) log.debug("acpi = %d", self.acpi) log.debug("apic = %d", self.apic) + log.debug("vga_pt = %d", self.vga_pt) rc = xc.hvm_build(domid = self.vm.getDomid(), image = self.loader, @@ -876,7 +891,8 @@ class HVMImageHandler(ImageHandler): target = mem_mb, vcpus = self.vm.getVCpuCount(), acpi = self.acpi, - apic = self.apic) + apic = self.apic, + vga_pt = self.vga_pt) rc['notes'] = { 'SUSPEND_CANCEL': 1 } rc['store_mfn'] = xc.hvm_get_param(self.vm.getDomid(), diff -rupN --ignore-blank-lines-X diffignore a/tools/python/xen/xend/XendConfig.py b/tools/python/xen/xend/XendConfig.py --- a/tools/python/xen/xend/XendConfig.py 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/python/xen/xend/XendConfig.py 2009-06-07 19:04:12.000000000 -0400 @@ -171,6 +171,9 @@ XENAPI_PLATFORM_CFG_TYPES = { 'pci_msitranslate': int, 'pci_power_mgmt': int, 'xen_platform_pci': int, + 'vga_passthrough': int, + 'dom0_input': str, + 'intel': int, } # Xen API console 'other_config' keys. diff -rupN --ignore-blank-lines-X diffignore a/tools/python/xen/xm/create.py b/tools/python/xen/xm/create.py --- a/tools/python/xen/xm/create.py 2009-06-07 20:44:47.000000000 -0400 +++ b/tools/python/xen/xm/create.py 2009-06-07 19:04:12.000000000 -0400 @@ -618,6 +618,18 @@ gopts.var('xen_platform_pci', val='0|1', fn=set_int, default=1, use="Is xen_platform_pci used?") +gopts.var('vga_passthrough', val='0|1', + fn=set_int, default=None, + use="Enable the passthrough for the graphic card.") + +gopts.var('dom0_input', val='DOM0_INPUT', + fn=set_value, default=None, + use="Input arguments for dom0 driver") + +gopts.var('intel', val='INTEL', + fn=set_int, default=None, + use="Use Intel GFX.") + def err(msg): """Print an error to stderr and exit. """ @@ -932,7 +944,8 @@ def configure_hvm(config_image, vals): 'acpi', 'apic', 'usb', 'usbdevice', 'keymap', 'pci', 'hpet', 'guest_os_type', 'hap', 'opengl', 'cpuid', 'cpuid_check', 'viridian', 'xen_extended_power_mgmt', 'pci_msitranslate', - 'vpt_align', 'pci_power_mgmt', 'xen_platform_pci' ] + 'vpt_align', 'pci_power_mgmt', 'xen_platform_pci', + 'vga_passthrough', 'dom0_input', 'intel' ] for a in args: if a in vals.__dict__ and vals.__dict__[a] is not None: [-- Attachment #5: Type: text/plain, Size: 138 bytes --] _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB 2009-08-25 17:25 ` nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB djmagee [not found] ` <f4527be0908251056s295eaf0fle31922e7d90faa1c@mail.gmail.com> @ 2009-08-26 8:26 ` Han, Weidong 2009-08-26 9:21 ` nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Teo En Ming (Zhang Enming) 1 sibling, 1 reply; 25+ messages in thread From: Han, Weidong @ 2009-08-26 8:26 UTC (permalink / raw) To: 'djmagee@mageenet.net', 'enming.teo@asiasoftsea.net' Cc: 'xen-devel@lists.xensource.com' I suppose you just use the patch posted in mailing list before. nVidia Geforce 8400 passthrough needs extra hacks. We can make it work in our experiments with 1:1 map of its MMIO BARs. We are working on gfx passthrough on latest xen-unstable. Firstly, we want to cook a simple patch including generic changes to support passthrough of virtualization friendly gfx cards, such as Nvidia FX3800. This patch is basically done. Then, we will add some hacks for more gfx cards passthrough, such as iGFX and some Nvidia and ATI cards. Regards, Weidong djmagee@mageenet.net wrote: > As I've pointed out on this list before, there are not enough PCIe > lanes on the DQ45CB to drive both the internal graphics adapter and > the add-on adapter at the same time. I believe the onboard one may > be able to operate in some sort of VGA only mode when there is a card > installed and used as the primary adapter. > > I have a similar setup, using the same motherboard, and an ATI 4770. > I used the VGA passthrough patches from an earlier posting (I saw you > followed the same set of instructions), with limited success. I've > been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made > modifications to the dom0 portion of the patches so they would apply > to my xenified 2.6.29 kernel. These patches include code that will > copy the VGA bios to the guest bios instead of using the emulated vga > bios. I've had very little success, however. I have only tried > passing through the ATI adapter. In all instances, the guest bios > messages appear on my monitor, so this much works. In some cases, > the guest essentially stops there; xm list show's about 2sec CPU > usage and nothing ever happens after that point. In other cases, the > guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot all > the way, but in very low res/color mode, and cannot properly > initialize the video device. Once or twice, it actually did > recognize the device and had a reasonable default color/resolution > combination. In all cases where the guest actually boots, the system > eventually freezes. In some cases, I get endless streams of iommu > page faults. > > I have 8GB ram installed. In all cases I've limited dom0 memory to > 2GB. In all cases, my guest has been assigned 2GB of memory. > > I have a dual core e6600. I've tried allowing dom0 to use both > cores, offlining one core (using xend dom0_vcpu setting) after boot, > and restricting dom0 to only one core using the dom0_max_vcpus xen > hypervisor parameter. In all of these cases I've tried both one and > two vcpus for the guest. My success with VGA passthrough seems > somewhat random and no combination of cpu assignment seems to have > any effect. > > I have not tried with the 2.6.18-xen kernel as I haven't gotten it to > boot on my hardware; it can never find my volume group, even if I > create a initrd with all of the required modules, or build those > drivers into the kernel. I have not spent more than maybe a half an > hour on this problem; I suspect it may have something to do with the > version of mkinitrd I'm using (from Fedora 9 x64). > > If anyone else has any insight or similar experience I'd also love to > hear it. > > Doug Magee > djmagee@mageenet.com > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo > En Ming (Zhang Enming) > Sent: Tuesday, August 25, 2009 11:57 AM > To: enming.teo@asiasoftsea.net > Cc: xen-devel@lists.xensource.com > Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel > Desktop BoardDQ45CB > > I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as > suggested > by Weidong. > > > On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >> Dear All, >> >> I have managed to do PCI-e VGA passthrough with the open source Xen >> but the work is still in progress because although Windows XP guest >> can see the REAL PCI-e x16 graphics card instead of an emulated >> graphics driver, it cannot be initialized yet. >> >> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >> shared their knowledge with me along the way. >> >> System Configuration: >> >> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >> Onboard Intel GMA 4500 Graphics (IGD) >> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >> >> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >> Primary Video Adapter in BIOS: IGD >> >> Please see the screenshots and my blog at the link here: >> >> http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400-gs-pci-express-x16.html >> >> > > > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 8:26 ` Han, Weidong @ 2009-08-26 9:21 ` Teo En Ming (Zhang Enming) 2009-08-26 10:56 ` Han, Weidong 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-26 9:21 UTC (permalink / raw) To: weidong.han, djmagee; +Cc: xen-devel Hi Weidong, Could you share with us the hack codes for making Geforce 8400 GS work and also how to let the Windows HVM guest boot up using the real BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com Alma Maters: Singapore Polytechnic, National University of Singapore -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, Weidong Sent: Wednesday, August 26, 2009 4:27 PM To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' Cc: 'xen-devel@lists.xensource.com' Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with IntelDesktop BoardDQ45CB I suppose you just use the patch posted in mailing list before. nVidia Geforce 8400 passthrough needs extra hacks. We can make it work in our experiments with 1:1 map of its MMIO BARs. We are working on gfx passthrough on latest xen-unstable. Firstly, we want to cook a simple patch including generic changes to support passthrough of virtualization friendly gfx cards, such as Nvidia FX3800. This patch is basically done. Then, we will add some hacks for more gfx cards passthrough, such as iGFX and some Nvidia and ATI cards. Regards, Weidong djmagee@mageenet.net wrote: > As I've pointed out on this list before, there are not enough PCIe > lanes on the DQ45CB to drive both the internal graphics adapter and > the add-on adapter at the same time. I believe the onboard one may > be able to operate in some sort of VGA only mode when there is a card > installed and used as the primary adapter. > > I have a similar setup, using the same motherboard, and an ATI 4770. > I used the VGA passthrough patches from an earlier posting (I saw you > followed the same set of instructions), with limited success. I've > been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made > modifications to the dom0 portion of the patches so they would apply > to my xenified 2.6.29 kernel. These patches include code that will > copy the VGA bios to the guest bios instead of using the emulated vga > bios. I've had very little success, however. I have only tried > passing through the ATI adapter. In all instances, the guest bios > messages appear on my monitor, so this much works. In some cases, > the guest essentially stops there; xm list show's about 2sec CPU > usage and nothing ever happens after that point. In other cases, the > guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot all > the way, but in very low res/color mode, and cannot properly > initialize the video device. Once or twice, it actually did > recognize the device and had a reasonable default color/resolution > combination. In all cases where the guest actually boots, the system > eventually freezes. In some cases, I get endless streams of iommu > page faults. > > I have 8GB ram installed. In all cases I've limited dom0 memory to > 2GB. In all cases, my guest has been assigned 2GB of memory. > > I have a dual core e6600. I've tried allowing dom0 to use both > cores, offlining one core (using xend dom0_vcpu setting) after boot, > and restricting dom0 to only one core using the dom0_max_vcpus xen > hypervisor parameter. In all of these cases I've tried both one and > two vcpus for the guest. My success with VGA passthrough seems > somewhat random and no combination of cpu assignment seems to have > any effect. > > I have not tried with the 2.6.18-xen kernel as I haven't gotten it to > boot on my hardware; it can never find my volume group, even if I > create a initrd with all of the required modules, or build those > drivers into the kernel. I have not spent more than maybe a half an > hour on this problem; I suspect it may have something to do with the > version of mkinitrd I'm using (from Fedora 9 x64). > > If anyone else has any insight or similar experience I'd also love to > hear it. > > Doug Magee > djmagee@mageenet.com > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo > En Ming (Zhang Enming) > Sent: Tuesday, August 25, 2009 11:57 AM > To: enming.teo@asiasoftsea.net > Cc: xen-devel@lists.xensource.com > Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel > Desktop BoardDQ45CB > > I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as > suggested > by Weidong. > > > On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >> Dear All, >> >> I have managed to do PCI-e VGA passthrough with the open source Xen >> but the work is still in progress because although Windows XP guest >> can see the REAL PCI-e x16 graphics card instead of an emulated >> graphics driver, it cannot be initialized yet. >> >> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >> shared their knowledge with me along the way. >> >> System Configuration: >> >> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >> Onboard Intel GMA 4500 Graphics (IGD) >> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >> >> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >> Primary Video Adapter in BIOS: IGD >> >> Please see the screenshots and my blog at the link here: >> >> http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 -gs-pci-express-x16.html >> >> > > > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 9:21 ` nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Teo En Ming (Zhang Enming) @ 2009-08-26 10:56 ` Han, Weidong 2009-08-26 11:24 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Han, Weidong @ 2009-08-26 10:56 UTC (permalink / raw) To: 'enming.teo@asiasoftsea.net', 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' Teo En Ming (Zhang Enming) wrote: > Hi Weidong, > > Could you share with us the hack codes for making Geforce 8400 GS > work and also how to let the Windows HVM guest boot up using the real > BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? > What patch are you using now? Using real VGA bios of gfx card to replace emulatd VGA bios is the prerequisite of gfx passthrough. You can find it in posted gfx passthrough patches or XCI. For hack of making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of the card. Currently our code is experimental, we will send out in mailing list after cleanup and more tests. Regards, Weidong > Thank you. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > Alma Maters: Singapore Polytechnic, National University of Singapore > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, > Weidong Sent: Wednesday, August 26, 2009 4:27 PM > To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 > VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with > IntelDesktop BoardDQ45CB > > I suppose you just use the patch posted in mailing list before. nVidia > Geforce 8400 passthrough needs extra hacks. We can make it work in our > experiments with 1:1 map of its MMIO BARs. > > We are working on gfx passthrough on latest xen-unstable. Firstly, we > want to cook a simple patch including generic changes to support > passthrough of virtualization friendly gfx cards, such as Nvidia > FX3800. This patch is basically done. Then, we will add some hacks > for more gfx cards passthrough, such as iGFX and some Nvidia and ATI > cards. > > Regards, > Weidong > > djmagee@mageenet.net wrote: >> As I've pointed out on this list before, there are not enough PCIe >> lanes on the DQ45CB to drive both the internal graphics adapter and >> the add-on adapter at the same time. I believe the onboard one may >> be able to operate in some sort of VGA only mode when there is a card >> installed and used as the primary adapter. >> >> I have a similar setup, using the same motherboard, and an ATI 4770. >> I used the VGA passthrough patches from an earlier posting (I saw you >> followed the same set of instructions), with limited success. I've >> been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made >> modifications to the dom0 portion of the patches so they would apply >> to my xenified 2.6.29 kernel. These patches include code that will >> copy the VGA bios to the guest bios instead of using the emulated vga >> bios. I've had very little success, however. I have only tried >> passing through the ATI adapter. In all instances, the guest bios >> messages appear on my monitor, so this much works. In some cases, >> the guest essentially stops there; xm list show's about 2sec CPU >> usage and nothing ever happens after that point. In other cases, the >> guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot all >> the way, but in very low res/color mode, and cannot properly >> initialize the video device. Once or twice, it actually did >> recognize the device and had a reasonable default color/resolution >> combination. In all cases where the guest actually boots, the system >> eventually freezes. In some cases, I get endless streams of iommu >> page faults. >> >> I have 8GB ram installed. In all cases I've limited dom0 memory to >> 2GB. In all cases, my guest has been assigned 2GB of memory. >> >> I have a dual core e6600. I've tried allowing dom0 to use both >> cores, offlining one core (using xend dom0_vcpu setting) after boot, >> and restricting dom0 to only one core using the dom0_max_vcpus xen >> hypervisor parameter. In all of these cases I've tried both one and >> two vcpus for the guest. My success with VGA passthrough seems >> somewhat random and no combination of cpu assignment seems to have >> any effect. >> >> I have not tried with the 2.6.18-xen kernel as I haven't gotten it to >> boot on my hardware; it can never find my volume group, even if I >> create a initrd with all of the required modules, or build those >> drivers into the kernel. I have not spent more than maybe a half an >> hour on this problem; I suspect it may have something to do with the >> version of mkinitrd I'm using (from Fedora 9 x64). >> >> If anyone else has any insight or similar experience I'd also love >> to hear it. >> >> Doug Magee >> djmagee@mageenet.com >> >> -----Original Message----- >> From: xen-devel-bounces@lists.xensource.com >> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >> To: enming.teo@asiasoftsea.net >> Cc: xen-devel@lists.xensource.com >> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel >> Desktop BoardDQ45CB >> >> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >> suggested by Weidong. >> >> >> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>> Dear All, >>> >>> I have managed to do PCI-e VGA passthrough with the open source Xen >>> but the work is still in progress because although Windows XP guest >>> can see the REAL PCI-e x16 graphics card instead of an emulated >>> graphics driver, it cannot be initialized yet. >>> >>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>> shared their knowledge with me along the way. >>> >>> System Configuration: >>> >>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>> Onboard Intel GMA 4500 Graphics (IGD) >>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>> >>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>> Primary Video Adapter in BIOS: IGD >>> >>> Please see the screenshots and my blog at the link here: >>> >>> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 > -gs-pci-express-x16.html >>> >>> >> >> >> >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 10:56 ` Han, Weidong @ 2009-08-26 11:24 ` Teo En Ming (Zhang Enming) 2009-08-26 11:31 ` Teo En Ming (Zhang Enming) 2009-08-27 2:01 ` nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Han, Weidong 0 siblings, 2 replies; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-26 11:24 UTC (permalink / raw) To: weidong.han, djmagee; +Cc: xen-devel Hi Weidong, Currently I am using the entire Xen 3.5-unstable branch and not using the posted pci-stub patch changeset 19893 to patch Xen 3.4.1-testing. I am using Xen 3.5-unstable in conjunction with pvops dom 0 kernel 2.6.31-rc6. Here are my questions: 1) How do I extract out the VGA BIOS/firmware from the firmware/EEPROM chip on the Geforce 8400 GS VGA card? If it is not possible to extract, where can I download the firmware? NVIDIA does not provide display card firmware/BIOS files for public download on its website. 2) How do I get the Geforce 8400 GS VGA BIOS to load instead of the emulated VGA BIOS? Which configuration files/source codes do I have to configure/patch? 3) What are physical MMIO, virtual MMIO Bars, dsdt.asl and 1:1 map? I am sorry but I am not a graphics card hardware engineer so I do not understand those terms. I would appreciate it if you could write out some steps that I could carry out or point me to already available documentation. 4) I don't mind using experimental code. Please post it on the xen-devel mailing list so that I could try it out. Thank you very much. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: Han, Weidong [mailto:weidong.han@intel.com] Sent: Wednesday, August 26, 2009 6:56 PM To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Teo En Ming (Zhang Enming) wrote: > Hi Weidong, > > Could you share with us the hack codes for making Geforce 8400 GS > work and also how to let the Windows HVM guest boot up using the real > BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? > What patch are you using now? Using real VGA bios of gfx card to replace emulatd VGA bios is the prerequisite of gfx passthrough. You can find it in posted gfx passthrough patches or XCI. For hack of making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of the card. Currently our code is experimental, we will send out in mailing list after cleanup and more tests. Regards, Weidong > Thank you. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > Alma Maters: Singapore Polytechnic, National University of Singapore > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, > Weidong Sent: Wednesday, August 26, 2009 4:27 PM > To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 > VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with > IntelDesktop BoardDQ45CB > > I suppose you just use the patch posted in mailing list before. nVidia > Geforce 8400 passthrough needs extra hacks. We can make it work in our > experiments with 1:1 map of its MMIO BARs. > > We are working on gfx passthrough on latest xen-unstable. Firstly, we > want to cook a simple patch including generic changes to support > passthrough of virtualization friendly gfx cards, such as Nvidia > FX3800. This patch is basically done. Then, we will add some hacks > for more gfx cards passthrough, such as iGFX and some Nvidia and ATI > cards. > > Regards, > Weidong > > djmagee@mageenet.net wrote: >> As I've pointed out on this list before, there are not enough PCIe >> lanes on the DQ45CB to drive both the internal graphics adapter and >> the add-on adapter at the same time. I believe the onboard one may >> be able to operate in some sort of VGA only mode when there is a card >> installed and used as the primary adapter. >> >> I have a similar setup, using the same motherboard, and an ATI 4770. >> I used the VGA passthrough patches from an earlier posting (I saw you >> followed the same set of instructions), with limited success. I've >> been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made >> modifications to the dom0 portion of the patches so they would apply >> to my xenified 2.6.29 kernel. These patches include code that will >> copy the VGA bios to the guest bios instead of using the emulated vga >> bios. I've had very little success, however. I have only tried >> passing through the ATI adapter. In all instances, the guest bios >> messages appear on my monitor, so this much works. In some cases, >> the guest essentially stops there; xm list show's about 2sec CPU >> usage and nothing ever happens after that point. In other cases, the >> guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot all >> the way, but in very low res/color mode, and cannot properly >> initialize the video device. Once or twice, it actually did >> recognize the device and had a reasonable default color/resolution >> combination. In all cases where the guest actually boots, the system >> eventually freezes. In some cases, I get endless streams of iommu >> page faults. >> >> I have 8GB ram installed. In all cases I've limited dom0 memory to >> 2GB. In all cases, my guest has been assigned 2GB of memory. >> >> I have a dual core e6600. I've tried allowing dom0 to use both >> cores, offlining one core (using xend dom0_vcpu setting) after boot, >> and restricting dom0 to only one core using the dom0_max_vcpus xen >> hypervisor parameter. In all of these cases I've tried both one and >> two vcpus for the guest. My success with VGA passthrough seems >> somewhat random and no combination of cpu assignment seems to have >> any effect. >> >> I have not tried with the 2.6.18-xen kernel as I haven't gotten it to >> boot on my hardware; it can never find my volume group, even if I >> create a initrd with all of the required modules, or build those >> drivers into the kernel. I have not spent more than maybe a half an >> hour on this problem; I suspect it may have something to do with the >> version of mkinitrd I'm using (from Fedora 9 x64). >> >> If anyone else has any insight or similar experience I'd also love >> to hear it. >> >> Doug Magee >> djmagee@mageenet.com >> >> -----Original Message----- >> From: xen-devel-bounces@lists.xensource.com >> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >> To: enming.teo@asiasoftsea.net >> Cc: xen-devel@lists.xensource.com >> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel >> Desktop BoardDQ45CB >> >> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >> suggested by Weidong. >> >> >> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>> Dear All, >>> >>> I have managed to do PCI-e VGA passthrough with the open source Xen >>> but the work is still in progress because although Windows XP guest >>> can see the REAL PCI-e x16 graphics card instead of an emulated >>> graphics driver, it cannot be initialized yet. >>> >>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>> shared their knowledge with me along the way. >>> >>> System Configuration: >>> >>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>> Onboard Intel GMA 4500 Graphics (IGD) >>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>> >>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>> Primary Video Adapter in BIOS: IGD >>> >>> Please see the screenshots and my blog at the link here: >>> >>> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 > -gs-pci-express-x16.html >>> >>> >> >> >> >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 11:24 ` Teo En Ming (Zhang Enming) @ 2009-08-26 11:31 ` Teo En Ming (Zhang Enming) 2009-08-26 11:37 ` Vincent Hanquez 2009-08-27 2:01 ` nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Han, Weidong 1 sibling, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-26 11:31 UTC (permalink / raw) To: enming.teo, weidong.han, djmagee; +Cc: xen-devel When I google about XCI, it points me to http://wiki.xensource.com/xenwiki/General_Project_Information Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Wednesday, August 26, 2009 7:25 PM To: weidong.han@intel.com; djmagee@mageenet.net Cc: xen-devel@lists.xensource.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machinewith Intel Desktop Board DQ45CB Hi Weidong, Currently I am using the entire Xen 3.5-unstable branch and not using the posted pci-stub patch changeset 19893 to patch Xen 3.4.1-testing. I am using Xen 3.5-unstable in conjunction with pvops dom 0 kernel 2.6.31-rc6. Here are my questions: 1) How do I extract out the VGA BIOS/firmware from the firmware/EEPROM chip on the Geforce 8400 GS VGA card? If it is not possible to extract, where can I download the firmware? NVIDIA does not provide display card firmware/BIOS files for public download on its website. 2) How do I get the Geforce 8400 GS VGA BIOS to load instead of the emulated VGA BIOS? Which configuration files/source codes do I have to configure/patch? 3) What are physical MMIO, virtual MMIO Bars, dsdt.asl and 1:1 map? I am sorry but I am not a graphics card hardware engineer so I do not understand those terms. I would appreciate it if you could write out some steps that I could carry out or point me to already available documentation. 4) I don't mind using experimental code. Please post it on the xen-devel mailing list so that I could try it out. Thank you very much. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: Han, Weidong [mailto:weidong.han@intel.com] Sent: Wednesday, August 26, 2009 6:56 PM To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Teo En Ming (Zhang Enming) wrote: > Hi Weidong, > > Could you share with us the hack codes for making Geforce 8400 GS > work and also how to let the Windows HVM guest boot up using the real > BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? > What patch are you using now? Using real VGA bios of gfx card to replace emulatd VGA bios is the prerequisite of gfx passthrough. You can find it in posted gfx passthrough patches or XCI. For hack of making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of the card. Currently our code is experimental, we will send out in mailing list after cleanup and more tests. Regards, Weidong > Thank you. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > Alma Maters: Singapore Polytechnic, National University of Singapore > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, > Weidong Sent: Wednesday, August 26, 2009 4:27 PM > To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 > VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with > IntelDesktop BoardDQ45CB > > I suppose you just use the patch posted in mailing list before. nVidia > Geforce 8400 passthrough needs extra hacks. We can make it work in our > experiments with 1:1 map of its MMIO BARs. > > We are working on gfx passthrough on latest xen-unstable. Firstly, we > want to cook a simple patch including generic changes to support > passthrough of virtualization friendly gfx cards, such as Nvidia > FX3800. This patch is basically done. Then, we will add some hacks > for more gfx cards passthrough, such as iGFX and some Nvidia and ATI > cards. > > Regards, > Weidong > > djmagee@mageenet.net wrote: >> As I've pointed out on this list before, there are not enough PCIe >> lanes on the DQ45CB to drive both the internal graphics adapter and >> the add-on adapter at the same time. I believe the onboard one may >> be able to operate in some sort of VGA only mode when there is a card >> installed and used as the primary adapter. >> >> I have a similar setup, using the same motherboard, and an ATI 4770. >> I used the VGA passthrough patches from an earlier posting (I saw you >> followed the same set of instructions), with limited success. I've >> been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made >> modifications to the dom0 portion of the patches so they would apply >> to my xenified 2.6.29 kernel. These patches include code that will >> copy the VGA bios to the guest bios instead of using the emulated vga >> bios. I've had very little success, however. I have only tried >> passing through the ATI adapter. In all instances, the guest bios >> messages appear on my monitor, so this much works. In some cases, >> the guest essentially stops there; xm list show's about 2sec CPU >> usage and nothing ever happens after that point. In other cases, the >> guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot all >> the way, but in very low res/color mode, and cannot properly >> initialize the video device. Once or twice, it actually did >> recognize the device and had a reasonable default color/resolution >> combination. In all cases where the guest actually boots, the system >> eventually freezes. In some cases, I get endless streams of iommu >> page faults. >> >> I have 8GB ram installed. In all cases I've limited dom0 memory to >> 2GB. In all cases, my guest has been assigned 2GB of memory. >> >> I have a dual core e6600. I've tried allowing dom0 to use both >> cores, offlining one core (using xend dom0_vcpu setting) after boot, >> and restricting dom0 to only one core using the dom0_max_vcpus xen >> hypervisor parameter. In all of these cases I've tried both one and >> two vcpus for the guest. My success with VGA passthrough seems >> somewhat random and no combination of cpu assignment seems to have >> any effect. >> >> I have not tried with the 2.6.18-xen kernel as I haven't gotten it to >> boot on my hardware; it can never find my volume group, even if I >> create a initrd with all of the required modules, or build those >> drivers into the kernel. I have not spent more than maybe a half an >> hour on this problem; I suspect it may have something to do with the >> version of mkinitrd I'm using (from Fedora 9 x64). >> >> If anyone else has any insight or similar experience I'd also love >> to hear it. >> >> Doug Magee >> djmagee@mageenet.com >> >> -----Original Message----- >> From: xen-devel-bounces@lists.xensource.com >> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >> To: enming.teo@asiasoftsea.net >> Cc: xen-devel@lists.xensource.com >> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel >> Desktop BoardDQ45CB >> >> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >> suggested by Weidong. >> >> >> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>> Dear All, >>> >>> I have managed to do PCI-e VGA passthrough with the open source Xen >>> but the work is still in progress because although Windows XP guest >>> can see the REAL PCI-e x16 graphics card instead of an emulated >>> graphics driver, it cannot be initialized yet. >>> >>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>> shared their knowledge with me along the way. >>> >>> System Configuration: >>> >>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>> Onboard Intel GMA 4500 Graphics (IGD) >>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>> >>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>> Primary Video Adapter in BIOS: IGD >>> >>> Please see the screenshots and my blog at the link here: >>> >>> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 > -gs-pci-express-x16.html >>> >>> >> >> >> >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 11:31 ` Teo En Ming (Zhang Enming) @ 2009-08-26 11:37 ` Vincent Hanquez 2009-08-26 11:49 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Vincent Hanquez @ 2009-08-26 11:37 UTC (permalink / raw) To: enming.teo; +Cc: djmagee, xen-devel, weidong.han Teo En Ming (Zhang Enming) wrote: > When I google about XCI, it points me to > http://wiki.xensource.com/xenwiki/General_Project_Information > you might want to have a look at: http://wiki.xensource.com/xenwiki/XCI -- Vincent ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 11:37 ` Vincent Hanquez @ 2009-08-26 11:49 ` Teo En Ming (Zhang Enming) 2009-08-26 12:17 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-26 11:49 UTC (permalink / raw) To: vincent.hanquez; +Cc: weidong.han, xen-devel, djmagee Hi Vincent, Thank you for your referral. Upon further googling, I also found this: http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00626.html Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Vincent Hanquez Sent: Wednesday, August 26, 2009 7:38 PM To: enming.teo@asiasoftsea.net Cc: djmagee@mageenet.net; xen-devel@lists.xensource.com; weidong.han@intel.com Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGAPass Throughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB Teo En Ming (Zhang Enming) wrote: > When I google about XCI, it points me to > http://wiki.xensource.com/xenwiki/General_Project_Information > you might want to have a look at: http://wiki.xensource.com/xenwiki/XCI -- Vincent _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 11:49 ` Teo En Ming (Zhang Enming) @ 2009-08-26 12:17 ` Teo En Ming (Zhang Enming) 2009-08-26 13:33 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-26 12:17 UTC (permalink / raw) To: enming.teo, vincent.hanquez; +Cc: djmagee, xen-devel, weidong.han Found this topic entitled "good progress on pci vga passthrough in xen-3.4-rc1" Link: http://www.nabble.com/good-progress-on-pci-vga-passthrough-in-xen-3.4-rc1-td 23015810.html Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Wednesday, August 26, 2009 7:49 PM To: vincent.hanquez@eu.citrix.com Cc: weidong.han@intel.com; xen-devel@lists.xensource.com; djmagee@mageenet.net Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machinewith Intel Desktop Board DQ45CB Hi Vincent, Thank you for your referral. Upon further googling, I also found this: http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00626.html Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Vincent Hanquez Sent: Wednesday, August 26, 2009 7:38 PM To: enming.teo@asiasoftsea.net Cc: djmagee@mageenet.net; xen-devel@lists.xensource.com; weidong.han@intel.com Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGAPass Throughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB Teo En Ming (Zhang Enming) wrote: > When I google about XCI, it points me to > http://wiki.xensource.com/xenwiki/General_Project_Information > you might want to have a look at: http://wiki.xensource.com/xenwiki/XCI -- Vincent _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 12:17 ` Teo En Ming (Zhang Enming) @ 2009-08-26 13:33 ` Teo En Ming (Zhang Enming) 2009-08-26 13:37 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-26 13:33 UTC (permalink / raw) To: enming.teo, vincent.hanquez; +Cc: weidong.han, xen-devel, djmagee OK, from what I have read so far, I have to port the code for PCI/PCI-e vga graphics pass through from XCI to patch Xen 3.5-unstable so that booting the real vga bios of the real vga card is possible instead of the emulated vga bios. Is that correct? Xen 3.5-unstable code supports PCI pass through but doesn't have the code for booting the real vga bios? If so, what code from the XCI sources do I have to single out to patch Xen 3.5-unstable? Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Wednesday, August 26, 2009 8:18 PM To: enming.teo@asiasoftsea.net; vincent.hanquez@eu.citrix.com Cc: djmagee@mageenet.net; xen-devel@lists.xensource.com; weidong.han@intel.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel Desktop Board DQ45CB Found this topic entitled "good progress on pci vga passthrough in xen-3.4-rc1" Link: http://www.nabble.com/good-progress-on-pci-vga-passthrough-in-xen-3.4-rc1-td 23015810.html Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Wednesday, August 26, 2009 7:49 PM To: vincent.hanquez@eu.citrix.com Cc: weidong.han@intel.com; xen-devel@lists.xensource.com; djmagee@mageenet.net Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machinewith Intel Desktop Board DQ45CB Hi Vincent, Thank you for your referral. Upon further googling, I also found this: http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00626.html Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Vincent Hanquez Sent: Wednesday, August 26, 2009 7:38 PM To: enming.teo@asiasoftsea.net Cc: djmagee@mageenet.net; xen-devel@lists.xensource.com; weidong.han@intel.com Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGAPass Throughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB Teo En Ming (Zhang Enming) wrote: > When I google about XCI, it points me to > http://wiki.xensource.com/xenwiki/General_Project_Information > you might want to have a look at: http://wiki.xensource.com/xenwiki/XCI -- Vincent _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 13:33 ` Teo En Ming (Zhang Enming) @ 2009-08-26 13:37 ` Teo En Ming (Zhang Enming) 2009-08-26 21:08 ` VGA passthrough to HVM Tim Moore 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-26 13:37 UTC (permalink / raw) To: enming.teo, vincent.hanquez; +Cc: djmagee, xen-devel, weidong.han There's very little documentation with XCI. So much so that users of XCI doesn't know what to do with it. See here: Topic: Xen XCI Compiled and Installed - Now what? Link: http://www.nabble.com/Xen-XCI-Compiled-and-Installed---Now-what--td24657428. html Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Wednesday, August 26, 2009 9:33 PM To: enming.teo@asiasoftsea.net; vincent.hanquez@eu.citrix.com Cc: weidong.han@intel.com; xen-devel@lists.xensource.com; djmagee@mageenet.net Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel Desktop Board DQ45CB OK, from what I have read so far, I have to port the code for PCI/PCI-e vga graphics pass through from XCI to patch Xen 3.5-unstable so that booting the real vga bios of the real vga card is possible instead of the emulated vga bios. Is that correct? Xen 3.5-unstable code supports PCI pass through but doesn't have the code for booting the real vga bios? If so, what code from the XCI sources do I have to single out to patch Xen 3.5-unstable? Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Wednesday, August 26, 2009 8:18 PM To: enming.teo@asiasoftsea.net; vincent.hanquez@eu.citrix.com Cc: djmagee@mageenet.net; xen-devel@lists.xensource.com; weidong.han@intel.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel Desktop Board DQ45CB Found this topic entitled "good progress on pci vga passthrough in xen-3.4-rc1" Link: http://www.nabble.com/good-progress-on-pci-vga-passthrough-in-xen-3.4-rc1-td 23015810.html Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Wednesday, August 26, 2009 7:49 PM To: vincent.hanquez@eu.citrix.com Cc: weidong.han@intel.com; xen-devel@lists.xensource.com; djmagee@mageenet.net Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machinewith Intel Desktop Board DQ45CB Hi Vincent, Thank you for your referral. Upon further googling, I also found this: http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00626.html Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Vincent Hanquez Sent: Wednesday, August 26, 2009 7:38 PM To: enming.teo@asiasoftsea.net Cc: djmagee@mageenet.net; xen-devel@lists.xensource.com; weidong.han@intel.com Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGAPass Throughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB Teo En Ming (Zhang Enming) wrote: > When I google about XCI, it points me to > http://wiki.xensource.com/xenwiki/General_Project_Information > you might want to have a look at: http://wiki.xensource.com/xenwiki/XCI -- Vincent _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: VGA passthrough to HVM 2009-08-26 13:37 ` Teo En Ming (Zhang Enming) @ 2009-08-26 21:08 ` Tim Moore 2009-08-27 2:40 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Tim Moore @ 2009-08-26 21:08 UTC (permalink / raw) To: 'enming.teo@asiasoftsea.net' Cc: 'Xci-devel@lists.xensource.com', xen-devel Hi Teo, We have all been eagerly awaiting the PCI Passthrough for VGA patches to be released by the Xen-Devel community .. but from the sounds of the guys at Intel - it's not ready for public testing and it will most probably generate more questions for them to answer. Ultimately this is a distraction and will only delay the release of this functionality. The closest you will get at the moment is the patches which form part of the XCI build tree and support is limited to the Intel chipset and onboard Intel graphics controller. I believe Ian Jackson also has some success with ATI but NVidia cards are known not to work. There are other examples available including the "Neo 1:1 patch" which is based on an older version of Xen (I think 3.2). NOTE: This ONLY applies to HVM, VGA Passthrough using xen-unstable and Linux non-HVM DomU is working, I have success with many Nvidia cards using this setup. You can use this to prove your Xen in preparation for the HVM patches and to ensure that your VT-d and Dom0 are sound. Peter J. van der Maas inspired us all with his announcement, but no one has ever been able to reproduce: http://www.nabble.com/Successful-PCIe-Graphics-VT-d-Passthrough-to-Win32-DomU,-Q35-chipset-td21671745.html Myself and Beng Heng also got to the same point as yourself, with issues on vBar<>pBar mapping and VGA BIOS. In all cases, the VGA card is visible in Device Manager in the DomU but the driver does not load. On My 8800GTS, the VGA Card FAN spins down the same as when booting under native XP but still the driver does not load. Here are some good links with relevant information. http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/presentation.pdf http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/report.pdf http://www.eecs.umich.edu/~bengheng/pubs/vgapt_techreport.pdf http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00579.html This is where I admit defeat and hand over to the brilliance of the Xen developers! My last state is captured here: http://markmail.org/message/y476tymnhu2wrc47 The most success I have evidence of is from Alex Bebbington here: http://markmail.org/message/mjiv437pva2yrhvp If you look closely, this feature is on the Offical Xen Roadmap for Xen 4.0 captured here: http://wiki.xensource.com/xenwiki/XenRoadMap {XCI} There is sufficient information for the XCI project to build and create a demo install for the enthusiast: http://wiki.xensource.com/xenwiki/XCI http://xenbits.xen.org/xenclient/ I was downloading and building XCI before the above information was available by reading the XCI build tree and patches. The code is quite well structured along with containing some additional information on the xenvm and xenops commands: http://xenbits.xen.org/gitweb?p=xenclient/toolstack.git;a=blob;f=xenvm/xenvm.readme Hopefully this pulls together all the information that is available on Xen VGA Passthrough. Rgds, Tim Moore ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: VGA passthrough to HVM 2009-08-26 21:08 ` VGA passthrough to HVM Tim Moore @ 2009-08-27 2:40 ` Teo En Ming (Zhang Enming) 2009-08-27 2:59 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-27 2:40 UTC (permalink / raw) To: timothy.moore; +Cc: Xci-devel, xen-devel Dear Timothy, Thank you for compiling all the links to the currently available work on Xen PCI/PCI-E/PEG VGA pass through to Windows HVM guests. When I looked at Peter J. van der Maas's reported success, I have some doubts. Intel Desktop Board DQ35JO only has an onboard graphics and an external PCI-E x16 slot. How was Peter J. van der Maas able to use two external display cards? Assuming that his nVidia GeForce2 MX200 is a PCI card, and his nVidia 9500GT is a PCI-e x16 card, this resolves the above doubt. Even then, when you plug in both PCI and PEG graphics card at the same time, as is with the case on my Intel Desktop Board DQ45CB, only the PCI VGA card will work and display to the monitor. The PCI-E x16 display card will not work. When the PCI display card is plugged in, my motherboard DQ45CB can't even detect how many lanes my PEG display card is using (should be 16 lanes). Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Tim Moore Sent: Thursday, August 27, 2009 5:09 AM To: 'enming.teo@asiasoftsea.net' Cc: 'Xci-devel@lists.xensource.com'; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Hi Teo, We have all been eagerly awaiting the PCI Passthrough for VGA patches to be released by the Xen-Devel community .. but from the sounds of the guys at Intel - it's not ready for public testing and it will most probably generate more questions for them to answer. Ultimately this is a distraction and will only delay the release of this functionality. The closest you will get at the moment is the patches which form part of the XCI build tree and support is limited to the Intel chipset and onboard Intel graphics controller. I believe Ian Jackson also has some success with ATI but NVidia cards are known not to work. There are other examples available including the "Neo 1:1 patch" which is based on an older version of Xen (I think 3.2). NOTE: This ONLY applies to HVM, VGA Passthrough using xen-unstable and Linux non-HVM DomU is working, I have success with many Nvidia cards using this setup. You can use this to prove your Xen in preparation for the HVM patches and to ensure that your VT-d and Dom0 are sound. Peter J. van der Maas inspired us all with his announcement, but no one has ever been able to reproduce: http://www.nabble.com/Successful-PCIe-Graphics-VT-d-Passthrough-to-Win32-Dom U,-Q35-chipset-td21671745.html Myself and Beng Heng also got to the same point as yourself, with issues on vBar<>pBar mapping and VGA BIOS. In all cases, the VGA card is visible in Device Manager in the DomU but the driver does not load. On My 8800GTS, the VGA Card FAN spins down the same as when booting under native XP but still the driver does not load. Here are some good links with relevant information. http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/presentation.pdf http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/report.pdf http://www.eecs.umich.edu/~bengheng/pubs/vgapt_techreport.pdf http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00579.html This is where I admit defeat and hand over to the brilliance of the Xen developers! My last state is captured here: http://markmail.org/message/y476tymnhu2wrc47 The most success I have evidence of is from Alex Bebbington here: http://markmail.org/message/mjiv437pva2yrhvp If you look closely, this feature is on the Offical Xen Roadmap for Xen 4.0 captured here: http://wiki.xensource.com/xenwiki/XenRoadMap {XCI} There is sufficient information for the XCI project to build and create a demo install for the enthusiast: http://wiki.xensource.com/xenwiki/XCI http://xenbits.xen.org/xenclient/ I was downloading and building XCI before the above information was available by reading the XCI build tree and patches. The code is quite well structured along with containing some additional information on the xenvm and xenops commands: http://xenbits.xen.org/gitweb?p=xenclient/toolstack.git;a=blob;f=xenvm/xenvm .readme Hopefully this pulls together all the information that is available on Xen VGA Passthrough. Rgds, Tim Moore _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: VGA passthrough to HVM 2009-08-27 2:40 ` Teo En Ming (Zhang Enming) @ 2009-08-27 2:59 ` Teo En Ming (Zhang Enming) 2009-08-28 2:19 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-27 2:59 UTC (permalink / raw) To: enming.teo, timothy.moore; +Cc: Xci-devel, xen-devel Here are my explanations on why we couldn't reproduce Peter van der Maas's work: 1) Our setups are vastly different 2) Peter van der Maas is using Intel DQ35JO while I am using Intel DQ45CB with different chipsets (Q35 vs Q45) 3) We are using different display card brand/model and different primary/secondary video controller setups 4) Peter van der Maas was using Xen unstable as of Jan 2009, while I am using Xen 3.5-unstable as of August 2009 5) Peter van der Maas was using Linux-2.6.18.8-xen Domain 0 NON-paravirt-ops kernel. He mentioned using pciback which I cannot use at all. I am using Jeremy Fitzhardinge's 2.6.31-rc6 Dom 0 pv-ops kernel. I can only use pci-stub compiled in or as module. 6) Peter van der Maas was using Windows XP Professional SP3 32-bit. I am using Windows XP Home SP3 32-bit. We have too many variables in our experimental setups. In order to reproduce and validate another scientist's work, we should have exactly the same setups. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 10:40 AM To: timothy.moore@expidas.net Cc: Xci-devel@lists.xensource.com; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Dear Timothy, Thank you for compiling all the links to the currently available work on Xen PCI/PCI-E/PEG VGA pass through to Windows HVM guests. When I looked at Peter J. van der Maas's reported success, I have some doubts. Intel Desktop Board DQ35JO only has an onboard graphics and an external PCI-E x16 slot. How was Peter J. van der Maas able to use two external display cards? Assuming that his nVidia GeForce2 MX200 is a PCI card, and his nVidia 9500GT is a PCI-e x16 card, this resolves the above doubt. Even then, when you plug in both PCI and PEG graphics card at the same time, as is with the case on my Intel Desktop Board DQ45CB, only the PCI VGA card will work and display to the monitor. The PCI-E x16 display card will not work. When the PCI display card is plugged in, my motherboard DQ45CB can't even detect how many lanes my PEG display card is using (should be 16 lanes). Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Tim Moore Sent: Thursday, August 27, 2009 5:09 AM To: 'enming.teo@asiasoftsea.net' Cc: 'Xci-devel@lists.xensource.com'; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Hi Teo, We have all been eagerly awaiting the PCI Passthrough for VGA patches to be released by the Xen-Devel community .. but from the sounds of the guys at Intel - it's not ready for public testing and it will most probably generate more questions for them to answer. Ultimately this is a distraction and will only delay the release of this functionality. The closest you will get at the moment is the patches which form part of the XCI build tree and support is limited to the Intel chipset and onboard Intel graphics controller. I believe Ian Jackson also has some success with ATI but NVidia cards are known not to work. There are other examples available including the "Neo 1:1 patch" which is based on an older version of Xen (I think 3.2). NOTE: This ONLY applies to HVM, VGA Passthrough using xen-unstable and Linux non-HVM DomU is working, I have success with many Nvidia cards using this setup. You can use this to prove your Xen in preparation for the HVM patches and to ensure that your VT-d and Dom0 are sound. Peter J. van der Maas inspired us all with his announcement, but no one has ever been able to reproduce: http://www.nabble.com/Successful-PCIe-Graphics-VT-d-Passthrough-to-Win32-Dom U,-Q35-chipset-td21671745.html Myself and Beng Heng also got to the same point as yourself, with issues on vBar<>pBar mapping and VGA BIOS. In all cases, the VGA card is visible in Device Manager in the DomU but the driver does not load. On My 8800GTS, the VGA Card FAN spins down the same as when booting under native XP but still the driver does not load. Here are some good links with relevant information. http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/presentation.pdf http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/report.pdf http://www.eecs.umich.edu/~bengheng/pubs/vgapt_techreport.pdf http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00579.html This is where I admit defeat and hand over to the brilliance of the Xen developers! My last state is captured here: http://markmail.org/message/y476tymnhu2wrc47 The most success I have evidence of is from Alex Bebbington here: http://markmail.org/message/mjiv437pva2yrhvp If you look closely, this feature is on the Offical Xen Roadmap for Xen 4.0 captured here: http://wiki.xensource.com/xenwiki/XenRoadMap {XCI} There is sufficient information for the XCI project to build and create a demo install for the enthusiast: http://wiki.xensource.com/xenwiki/XCI http://xenbits.xen.org/xenclient/ I was downloading and building XCI before the above information was available by reading the XCI build tree and patches. The code is quite well structured along with containing some additional information on the xenvm and xenops commands: http://xenbits.xen.org/gitweb?p=xenclient/toolstack.git;a=blob;f=xenvm/xenvm .readme Hopefully this pulls together all the information that is available on Xen VGA Passthrough. Rgds, Tim Moore _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: VGA passthrough to HVM 2009-08-27 2:59 ` Teo En Ming (Zhang Enming) @ 2009-08-28 2:19 ` Teo En Ming (Zhang Enming) 2009-08-28 2:51 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-28 2:19 UTC (permalink / raw) To: enming.teo, timothy.moore; +Cc: Xci-devel, xen-devel Dear Tim Moore, Could you help me verify that my findings here http://lists.xensource.com/archives/html/xen-devel/2009-08/msg01032.html are correct and whether anything more should be done? Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 11:00 AM To: enming.teo@asiasoftsea.net; timothy.moore@expidas.net Cc: Xci-devel@lists.xensource.com; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Here are my explanations on why we couldn't reproduce Peter van der Maas's work: 1) Our setups are vastly different 2) Peter van der Maas is using Intel DQ35JO while I am using Intel DQ45CB with different chipsets (Q35 vs Q45) 3) We are using different display card brand/model and different primary/secondary video controller setups 4) Peter van der Maas was using Xen unstable as of Jan 2009, while I am using Xen 3.5-unstable as of August 2009 5) Peter van der Maas was using Linux-2.6.18.8-xen Domain 0 NON-paravirt-ops kernel. He mentioned using pciback which I cannot use at all. I am using Jeremy Fitzhardinge's 2.6.31-rc6 Dom 0 pv-ops kernel. I can only use pci-stub compiled in or as module. 6) Peter van der Maas was using Windows XP Professional SP3 32-bit. I am using Windows XP Home SP3 32-bit. We have too many variables in our experimental setups. In order to reproduce and validate another scientist's work, we should have exactly the same setups. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 10:40 AM To: timothy.moore@expidas.net Cc: Xci-devel@lists.xensource.com; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Dear Timothy, Thank you for compiling all the links to the currently available work on Xen PCI/PCI-E/PEG VGA pass through to Windows HVM guests. When I looked at Peter J. van der Maas's reported success, I have some doubts. Intel Desktop Board DQ35JO only has an onboard graphics and an external PCI-E x16 slot. How was Peter J. van der Maas able to use two external display cards? Assuming that his nVidia GeForce2 MX200 is a PCI card, and his nVidia 9500GT is a PCI-e x16 card, this resolves the above doubt. Even then, when you plug in both PCI and PEG graphics card at the same time, as is with the case on my Intel Desktop Board DQ45CB, only the PCI VGA card will work and display to the monitor. The PCI-E x16 display card will not work. When the PCI display card is plugged in, my motherboard DQ45CB can't even detect how many lanes my PEG display card is using (should be 16 lanes). Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Tim Moore Sent: Thursday, August 27, 2009 5:09 AM To: 'enming.teo@asiasoftsea.net' Cc: 'Xci-devel@lists.xensource.com'; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Hi Teo, We have all been eagerly awaiting the PCI Passthrough for VGA patches to be released by the Xen-Devel community .. but from the sounds of the guys at Intel - it's not ready for public testing and it will most probably generate more questions for them to answer. Ultimately this is a distraction and will only delay the release of this functionality. The closest you will get at the moment is the patches which form part of the XCI build tree and support is limited to the Intel chipset and onboard Intel graphics controller. I believe Ian Jackson also has some success with ATI but NVidia cards are known not to work. There are other examples available including the "Neo 1:1 patch" which is based on an older version of Xen (I think 3.2). NOTE: This ONLY applies to HVM, VGA Passthrough using xen-unstable and Linux non-HVM DomU is working, I have success with many Nvidia cards using this setup. You can use this to prove your Xen in preparation for the HVM patches and to ensure that your VT-d and Dom0 are sound. Peter J. van der Maas inspired us all with his announcement, but no one has ever been able to reproduce: http://www.nabble.com/Successful-PCIe-Graphics-VT-d-Passthrough-to-Win32-Dom U,-Q35-chipset-td21671745.html Myself and Beng Heng also got to the same point as yourself, with issues on vBar<>pBar mapping and VGA BIOS. In all cases, the VGA card is visible in Device Manager in the DomU but the driver does not load. On My 8800GTS, the VGA Card FAN spins down the same as when booting under native XP but still the driver does not load. Here are some good links with relevant information. http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/presentation.pdf http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/report.pdf http://www.eecs.umich.edu/~bengheng/pubs/vgapt_techreport.pdf http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00579.html This is where I admit defeat and hand over to the brilliance of the Xen developers! My last state is captured here: http://markmail.org/message/y476tymnhu2wrc47 The most success I have evidence of is from Alex Bebbington here: http://markmail.org/message/mjiv437pva2yrhvp If you look closely, this feature is on the Offical Xen Roadmap for Xen 4.0 captured here: http://wiki.xensource.com/xenwiki/XenRoadMap {XCI} There is sufficient information for the XCI project to build and create a demo install for the enthusiast: http://wiki.xensource.com/xenwiki/XCI http://xenbits.xen.org/xenclient/ I was downloading and building XCI before the above information was available by reading the XCI build tree and patches. The code is quite well structured along with containing some additional information on the xenvm and xenops commands: http://xenbits.xen.org/gitweb?p=xenclient/toolstack.git;a=blob;f=xenvm/xenvm .readme Hopefully this pulls together all the information that is available on Xen VGA Passthrough. Rgds, Tim Moore _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.409 / Virus Database: 270.13.69/2328 - Release Date: 08/27/09 18:02:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: VGA passthrough to HVM 2009-08-28 2:19 ` Teo En Ming (Zhang Enming) @ 2009-08-28 2:51 ` Teo En Ming (Zhang Enming) 0 siblings, 0 replies; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-28 2:51 UTC (permalink / raw) To: enming.teo, timothy.moore; +Cc: Xci-devel, xen-devel I have a feeling that my findings are far too simplistic and it won't work. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Friday, August 28, 2009 10:20 AM To: enming.teo@asiasoftsea.net; timothy.moore@expidas.net Cc: Xci-devel@lists.xensource.com; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Dear Tim Moore, Could you help me verify that my findings here http://lists.xensource.com/archives/html/xen-devel/2009-08/msg01032.html are correct and whether anything more should be done? Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 11:00 AM To: enming.teo@asiasoftsea.net; timothy.moore@expidas.net Cc: Xci-devel@lists.xensource.com; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Here are my explanations on why we couldn't reproduce Peter van der Maas's work: 1) Our setups are vastly different 2) Peter van der Maas is using Intel DQ35JO while I am using Intel DQ45CB with different chipsets (Q35 vs Q45) 3) We are using different display card brand/model and different primary/secondary video controller setups 4) Peter van der Maas was using Xen unstable as of Jan 2009, while I am using Xen 3.5-unstable as of August 2009 5) Peter van der Maas was using Linux-2.6.18.8-xen Domain 0 NON-paravirt-ops kernel. He mentioned using pciback which I cannot use at all. I am using Jeremy Fitzhardinge's 2.6.31-rc6 Dom 0 pv-ops kernel. I can only use pci-stub compiled in or as module. 6) Peter van der Maas was using Windows XP Professional SP3 32-bit. I am using Windows XP Home SP3 32-bit. We have too many variables in our experimental setups. In order to reproduce and validate another scientist's work, we should have exactly the same setups. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 10:40 AM To: timothy.moore@expidas.net Cc: Xci-devel@lists.xensource.com; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Dear Timothy, Thank you for compiling all the links to the currently available work on Xen PCI/PCI-E/PEG VGA pass through to Windows HVM guests. When I looked at Peter J. van der Maas's reported success, I have some doubts. Intel Desktop Board DQ35JO only has an onboard graphics and an external PCI-E x16 slot. How was Peter J. van der Maas able to use two external display cards? Assuming that his nVidia GeForce2 MX200 is a PCI card, and his nVidia 9500GT is a PCI-e x16 card, this resolves the above doubt. Even then, when you plug in both PCI and PEG graphics card at the same time, as is with the case on my Intel Desktop Board DQ45CB, only the PCI VGA card will work and display to the monitor. The PCI-E x16 display card will not work. When the PCI display card is plugged in, my motherboard DQ45CB can't even detect how many lanes my PEG display card is using (should be 16 lanes). Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Tim Moore Sent: Thursday, August 27, 2009 5:09 AM To: 'enming.teo@asiasoftsea.net' Cc: 'Xci-devel@lists.xensource.com'; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] VGA passthrough to HVM Hi Teo, We have all been eagerly awaiting the PCI Passthrough for VGA patches to be released by the Xen-Devel community .. but from the sounds of the guys at Intel - it's not ready for public testing and it will most probably generate more questions for them to answer. Ultimately this is a distraction and will only delay the release of this functionality. The closest you will get at the moment is the patches which form part of the XCI build tree and support is limited to the Intel chipset and onboard Intel graphics controller. I believe Ian Jackson also has some success with ATI but NVidia cards are known not to work. There are other examples available including the "Neo 1:1 patch" which is based on an older version of Xen (I think 3.2). NOTE: This ONLY applies to HVM, VGA Passthrough using xen-unstable and Linux non-HVM DomU is working, I have success with many Nvidia cards using this setup. You can use this to prove your Xen in preparation for the HVM patches and to ensure that your VT-d and Dom0 are sound. Peter J. van der Maas inspired us all with his announcement, but no one has ever been able to reproduce: http://www.nabble.com/Successful-PCIe-Graphics-VT-d-Passthrough-to-Win32-Dom U,-Q35-chipset-td21671745.html Myself and Beng Heng also got to the same point as yourself, with issues on vBar<>pBar mapping and VGA BIOS. In all cases, the VGA card is visible in Device Manager in the DomU but the driver does not load. On My 8800GTS, the VGA Card FAN spins down the same as when booting under native XP but still the driver does not load. Here are some good links with relevant information. http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/presentation.pdf http://staff.science.uva.nl/~delaat/sne-2008-2009/p22/report.pdf http://www.eecs.umich.edu/~bengheng/pubs/vgapt_techreport.pdf http://lists.xensource.com/archives/html/xen-devel/2009-05/msg00579.html This is where I admit defeat and hand over to the brilliance of the Xen developers! My last state is captured here: http://markmail.org/message/y476tymnhu2wrc47 The most success I have evidence of is from Alex Bebbington here: http://markmail.org/message/mjiv437pva2yrhvp If you look closely, this feature is on the Offical Xen Roadmap for Xen 4.0 captured here: http://wiki.xensource.com/xenwiki/XenRoadMap {XCI} There is sufficient information for the XCI project to build and create a demo install for the enthusiast: http://wiki.xensource.com/xenwiki/XCI http://xenbits.xen.org/xenclient/ I was downloading and building XCI before the above information was available by reading the XCI build tree and patches. The code is quite well structured along with containing some additional information on the xenvm and xenops commands: http://xenbits.xen.org/gitweb?p=xenclient/toolstack.git;a=blob;f=xenvm/xenvm .readme Hopefully this pulls together all the information that is available on Xen VGA Passthrough. Rgds, Tim Moore _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.409 / Virus Database: 270.13.69/2328 - Release Date: 08/27/09 18:02:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.409 / Virus Database: 270.13.69/2328 - Release Date: 08/27/09 18:02:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.409 / Virus Database: 270.13.69/2328 - Release Date: 08/27/09 18:02:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-26 11:24 ` Teo En Ming (Zhang Enming) 2009-08-26 11:31 ` Teo En Ming (Zhang Enming) @ 2009-08-27 2:01 ` Han, Weidong 2009-08-27 3:18 ` Teo En Ming (Zhang Enming) 1 sibling, 1 reply; 25+ messages in thread From: Han, Weidong @ 2009-08-27 2:01 UTC (permalink / raw) To: 'enming.teo@asiasoftsea.net', 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' Teo En Ming (Zhang Enming) wrote: > Hi Weidong, > > Currently I am using the entire Xen 3.5-unstable branch and not using > the posted pci-stub patch changeset 19893 to patch Xen 3.4.1-testing. > I am using Xen 3.5-unstable in conjunction with pvops dom 0 kernel > 2.6.31-rc6. > > Here are my questions: > > 1) How do I extract out the VGA BIOS/firmware from the > firmware/EEPROM chip on the Geforce 8400 GS VGA card? If it is not > possible to extract, where can I download the firmware? NVIDIA does > not provide display card firmware/BIOS files for public download on > its website. There are 3 ways to load VGA bios to guest: 1. copy it from host 0xc0000, and map it to guest 0xc0000. This works for primary gfx card, XCI does like this. But some VGA bios re-execution doesn't work well. 2. dump VGA bios to a file first, then load it to guest: you can use NiBiTor to dump VGA bios of Nvidia gfx cards to a file, then load it into guest roms like what emulated VGA does in hvmloader. This works for all gfx cards, but it involves manual steps. 3. load VGA bios from expansion rom: Intel iGFX doesn't have expansion rom > > 2) How do I get the Geforce 8400 GS VGA BIOS to load instead of the > emulated VGA BIOS? Which configuration files/source codes do I have to > configure/patch? don't initialize emulated VGA in qemu, I I think you have seen the code. > > 3) What are physical MMIO, virtual MMIO Bars, dsdt.asl and 1:1 map? I > am sorry but I am not a graphics card hardware engineer so I do not > understand those terms. I would appreciate it if you could write out > some steps that I could carry out or point me to already available > documentation. When you use command 'lspci -v', you may see below messages under each device: ... Memory at d0000000 (32-bit, prefetchable) [size=256M] Memory at ff900000 (32-bit, non-prefetchable) [size=1M] ... They are MMIO BARs. 1:1 map means make these MMIO BARs in guest are equal to the values in host. > > 4) I don't mind using experimental code. Please post it on the > xen-devel mailing list so that I could try it out. we will post in near future. pls wait for a while. Regards, Weidong > > Thank you very much. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > > -----Original Message----- > From: Han, Weidong [mailto:weidong.han@intel.com] > Sent: Wednesday, August 26, 2009 6:56 PM > To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel > Desktop Board DQ45CB > > Teo En Ming (Zhang Enming) wrote: >> Hi Weidong, >> >> Could you share with us the hack codes for making Geforce 8400 GS >> work and also how to let the Windows HVM guest boot up using the real >> BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? >> > > What patch are you using now? Using real VGA bios of gfx card to > replace emulatd VGA bios is the prerequisite of gfx passthrough. You > can find it in posted gfx passthrough patches or XCI. For hack of > making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and > make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of > the card. Currently our code is experimental, we will send out in > mailing list after cleanup and more tests. > > Regards, > Weidong > >> Thank you. >> >> Regards, >> >> Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) >> BEng(Hons)(Mechanical Engineering) >> Technical Support Engineer >> Information Technology Department >> Asiasoft Online Pte Ltd >> Tampines Central 1 #04-01 Tampines Plaza >> Singapore 529541 >> Republic of Singapore >> Mobile: +65-9648-9798 >> MSN: teoenming@hotmail.com >> Alma Maters: Singapore Polytechnic, National University of Singapore >> >> -----Original Message----- >> From: xen-devel-bounces@lists.xensource.com >> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, >> Weidong Sent: Wednesday, August 26, 2009 4:27 PM >> To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' >> Cc: 'xen-devel@lists.xensource.com' >> Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 >> VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with >> IntelDesktop BoardDQ45CB >> >> I suppose you just use the patch posted in mailing list before. >> nVidia Geforce 8400 passthrough needs extra hacks. We can make it >> work in our experiments with 1:1 map of its MMIO BARs. >> >> We are working on gfx passthrough on latest xen-unstable. Firstly, we >> want to cook a simple patch including generic changes to support >> passthrough of virtualization friendly gfx cards, such as Nvidia >> FX3800. This patch is basically done. Then, we will add some hacks >> for more gfx cards passthrough, such as iGFX and some Nvidia and ATI >> cards. >> >> Regards, >> Weidong >> >> djmagee@mageenet.net wrote: >>> As I've pointed out on this list before, there are not enough PCIe >>> lanes on the DQ45CB to drive both the internal graphics adapter and >>> the add-on adapter at the same time. I believe the onboard one may >>> be able to operate in some sort of VGA only mode when there is a >>> card installed and used as the primary adapter. >>> >>> I have a similar setup, using the same motherboard, and an ATI 4770. >>> I used the VGA passthrough patches from an earlier posting (I saw >>> you followed the same set of instructions), with limited success. >>> I've been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made >>> modifications to the dom0 portion of the patches so they would apply >>> to my xenified 2.6.29 kernel. These patches include code that will >>> copy the VGA bios to the guest bios instead of using the emulated >>> vga bios. I've had very little success, however. I have only tried >>> passing through the ATI adapter. In all instances, the guest bios >>> messages appear on my monitor, so this much works. In some cases, >>> the guest essentially stops there; xm list show's about 2sec CPU >>> usage and nothing ever happens after that point. In other cases, >>> the guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot >>> all the way, but in very low res/color mode, and cannot properly >>> initialize the video device. Once or twice, it actually did >>> recognize the device and had a reasonable default color/resolution >>> combination. In all cases where the guest actually boots, the >>> system eventually freezes. In some cases, I get endless streams of >>> iommu page faults. >>> >>> I have 8GB ram installed. In all cases I've limited dom0 memory to >>> 2GB. In all cases, my guest has been assigned 2GB of memory. >>> >>> I have a dual core e6600. I've tried allowing dom0 to use both >>> cores, offlining one core (using xend dom0_vcpu setting) after boot, >>> and restricting dom0 to only one core using the dom0_max_vcpus xen >>> hypervisor parameter. In all of these cases I've tried both one and >>> two vcpus for the guest. My success with VGA passthrough seems >>> somewhat random and no combination of cpu assignment seems to have >>> any effect. >>> >>> I have not tried with the 2.6.18-xen kernel as I haven't gotten it >>> to boot on my hardware; it can never find my volume group, even if I >>> create a initrd with all of the required modules, or build those >>> drivers into the kernel. I have not spent more than maybe a half an >>> hour on this problem; I suspect it may have something to do with the >>> version of mkinitrd I'm using (from Fedora 9 x64). >>> >>> If anyone else has any insight or similar experience I'd also love >>> to hear it. >>> >>> Doug Magee >>> djmagee@mageenet.com >>> >>> -----Original Message----- >>> From: xen-devel-bounces@lists.xensource.com >>> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >>> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >>> To: enming.teo@asiasoftsea.net >>> Cc: xen-devel@lists.xensource.com >>> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >>> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel >>> Desktop BoardDQ45CB >>> >>> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >>> suggested by Weidong. >>> >>> >>> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>>> Dear All, >>>> >>>> I have managed to do PCI-e VGA passthrough with the open source Xen >>>> but the work is still in progress because although Windows XP guest >>>> can see the REAL PCI-e x16 graphics card instead of an emulated >>>> graphics driver, it cannot be initialized yet. >>>> >>>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>>> shared their knowledge with me along the way. >>>> >>>> System Configuration: >>>> >>>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>>> Onboard Intel GMA 4500 Graphics (IGD) >>>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>>> >>>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>>> Primary Video Adapter in BIOS: IGD >>>> >>>> Please see the screenshots and my blog at the link here: >>>> >>>> >> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 >> -gs-pci-express-x16.html >>>> >>>> >>> >>> >>> >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >> >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel >> >> No virus found in this incoming message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 >> >> No virus found in this outgoing message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 > > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-27 2:01 ` nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Han, Weidong @ 2009-08-27 3:18 ` Teo En Ming (Zhang Enming) 2009-08-27 6:03 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-27 3:18 UTC (permalink / raw) To: weidong.han, djmagee; +Cc: xen-devel Dear Weidong, I will go for option 2. > don't initialize emulated VGA in qemu, I I think you have seen the code. I have not seen the source code. Though I have learnt simple structured C before in Singapore Polytechnic some 11 years ago, I am not a programmer by profession. So I would still need some instructions on how I can prevent emulated VGA bios from loading and getting the vga bios of the real display card to load. Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, Weidong Sent: Thursday, August 27, 2009 10:01 AM To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine with IntelDesktop Board DQ45CB Teo En Ming (Zhang Enming) wrote: > Hi Weidong, > > Currently I am using the entire Xen 3.5-unstable branch and not using > the posted pci-stub patch changeset 19893 to patch Xen 3.4.1-testing. > I am using Xen 3.5-unstable in conjunction with pvops dom 0 kernel > 2.6.31-rc6. > > Here are my questions: > > 1) How do I extract out the VGA BIOS/firmware from the > firmware/EEPROM chip on the Geforce 8400 GS VGA card? If it is not > possible to extract, where can I download the firmware? NVIDIA does > not provide display card firmware/BIOS files for public download on > its website. There are 3 ways to load VGA bios to guest: 1. copy it from host 0xc0000, and map it to guest 0xc0000. This works for primary gfx card, XCI does like this. But some VGA bios re-execution doesn't work well. 2. dump VGA bios to a file first, then load it to guest: you can use NiBiTor to dump VGA bios of Nvidia gfx cards to a file, then load it into guest roms like what emulated VGA does in hvmloader. This works for all gfx cards, but it involves manual steps. 3. load VGA bios from expansion rom: Intel iGFX doesn't have expansion rom > > 2) How do I get the Geforce 8400 GS VGA BIOS to load instead of the > emulated VGA BIOS? Which configuration files/source codes do I have to > configure/patch? don't initialize emulated VGA in qemu, I I think you have seen the code. > > 3) What are physical MMIO, virtual MMIO Bars, dsdt.asl and 1:1 map? I > am sorry but I am not a graphics card hardware engineer so I do not > understand those terms. I would appreciate it if you could write out > some steps that I could carry out or point me to already available > documentation. When you use command 'lspci -v', you may see below messages under each device: ... Memory at d0000000 (32-bit, prefetchable) [size=256M] Memory at ff900000 (32-bit, non-prefetchable) [size=1M] ... They are MMIO BARs. 1:1 map means make these MMIO BARs in guest are equal to the values in host. > > 4) I don't mind using experimental code. Please post it on the > xen-devel mailing list so that I could try it out. we will post in near future. pls wait for a while. Regards, Weidong > > Thank you very much. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > > -----Original Message----- > From: Han, Weidong [mailto:weidong.han@intel.com] > Sent: Wednesday, August 26, 2009 6:56 PM > To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel > Desktop Board DQ45CB > > Teo En Ming (Zhang Enming) wrote: >> Hi Weidong, >> >> Could you share with us the hack codes for making Geforce 8400 GS >> work and also how to let the Windows HVM guest boot up using the real >> BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? >> > > What patch are you using now? Using real VGA bios of gfx card to > replace emulatd VGA bios is the prerequisite of gfx passthrough. You > can find it in posted gfx passthrough patches or XCI. For hack of > making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and > make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of > the card. Currently our code is experimental, we will send out in > mailing list after cleanup and more tests. > > Regards, > Weidong > >> Thank you. >> >> Regards, >> >> Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) >> BEng(Hons)(Mechanical Engineering) >> Technical Support Engineer >> Information Technology Department >> Asiasoft Online Pte Ltd >> Tampines Central 1 #04-01 Tampines Plaza >> Singapore 529541 >> Republic of Singapore >> Mobile: +65-9648-9798 >> MSN: teoenming@hotmail.com >> Alma Maters: Singapore Polytechnic, National University of Singapore >> >> -----Original Message----- >> From: xen-devel-bounces@lists.xensource.com >> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, >> Weidong Sent: Wednesday, August 26, 2009 4:27 PM >> To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' >> Cc: 'xen-devel@lists.xensource.com' >> Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 >> VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with >> IntelDesktop BoardDQ45CB >> >> I suppose you just use the patch posted in mailing list before. >> nVidia Geforce 8400 passthrough needs extra hacks. We can make it >> work in our experiments with 1:1 map of its MMIO BARs. >> >> We are working on gfx passthrough on latest xen-unstable. Firstly, we >> want to cook a simple patch including generic changes to support >> passthrough of virtualization friendly gfx cards, such as Nvidia >> FX3800. This patch is basically done. Then, we will add some hacks >> for more gfx cards passthrough, such as iGFX and some Nvidia and ATI >> cards. >> >> Regards, >> Weidong >> >> djmagee@mageenet.net wrote: >>> As I've pointed out on this list before, there are not enough PCIe >>> lanes on the DQ45CB to drive both the internal graphics adapter and >>> the add-on adapter at the same time. I believe the onboard one may >>> be able to operate in some sort of VGA only mode when there is a >>> card installed and used as the primary adapter. >>> >>> I have a similar setup, using the same motherboard, and an ATI 4770. >>> I used the VGA passthrough patches from an earlier posting (I saw >>> you followed the same set of instructions), with limited success. >>> I've been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made >>> modifications to the dom0 portion of the patches so they would apply >>> to my xenified 2.6.29 kernel. These patches include code that will >>> copy the VGA bios to the guest bios instead of using the emulated >>> vga bios. I've had very little success, however. I have only tried >>> passing through the ATI adapter. In all instances, the guest bios >>> messages appear on my monitor, so this much works. In some cases, >>> the guest essentially stops there; xm list show's about 2sec CPU >>> usage and nothing ever happens after that point. In other cases, >>> the guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot >>> all the way, but in very low res/color mode, and cannot properly >>> initialize the video device. Once or twice, it actually did >>> recognize the device and had a reasonable default color/resolution >>> combination. In all cases where the guest actually boots, the >>> system eventually freezes. In some cases, I get endless streams of >>> iommu page faults. >>> >>> I have 8GB ram installed. In all cases I've limited dom0 memory to >>> 2GB. In all cases, my guest has been assigned 2GB of memory. >>> >>> I have a dual core e6600. I've tried allowing dom0 to use both >>> cores, offlining one core (using xend dom0_vcpu setting) after boot, >>> and restricting dom0 to only one core using the dom0_max_vcpus xen >>> hypervisor parameter. In all of these cases I've tried both one and >>> two vcpus for the guest. My success with VGA passthrough seems >>> somewhat random and no combination of cpu assignment seems to have >>> any effect. >>> >>> I have not tried with the 2.6.18-xen kernel as I haven't gotten it >>> to boot on my hardware; it can never find my volume group, even if I >>> create a initrd with all of the required modules, or build those >>> drivers into the kernel. I have not spent more than maybe a half an >>> hour on this problem; I suspect it may have something to do with the >>> version of mkinitrd I'm using (from Fedora 9 x64). >>> >>> If anyone else has any insight or similar experience I'd also love >>> to hear it. >>> >>> Doug Magee >>> djmagee@mageenet.com >>> >>> -----Original Message----- >>> From: xen-devel-bounces@lists.xensource.com >>> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >>> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >>> To: enming.teo@asiasoftsea.net >>> Cc: xen-devel@lists.xensource.com >>> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >>> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel >>> Desktop BoardDQ45CB >>> >>> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >>> suggested by Weidong. >>> >>> >>> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>>> Dear All, >>>> >>>> I have managed to do PCI-e VGA passthrough with the open source Xen >>>> but the work is still in progress because although Windows XP guest >>>> can see the REAL PCI-e x16 graphics card instead of an emulated >>>> graphics driver, it cannot be initialized yet. >>>> >>>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>>> shared their knowledge with me along the way. >>>> >>>> System Configuration: >>>> >>>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>>> Onboard Intel GMA 4500 Graphics (IGD) >>>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>>> >>>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>>> Primary Video Adapter in BIOS: IGD >>>> >>>> Please see the screenshots and my blog at the link here: >>>> >>>> >> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 >> -gs-pci-express-x16.html >>>> >>>> >>> >>> >>> >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >> >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel >> >> No virus found in this incoming message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 >> >> No virus found in this outgoing message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 > > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-27 3:18 ` Teo En Ming (Zhang Enming) @ 2009-08-27 6:03 ` Teo En Ming (Zhang Enming) 2009-08-27 16:18 ` Teo En Ming (Zhang Enming) 0 siblings, 1 reply; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-27 6:03 UTC (permalink / raw) To: enming.teo, weidong.han, djmagee; +Cc: xen-devel I think I cannot use NiBiTor already as I don't have any native Windows operating system running on my computer. Instead, I should create a CD-ROM boot disc that contains nvflash.exe and also loads USB drivers. I will use nvflash.exe to extract out and save VGA BIOS to USB sticks. NVIDIA GeForce BIOS Backup Guide: http://www.mvktech.net/content/view/14/37/ Motherboard Flash Boot CD from Linux Mini HOWTO: http://www.nenie.org/misc/flashbootcd.html DOS USB Drivers: http://www.bootdisk.com/usb.htm Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 11:18 AM To: weidong.han@intel.com; djmagee@mageenet.net Cc: xen-devel@lists.xensource.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel Desktop Board DQ45CB Dear Weidong, I will go for option 2. > don't initialize emulated VGA in qemu, I I think you have seen the code. I have not seen the source code. Though I have learnt simple structured C before in Singapore Polytechnic some 11 years ago, I am not a programmer by profession. So I would still need some instructions on how I can prevent emulated VGA bios from loading and getting the vga bios of the real display card to load. Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, Weidong Sent: Thursday, August 27, 2009 10:01 AM To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine with IntelDesktop Board DQ45CB Teo En Ming (Zhang Enming) wrote: > Hi Weidong, > > Currently I am using the entire Xen 3.5-unstable branch and not using > the posted pci-stub patch changeset 19893 to patch Xen 3.4.1-testing. > I am using Xen 3.5-unstable in conjunction with pvops dom 0 kernel > 2.6.31-rc6. > > Here are my questions: > > 1) How do I extract out the VGA BIOS/firmware from the > firmware/EEPROM chip on the Geforce 8400 GS VGA card? If it is not > possible to extract, where can I download the firmware? NVIDIA does > not provide display card firmware/BIOS files for public download on > its website. There are 3 ways to load VGA bios to guest: 1. copy it from host 0xc0000, and map it to guest 0xc0000. This works for primary gfx card, XCI does like this. But some VGA bios re-execution doesn't work well. 2. dump VGA bios to a file first, then load it to guest: you can use NiBiTor to dump VGA bios of Nvidia gfx cards to a file, then load it into guest roms like what emulated VGA does in hvmloader. This works for all gfx cards, but it involves manual steps. 3. load VGA bios from expansion rom: Intel iGFX doesn't have expansion rom > > 2) How do I get the Geforce 8400 GS VGA BIOS to load instead of the > emulated VGA BIOS? Which configuration files/source codes do I have to > configure/patch? don't initialize emulated VGA in qemu, I I think you have seen the code. > > 3) What are physical MMIO, virtual MMIO Bars, dsdt.asl and 1:1 map? I > am sorry but I am not a graphics card hardware engineer so I do not > understand those terms. I would appreciate it if you could write out > some steps that I could carry out or point me to already available > documentation. When you use command 'lspci -v', you may see below messages under each device: ... Memory at d0000000 (32-bit, prefetchable) [size=256M] Memory at ff900000 (32-bit, non-prefetchable) [size=1M] ... They are MMIO BARs. 1:1 map means make these MMIO BARs in guest are equal to the values in host. > > 4) I don't mind using experimental code. Please post it on the > xen-devel mailing list so that I could try it out. we will post in near future. pls wait for a while. Regards, Weidong > > Thank you very much. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > > -----Original Message----- > From: Han, Weidong [mailto:weidong.han@intel.com] > Sent: Wednesday, August 26, 2009 6:56 PM > To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel > Desktop Board DQ45CB > > Teo En Ming (Zhang Enming) wrote: >> Hi Weidong, >> >> Could you share with us the hack codes for making Geforce 8400 GS >> work and also how to let the Windows HVM guest boot up using the real >> BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? >> > > What patch are you using now? Using real VGA bios of gfx card to > replace emulatd VGA bios is the prerequisite of gfx passthrough. You > can find it in posted gfx passthrough patches or XCI. For hack of > making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and > make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of > the card. Currently our code is experimental, we will send out in > mailing list after cleanup and more tests. > > Regards, > Weidong > >> Thank you. >> >> Regards, >> >> Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) >> BEng(Hons)(Mechanical Engineering) >> Technical Support Engineer >> Information Technology Department >> Asiasoft Online Pte Ltd >> Tampines Central 1 #04-01 Tampines Plaza >> Singapore 529541 >> Republic of Singapore >> Mobile: +65-9648-9798 >> MSN: teoenming@hotmail.com >> Alma Maters: Singapore Polytechnic, National University of Singapore >> >> -----Original Message----- >> From: xen-devel-bounces@lists.xensource.com >> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, >> Weidong Sent: Wednesday, August 26, 2009 4:27 PM >> To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' >> Cc: 'xen-devel@lists.xensource.com' >> Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 >> VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with >> IntelDesktop BoardDQ45CB >> >> I suppose you just use the patch posted in mailing list before. >> nVidia Geforce 8400 passthrough needs extra hacks. We can make it >> work in our experiments with 1:1 map of its MMIO BARs. >> >> We are working on gfx passthrough on latest xen-unstable. Firstly, we >> want to cook a simple patch including generic changes to support >> passthrough of virtualization friendly gfx cards, such as Nvidia >> FX3800. This patch is basically done. Then, we will add some hacks >> for more gfx cards passthrough, such as iGFX and some Nvidia and ATI >> cards. >> >> Regards, >> Weidong >> >> djmagee@mageenet.net wrote: >>> As I've pointed out on this list before, there are not enough PCIe >>> lanes on the DQ45CB to drive both the internal graphics adapter and >>> the add-on adapter at the same time. I believe the onboard one may >>> be able to operate in some sort of VGA only mode when there is a >>> card installed and used as the primary adapter. >>> >>> I have a similar setup, using the same motherboard, and an ATI 4770. >>> I used the VGA passthrough patches from an earlier posting (I saw >>> you followed the same set of instructions), with limited success. >>> I've been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made >>> modifications to the dom0 portion of the patches so they would apply >>> to my xenified 2.6.29 kernel. These patches include code that will >>> copy the VGA bios to the guest bios instead of using the emulated >>> vga bios. I've had very little success, however. I have only tried >>> passing through the ATI adapter. In all instances, the guest bios >>> messages appear on my monitor, so this much works. In some cases, >>> the guest essentially stops there; xm list show's about 2sec CPU >>> usage and nothing ever happens after that point. In other cases, >>> the guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot >>> all the way, but in very low res/color mode, and cannot properly >>> initialize the video device. Once or twice, it actually did >>> recognize the device and had a reasonable default color/resolution >>> combination. In all cases where the guest actually boots, the >>> system eventually freezes. In some cases, I get endless streams of >>> iommu page faults. >>> >>> I have 8GB ram installed. In all cases I've limited dom0 memory to >>> 2GB. In all cases, my guest has been assigned 2GB of memory. >>> >>> I have a dual core e6600. I've tried allowing dom0 to use both >>> cores, offlining one core (using xend dom0_vcpu setting) after boot, >>> and restricting dom0 to only one core using the dom0_max_vcpus xen >>> hypervisor parameter. In all of these cases I've tried both one and >>> two vcpus for the guest. My success with VGA passthrough seems >>> somewhat random and no combination of cpu assignment seems to have >>> any effect. >>> >>> I have not tried with the 2.6.18-xen kernel as I haven't gotten it >>> to boot on my hardware; it can never find my volume group, even if I >>> create a initrd with all of the required modules, or build those >>> drivers into the kernel. I have not spent more than maybe a half an >>> hour on this problem; I suspect it may have something to do with the >>> version of mkinitrd I'm using (from Fedora 9 x64). >>> >>> If anyone else has any insight or similar experience I'd also love >>> to hear it. >>> >>> Doug Magee >>> djmagee@mageenet.com >>> >>> -----Original Message----- >>> From: xen-devel-bounces@lists.xensource.com >>> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >>> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >>> To: enming.teo@asiasoftsea.net >>> Cc: xen-devel@lists.xensource.com >>> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >>> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel >>> Desktop BoardDQ45CB >>> >>> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >>> suggested by Weidong. >>> >>> >>> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>>> Dear All, >>>> >>>> I have managed to do PCI-e VGA passthrough with the open source Xen >>>> but the work is still in progress because although Windows XP guest >>>> can see the REAL PCI-e x16 graphics card instead of an emulated >>>> graphics driver, it cannot be initialized yet. >>>> >>>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>>> shared their knowledge with me along the way. >>>> >>>> System Configuration: >>>> >>>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>>> Onboard Intel GMA 4500 Graphics (IGD) >>>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>>> >>>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>>> Primary Video Adapter in BIOS: IGD >>>> >>>> Please see the screenshots and my blog at the link here: >>>> >>>> >> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 >> -gs-pci-express-x16.html >>>> >>>> >>> >>> >>> >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >> >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel >> >> No virus found in this incoming message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 >> >> No virus found in this outgoing message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 > > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-27 6:03 ` Teo En Ming (Zhang Enming) @ 2009-08-27 16:18 ` Teo En Ming (Zhang Enming) 2009-08-27 18:28 ` djmagee 2009-08-28 3:30 ` Han, Weidong 0 siblings, 2 replies; 25+ messages in thread From: Teo En Ming (Zhang Enming) @ 2009-08-27 16:18 UTC (permalink / raw) To: enming.teo, weidong.han, djmagee; +Cc: xen-devel Dear Weidong, I have read through the technical papers on VGA passthrough presented by the University of Amsterdam and the University of Michigan and thus gained a better understanding of VGA passthrough in Xen. For my preferred setup, I would not be able to adopt the guidance provided by the University of Amsterdam. I should adopt the guidance provided by the University of Michigan. I want dom0 to have access to Intel GMA 4500 onboard graphics while the Windows XP Home HVM domU to have access to PCI Express x16 graphics. I have taken the first step in examining the source code for disabling the Cirrus emulated vga card bootup in Xen 3.5-unstable. According to Beng Heng's UMICH technical paper, he referenced source file tools/ioemu-remote/hw/pc.c for disabling emulated vga device. I went to the xen-unstable mercurial reponsitory http://xenbits.xensource.com/xen-unstable.hg and found tools/ioemu/hw/pc.c Now, my objective is to disable the emulated/virtual vga card and load the bios of my nvidia geforce 8400 gs pcie x16 graphics card. Source code file http://xenbits.xensource.com/xen-unstable.hg?file/18b41609a980/tools/ioemu/h w/pc.c ========================================================================= Line 862 says that if variable cirrus_vga_enabled returns true (having a value of 1), the cirrus vga bios VGABIOS_CIRRUS_FILENAME will be loaded. Otherwise, an alternative vga bios file VGABIOS_FILENAME will be loaded. On line 30 VGABIOS_FILENAME is defined. On line 31 VGABIOS_CIRRUS_FILENAME is defined. On line 24 a header file vl.h is referenced. Source code file http://xenbits.xensource.com/xen-unstable.hg?file/7eefe6399bcd/tools/ioemu/v l.h =========================================================================== On line 177 the integer variable cirrus_vga_enabled is declared but I could not find any line where this variable is initialized to a value. But I found file vl.c. Source code file http://xenbits.xensource.com/xen-unstable.hg?file/dade7f0bdc8d/tools/ioemu/v l.c ============================================================================ On line 168 integer variable cirrus_vga_enabled is initialized to a value of 1. However, I want to disable cirrus vga and load my nvidia geforce 8400 gs vga bios. So I should change line 168 to "int cirrus_vga_enabled = 0;" in my case. Hence line 865 in file pc.c would be executed. After extracting the vga bios of my nvidia geforce 8400 gs pcie x16 graphics card using the nvflash.exe utility and saving the firmware file as nv-gf-8400-gs.bin, I should place this bios image file in the xen-unstable sources directory on my PC. It should be in the same directory where files vgabios.bin and vgabios-cirrus.bin are found. Then I should change line 30 in file pc.c from '#define VGABIOS_FILENAME "vgabios.bin"' to '#define VGABIOS_FILENAME nv-gf-8400-gs.bin' and then recompile and install the source code. Besides disabling the emulated cirrus vga card, may I know what other things do I need to take care of? Does the pre-existing xen 3.5-unstable source code already can take care of re-executing the geforce 8400 gs vga bios, passing through the framebuffer and vga I/O ports, and intercepting the values in the host bridge such that I do not have to hack the other relevant source code files? Thank you very much. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com Alma Maters: Singapore Polytechnic, National University of Singapore -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 2:03 PM To: enming.teo@asiasoftsea.net; weidong.han@intel.com; djmagee@mageenet.net Cc: xen-devel@lists.xensource.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel Desktop Board DQ45CB I think I cannot use NiBiTor already as I don't have any native Windows operating system running on my computer. Instead, I should create a CD-ROM boot disc that contains nvflash.exe and also loads USB drivers. I will use nvflash.exe to extract out and save VGA BIOS to USB sticks. NVIDIA GeForce BIOS Backup Guide: http://www.mvktech.net/content/view/14/37/ Motherboard Flash Boot CD from Linux Mini HOWTO: http://www.nenie.org/misc/flashbootcd.html DOS USB Drivers: http://www.bootdisk.com/usb.htm Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 11:18 AM To: weidong.han@intel.com; djmagee@mageenet.net Cc: xen-devel@lists.xensource.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel Desktop Board DQ45CB Dear Weidong, I will go for option 2. > don't initialize emulated VGA in qemu, I I think you have seen the code. I have not seen the source code. Though I have learnt simple structured C before in Singapore Polytechnic some 11 years ago, I am not a programmer by profession. So I would still need some instructions on how I can prevent emulated VGA bios from loading and getting the vga bios of the real display card to load. Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, Weidong Sent: Thursday, August 27, 2009 10:01 AM To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine with IntelDesktop Board DQ45CB Teo En Ming (Zhang Enming) wrote: > Hi Weidong, > > Currently I am using the entire Xen 3.5-unstable branch and not using > the posted pci-stub patch changeset 19893 to patch Xen 3.4.1-testing. > I am using Xen 3.5-unstable in conjunction with pvops dom 0 kernel > 2.6.31-rc6. > > Here are my questions: > > 1) How do I extract out the VGA BIOS/firmware from the > firmware/EEPROM chip on the Geforce 8400 GS VGA card? If it is not > possible to extract, where can I download the firmware? NVIDIA does > not provide display card firmware/BIOS files for public download on > its website. There are 3 ways to load VGA bios to guest: 1. copy it from host 0xc0000, and map it to guest 0xc0000. This works for primary gfx card, XCI does like this. But some VGA bios re-execution doesn't work well. 2. dump VGA bios to a file first, then load it to guest: you can use NiBiTor to dump VGA bios of Nvidia gfx cards to a file, then load it into guest roms like what emulated VGA does in hvmloader. This works for all gfx cards, but it involves manual steps. 3. load VGA bios from expansion rom: Intel iGFX doesn't have expansion rom > > 2) How do I get the Geforce 8400 GS VGA BIOS to load instead of the > emulated VGA BIOS? Which configuration files/source codes do I have to > configure/patch? don't initialize emulated VGA in qemu, I I think you have seen the code. > > 3) What are physical MMIO, virtual MMIO Bars, dsdt.asl and 1:1 map? I > am sorry but I am not a graphics card hardware engineer so I do not > understand those terms. I would appreciate it if you could write out > some steps that I could carry out or point me to already available > documentation. When you use command 'lspci -v', you may see below messages under each device: ... Memory at d0000000 (32-bit, prefetchable) [size=256M] Memory at ff900000 (32-bit, non-prefetchable) [size=1M] ... They are MMIO BARs. 1:1 map means make these MMIO BARs in guest are equal to the values in host. > > 4) I don't mind using experimental code. Please post it on the > xen-devel mailing list so that I could try it out. we will post in near future. pls wait for a while. Regards, Weidong > > Thank you very much. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > > -----Original Message----- > From: Han, Weidong [mailto:weidong.han@intel.com] > Sent: Wednesday, August 26, 2009 6:56 PM > To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel > Desktop Board DQ45CB > > Teo En Ming (Zhang Enming) wrote: >> Hi Weidong, >> >> Could you share with us the hack codes for making Geforce 8400 GS >> work and also how to let the Windows HVM guest boot up using the real >> BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? >> > > What patch are you using now? Using real VGA bios of gfx card to > replace emulatd VGA bios is the prerequisite of gfx passthrough. You > can find it in posted gfx passthrough patches or XCI. For hack of > making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and > make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of > the card. Currently our code is experimental, we will send out in > mailing list after cleanup and more tests. > > Regards, > Weidong > >> Thank you. >> >> Regards, >> >> Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) >> BEng(Hons)(Mechanical Engineering) >> Technical Support Engineer >> Information Technology Department >> Asiasoft Online Pte Ltd >> Tampines Central 1 #04-01 Tampines Plaza >> Singapore 529541 >> Republic of Singapore >> Mobile: +65-9648-9798 >> MSN: teoenming@hotmail.com >> Alma Maters: Singapore Polytechnic, National University of Singapore >> >> -----Original Message----- >> From: xen-devel-bounces@lists.xensource.com >> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, >> Weidong Sent: Wednesday, August 26, 2009 4:27 PM >> To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' >> Cc: 'xen-devel@lists.xensource.com' >> Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 >> VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with >> IntelDesktop BoardDQ45CB >> >> I suppose you just use the patch posted in mailing list before. >> nVidia Geforce 8400 passthrough needs extra hacks. We can make it >> work in our experiments with 1:1 map of its MMIO BARs. >> >> We are working on gfx passthrough on latest xen-unstable. Firstly, we >> want to cook a simple patch including generic changes to support >> passthrough of virtualization friendly gfx cards, such as Nvidia >> FX3800. This patch is basically done. Then, we will add some hacks >> for more gfx cards passthrough, such as iGFX and some Nvidia and ATI >> cards. >> >> Regards, >> Weidong >> >> djmagee@mageenet.net wrote: >>> As I've pointed out on this list before, there are not enough PCIe >>> lanes on the DQ45CB to drive both the internal graphics adapter and >>> the add-on adapter at the same time. I believe the onboard one may >>> be able to operate in some sort of VGA only mode when there is a >>> card installed and used as the primary adapter. >>> >>> I have a similar setup, using the same motherboard, and an ATI 4770. >>> I used the VGA passthrough patches from an earlier posting (I saw >>> you followed the same set of instructions), with limited success. >>> I've been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made >>> modifications to the dom0 portion of the patches so they would apply >>> to my xenified 2.6.29 kernel. These patches include code that will >>> copy the VGA bios to the guest bios instead of using the emulated >>> vga bios. I've had very little success, however. I have only tried >>> passing through the ATI adapter. In all instances, the guest bios >>> messages appear on my monitor, so this much works. In some cases, >>> the guest essentially stops there; xm list show's about 2sec CPU >>> usage and nothing ever happens after that point. In other cases, >>> the guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot >>> all the way, but in very low res/color mode, and cannot properly >>> initialize the video device. Once or twice, it actually did >>> recognize the device and had a reasonable default color/resolution >>> combination. In all cases where the guest actually boots, the >>> system eventually freezes. In some cases, I get endless streams of >>> iommu page faults. >>> >>> I have 8GB ram installed. In all cases I've limited dom0 memory to >>> 2GB. In all cases, my guest has been assigned 2GB of memory. >>> >>> I have a dual core e6600. I've tried allowing dom0 to use both >>> cores, offlining one core (using xend dom0_vcpu setting) after boot, >>> and restricting dom0 to only one core using the dom0_max_vcpus xen >>> hypervisor parameter. In all of these cases I've tried both one and >>> two vcpus for the guest. My success with VGA passthrough seems >>> somewhat random and no combination of cpu assignment seems to have >>> any effect. >>> >>> I have not tried with the 2.6.18-xen kernel as I haven't gotten it >>> to boot on my hardware; it can never find my volume group, even if I >>> create a initrd with all of the required modules, or build those >>> drivers into the kernel. I have not spent more than maybe a half an >>> hour on this problem; I suspect it may have something to do with the >>> version of mkinitrd I'm using (from Fedora 9 x64). >>> >>> If anyone else has any insight or similar experience I'd also love >>> to hear it. >>> >>> Doug Magee >>> djmagee@mageenet.com >>> >>> -----Original Message----- >>> From: xen-devel-bounces@lists.xensource.com >>> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >>> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >>> To: enming.teo@asiasoftsea.net >>> Cc: xen-devel@lists.xensource.com >>> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >>> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel >>> Desktop BoardDQ45CB >>> >>> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >>> suggested by Weidong. >>> >>> >>> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>>> Dear All, >>>> >>>> I have managed to do PCI-e VGA passthrough with the open source Xen >>>> but the work is still in progress because although Windows XP guest >>>> can see the REAL PCI-e x16 graphics card instead of an emulated >>>> graphics driver, it cannot be initialized yet. >>>> >>>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>>> shared their knowledge with me along the way. >>>> >>>> System Configuration: >>>> >>>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>>> Onboard Intel GMA 4500 Graphics (IGD) >>>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>>> >>>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>>> Primary Video Adapter in BIOS: IGD >>>> >>>> Please see the screenshots and my blog at the link here: >>>> >>>> >> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 >> -gs-pci-express-x16.html >>>> >>>> >>> >>> >>> >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >> >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel >> >> No virus found in this incoming message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 >> >> No virus found in this outgoing message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 > > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-27 16:18 ` Teo En Ming (Zhang Enming) @ 2009-08-27 18:28 ` djmagee 2009-08-28 3:30 ` Han, Weidong 1 sibling, 0 replies; 25+ messages in thread From: djmagee @ 2009-08-27 18:28 UTC (permalink / raw) To: enming.teo, weidong.han; +Cc: xen-devel Tao, The DQ45CB does not support the configuration you are trying to use; it does not have enough PCIe lanes. See the following reference: http://download.intel.com/support/motherboards/desktop/dq45cb/sb/e42785001us.pdf In the block diagram on page 13, you'll see that motherboard uses 8 PCIe lanes to drive the two DVI outputs on the motherboard. Those same lanes are used to drive an add-on PCIex16 graphics card. They can't be used for both purposes at the same time. On page 18, this is spelled out. Under the "Intel Q45 Graphics Subsystem" heading it reads, "The Intel Q45 Express chipset contains two separate, mutually exclusive graphics options. Either the Intel Graphics Media Accelerator 4500 (Intel GMA 4500) graphics controller (contained within the 82Q45 GMCH) is used, or a PCI Express x16 add-in card can be used. When a PCI Express x16 add-in card is installed, the Intel GMA 4500 graphics controller is disabled." I believe at the very best, you can realize very basic, analog only output using a DVI-VGA adapter on the primary output, but I'm not even sure about that. Doug Magee -----Original Message----- From: Teo En Ming (Zhang Enming) [mailto:enming.teo@asiasoftsea.net] Sent: Thursday, August 27, 2009 12:18 PM To: enming.teo@asiasoftsea.net; weidong.han@intel.com; djmagee@mageenet.net Cc: xen-devel@lists.xensource.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Dear Weidong, I have read through the technical papers on VGA passthrough presented by the University of Amsterdam and the University of Michigan and thus gained a better understanding of VGA passthrough in Xen. For my preferred setup, I would not be able to adopt the guidance provided by the University of Amsterdam. I should adopt the guidance provided by the University of Michigan. I want dom0 to have access to Intel GMA 4500 onboard graphics while the Windows XP Home HVM domU to have access to PCI Express x16 graphics. I have taken the first step in examining the source code for disabling the Cirrus emulated vga card bootup in Xen 3.5-unstable. According to Beng Heng's UMICH technical paper, he referenced source file tools/ioemu-remote/hw/pc.c for disabling emulated vga device. I went to the xen-unstable mercurial reponsitory http://xenbits.xensource.com/xen-unstable.hg and found tools/ioemu/hw/pc.c Now, my objective is to disable the emulated/virtual vga card and load the bios of my nvidia geforce 8400 gs pcie x16 graphics card. Source code file http://xenbits.xensource.com/xen-unstable.hg?file/18b41609a980/tools/ioemu/h w/pc.c ========================================================================= Line 862 says that if variable cirrus_vga_enabled returns true (having a value of 1), the cirrus vga bios VGABIOS_CIRRUS_FILENAME will be loaded. Otherwise, an alternative vga bios file VGABIOS_FILENAME will be loaded. On line 30 VGABIOS_FILENAME is defined. On line 31 VGABIOS_CIRRUS_FILENAME is defined. On line 24 a header file vl.h is referenced. Source code file http://xenbits.xensource.com/xen-unstable.hg?file/7eefe6399bcd/tools/ioemu/v l.h =========================================================================== On line 177 the integer variable cirrus_vga_enabled is declared but I could not find any line where this variable is initialized to a value. But I found file vl.c. Source code file http://xenbits.xensource.com/xen-unstable.hg?file/dade7f0bdc8d/tools/ioemu/v l.c ============================================================================ On line 168 integer variable cirrus_vga_enabled is initialized to a value of 1. However, I want to disable cirrus vga and load my nvidia geforce 8400 gs vga bios. So I should change line 168 to "int cirrus_vga_enabled = 0;" in my case. Hence line 865 in file pc.c would be executed. After extracting the vga bios of my nvidia geforce 8400 gs pcie x16 graphics card using the nvflash.exe utility and saving the firmware file as nv-gf-8400-gs.bin, I should place this bios image file in the xen-unstable sources directory on my PC. It should be in the same directory where files vgabios.bin and vgabios-cirrus.bin are found. Then I should change line 30 in file pc.c from '#define VGABIOS_FILENAME "vgabios.bin"' to '#define VGABIOS_FILENAME nv-gf-8400-gs.bin' and then recompile and install the source code. Besides disabling the emulated cirrus vga card, may I know what other things do I need to take care of? Does the pre-existing xen 3.5-unstable source code already can take care of re-executing the geforce 8400 gs vga bios, passing through the framebuffer and vga I/O ports, and intercepting the values in the host bridge such that I do not have to hack the other relevant source code files? Thank you very much. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com Alma Maters: Singapore Polytechnic, National University of Singapore -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 2:03 PM To: enming.teo@asiasoftsea.net; weidong.han@intel.com; djmagee@mageenet.net Cc: xen-devel@lists.xensource.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel Desktop Board DQ45CB I think I cannot use NiBiTor already as I don't have any native Windows operating system running on my computer. Instead, I should create a CD-ROM boot disc that contains nvflash.exe and also loads USB drivers. I will use nvflash.exe to extract out and save VGA BIOS to USB sticks. NVIDIA GeForce BIOS Backup Guide: http://www.mvktech.net/content/view/14/37/ Motherboard Flash Boot CD from Linux Mini HOWTO: http://www.nenie.org/misc/flashbootcd.html DOS USB Drivers: http://www.bootdisk.com/usb.htm Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En Ming (Zhang Enming) Sent: Thursday, August 27, 2009 11:18 AM To: weidong.han@intel.com; djmagee@mageenet.net Cc: xen-devel@lists.xensource.com Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel Desktop Board DQ45CB Dear Weidong, I will go for option 2. > don't initialize emulated VGA in qemu, I I think you have seen the code. I have not seen the source code. Though I have learnt simple structured C before in Singapore Polytechnic some 11 years ago, I am not a programmer by profession. So I would still need some instructions on how I can prevent emulated VGA bios from loading and getting the vga bios of the real display card to load. Thank you. Regards, Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) BEng(Hons)(Mechanical Engineering) Technical Support Engineer Information Technology Department Asiasoft Online Pte Ltd Tampines Central 1 #04-01 Tampines Plaza Singapore 529541 Republic of Singapore Mobile: +65-9648-9798 MSN: teoenming@hotmail.com -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, Weidong Sent: Thursday, August 27, 2009 10:01 AM To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA PassThrough to Windows XP Home 32-bit HVM Virtual Machine with IntelDesktop Board DQ45CB Teo En Ming (Zhang Enming) wrote: > Hi Weidong, > > Currently I am using the entire Xen 3.5-unstable branch and not using > the posted pci-stub patch changeset 19893 to patch Xen 3.4.1-testing. > I am using Xen 3.5-unstable in conjunction with pvops dom 0 kernel > 2.6.31-rc6. > > Here are my questions: > > 1) How do I extract out the VGA BIOS/firmware from the > firmware/EEPROM chip on the Geforce 8400 GS VGA card? If it is not > possible to extract, where can I download the firmware? NVIDIA does > not provide display card firmware/BIOS files for public download on > its website. There are 3 ways to load VGA bios to guest: 1. copy it from host 0xc0000, and map it to guest 0xc0000. This works for primary gfx card, XCI does like this. But some VGA bios re-execution doesn't work well. 2. dump VGA bios to a file first, then load it to guest: you can use NiBiTor to dump VGA bios of Nvidia gfx cards to a file, then load it into guest roms like what emulated VGA does in hvmloader. This works for all gfx cards, but it involves manual steps. 3. load VGA bios from expansion rom: Intel iGFX doesn't have expansion rom > > 2) How do I get the Geforce 8400 GS VGA BIOS to load instead of the > emulated VGA BIOS? Which configuration files/source codes do I have to > configure/patch? don't initialize emulated VGA in qemu, I I think you have seen the code. > > 3) What are physical MMIO, virtual MMIO Bars, dsdt.asl and 1:1 map? I > am sorry but I am not a graphics card hardware engineer so I do not > understand those terms. I would appreciate it if you could write out > some steps that I could carry out or point me to already available > documentation. When you use command 'lspci -v', you may see below messages under each device: ... Memory at d0000000 (32-bit, prefetchable) [size=256M] Memory at ff900000 (32-bit, non-prefetchable) [size=1M] ... They are MMIO BARs. 1:1 map means make these MMIO BARs in guest are equal to the values in host. > > 4) I don't mind using experimental code. Please post it on the > xen-devel mailing list so that I could try it out. we will post in near future. pls wait for a while. Regards, Weidong > > Thank you very much. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > > -----Original Message----- > From: Han, Weidong [mailto:weidong.han@intel.com] > Sent: Wednesday, August 26, 2009 6:56 PM > To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel > Desktop Board DQ45CB > > Teo En Ming (Zhang Enming) wrote: >> Hi Weidong, >> >> Could you share with us the hack codes for making Geforce 8400 GS >> work and also how to let the Windows HVM guest boot up using the real >> BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? >> > > What patch are you using now? Using real VGA bios of gfx card to > replace emulatd VGA bios is the prerequisite of gfx passthrough. You > can find it in posted gfx passthrough patches or XCI. For hack of > making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and > make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of > the card. Currently our code is experimental, we will send out in > mailing list after cleanup and more tests. > > Regards, > Weidong > >> Thank you. >> >> Regards, >> >> Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) >> BEng(Hons)(Mechanical Engineering) >> Technical Support Engineer >> Information Technology Department >> Asiasoft Online Pte Ltd >> Tampines Central 1 #04-01 Tampines Plaza >> Singapore 529541 >> Republic of Singapore >> Mobile: +65-9648-9798 >> MSN: teoenming@hotmail.com >> Alma Maters: Singapore Polytechnic, National University of Singapore >> >> -----Original Message----- >> From: xen-devel-bounces@lists.xensource.com >> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, >> Weidong Sent: Wednesday, August 26, 2009 4:27 PM >> To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' >> Cc: 'xen-devel@lists.xensource.com' >> Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 >> VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with >> IntelDesktop BoardDQ45CB >> >> I suppose you just use the patch posted in mailing list before. >> nVidia Geforce 8400 passthrough needs extra hacks. We can make it >> work in our experiments with 1:1 map of its MMIO BARs. >> >> We are working on gfx passthrough on latest xen-unstable. Firstly, we >> want to cook a simple patch including generic changes to support >> passthrough of virtualization friendly gfx cards, such as Nvidia >> FX3800. This patch is basically done. Then, we will add some hacks >> for more gfx cards passthrough, such as iGFX and some Nvidia and ATI >> cards. >> >> Regards, >> Weidong >> >> djmagee@mageenet.net wrote: >>> As I've pointed out on this list before, there are not enough PCIe >>> lanes on the DQ45CB to drive both the internal graphics adapter and >>> the add-on adapter at the same time. I believe the onboard one may >>> be able to operate in some sort of VGA only mode when there is a >>> card installed and used as the primary adapter. >>> >>> I have a similar setup, using the same motherboard, and an ATI 4770. >>> I used the VGA passthrough patches from an earlier posting (I saw >>> you followed the same set of instructions), with limited success. >>> I've been using 3.4-testing, and 'xenified' 2.6.29.6 kernel. I made >>> modifications to the dom0 portion of the patches so they would apply >>> to my xenified 2.6.29 kernel. These patches include code that will >>> copy the VGA bios to the guest bios instead of using the emulated >>> vga bios. I've had very little success, however. I have only tried >>> passing through the ATI adapter. In all instances, the guest bios >>> messages appear on my monitor, so this much works. In some cases, >>> the guest essentially stops there; xm list show's about 2sec CPU >>> usage and nothing ever happens after that point. In other cases, >>> the guest (win xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot >>> all the way, but in very low res/color mode, and cannot properly >>> initialize the video device. Once or twice, it actually did >>> recognize the device and had a reasonable default color/resolution >>> combination. In all cases where the guest actually boots, the >>> system eventually freezes. In some cases, I get endless streams of >>> iommu page faults. >>> >>> I have 8GB ram installed. In all cases I've limited dom0 memory to >>> 2GB. In all cases, my guest has been assigned 2GB of memory. >>> >>> I have a dual core e6600. I've tried allowing dom0 to use both >>> cores, offlining one core (using xend dom0_vcpu setting) after boot, >>> and restricting dom0 to only one core using the dom0_max_vcpus xen >>> hypervisor parameter. In all of these cases I've tried both one and >>> two vcpus for the guest. My success with VGA passthrough seems >>> somewhat random and no combination of cpu assignment seems to have >>> any effect. >>> >>> I have not tried with the 2.6.18-xen kernel as I haven't gotten it >>> to boot on my hardware; it can never find my volume group, even if I >>> create a initrd with all of the required modules, or build those >>> drivers into the kernel. I have not spent more than maybe a half an >>> hour on this problem; I suspect it may have something to do with the >>> version of mkinitrd I'm using (from Fedora 9 x64). >>> >>> If anyone else has any insight or similar experience I'd also love >>> to hear it. >>> >>> Doug Magee >>> djmagee@mageenet.com >>> >>> -----Original Message----- >>> From: xen-devel-bounces@lists.xensource.com >>> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >>> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >>> To: enming.teo@asiasoftsea.net >>> Cc: xen-devel@lists.xensource.com >>> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >>> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel >>> Desktop BoardDQ45CB >>> >>> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >>> suggested by Weidong. >>> >>> >>> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>>> Dear All, >>>> >>>> I have managed to do PCI-e VGA passthrough with the open source Xen >>>> but the work is still in progress because although Windows XP guest >>>> can see the REAL PCI-e x16 graphics card instead of an emulated >>>> graphics driver, it cannot be initialized yet. >>>> >>>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>>> shared their knowledge with me along the way. >>>> >>>> System Configuration: >>>> >>>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>>> Onboard Intel GMA 4500 Graphics (IGD) >>>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>>> >>>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>>> Primary Video Adapter in BIOS: IGD >>>> >>>> Please see the screenshots and my blog at the link here: >>>> >>>> >> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 >> -gs-pci-express-x16.html >>>> >>>> >>> >>> >>> >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >> >> >> _______________________________________________ >> Xen-devel mailing list >> Xen-devel@lists.xensource.com >> http://lists.xensource.com/xen-devel >> >> No virus found in this incoming message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 >> >> No virus found in this outgoing message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 > > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: > 08/25/09 18:07:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/25/09 18:07:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: 08/26/09 12:16:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
* RE: nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB 2009-08-27 16:18 ` Teo En Ming (Zhang Enming) 2009-08-27 18:28 ` djmagee @ 2009-08-28 3:30 ` Han, Weidong 1 sibling, 0 replies; 25+ messages in thread From: Han, Weidong @ 2009-08-28 3:30 UTC (permalink / raw) To: 'enming.teo@asiasoftsea.net', 'djmagee@mageenet.net' Cc: 'xen-devel@lists.xensource.com' It's easy to disable emulated VGA, you just add a check in hw/pc.c: + if (gfx_passthru == GFX_NO_PASSTHRU) { + if (cirrus_vga_enabled) { ...... That means: if gfx passthroughed, disable emulated VGA. For another question about loading VGA bios, given you have saved the VGA BIOS of your card to vgabios-pt.bin, you should first put it at xen-unstable.hg/tools/firmware/vgabios/, and also you needs following changes, at last, you still need to pass BDF of gfx to rombios per firmware spec. diff -r 494be76c1ad9 tools/firmware/hvmloader/Makefile --- a/tools/firmware/hvmloader/Makefile Thu Aug 27 16:54:33 2009 +0800 +++ b/tools/firmware/hvmloader/Makefile Thu Aug 27 17:22:01 2009 +0800 @@ -50,6 +50,7 @@ roms.h: ../rombios/BIOS-bochs-latest ../ roms.h: ../rombios/BIOS-bochs-latest ../vgabios/VGABIOS-lgpl-latest.bin \ ../vgabios/VGABIOS-lgpl-latest.cirrus.bin ../etherboot/eb-roms.h sh ./mkhex rombios ../rombios/BIOS-bochs-latest > roms.h + sh ./mkhex vgabios_pt ../vgabios/vgabios-pt.bin >> roms.h sh ./mkhex vgabios_stdvga ../vgabios/VGABIOS-lgpl-latest.bin >> roms.h sh ./mkhex vgabios_cirrusvga \ ../vgabios/VGABIOS-lgpl-latest.cirrus.bin >> roms.h diff -r 494be76c1ad9 tools/firmware/hvmloader/hvmloader.c --- a/tools/firmware/hvmloader/hvmloader.c Thu Aug 27 16:54:33 2009 +0800 +++ b/tools/firmware/hvmloader/hvmloader.c Thu Aug 27 17:23:00 2009 +0800 @@ -688,9 +688,9 @@ int main(void) vgabios_sz = round_option_rom(sizeof(vgabios_stdvga)); break; case VGA_pt: - printf("Loading Gfx Video BIOS from 0xC0000 ...\n"); - vgabios_sz = - round_option_rom((*(uint8_t *)(VGABIOS_PHYSICAL_ADDRESS+2)) * 512); + printf("Loading Gfx Video BIOS from file ...\n"); + memcpy((void *)VGABIOS_PHYSICAL_ADDRESS, vgabios_pt, sizeof(vgabios_pt)); + vgabios_sz = round_option_rom(sizeof(vgabios_pt)); break; default: printf("No emulated VGA adaptor ...\n"); Teo En Ming (Zhang Enming) wrote: > Dear Weidong, > > I have read through the technical papers on VGA passthrough presented > by the University of Amsterdam and the University of Michigan and > thus gained a better understanding of VGA passthrough in Xen. > > For my preferred setup, I would not be able to adopt the guidance > provided by the University of Amsterdam. I should adopt the guidance > provided by the University of Michigan. I want dom0 to have access to > Intel GMA 4500 onboard graphics while the Windows XP Home HVM domU to > have access to PCI Express x16 graphics. > > I have taken the first step in examining the source code for > disabling the Cirrus emulated vga card bootup in Xen 3.5-unstable. > > According to Beng Heng's UMICH technical paper, he referenced source > file tools/ioemu-remote/hw/pc.c for disabling emulated vga device. > > I went to the xen-unstable mercurial reponsitory > http://xenbits.xensource.com/xen-unstable.hg and found > tools/ioemu/hw/pc.c > > Now, my objective is to disable the emulated/virtual vga card and > load the bios of my nvidia geforce 8400 gs pcie x16 graphics card. > > Source code file > http://xenbits.xensource.com/xen-unstable.hg?file/18b41609a980/tools/ioemu/h > w/pc.c > > ========================================================================= > > Line 862 says that if variable cirrus_vga_enabled returns true > (having a value of 1), the cirrus vga bios VGABIOS_CIRRUS_FILENAME > will be loaded. Otherwise, an alternative vga bios file > VGABIOS_FILENAME will be loaded. > > On line 30 VGABIOS_FILENAME is defined. > > On line 31 VGABIOS_CIRRUS_FILENAME is defined. > > On line 24 a header file vl.h is referenced. > > Source code file > http://xenbits.xensource.com/xen-unstable.hg?file/7eefe6399bcd/tools/ioemu/v > l.h > > =========================================================================== > > On line 177 the integer variable cirrus_vga_enabled is declared but I > could not find any line where this variable is initialized to a > value. But I found file vl.c. > > Source code file > http://xenbits.xensource.com/xen-unstable.hg?file/dade7f0bdc8d/tools/ioemu/v > l.c > > ============================================================================ > > On line 168 integer variable cirrus_vga_enabled is initialized to a > value of > 1. However, I want to disable cirrus vga and load my nvidia geforce > 8400 gs vga bios. > > So I should change line 168 to "int cirrus_vga_enabled = 0;" in my > case. > > Hence line 865 in file pc.c would be executed. > > After extracting the vga bios of my nvidia geforce 8400 gs pcie x16 > graphics card using the nvflash.exe utility and saving the firmware > file as nv-gf-8400-gs.bin, I should place this bios image file in the > xen-unstable sources directory on my PC. It should be in the same > directory where files vgabios.bin and vgabios-cirrus.bin are found. > > Then I should change line 30 in file pc.c from '#define > VGABIOS_FILENAME "vgabios.bin"' to '#define VGABIOS_FILENAME > nv-gf-8400-gs.bin' and then recompile and install the source code. > > Besides disabling the emulated cirrus vga card, may I know what other > things do I need to take care of? Does the pre-existing xen > 3.5-unstable source code already can take care of re-executing the > geforce 8400 gs vga bios, passing through the framebuffer and vga I/O > ports, and intercepting the values in the host bridge such that I do > not have to hack the other relevant source code files? > > Thank you very much. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > Alma Maters: Singapore Polytechnic, National University of Singapore > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En > Ming (Zhang Enming) > Sent: Thursday, August 27, 2009 2:03 PM > To: enming.teo@asiasoftsea.net; weidong.han@intel.com; > djmagee@mageenet.net Cc: xen-devel@lists.xensource.com > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel > Desktop Board DQ45CB > > I think I cannot use NiBiTor already as I don't have any native > Windows operating system running on my computer. > > Instead, I should create a CD-ROM boot disc that contains nvflash.exe > and also loads USB drivers. > > I will use nvflash.exe to extract out and save VGA BIOS to USB sticks. > > NVIDIA GeForce BIOS Backup Guide: > http://www.mvktech.net/content/view/14/37/ > > Motherboard Flash Boot CD from Linux Mini HOWTO: > http://www.nenie.org/misc/flashbootcd.html > > DOS USB Drivers: http://www.bootdisk.com/usb.htm > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Teo En > Ming (Zhang Enming) > Sent: Thursday, August 27, 2009 11:18 AM > To: weidong.han@intel.com; djmagee@mageenet.net > Cc: xen-devel@lists.xensource.com > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > PassThrough to Windows XP Home 32-bit HVM Virtual Machine withIntel > Desktop Board DQ45CB > > Dear Weidong, > > I will go for option 2. > >> don't initialize emulated VGA in qemu, I I think you have seen the >> code. > > I have not seen the source code. Though I have learnt simple > structured C before in Singapore Polytechnic some 11 years ago, I am > not a programmer by profession. So I would still need some > instructions on how I can prevent emulated VGA bios from loading and > getting the vga bios of the real display card to load. > > Thank you. > > Regards, > > Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) > BEng(Hons)(Mechanical Engineering) > Technical Support Engineer > Information Technology Department > Asiasoft Online Pte Ltd > Tampines Central 1 #04-01 Tampines Plaza > Singapore 529541 > Republic of Singapore > Mobile: +65-9648-9798 > MSN: teoenming@hotmail.com > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, > Weidong Sent: Thursday, August 27, 2009 10:01 AM > To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' > Cc: 'xen-devel@lists.xensource.com' > Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA > PassThrough to Windows XP Home 32-bit HVM Virtual Machine with > IntelDesktop Board DQ45CB > > Teo En Ming (Zhang Enming) wrote: >> Hi Weidong, >> >> Currently I am using the entire Xen 3.5-unstable branch and not using >> the posted pci-stub patch changeset 19893 to patch Xen 3.4.1-testing. >> I am using Xen 3.5-unstable in conjunction with pvops dom 0 kernel >> 2.6.31-rc6. >> >> Here are my questions: >> >> 1) How do I extract out the VGA BIOS/firmware from the >> firmware/EEPROM chip on the Geforce 8400 GS VGA card? If it is not >> possible to extract, where can I download the firmware? NVIDIA does >> not provide display card firmware/BIOS files for public download on >> its website. > > There are 3 ways to load VGA bios to guest: > 1. copy it from host 0xc0000, and map it to guest 0xc0000. This works > for primary gfx card, XCI does like this. But some VGA bios > re-execution doesn't work well. > 2. dump VGA bios to a file first, then load it to guest: you can use > NiBiTor to dump VGA bios of Nvidia gfx cards to a file, then load it > into guest roms like what emulated VGA does in hvmloader. This works > for all gfx cards, but it involves manual steps. > > 3. load VGA bios from expansion rom: Intel iGFX doesn't have > expansion rom > >> >> 2) How do I get the Geforce 8400 GS VGA BIOS to load instead of the >> emulated VGA BIOS? Which configuration files/source codes do I have >> to configure/patch? > > don't initialize emulated VGA in qemu, I I think you have seen the > code. > >> >> 3) What are physical MMIO, virtual MMIO Bars, dsdt.asl and 1:1 map? I >> am sorry but I am not a graphics card hardware engineer so I do not >> understand those terms. I would appreciate it if you could write out >> some steps that I could carry out or point me to already available >> documentation. > > When you use command 'lspci -v', you may see below messages under each > device: > > ... > Memory at d0000000 (32-bit, prefetchable) [size=256M] > Memory at ff900000 (32-bit, non-prefetchable) [size=1M] > ... > > They are MMIO BARs. 1:1 map means make these MMIO BARs in guest are > equal to the values in host. > >> >> 4) I don't mind using experimental code. Please post it on the >> xen-devel mailing list so that I could try it out. > > we will post in near future. pls wait for a while. > > Regards, > Weidong > >> >> Thank you very much. >> >> Regards, >> >> Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) >> BEng(Hons)(Mechanical Engineering) >> Technical Support Engineer >> Information Technology Department >> Asiasoft Online Pte Ltd >> Tampines Central 1 #04-01 Tampines Plaza >> Singapore 529541 >> Republic of Singapore >> Mobile: +65-9648-9798 >> MSN: teoenming@hotmail.com >> >> -----Original Message----- >> From: Han, Weidong [mailto:weidong.han@intel.com] >> Sent: Wednesday, August 26, 2009 6:56 PM >> To: 'enming.teo@asiasoftsea.net'; 'djmagee@mageenet.net' >> Cc: 'xen-devel@lists.xensource.com' >> Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >> Pass Through to Windows XP Home 32-bit HVM Virtual Machine with >> Intel Desktop Board DQ45CB >> >> Teo En Ming (Zhang Enming) wrote: >>> Hi Weidong, >>> >>> Could you share with us the hack codes for making Geforce 8400 GS >>> work and also how to let the Windows HVM guest boot up using the >>> real BIOS of Geforce 8400 GS instead of an emulated VGA BIOS? >>> >> >> What patch are you using now? Using real VGA bios of gfx card to >> replace emulatd VGA bios is the prerequisite of gfx passthrough. You >> can find it in posted gfx passthrough patches or XCI. For hack of >> making Geforce 8400, we reserve physical MMIO BARs in dsdt.asl, and >> make it 1:1 map between physical MMIO BARs and virtual MMIO BARs of >> the card. Currently our code is experimental, we will send out in >> mailing list after cleanup and more tests. >> >> Regards, >> Weidong >> >>> Thank you. >>> >>> Regards, >>> >>> Mr. Teo En Ming (Zhang Enming) Dip(Mechatronics Engineering) >>> BEng(Hons)(Mechanical Engineering) >>> Technical Support Engineer >>> Information Technology Department >>> Asiasoft Online Pte Ltd >>> Tampines Central 1 #04-01 Tampines Plaza >>> Singapore 529541 >>> Republic of Singapore >>> Mobile: +65-9648-9798 >>> MSN: teoenming@hotmail.com >>> Alma Maters: Singapore Polytechnic, National University of Singapore >>> >>> -----Original Message----- >>> From: xen-devel-bounces@lists.xensource.com >>> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Han, >>> Weidong Sent: Wednesday, August 26, 2009 4:27 PM >>> To: 'djmagee@mageenet.net'; 'enming.teo@asiasoftsea.net' >>> Cc: 'xen-devel@lists.xensource.com' >>> Subject: RE: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 >>> VGAPassThroughto Windows XP Home 32-bit HVM Virtual Machine with >>> IntelDesktop BoardDQ45CB >>> >>> I suppose you just use the patch posted in mailing list before. >>> nVidia Geforce 8400 passthrough needs extra hacks. We can make it >>> work in our experiments with 1:1 map of its MMIO BARs. >>> >>> We are working on gfx passthrough on latest xen-unstable. Firstly, >>> we want to cook a simple patch including generic changes to support >>> passthrough of virtualization friendly gfx cards, such as Nvidia >>> FX3800. This patch is basically done. Then, we will add some hacks >>> for more gfx cards passthrough, such as iGFX and some Nvidia and >>> ATI cards. >>> >>> Regards, >>> Weidong >>> >>> djmagee@mageenet.net wrote: >>>> As I've pointed out on this list before, there are not enough PCIe >>>> lanes on the DQ45CB to drive both the internal graphics adapter and >>>> the add-on adapter at the same time. I believe the onboard one may >>>> be able to operate in some sort of VGA only mode when there is a >>>> card installed and used as the primary adapter. >>>> >>>> I have a similar setup, using the same motherboard, and an ATI >>>> 4770. I used the VGA passthrough patches from an earlier posting >>>> (I saw you followed the same set of instructions), with limited >>>> success. I've been using 3.4-testing, and 'xenified' 2.6.29.6 >>>> kernel. I made modifications to the dom0 portion of the patches >>>> so they would apply to my xenified 2.6.29 kernel. These patches >>>> include code that will copy the VGA bios to the guest bios instead >>>> of using the emulated vga bios. I've had very little success, >>>> however. I have only tried passing through the ATI adapter. In >>>> all instances, the guest bios messages appear on my monitor, so >>>> this much works. In some cases, the guest essentially stops >>>> there; xm list show's about 2sec CPU usage and nothing ever >>>> happens after that point. In other cases, the guest (win >>>> xp/vista/7, as well as KNOPPIX 5.3.1 DVD) will boot all the way, >>>> but in very low res/color mode, and cannot properly initialize the >>>> video device. Once or twice, it actually did recognize the device >>>> and had a reasonable default color/resolution combination. In all >>>> cases where the guest actually boots, the system eventually >>>> freezes. In some cases, I get endless streams of iommu page >>>> faults. >>>> >>>> I have 8GB ram installed. In all cases I've limited dom0 memory to >>>> 2GB. In all cases, my guest has been assigned 2GB of memory. >>>> >>>> I have a dual core e6600. I've tried allowing dom0 to use both >>>> cores, offlining one core (using xend dom0_vcpu setting) after >>>> boot, and restricting dom0 to only one core using the >>>> dom0_max_vcpus xen hypervisor parameter. In all of these cases >>>> I've tried both one and two vcpus for the guest. My success with >>>> VGA passthrough seems somewhat random and no combination of cpu >>>> assignment seems to have any effect. >>>> >>>> I have not tried with the 2.6.18-xen kernel as I haven't gotten it >>>> to boot on my hardware; it can never find my volume group, even if >>>> I create a initrd with all of the required modules, or build those >>>> drivers into the kernel. I have not spent more than maybe a half >>>> an hour on this problem; I suspect it may have something to do >>>> with the version of mkinitrd I'm using (from Fedora 9 x64). >>>> >>>> If anyone else has any insight or similar experience I'd also love >>>> to hear it. >>>> >>>> Doug Magee >>>> djmagee@mageenet.com >>>> >>>> -----Original Message----- >>>> From: xen-devel-bounces@lists.xensource.com >>>> [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Mr. Teo >>>> En Ming (Zhang Enming) Sent: Tuesday, August 25, 2009 11:57 AM >>>> To: enming.teo@asiasoftsea.net >>>> Cc: xen-devel@lists.xensource.com >>>> Subject: Re: [Xen-devel] nVidia Geforce 8400 GS PCI Express x16 VGA >>>> PassThroughto Windows XP Home 32-bit HVM Virtual Machine with >>>> Intel Desktop BoardDQ45CB >>>> >>>> I have uninstalled Xen 3.4.1 and installed Xen 3.5-unstable as >>>> suggested by Weidong. >>>> >>>> >>>> On 08/25/2009 11:47 PM, Mr. Teo En Ming (Zhang Enming) wrote: >>>>> Dear All, >>>>> >>>>> I have managed to do PCI-e VGA passthrough with the open source >>>>> Xen but the work is still in progress because although Windows XP >>>>> guest can see the REAL PCI-e x16 graphics card instead of an >>>>> emulated graphics driver, it cannot be initialized yet. >>>>> >>>>> Thanks to Intel Engineer Han Weidong, Pasi Kärkkäinen, Boris >>>>> Derzhavets, Marc, Caz Yokoyama, and others who have helped me and >>>>> shared their knowledge with me along the way. >>>>> >>>>> System Configuration: >>>>> >>>>> Intel Desktop Board DQ45CB with BIOS upgraded to 0093 >>>>> Onboard Intel GMA 4500 Graphics (IGD) >>>>> nVidia Geforce 8400 GS PCI Express x16 Graphics Card >>>>> >>>>> Fedora 11 Linux 64-bit Xen paravirt operations Domain 0 Host >>>>> Operating System Xen 3.5 Unstable/Development Type 1 Hypervisor >>>>> Jeremy Fitzhardinge's Xen paravirt-ops domain 0 Kernel 2.6.31-rc6 >>>>> Primary Video Adapter in BIOS: IGD >>>>> >>>>> Please see the screenshots and my blog at the link here: >>>>> >>>>> >>> >> > http://teo-en-ming-aka-zhang-enming.blogspot.com/2009/08/nvidia-geforce-8400 >>> -gs-pci-express-x16.html >>>>> >>>>> >>>> >>>> >>>> >>>> >>>> _______________________________________________ >>>> Xen-devel mailing list >>>> Xen-devel@lists.xensource.com >>>> http://lists.xensource.com/xen-devel >>>> >>>> _______________________________________________ >>>> Xen-devel mailing list >>>> Xen-devel@lists.xensource.com >>>> http://lists.xensource.com/xen-devel >>> >>> >>> _______________________________________________ >>> Xen-devel mailing list >>> Xen-devel@lists.xensource.com >>> http://lists.xensource.com/xen-devel >>> >>> No virus found in this incoming message. >>> Checked by AVG - www.avg.com >>> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >>> 08/25/09 18:07:00 >>> >>> No virus found in this outgoing message. >>> Checked by AVG - www.avg.com >>> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >>> 08/25/09 18:07:00 >> >> >> No virus found in this incoming message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 >> >> No virus found in this outgoing message. >> Checked by AVG - www.avg.com >> Version: 8.5.392 / Virus Database: 270.13.65/2324 - Release Date: >> 08/25/09 18:07:00 > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: > 08/25/09 18:07:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.67/2326 - Release Date: > 08/26/09 12:16:00 > > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: > 08/26/09 12:16:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: > 08/26/09 12:16:00 > > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: > 08/26/09 12:16:00 > > No virus found in this outgoing message. > Checked by AVG - www.avg.com > Version: 8.5.392 / Virus Database: 270.13.69/2328 - Release Date: > 08/26/09 12:16:00 ^ permalink raw reply [flat|nested] 25+ messages in thread
end of thread, other threads:[~2009-08-28 3:30 UTC | newest]
Thread overview: 25+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-08-25 15:47 nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Mr. Teo En Ming (Zhang Enming)
2009-08-25 15:56 ` Mr. Teo En Ming (Zhang Enming)
2009-08-25 17:25 ` nVidia Geforce 8400 GS PCI Express x16 VGA PassThroughto Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop BoardDQ45CB djmagee
[not found] ` <f4527be0908251056s295eaf0fle31922e7d90faa1c@mail.gmail.com>
2009-08-25 19:57 ` djmagee
2009-08-26 8:26 ` Han, Weidong
2009-08-26 9:21 ` nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Teo En Ming (Zhang Enming)
2009-08-26 10:56 ` Han, Weidong
2009-08-26 11:24 ` Teo En Ming (Zhang Enming)
2009-08-26 11:31 ` Teo En Ming (Zhang Enming)
2009-08-26 11:37 ` Vincent Hanquez
2009-08-26 11:49 ` Teo En Ming (Zhang Enming)
2009-08-26 12:17 ` Teo En Ming (Zhang Enming)
2009-08-26 13:33 ` Teo En Ming (Zhang Enming)
2009-08-26 13:37 ` Teo En Ming (Zhang Enming)
2009-08-26 21:08 ` VGA passthrough to HVM Tim Moore
2009-08-27 2:40 ` Teo En Ming (Zhang Enming)
2009-08-27 2:59 ` Teo En Ming (Zhang Enming)
2009-08-28 2:19 ` Teo En Ming (Zhang Enming)
2009-08-28 2:51 ` Teo En Ming (Zhang Enming)
2009-08-27 2:01 ` nVidia Geforce 8400 GS PCI Express x16 VGA Pass Through to Windows XP Home 32-bit HVM Virtual Machine with Intel Desktop Board DQ45CB Han, Weidong
2009-08-27 3:18 ` Teo En Ming (Zhang Enming)
2009-08-27 6:03 ` Teo En Ming (Zhang Enming)
2009-08-27 16:18 ` Teo En Ming (Zhang Enming)
2009-08-27 18:28 ` djmagee
2009-08-28 3:30 ` Han, Weidong
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.