On Fri, 2011-08-05 at 13:16 +0100, Keir Fraser wrote:
> On 05/08/2011 04:20, "Bei Guan" <gbtju85@xxxxxxxxx> wrote:
>
> > Hi All,
> >
> > Thank you for all your comments. I has removed the patch for xend and
> > modified
> > some other little stuff according to your comments.
> > I have rebased my patches on the top of the latest Xen-unstable version
> > (changeset: 23756:0f36c2eec2e).
> > Is there any more comment? Thank you very much.
>
> Apart from the pci_setup hook I just emailed about, hopefully we can avoid
> code duplication in the init_vm86_tss hook as well. I'm not sure why we even
> need the hook -- I'll need to check what it's for with Ian Campbell (cc'ed).
I dunno ;-) I had assumed it was a rombios specific thing when I
refactored for SeaBIOS (which doesn't implement the hook).
Thinking about it again is this VM86 TSS a hangover from the days where
we used to "emulate" realmode using vm86 in the HVM guest context (IIRC
that was called vmxassist?). Perhaps the hook can be simply dropped?
Aha:
changeset: 18891:6595393a3d28
user: Keir Fraser <keir.fraser@xxxxxxxxxx>
date: Tue Dec 09 16:28:02 2008 +0000
files: [...]
description:
Use virtual 8086 mode for VMX guests with CR0.PE == 0
When a VMX guest tries to enter real mode, put it in virtual 8086 mode
instead, if that's possible. Handle all errors and corner cases by
falling back to the real-mode emulator.
This is similar to the old VMXASSIST system except it uses Xen's
x86_emulate emulator instead of having a partial emulator in the guest
firmware. It more than doubles the speed of real-mode operation on
VMX.
Signed-off-by: Tim Deegan <Tim.Deegan@xxxxxxxxxx>
Looks like it is actually needed for all BIOS types?
Ian.
>
> - Keir
>
> > In the patch for ovmf.c, we add a method ovmf_pci_setup(), which is copied
> > form method pci_setup() in hvmloader/pci.c except for all the VGA related
> > stuff. This method is assigned to bios->pci_setup(). The reason for adding
> > such a method is stated here:
> > http://sourceforge.net/mailarchive/message.php?msg_id=27854247
> >
> >
> > The patches are also available at
> > https://github.com/GavinGuan/gavinguan/blob/master/xen/ovmf-support/ovmf_xen_s
> > upport.patch
> > https://github.com/GavinGuan/gavinguan/blob/master/xen/ovmf-support/ovmf_xl.pa
> > tch
> > https://github.com/GavinGuan/gavinguan/blob/master/xen/ovmf-support/ovmf_firmw
> > are.patch
> >
> >
> > # HG changeset patch
> > # User gbtju85@xxxxxxxxx
> > #
> >
> > Enable Xen-unstable hvmloader to load OVMF BIOS.
> > It supports OVMF BIOS in IA32 and X86 environment.
> >
> > Usage:
> > Add an option field in HVM config file.
> > # OVMF support. When enabled, hvmloader can load OVMF bios of
> > IA32("ovmf-ia32") and X64("ovmf-x64")
> > hvmbios = "ovmf-ia32"
> > #hvmbios = "ovmf-x64"
> >
> > Note:
> > Enable the HVM guest ACPI: acpi=1
> > Use the OVMF to boot into a UEFI-aware OS, such as
> > ubuntu-10.10-desktop-amd64.iso. Just set the "disk" option like this:
> > disk = [ 'file:/root/<img_name>.img,ioemu:hda,w',
> > 'file:/root/ubuntu-10.10-desktop-amd64.iso,hdc:cdrom,r' ]
> >
> > diff -r 0f36c2eec2e1 tools/firmware/hvmloader/Makefile
> > --- a/tools/firmware/hvmloader/Makefile Thu Jul 28 15:40:54 2011 +0100
> > +++ b/tools/firmware/hvmloader/Makefile Fri Aug 05 17:58:27 2011 +0800
> > @@ -43,6 +43,19 @@
> > CFLAGS += -DENABLE_ROMBIOS
> > ROMBIOS_ROM := $(ROMBIOS_DIR)/BIOS-bochs-latest
> > endif
> > +OVMF_DIR := ../ovmf
> > +OVMF32_ROM := $(OVMF_DIR)/ovmf-ia32.bin
> > +OVMF64_ROM := $(OVMF_DIR)/ovmf-x64.bin
> > +OVMF32_CIRRUS_VGA_ROM := $(OVMF_DIR)/ovmf-ia32-cirrus-vga.bin
> > +OVMF64_CIRRUS_VGA_ROM := $(OVMF_DIR)/ovmf-x64-cirrus-vga.bin
> > +
> > +ifneq ($(OVMF32_ROM),)
> > +OBJS += ovmf.o
> > +endif
> > +
> > +ifneq ($(OVMF64_ROM),)
> > +OBJS += ovmf.o
> > +endif
> >
> > ifneq ($(SEABIOS_DIR),)
> > OBJS += seabios.o
> > @@ -69,7 +82,7 @@
> > $(OBJCOPY) hvmloader.tmp hvmloader
> > rm -f hvmloader.tmp
> >
> > -roms.inc: $(ROMBIOS_ROM) $(SEABIOS_ROM) $(STDVGA_ROM) $(CIRRUSVGA_ROM)
> > ../etherboot/eb-roms.h
> > +roms.inc: $(ROMBIOS_ROM) $(SEABIOS_ROM) $(STDVGA_ROM) $(CIRRUSVGA_ROM)
> > $(OVMF32_ROM) $(OVMF64_ROM) $(OVMF32_CIRRUS_VGA_ROM)
> > $(OVMF64_CIRRUS_VGA_ROM)
> > ../etherboot/eb-roms.h
> > echo "/* Autogenerated file. DO NOT EDIT */" > $@.new
> >
> > ifneq ($(ROMBIOS_ROM),)
> > @@ -84,6 +97,30 @@
> > echo "#endif" >> $@.new
> > endif
> >
> > +ifneq ($(OVMF32_ROM),)
> > + echo "#ifdef ROM_INCLUDE_OVMF32" >> $@.new
> > + sh ./mkhex ovmf32 $(OVMF32_ROM) >> $@.new
> > + echo "#endif" >> $@.new
> > +endif
> > +
> > +ifneq ($(OVMF64_ROM),)
> > + echo "#ifdef ROM_INCLUDE_OVMF64" >> $@.new
> > + sh ./mkhex ovmf64 $(OVMF64_ROM) >> $@.new
> > + echo "#endif" >> $@.new
> > +endif
> > +
> > +ifneq ($(OVMF32_CIRRUS_VGA_ROM),)
> > + echo "#ifdef ROM_INCLUDE_OVMF32_CIRRUS_VGA" >> $@.new
> > + sh ./mkhex ovmf32_cirrus_vga $(OVMF32_CIRRUS_VGA_ROM) >> $@.new
> > + echo "#endif" >> $@.new
> > +endif
> > +
> > +ifneq ($(OVMF64_CIRRUS_VGA_ROM),)
> > + echo "#ifdef ROM_INCLUDE_OVMF64_CIRRUS_VGA" >> $@.new
> > + sh ./mkhex ovmf64_cirrus_vga $(OVMF64_CIRRUS_VGA_ROM) >> $@.new
> > + echo "#endif" >> $@.new
> > +endif
> > +
> > ifneq ($(STDVGA_ROM),)
> > echo "#ifdef ROM_INCLUDE_VGABIOS" >> $@.new
> > sh ./mkhex vgabios_stdvga $(STDVGA_ROM) >> $@.new
> > diff -r 0f36c2eec2e1 tools/firmware/hvmloader/config.h
> > --- a/tools/firmware/hvmloader/config.h Thu Jul 28 15:40:54 2011 +0100
> > +++ b/tools/firmware/hvmloader/config.h Fri Aug 05 17:58:27 2011 +0800
> > @@ -3,7 +3,7 @@
> >
> > #include <stdint.h>
> >
> > -enum virtual_vga { VGA_none, VGA_std, VGA_cirrus, VGA_pt };
> > +enum virtual_vga { VGA_none, VGA_std, VGA_cirrus, VGA_pt, VGA_custom };
> > extern enum virtual_vga virtual_vga;
> >
> > struct bios_config {
> > @@ -27,6 +27,7 @@
> >
> > void (*vm86_setup)(void);
> > void (*e820_setup)(void);
> > + void (*pci_setup)(void);
> >
> > void (*acpi_build_tables)(void);
> > void (*create_mp_tables)(void);
> > @@ -36,6 +37,8 @@
> >
> > extern struct bios_config rombios_config;
> > extern struct bios_config seabios_config;
> > +extern struct bios_config ovmf32_config;
> > +extern struct bios_config ovmf64_config;
> >
> > #define PAGE_SHIFT 12
> > #define PAGE_SIZE (1ul << PAGE_SHIFT)
> > diff -r 0f36c2eec2e1 tools/firmware/hvmloader/hvmloader.c
> > --- a/tools/firmware/hvmloader/hvmloader.c Thu Jul 28 15:40:54 2011 +0100
> > +++ b/tools/firmware/hvmloader/hvmloader.c Fri Aug 05 17:58:27 2011 +0800
> > @@ -361,6 +361,8 @@
> > #ifdef ENABLE_SEABIOS
> > { "seabios", &seabios_config, },
> > #endif
> > + { "ovmf-ia32", &ovmf32_config, },
> > + { "ovmf-x64", &ovmf64_config, },
> > { NULL, NULL }
> > };
> >
> > @@ -420,7 +422,11 @@
> > printf("CPU speed is %u MHz\n", get_cpu_mhz());
> >
> > apic_setup();
> > - pci_setup();
> > + if (bios->pci_setup) {
> > + bios->pci_setup();
> > + } else {
> > + pci_setup();
> > + }
> >
> > smp_initialise();
> >
> > @@ -471,6 +477,8 @@
> > vgabios_sz = round_option_rom(
> > (*(uint8_t *)(VGABIOS_PHYSICAL_ADDRESS+2)) * 512);
> > break;
> > + case VGA_custom:
> > + break;
> > default:
> > printf("No emulated VGA adaptor ...\n");
> > break;
> > diff -r 0f36c2eec2e1 tools/firmware/hvmloader/ovmf.c
> > --- /dev/null Thu Jan 01 00:00:00 1970 +0000
> > +++ b/tools/firmware/hvmloader/ovmf.c Fri Aug 05 17:58:27 2011 +0800
> > @@ -0,0 +1,394 @@
> > +/*
> > + * HVM OVMF UEFI support.
> > + *
> > + * Bei Guan, gbtju85@xxxxxxxxx
> > + * Andrei Warkentin, andreiw@xxxxxxxxxxxx
> > + * Leendert van Doorn, leendert@xxxxxxxxxxxxxx
> > + * Copyright (c) 2005, International Business Machines Corporation.
> > + * Copyright (c) 2006, Keir Fraser, XenSource Inc.
> > + * Copyright (c) 2011, Citrix Inc.
> > + *
> > + * This program is free software; you can redistribute it and/or modify it
> > + * under the terms and conditions of the GNU General Public License,
> > + * version 2, as published by the Free Software Foundation.
> > + *
> > + * This program is distributed in the hope it will be useful, but WITHOUT
> > + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
> > + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
> > for
> > + * more details.
> > + *
> > + * You should have received a copy of the GNU General Public License along
> > with
> > + * this program; if not, write to the Free Software Foundation, Inc., 59
> > Temple
> > + * Place - Suite 330, Boston, MA 02111-1307 USA.
> > + */
> > +
> > +#include "config.h"
> > +#include "smbios_types.h"
> > +#include "acpi/acpi2_0.h"
> > +#include "apic_regs.h"
> > +#include "../rombios/config.h"
> > +#include "util.h"
> > +#include "pci_regs.h"
> > +#include "hypercall.h"
> > +
> > +#include <xen/hvm/params.h>
> > +#include <xen/hvm/ioreq.h>
> > +#include <xen/memory.h>
> > +
> > +#define ROM_INCLUDE_OVMF32
> > +#define ROM_INCLUDE_OVMF64
> > +#define ROM_INCLUDE_OVMF32_CIRRUS_VGA
> > +#define ROM_INCLUDE_OVMF64_CIRRUS_VGA
> > +#include "roms.inc"
> > +
> > +#define OVMF_BEGIN 0xFFF00000ULL
> > +#define OVMF_SIZE 0x00100000ULL
> > +#define OVMF_MAXOFFSET 0x000FFFFFULL
> > +#define OVMF_END (OVMF_BEGIN + OVMF_SIZE)
> > +#define LOWCHUNK_BEGIN 0x000F0000
> > +#define LOWCHUNK_SIZE 0x00010000
> > +#define LOWCHUNK_MAXOFFSET 0x0000FFFF
> > +#define LOWCHUNK_END (OVMF_BEGIN + OVMF_SIZE)
> > +
> > +/*
> > + * Set up an empty TSS area for virtual 8086 mode to use.
> > + * The only important thing is that it musn't have any bits set
> > + * in the interrupt redirection bitmap, so all zeros will do.
> > + */
> > +static void ovmf_init_vm86_tss(void)
> > +{
> > + void *tss;
> > + struct xen_hvm_param p;
> > +
> > + tss = mem_alloc(128, 128);
> > + memset(tss, 0, 128);
> > + p.domid = DOMID_SELF;
> > + p.index = HVM_PARAM_VM86_TSS;
> > + p.value = virt_to_phys(tss);
> > + hypercall_hvm_op(HVMOP_set_param, &p);
> > + printf("vm86 TSS at %08lx\n", virt_to_phys(tss));
> > +}
> > +
> > +static void ovmf_load(const struct bios_config *config)
> > +{
> > + xen_pfn_t mfn;
> > + uint64_t addr = OVMF_BEGIN;
> > +
> > + virtual_vga = VGA_custom;
> > +
> > + /* Copy video ROM. */
> > + if (config == &ovmf32_config) {
> > + memcpy((void *)VGABIOS_PHYSICAL_ADDRESS,
> > + ovmf32_cirrus_vga, sizeof(ovmf32_cirrus_vga));
> > + printf("OVMF32 Cirrus [0x%x-0x%x]\n", VGABIOS_PHYSICAL_ADDRESS,
> > + VGABIOS_PHYSICAL_ADDRESS + sizeof(ovmf32_cirrus_vga));
> > + } else if (config == &ovmf64_config) {
> > + memcpy((void *)VGABIOS_PHYSICAL_ADDRESS,
> > + ovmf64_cirrus_vga, sizeof(ovmf64_cirrus_vga));
> > + printf("OVMF64 Cirrus [0x%x-0x%x]\n", VGABIOS_PHYSICAL_ADDRESS,
> > + VGABIOS_PHYSICAL_ADDRESS + sizeof(ovmf64_cirrus_vga));
> > + }
> > +
> > + /* Copy low-reset vector portion. */
> > + memcpy((void *) LOWCHUNK_BEGIN, (uint8_t *) config->image
> > + + OVMF_SIZE
> > + - LOWCHUNK_SIZE,
> > + LOWCHUNK_SIZE);
> > +
> > + /* Ensure we have backing page prior to moving FD. */
> > + while ((addr >> PAGE_SHIFT) != (OVMF_END >> PAGE_SHIFT)) {
> > + mfn = (uint32_t) (addr >> PAGE_SHIFT);
> > + addr += PAGE_SIZE;
> > +
> > + mem_hole_populate_ram(mfn, 1);
> > + }
> > +
> > + printf("Initialized FD backing pages...\n");
> > +
> > + /* Copy FD. */
> > + memcpy((void *) OVMF_BEGIN, config->image, OVMF_SIZE);
> > + printf("Load complete!\n");
> > +}
> > +
> > +/*
> > + * Ideally this function should just adjust the low memory size so MMIO
> > fits,
> > + * everything else should be done in UEFI code
> > + */
> > +static void ovmf_pci_setup(void)
> > +{
> > + uint32_t base, devfn, bar_reg, bar_data, bar_sz, cmd, mmio_total = 0;
> > + uint16_t class, vendor_id, device_id;
> > + unsigned int bar, pin, link, isa_irq;
> > +
> > + /* Resources assignable to PCI devices via BARs. */
> > + struct resource {
> > + uint32_t base, max;
> > + } *resource, mem_resource, io_resource;
> > +
> > + /* Create a list of device BARs in descending order of size. */
> > + struct bars {
> > + uint32_t devfn, bar_reg, bar_sz;
> > + } *bars = (struct bars *)SCRATCH_PHYSICAL_ADDRESS;
> > + unsigned int i, nr_bars = 0;
> > +
> > + /* Program PCI-ISA bridge with appropriate link routes. */
> > + isa_irq = 0;
> > + for ( link = 0; link < 4; link++ )
> > + {
> > + do { isa_irq = (isa_irq + 1) & 15;
> > + } while ( !(PCI_ISA_IRQ_MASK & (1U << isa_irq)) );
> > + pci_writeb(PCI_ISA_DEVFN, 0x60 + link, isa_irq);
> > + printf("PCI-ISA link %u routed to IRQ%u\n", link, isa_irq);
> > + }
> > +
> > + /* Program ELCR to match PCI-wired IRQs. */
> > + outb(0x4d0, (uint8_t)(PCI_ISA_IRQ_MASK >> 0));
> > + outb(0x4d1, (uint8_t)(PCI_ISA_IRQ_MASK >> 8));
> > +
> > + /* Scan the PCI bus and map resources. */
> > + for ( devfn = 0; devfn < 256; devfn++ )
> > + {
> > + class = pci_readw(devfn, PCI_CLASS_DEVICE);
> > + vendor_id = pci_readw(devfn, PCI_VENDOR_ID);
> > + device_id = pci_readw(devfn, PCI_DEVICE_ID);
> > + if ( (vendor_id == 0xffff) && (device_id == 0xffff) )
> > + continue;
> > +
> > + ASSERT((devfn != PCI_ISA_DEVFN) ||
> > + ((vendor_id == 0x8086) && (device_id == 0x7000)));
> > +
> > + switch ( class )
> > + {
> > + case 0x0680:
> > + /* PIIX4 ACPI PM. Special device with special PCI config space.
> > */
> > + ASSERT((vendor_id == 0x8086) && (device_id == 0x7113));
> > + pci_writew(devfn, 0x20, 0x0000); /* No smb bus IO enable */
> > + pci_writew(devfn, 0xd2, 0x0000); /* No smb bus IO enable */
> > + pci_writew(devfn, 0x22, 0x0000);
> > + pci_writew(devfn, 0x3c, 0x0009); /* Hardcoded IRQ9 */
> > + pci_writew(devfn, 0x3d, 0x0001);
> > + pci_writel(devfn, 0x40, ACPI_PM1A_EVT_BLK_ADDRESS_V1 | 1);
> > + pci_writeb(devfn, 0x80, 0x01); /* enable PM io space */
> > + break;
> > + case 0x0101:
> > + if ( vendor_id == 0x8086 )
> > + {
> > + /* Intel ICHs since PIIX3: enable IDE legacy mode. */
> > + pci_writew(devfn, 0x40, 0x8000); /* enable IDE0 */
> > + pci_writew(devfn, 0x42, 0x8000); /* enable IDE1 */
> > + }
> > + break;
> > + }
> > +
> > + /* Map the I/O memory and port resources. */
> > + for ( bar = 0; bar < 7; bar++ )
> > + {
> > + bar_reg = PCI_BASE_ADDRESS_0 + 4*bar;
> > + if ( bar == 6 )
> > + bar_reg = PCI_ROM_ADDRESS;
> > +
> > + bar_data = pci_readl(devfn, bar_reg);
> > + pci_writel(devfn, bar_reg, ~0);
> > + bar_sz = pci_readl(devfn, bar_reg);
> > + pci_writel(devfn, bar_reg, bar_data);
> > + if ( bar_sz == 0 )
> > + continue;
> > +
> > + bar_sz &= (((bar_data & PCI_BASE_ADDRESS_SPACE) ==
> > + PCI_BASE_ADDRESS_SPACE_MEMORY) ?
> > + PCI_BASE_ADDRESS_MEM_MASK :
> > + (PCI_BASE_ADDRESS_IO_MASK & 0xffff));
> > + bar_sz &= ~(bar_sz - 1);
> > +
> > + for ( i = 0; i < nr_bars; i++ )
> > + if ( bars[i].bar_sz < bar_sz )
> > + break;
> > +
> > + if ( i != nr_bars )
> > + memmove(&bars[i+1], &bars[i], (nr_bars-i) * sizeof(*bars));
> > +
> > + bars[i].devfn = devfn;
> > + bars[i].bar_reg = bar_reg;
> > + bars[i].bar_sz = bar_sz;
> > +
> > + if ( (bar_data & PCI_BASE_ADDRESS_SPACE) ==
> > + PCI_BASE_ADDRESS_SPACE_MEMORY )
> > + mmio_total += bar_sz;
> > +
> > + nr_bars++;
> > +
> > + /* Skip the upper-half of the address for a 64-bit BAR. */
> > + if ( (bar_data & (PCI_BASE_ADDRESS_SPACE |
> > + PCI_BASE_ADDRESS_MEM_TYPE_MASK)) ==
> > + (PCI_BASE_ADDRESS_SPACE_MEMORY |
> > + PCI_BASE_ADDRESS_MEM_TYPE_64) )
> > + bar++;
> > + }
> > +
> > + /* Map the interrupt. */
> > + pin = pci_readb(devfn, PCI_INTERRUPT_PIN);
> > + if ( pin != 0 )
> > + {
> > + /* This is the barber's pole mapping used by Xen. */
> > + link = ((pin - 1) + (devfn >> 3)) & 3;
> > + isa_irq = pci_readb(PCI_ISA_DEVFN, 0x60 + link);
> > + pci_writeb(devfn, PCI_INTERRUPT_LINE, isa_irq);
> > + printf("pci dev %02x:%x INT%c->IRQ%u\n",
> > + devfn>>3, devfn&7, 'A'+pin-1, isa_irq);
> > + }
> > +
> > + /* Enable bus mastering. */
> > + cmd = pci_readw(devfn, PCI_COMMAND);
> > + cmd |= PCI_COMMAND_MASTER;
> > + pci_writew(devfn, PCI_COMMAND, cmd);
> > + }
> > +
> > + while ( (mmio_total > (pci_mem_end - pci_mem_start)) &&
> > + ((pci_mem_start << 1) != 0) )
> > + pci_mem_start <<= 1;
> > +
> > + while ( (pci_mem_start >> PAGE_SHIFT) < hvm_info->low_mem_pgend )
> > + {
> > + struct xen_add_to_physmap xatp;
> > + if ( hvm_info->high_mem_pgend == 0 )
> > + hvm_info->high_mem_pgend = 1ull << (32 - PAGE_SHIFT);
> > + xatp.domid = DOMID_SELF;
> > + xatp.space = XENMAPSPACE_gmfn;
> > + xatp.idx = --hvm_info->low_mem_pgend;
> > + xatp.gpfn = hvm_info->high_mem_pgend++;
> > + if ( hypercall_memory_op(XENMEM_add_to_physmap, &xatp) != 0 )
> > + BUG();
> > + }
> > +
> > + mem_resource.base = pci_mem_start;
> > + mem_resource.max = pci_mem_end;
> > + io_resource.base = 0xc000;
> > + io_resource.max = 0x10000;
> > +
> > + /* Assign iomem and ioport resources in descending order of size. */
> > + for ( i = 0; i < nr_bars; i++ )
> > + {
> > + devfn = bars[i].devfn;
> > + bar_reg = bars[i].bar_reg;
> > + bar_sz = bars[i].bar_sz;
> > +
> > + bar_data = pci_readl(devfn, bar_reg);
> > +
> > + if ( (bar_data & PCI_BASE_ADDRESS_SPACE) ==
> > + PCI_BASE_ADDRESS_SPACE_MEMORY )
> > + {
> > + resource = &mem_resource;
> > + bar_data &= ~PCI_BASE_ADDRESS_MEM_MASK;
> > + }
> > + else
> > + {
> > + resource = &io_resource;
> > + bar_data &= ~PCI_BASE_ADDRESS_IO_MASK;
> > + }
> > +
> > + base = (resource->base + bar_sz - 1) & ~(bar_sz - 1);
> > + bar_data |= base;
> > + base += bar_sz;
> > +
> > + if ( (base < resource->base) || (base > resource->max) )
> > + {
> > + printf("pci dev %02x:%x bar %02x size %08x: no space for "
> > + "resource!\n", devfn>>3, devfn&7, bar_reg, bar_sz);
> > + continue;
> > + }
> > +
> > + resource->base = base;
> > +
> > + pci_writel(devfn, bar_reg, bar_data);
> > + printf("pci dev %02x:%x bar %02x size %08x: %08x\n",
> > + devfn>>3, devfn&7, bar_reg, bar_sz, bar_data);
> > +
> > + /* Now enable the memory or I/O mapping. */
> > + cmd = pci_readw(devfn, PCI_COMMAND);
> > + if ( (bar_reg == PCI_ROM_ADDRESS) ||
> > + ((bar_data & PCI_BASE_ADDRESS_SPACE) ==
> > + PCI_BASE_ADDRESS_SPACE_MEMORY) )
> > + cmd |= PCI_COMMAND_MEMORY;
> > + else
> > + cmd |= PCI_COMMAND_IO;
> > + pci_writew(devfn, PCI_COMMAND, cmd);
> > + }
> > +}
> > +
> > +static void ovmf_acpi_build_tables(void)
> > +{
> > + acpi_build_tables(ACPI_PHYSICAL_ADDRESS);
> > +}
> > +
> > +static void ovmf_create_smbios_tables(void)
> > +{
> > + hvm_write_smbios_tables(SMBIOS_PHYSICAL_ADDRESS,
> > + SMBIOS_PHYSICAL_ADDRESS + sizeof(struct
> > smbios_entry_point),
> > + SMBIOS_PHYSICAL_END);
> > +}
> > +
> > +struct bios_config ovmf32_config = {
> > + .name = "OVMF-IA32",
> > +
> > + .image = ovmf32,
> > + .image_size = sizeof(ovmf32),
> > +
> > + .bios_address = 0,
> > +
> > + .load_roms = 0,
> > +
> > + .optionrom_start = 0,
> > + .optionrom_end = 0,
> > +
> > + .bios_load = ovmf_load,
> > +
> > + .bios_info_setup = NULL,
> > + .bios_info_finish = NULL,
> > +
> > + .vm86_setup = ovmf_init_vm86_tss,
> > + .e820_setup = NULL,
> > + .pci_setup = ovmf_pci_setup,
> > +
> > + .acpi_build_tables = ovmf_acpi_build_tables,
> > + .create_mp_tables = NULL,
> > + .create_smbios_tables = ovmf_create_smbios_tables,
> > + .create_pir_tables = NULL,
> > +};
> > +
> > +struct bios_config ovmf64_config = {
> > + .name = "OVMF-X64",
> > +
> > + .image = ovmf64,
> > + .image_size = sizeof(ovmf64),
> > +
> > + .bios_address = 0,
> > +
> > + .load_roms = 0,
> > +
> > + .optionrom_start = 0,
> > + .optionrom_end = 0,
> > +
> > + .bios_load = ovmf_load,
> > +
> > + .bios_info_setup = NULL,
> > + .bios_info_finish = NULL,
> > +
> > + .vm86_setup = ovmf_init_vm86_tss,
> > + .e820_setup = NULL,
> > + .pci_setup = ovmf_pci_setup,
> > +
> > + .acpi_build_tables = ovmf_acpi_build_tables,
> > + .create_mp_tables = NULL,
> > + .create_smbios_tables = ovmf_create_smbios_tables,
> > + .create_pir_tables = NULL,
> > +};
> > +
> > +/*
> > + * Local variables:
> > + * mode: C
> > + * c-set-style: "BSD"
> > + * c-basic-offset: 4
> > + * tab-width: 4
> > + * indent-tabs-mode: nil
> > + * End:
> > + */
> > diff -r 0f36c2eec2e1 tools/firmware/ovmf/ovmf-ia32-cirrus-vga.bin
> > Binary file tools/firmware/ovmf/ovmf-ia32-cirrus-vga.bin has changed
> > diff -r 0f36c2eec2e1 tools/firmware/ovmf/ovmf-ia32.bin
> > Binary file tools/firmware/ovmf/ovmf-ia32.bin has changed
> > diff -r 0f36c2eec2e1 tools/firmware/ovmf/ovmf-x64-cirrus-vga.bin
> > Binary file tools/firmware/ovmf/ovmf-x64-cirrus-vga.bin has changed
> > diff -r 0f36c2eec2e1 tools/firmware/ovmf/ovmf-x64.bin
> > Binary file tools/firmware/ovmf/ovmf-x64.bin has changed
> >
> >
> >
> >
> >
> >
> > # HG changeset patch
> > # User gbtju85@xxxxxxxxx
> > #
> >
> > Xen: Expose hvmloader/bios in libxl.
> >
> > Exposes the hvmloader/bios xenstore key in libxl, so firmware loaded
> > can be overriden (choices: rombios, seabios, ovmf-ia32, ovmf-x64).
> >
> > Sign-off-by: Bei Guan <gbtju85@xxxxxxxxx>
> >
> > diff -r 0f36c2eec2e1 tools/libxl/libxl.idl
> > --- a/tools/libxl/libxl.idl Thu Jul 28 15:40:54 2011 +0100
> > +++ b/tools/libxl/libxl.idl Fri Aug 05 18:13:37 2011 +0800
> > @@ -137,6 +137,7 @@
> >
> > libxl_domain_create_info = Struct("domain_create_info",[
> > ("type", libxl_domain_type),
> > + ("hvmbios", string),
> > ("hap", bool),
> > ("oos", bool),
> > ("ssidref", uint32),
> > diff -r 0f36c2eec2e1 tools/libxl/libxl_create.c
> > --- a/tools/libxl/libxl_create.c Thu Jul 28 15:40:54 2011 +0100
> > +++ b/tools/libxl/libxl_create.c Fri Aug 05 18:13:37 2011 +0800
> > @@ -407,6 +407,10 @@
> > if (info->poolname)
> > xs_write(ctx->xsh, t, libxl__sprintf(gc, "%s/pool_name", vm_path),
> > info->poolname, strlen(info->poolname));
> >
> > + if (info->hvmbios){
> > + xs_write(ctx->xsh, t, libxl__sprintf(gc, "%s/hvmloader/bios",
> > dom_path), info->hvmbios, strlen(info->hvmbios));
> > + }
> > +
> > libxl__xs_writev(gc, t, dom_path, info->xsdata);
> > libxl__xs_writev(gc, t, libxl__sprintf(gc, "%s/platform", dom_path),
> > info->platformdata);
> >
> > diff -r 0f36c2eec2e1 tools/libxl/libxl_dm.c
> > --- a/tools/libxl/libxl_dm.c Thu Jul 28 15:40:54 2011 +0100
> > +++ b/tools/libxl/libxl_dm.c Fri Aug 05 18:13:37 2011 +0800
> > @@ -804,6 +804,7 @@
> > char *vm_path;
> > char **pass_stuff;
> > const char *dm;
> > + char *custom_bios;
> >
> > if (info->device_model_stubdomain) {
> > libxl_device_vfb vfb;
> > @@ -835,10 +836,13 @@
> > goto out;
> > }
> >
> > - path = libxl__sprintf(gc, "/local/domain/%d/hvmloader", info->domid);
> > - xs_mkdir(ctx->xsh, XBT_NULL, path);
> > - libxl__xs_write(gc, XBT_NULL, libxl__sprintf(gc, "%s/bios", path),
> > - "%s", libxl__domain_bios(gc, info));
> > + custom_bios = libxl__xs_read(gc, XBT_NULL, libxl__sprintf(gc,
> > "/local/domain/%d/hvmloader/bios", info->domid));
> > + if (!custom_bios) {
> > + path = libxl__sprintf(gc, "/local/domain/%d/hvmloader",
> > info->domid);
> > + xs_mkdir(ctx->xsh, XBT_NULL, path);
> > + libxl__xs_write(gc, XBT_NULL, libxl__sprintf(gc, "%s/bios", path),
> > + "%s", libxl__domain_bios(gc, info));
> > + }
> >
> > path = libxl__sprintf(gc, "/local/domain/0/device-model/%d",
> > info->domid);
> > xs_mkdir(ctx->xsh, XBT_NULL, path);
> > diff -r 0f36c2eec2e1 tools/libxl/xl_cmdimpl.c
> > --- a/tools/libxl/xl_cmdimpl.c Thu Jul 28 15:40:54 2011 +0100
> > +++ b/tools/libxl/xl_cmdimpl.c Fri Aug 05 18:13:37 2011 +0800
> > @@ -567,6 +567,10 @@
> > }
> > }
> >
> > + if (!xlu_cfg_get_string (config, "hvmbios", &buf)){
> > + c_info->hvmbios = strdup(buf);
> > + }
> > +
> > c_info->type = LIBXL_DOMAIN_TYPE_PV;
> > if (!xlu_cfg_get_string (config, "builder", &buf) &&
> > !strncmp(buf, "hvm", strlen(buf)))
> >
> >
> >
> >
> > Thanks,
> > Bei Guan
> >
> >
> >
> >
> > 2011/8/4 Ian Jackson <Ian.Jackson@xxxxxxxxxxxxx>
> >> Keir Fraser writes ("Re: [Xen-devel] [PATCH 2 of 3] Enable UEFI BIOS(OVMF)
> >> support in Xen-unstable HVM"):
> >>> I'd probably prefer to let xend grow obsolete quietly than add further
> >>> features. Noone's maintaining it to fix it up if new features break it. No
> >>> strong opinion on that though, it's up to the tools maintainers.
> >>
> >> I would agree. I would want a compelling reason (or strong community
> >> support) to accept new feature patches to xend.
> >>
> >> Ian.
> >
> >
>
>
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
Home