Re: [Qemu-devel] [PATCH 27/51] piix: fix 32bit pci hole

[Laszlo Ersek]

On 02/21/14 09:17, Michael Roth wrote:
> From: Gerd Hoffmann <address@hidden>
> 
> Make the 32bit pci hole start at end of ram, so all possible address
> space is covered.
> 
> We used to try and make addresses aligned so they are easier to cover
> with MTRRs, but since they are cosmetic on KVM, this is probably not
> worth worrying about.
> Of course the firmware can use less than that.  Leaving space unused is
> no problem, mapping pci bars outside the hole causes problems though.
> 
> Signed-off-by: Gerd Hoffmann <address@hidden>
> Signed-off-by: Laszlo Ersek <address@hidden>
> Reviewed-by: Michael S. Tsirkin <address@hidden>
> Signed-off-by: Michael S. Tsirkin <address@hidden>
> (cherry picked from commit ddaaefb4dd427d6d2e41c1cfbe0cd8d8e8d6aad9)
> 
> Signed-off-by: Michael Roth <address@hidden>
> ---
>  hw/i386/pc_piix.c    |    1 +
>  hw/pci-host/piix.c   |   11 ++---------
>  include/hw/i386/pc.h |    1 +
>  3 files changed, 4 insertions(+), 9 deletions(-)
> 
> diff --git a/hw/i386/pc_piix.c b/hw/i386/pc_piix.c
> index 29b47d4..cc9b273 100644
> --- a/hw/i386/pc_piix.c
> +++ b/hw/i386/pc_piix.c
> @@ -149,6 +149,7 @@ static void pc_init1(QEMUMachineInitArgs *args,
>      if (pci_enabled) {
>          pci_bus = i440fx_init(&i440fx_state, &piix3_devfn, &isa_bus, gsi,
>                                system_memory, system_io, args->ram_size,
> +                              below_4g_mem_size,
>                                above_4g_mem_size,
>                                pci_memory, ram_memory);
>      } else {
> diff --git a/hw/pci-host/piix.c b/hw/pci-host/piix.c
> index 63be7f6..4229d09 100644
> --- a/hw/pci-host/piix.c
> +++ b/hw/pci-host/piix.c
> @@ -311,6 +311,7 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state,
>                      MemoryRegion *address_space_mem,
>                      MemoryRegion *address_space_io,
>                      ram_addr_t ram_size,
> +                    ram_addr_t below_4g_mem_size,
>                      ram_addr_t above_4g_mem_size,
>                      MemoryRegion *pci_address_space,
>                      MemoryRegion *ram_memory)
> @@ -340,15 +341,7 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state,
>      f->ram_memory = ram_memory;
>  
>      i440fx = I440FX_PCI_HOST_BRIDGE(dev);
> -    /* Set PCI window size the way seabios has always done it. */
> -    /* Power of 2 so bios can cover it with a single MTRR */
> -    if (ram_size <= 0x80000000) {
> -        i440fx->pci_info.w32.begin = 0x80000000;
> -    } else if (ram_size <= 0xc0000000) {
> -        i440fx->pci_info.w32.begin = 0xc0000000;
> -    } else {
> -        i440fx->pci_info.w32.begin = 0xe0000000;
> -    }
> +    i440fx->pci_info.w32.begin = below_4g_mem_size;
>  
>      /* setup pci memory mapping */
>      pc_pci_as_mapping_init(OBJECT(f), f->system_memory,
> diff --git a/include/hw/i386/pc.h b/include/hw/i386/pc.h
> index 8ea1a98..2a4a094 100644
> --- a/include/hw/i386/pc.h
> +++ b/include/hw/i386/pc.h
> @@ -179,6 +179,7 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state, int 
> *piix_devfn,
>                      MemoryRegion *address_space_mem,
>                      MemoryRegion *address_space_io,
>                      ram_addr_t ram_size,
> +                    ram_addr_t below_4g_mem_size,
>                      ram_addr_t above_4g_mem_size,
>                      MemoryRegion *pci_memory,
>                      MemoryRegion *ram_memory);
> 

Env:
- machine type: pc-i440fx-1.7
- OVMF guest firmware with ACPI download patch
- guest OS: Fedora 19

Test results:

----------------+-----------------------------------+-------------------
 guest RAM size |                v1.7.0             | v1.7.0 + this set
----------------+-----------------------------------+-------------------
        1024 MB |                 PASS              |       PASS
----------------+-----------------------------------+-------------------
        2560 MB | in Linux: black screen then crash |      *PASS*
----------------+-----------------------------------+-------------------

I repeated the test with a 2.5GB Windows 2012 R2 guest as well (the
problem in v1.7.0 affects it the same way), and the patch fixes the
problem for this guest as well.

Much appreciated!

Tested-by: Laszlo Ersek <address@hidden>

Laszlo