[Qemu-devel] [PATCH for-2.0] hw/i386: fix acpi tables generation for big endian machines

[Qemu-devel] [PATCH for-2.0] hw/i386: fix acpi tables generation for big endian machines

[Marcel Apfelbaum]

The acpi tables are not corrected loaded into guest's memory
for big-endian hosts because of a linker-loader field
swapped incorrectly. Fixed that.

Signed-off-by: Marcel Apfelbaum <marcel.a@redhat.com>
---
Note that the acpi test still fails because of
an iasl issue, I will send a patch separately
that disables disassembly validation for big
endian machines until we'll have an iasl fix. 

 hw/i386/bios-linker-loader.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/hw/i386/bios-linker-loader.c b/hw/i386/bios-linker-loader.c
index aa56184..0e1cfb7 100644
--- a/hw/i386/bios-linker-loader.c
+++ b/hw/i386/bios-linker-loader.c
@@ -109,9 +109,8 @@ void bios_linker_loader_alloc(GArray *linker,
     strncpy(entry.alloc.file, file, sizeof entry.alloc.file - 1);
     entry.command = cpu_to_le32(BIOS_LINKER_LOADER_COMMAND_ALLOCATE);
     entry.alloc.align = cpu_to_le32(alloc_align);
-    entry.alloc.zone = cpu_to_le32(alloc_fseg ?
-                                    BIOS_LINKER_LOADER_ALLOC_ZONE_FSEG :
-                                    BIOS_LINKER_LOADER_ALLOC_ZONE_HIGH);
+    entry.alloc.zone = alloc_fseg ? BIOS_LINKER_LOADER_ALLOC_ZONE_FSEG :
+                                    BIOS_LINKER_LOADER_ALLOC_ZONE_HIGH;
 
     /* Alloc entries must come first, so prepend them */
     g_array_prepend_vals(linker, &entry, sizeof entry);
-- 
1.8.3.1

Re: [Qemu-devel] [PATCH for-2.0] hw/i386: fix acpi tables generation for big endian machines

[Michael S. Tsirkin]

On Thu, Mar 20, 2014 at 06:40:50PM +0200, Marcel Apfelbaum wrote:
> The acpi tables are not corrected loaded into guest's memory
> for big-endian hosts because of a linker-loader field
> swapped incorrectly. Fixed that.
> 
> Signed-off-by: Marcel Apfelbaum <marcel.a@redhat.com>

Applied, thanks!

> ---
> Note that the acpi test still fails because of
> an iasl issue, I will send a patch separately
> that disables disassembly validation for big
> endian machines until we'll have an iasl fix. 
> 
>  hw/i386/bios-linker-loader.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/hw/i386/bios-linker-loader.c b/hw/i386/bios-linker-loader.c
> index aa56184..0e1cfb7 100644
> --- a/hw/i386/bios-linker-loader.c
> +++ b/hw/i386/bios-linker-loader.c
> @@ -109,9 +109,8 @@ void bios_linker_loader_alloc(GArray *linker,
>      strncpy(entry.alloc.file, file, sizeof entry.alloc.file - 1);
>      entry.command = cpu_to_le32(BIOS_LINKER_LOADER_COMMAND_ALLOCATE);
>      entry.alloc.align = cpu_to_le32(alloc_align);
> -    entry.alloc.zone = cpu_to_le32(alloc_fseg ?
> -                                    BIOS_LINKER_LOADER_ALLOC_ZONE_FSEG :
> -                                    BIOS_LINKER_LOADER_ALLOC_ZONE_HIGH);
> +    entry.alloc.zone = alloc_fseg ? BIOS_LINKER_LOADER_ALLOC_ZONE_FSEG :
> +                                    BIOS_LINKER_LOADER_ALLOC_ZONE_HIGH;
>  
>      /* Alloc entries must come first, so prepend them */
>      g_array_prepend_vals(linker, &entry, sizeof entry);
> -- 
> 1.8.3.1