Re: [PATCH] OvmfPkg/Gop: clear the screen to black in SetMode()

[Laszlo Ersek <lersek@redhat.com>]

Hi Ray,

On 03/13/18 11:05, Ruiyu Ni wrote:
> Today's implementation forgot to clear the screen to black in
> SetMode(). It causes SCT SetMode() test fails.
> 
> The patch adds the clear screen operation in SetMode() to fix
> the SCT failure.
> 
> Contributed-under: TianoCore Contribution Agreement 1.1
> Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
> Cc: Jordan Justen <jordan.l.justen@intel.com>
> Cc: Laszlo Ersek <lersek@redhat.com>
> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
> ---
>  OvmfPkg/QemuVideoDxe/Gop.c | 23 ++++++++++++++++++-----
>  1 file changed, 18 insertions(+), 5 deletions(-)
> 
> diff --git a/OvmfPkg/QemuVideoDxe/Gop.c b/OvmfPkg/QemuVideoDxe/Gop.c
> index 512fd27acb..f573f7011a 100644
> --- a/OvmfPkg/QemuVideoDxe/Gop.c
> +++ b/OvmfPkg/QemuVideoDxe/Gop.c
> @@ -1,7 +1,7 @@
>  /** @file
>    Graphics Output Protocol functions for the QEMU video controller.
>  
> -  Copyright (c) 2007 - 2017, Intel Corporation. All rights reserved.<BR>
> +  Copyright (c) 2007 - 2018, Intel Corporation. All rights reserved.<BR>
>  
>    This program and the accompanying materials
>    are licensed and made available under the terms and conditions of the BSD License
> @@ -196,9 +196,10 @@ Routine Description:
>  
>  --*/
>  {
> -  QEMU_VIDEO_PRIVATE_DATA    *Private;
> -  QEMU_VIDEO_MODE_DATA       *ModeData;
> -  RETURN_STATUS              Status;
> +  QEMU_VIDEO_PRIVATE_DATA       *Private;
> +  QEMU_VIDEO_MODE_DATA          *ModeData;
> +  RETURN_STATUS                 Status;
> +  EFI_GRAPHICS_OUTPUT_BLT_PIXEL Black;
>  
>    Private = QEMU_VIDEO_PRIVATE_DATA_FROM_GRAPHICS_OUTPUT_THIS (This);
>  
> @@ -271,7 +272,19 @@ Routine Description:
>    }
>    ASSERT (Status == RETURN_SUCCESS);
>  
> -  return EFI_SUCCESS;
> +  //
> +  // Per UEFI Spec, need to clear the visible portions of the output display to black.
> +  //
> +  ZeroMem (&Black, sizeof (Black));
> +  return FrameBufferBlt (
> +           Private->FrameBufferBltConfigure,
> +           &Black,
> +           EfiBltVideoFill,
> +           0, 0,
> +           0, 0,
> +           This->Mode->Info->HorizontalResolution, This->Mode->Info->VerticalResolution,
> +           0
> +           );
>  }
>  
>  EFI_STATUS
> 

Thanks for the patch!

What would you think of the following update: FrameBufferBlt() is
documented to return RETURN_INVALID_PARAMETER for "Invalid parameter
were passed in", and RETURN_SUCCESS otherwise ("The Blt operation was
performed successfully"). I suggest that we keep the "return
EFI_SUCCESS" in the end, just ASSERT() that FrameBufferBlt() succeeds.

(Even if there was a third possibility, i.e. for FrameBufferBlt() to
fail for a different reason, I think that should still not report that
SetMode() failed.)

If you agree with the idea, please update the patch before pushing it.
If you disagree with it, I don't insist.

Reviewed-by: Laszlo Ersek <lersek@redhat.com>

Thanks
Laszlo