Re: [PATCH RFC v2] drm/verisilicon: Switch to drm_fb_dma_get_addr() for framebuffer addresses

[Icenowy Zheng]

在 2026-08-07五的 19:46 +0800，Chen-Yu Tsai写道：
> The verisilicon driver has a custom framebuffer address calculating
> helper that the common drm_fb_dma_get_addr() can substitute.
> 
> Differences from drm_fb_dma_get_addr():
> 
> - Uses drm_format_info_min_pitch() to calculate the horizontal
> offset;
>   however the driver does not support any of the blocked formats, so

Technically this DC, advertised as part of the "Vivante" product line
(considering Vivante Corporation is acquired by VeriSilicon), seems to
support DRM_FORMAT_MOD_VIVANTE_SUPER_TILED modifier (the TH1520
documentation says the DC supports
`SuperTileX8x8/SuperTileX8x4/SuperTileY4x8`, although I think
DRM_FORMAT_MOD_VIVANTE_SUPER_TILED is just one of these tiling).

However, as my accessible SoCs with such DC have no GC-series 3D GPUs
(TH1520 does have a 2D-only GC620 GPU), I think it's quite difficult to
get this piece of thing right and it should be low-priority.

>   this just ends up being the same as in drm_fb_dma_get_addr():
>   "cpp[plane] * y"
> 
> - Uses clipped source coordinates instead of non-clipped coordinates
>   as in drm_fb_dma_get_addr();
> 
>   For the primary plane this doesn't matter, since the primary plane
>   must match the output, i.e. it cannot be clipped. Also this driver
>   doesn't support scaling.
> 
>   For the cursor plane this seems wrong, as the clipping seems to be
>   done by the hardware, and thus the buffer address should be
> unclipped.

Yes this is right and the current state of the cursor plane is broken.

However another error compensates this error so I didn't catch it when
developing -- the [XY]_OFF fields aren't properly written because I
forgot to shift the values for them (and then the value gets masked by
regmap_update_bits()), which prevents the HW clipping to happen, and
the normal-state arrow cursor happens to have no non-transparent pixels
before the hotspot. When testing with `X -retro`, the retro X cursor
gets quite glitchy with the current code; and when this patch is
applied w/o the offset fix, the cursor isn't clipped at all.

Both errors deserve fixes, I will then send the fix for the offset
writing problem.

Thanks,
Icenowy

> 
> As such, it should be fine to use the common helper and drop the
> custom
> code.
> 
> Signed-off-by: Chen-Yu Tsai <wenst@xxxxxxxxxxxx>
> ---
> Changes since v1:
> - Fixed compile issues
> 
> This is only compile tested. I do not have the hardware.
> ---
>  drivers/gpu/drm/verisilicon/vs_cursor_plane.c |  4 +++-
>  drivers/gpu/drm/verisilicon/vs_plane.c        | 20 -----------------
> --
>  .../gpu/drm/verisilicon/vs_primary_plane.c    |  7 ++++++-
>  3 files changed, 9 insertions(+), 22 deletions(-)
> 
> diff --git a/drivers/gpu/drm/verisilicon/vs_cursor_plane.c
> b/drivers/gpu/drm/verisilicon/vs_cursor_plane.c
> index fa4f601dd0c8..59778433ae84 100644
> --- a/drivers/gpu/drm/verisilicon/vs_cursor_plane.c
> +++ b/drivers/gpu/drm/verisilicon/vs_cursor_plane.c
> @@ -12,6 +12,7 @@
>  #include <drm/drm_atomic.h>
>  #include <drm/drm_atomic_helper.h>
>  #include <drm/drm_crtc.h>
> +#include <drm/drm_fb_dma_helper.h>
>  #include <drm/drm_fourcc.h>
>  #include <drm/drm_framebuffer.h>
>  #include <drm/drm_gem_atomic_helper.h>
> @@ -176,7 +177,8 @@ static void vs_cursor_plane_atomic_update(struct
> drm_plane *plane,
>  		break;
>  	}
>  
> -	dma_addr = vs_fb_get_dma_addr(fb, &state->src);
> +	/* hardware handles clipping as seen below */
> +	dma_addr = drm_fb_dma_get_gem_addr(fb, state, 0);
>  
>  	regmap_write(dc->regs, VSDC_CURSOR_ADDRESS(output),
>  		     lower_32_bits(dma_addr));
> diff --git a/drivers/gpu/drm/verisilicon/vs_plane.c
> b/drivers/gpu/drm/verisilicon/vs_plane.c
> index d81f7b8f4c65..38b8b536eccb 100644
> --- a/drivers/gpu/drm/verisilicon/vs_plane.c
> +++ b/drivers/gpu/drm/verisilicon/vs_plane.c
> @@ -107,26 +107,6 @@ int drm_format_to_vs_format(u32 drm_format,
> struct vs_format *vs_format)
>  	return 0;
>  }
>  
> -dma_addr_t vs_fb_get_dma_addr(struct drm_framebuffer *fb,
> -			      const struct drm_rect *src_rect)
> -{
> -	struct drm_gem_dma_object *gem;
> -	dma_addr_t dma_addr;
> -
> -	/* Get the physical address of the buffer in memory */
> -	gem = drm_fb_dma_get_gem_obj(fb, 0);
> -
> -	/* Compute the start of the displayed memory */
> -	dma_addr = gem->dma_addr + fb->offsets[0];
> -
> -	/* Fixup framebuffer address for src coordinates */
> -	dma_addr += drm_format_info_min_pitch(fb->format, 0,
> -					      src_rect->x1 >> 16);
> -	dma_addr += (src_rect->y1 >> 16) * fb->pitches[0];
> -
> -	return dma_addr;
> -}
> -
>  struct drm_plane_state *vs_plane_duplicate_state(struct drm_plane
> *plane)
>  {
>  	struct vs_plane_state *vs_state, *vs_state_old;
> diff --git a/drivers/gpu/drm/verisilicon/vs_primary_plane.c
> b/drivers/gpu/drm/verisilicon/vs_primary_plane.c
> index 1f2be41ae496..2750016a7f2c 100644
> --- a/drivers/gpu/drm/verisilicon/vs_primary_plane.c
> +++ b/drivers/gpu/drm/verisilicon/vs_primary_plane.c
> @@ -8,6 +8,7 @@
>  #include <drm/drm_atomic.h>
>  #include <drm/drm_atomic_helper.h>
>  #include <drm/drm_crtc.h>
> +#include <drm/drm_fb_dma_helper.h>
>  #include <drm/drm_fourcc.h>
>  #include <drm/drm_framebuffer.h>
>  #include <drm/drm_gem_atomic_helper.h>
> @@ -126,7 +127,11 @@ static void
> vs_primary_plane_atomic_update(struct drm_plane *plane,
>  			   VSDC_FB_CONFIG_UV_SWIZZLE_EN,
>  			   vs_state->format.uv_swizzle);
>  
> -	dma_addr = vs_fb_get_dma_addr(fb, &state->src);
> +	/*
> +	 * Primary plane cannot be moved, no clipping is involved,
> +	 * so the non-clipped framebuffer address can be used.
> +	 */
> +	dma_addr = drm_fb_dma_get_gem_addr(fb, state, 0);
>  
>  	regmap_write(dc->regs, VSDC_FB_ADDRESS(output),
>  		     lower_32_bits(dma_addr));