drm/meson: plane: add support for AFBC mode for OSD1 plane

This adds all the OSD configuration plumbing to support the AFBC decoders
path to display of the OSD1 plane.

The Amlogic GXM and G12A AFBC decoders are integrated very differently.

The Amlogic GXM has a direct output path to the OSD1 VIU pixel input,
because the GXM AFBC decoder seem to be a custom IP developed by Amlogic.

On the other side, the Amlogic G12A AFBC decoder seems to be an external
IP that emit pixels on an AXI master hooked to a "Mali Unpack" block
feeding the OSD1 VIU pixel input.
This uses a weird "0x1000000" internal HW physical address on both
sides to transfer the pixels.

For Amlogic GXM, the supported pixel formats are the same as the normal
linear OSD1 mode.

On the other side, Amlogic added support for all AFBC v1.2 formats for
the G12A AFBC integration.

For simplicity, we stick to the already supported formats for now.

Signed-off-by: Neil Armstrong <[e-mail]>
Reviewed-by: Kevin Hilman <[e-mail]>
Link: https://patchwork.freedesktop.org/patch/msgid/[e-mail]

drivers/gpu/drm/meson/meson_crtc.c

···
 	if (priv->viu.osd1_enabled && priv->viu.osd1_commit) {
 		writel_relaxed(priv->viu.osd1_ctrl_stat,
 				priv->io_base + _REG(VIU_OSD1_CTRL_STAT));
+		writel_relaxed(priv->viu.osd1_ctrl_stat2,
+				priv->io_base + _REG(VIU_OSD1_CTRL_STAT2));
 		writel_relaxed(priv->viu.osd1_blk0_cfg[0],
 				priv->io_base + _REG(VIU_OSD1_BLK0_CFG_W0));
 		writel_relaxed(priv->viu.osd1_blk0_cfg[1],

drivers/gpu/drm/meson/meson_drv.h

···
 		bool osd1_enabled;
 		bool osd1_interlace;
 		bool osd1_commit;
+		bool osd1_afbcd;
 		uint32_t osd1_ctrl_stat;
+		uint32_t osd1_ctrl_stat2;
 		uint32_t osd1_blk0_cfg[5];
+		uint32_t osd1_blk1_cfg4;
+		uint32_t osd1_blk2_cfg4;
 		uint32_t osd1_addr;
 		uint32_t osd1_stride;
 		uint32_t osd1_height;

drivers/gpu/drm/meson/meson_plane.c

···
 #include "meson_plane.h"
 #include "meson_registers.h"
 #include "meson_viu.h"
+#include "meson_osd_afbcd.h"
 
 /* OSD_SCI_WH_M1 */
 #define SCI_WH_M1_W(w)			FIELD_PREP(GENMASK(28, 16), w)
···
 						   false, true);
 }
 
+#define MESON_MOD_AFBC_VALID_BITS (AFBC_FORMAT_MOD_BLOCK_SIZE_16x16 |	\
+				   AFBC_FORMAT_MOD_BLOCK_SIZE_32x8 |	\
+				   AFBC_FORMAT_MOD_YTR |		\
+				   AFBC_FORMAT_MOD_SPARSE |		\
+				   AFBC_FORMAT_MOD_SPLIT)
+
 /* Takes a fixed 16.16 number and converts it to integer. */
 static inline int64_t fixed16_to_int(int64_t value)
 {
 	return value >> 16;
+}
+
+static u32 meson_g12a_afbcd_line_stride(struct meson_drm *priv)
+{
+	u32 line_stride = 0;
+
+	switch (priv->afbcd.format) {
+	case DRM_FORMAT_RGB565:
+		line_stride = ((priv->viu.osd1_width << 4) + 127) >> 7;
+		break;
+	case DRM_FORMAT_RGB888:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_ABGR8888:
+		line_stride = ((priv->viu.osd1_width << 5) + 127) >> 7;
+		break;
+	}
+
+	return ((line_stride + 1) >> 1) << 1;
 }
 
 static void meson_plane_atomic_update(struct drm_plane *plane,
···
 	 */
 	spin_lock_irqsave(&priv->drm->event_lock, flags);
 
+	/* Check if AFBC decoder is required for this buffer */
+	if ((meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXM) ||
+	     meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A)) &&
+	    fb->modifier & DRM_FORMAT_MOD_ARM_AFBC(MESON_MOD_AFBC_VALID_BITS))
+		priv->viu.osd1_afbcd = true;
+	else
+		priv->viu.osd1_afbcd = false;
+
 	/* Enable OSD and BLK0, set max global alpha */
 	priv->viu.osd1_ctrl_stat = OSD_ENABLE |
 				   (0xFF << OSD_GLOBAL_ALPHA_SHIFT) |
 				   OSD_BLK0_ENABLE;
 
+	priv->viu.osd1_ctrl_stat2 = readl(priv->io_base +
+					  _REG(VIU_OSD1_CTRL_STAT2));
+
 	canvas_id_osd1 = priv->canvas_id_osd1;
 
 	/* Set up BLK0 to point to the right canvas */
-	priv->viu.osd1_blk0_cfg[0] = ((canvas_id_osd1 << OSD_CANVAS_SEL) |
-				      OSD_ENDIANNESS_LE);
+	priv->viu.osd1_blk0_cfg[0] = canvas_id_osd1 << OSD_CANVAS_SEL;
+
+	if (priv->viu.osd1_afbcd) {
+		if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A)) {
+			/* This is the internal decoding memory address */
+			priv->viu.osd1_blk1_cfg4 = MESON_G12A_AFBCD_OUT_ADDR;
+			priv->viu.osd1_blk0_cfg[0] |= OSD_ENDIANNESS_BE;
+			priv->viu.osd1_ctrl_stat2 |= OSD_PENDING_STAT_CLEAN;
+			priv->viu.osd1_ctrl_stat |= VIU_OSD1_CFG_SYN_EN;
+		}
+
+		if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXM)) {
+			priv->viu.osd1_blk0_cfg[0] |= OSD_ENDIANNESS_LE;
+			priv->viu.osd1_ctrl_stat2 |= OSD_DPATH_MALI_AFBCD;
+		}
+	} else {
+		priv->viu.osd1_blk0_cfg[0] |= OSD_ENDIANNESS_LE;
+
+		if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXM))
+			priv->viu.osd1_ctrl_stat2 &= ~OSD_DPATH_MALI_AFBCD;
+	}
 
 	/* On GXBB, Use the old non-HDR RGB2YUV converter */
 	if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXBB))
 		priv->viu.osd1_blk0_cfg[0] |= OSD_OUTPUT_COLOR_RGB;
 
+	if (priv->viu.osd1_afbcd &&
+	    meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A)) {
+		priv->viu.osd1_blk0_cfg[0] |= OSD_MALI_SRC_EN |
+			priv->afbcd.ops->fmt_to_blk_mode(fb->modifier,
+							  fb->format->format);
+	} else {
+		switch (fb->format->format) {
+		case DRM_FORMAT_XRGB8888:
+		case DRM_FORMAT_ARGB8888:
+			priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_32 |
+						OSD_COLOR_MATRIX_32_ARGB;
+			break;
+		case DRM_FORMAT_XBGR8888:
+		case DRM_FORMAT_ABGR8888:
+			priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_32 |
+						OSD_COLOR_MATRIX_32_ABGR;
+			break;
+		case DRM_FORMAT_RGB888:
+			priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_24 |
+						OSD_COLOR_MATRIX_24_RGB;
+			break;
+		case DRM_FORMAT_RGB565:
+			priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_16 |
+						OSD_COLOR_MATRIX_16_RGB565;
+			break;
+		};
+	}
+
 	switch (fb->format->format) {
 	case DRM_FORMAT_XRGB8888:
-		/* For XRGB, replace the pixel's alpha by 0xFF */
-		writel_bits_relaxed(OSD_REPLACE_EN, OSD_REPLACE_EN,
-				    priv->io_base + _REG(VIU_OSD1_CTRL_STAT2));
-		priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_32 |
-					      OSD_COLOR_MATRIX_32_ARGB;
-		break;
 	case DRM_FORMAT_XBGR8888:
 		/* For XRGB, replace the pixel's alpha by 0xFF */
-		writel_bits_relaxed(OSD_REPLACE_EN, OSD_REPLACE_EN,
-				    priv->io_base + _REG(VIU_OSD1_CTRL_STAT2));
-		priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_32 |
-					      OSD_COLOR_MATRIX_32_ABGR;
+		priv->viu.osd1_ctrl_stat2 |= OSD_REPLACE_EN;
 		break;
 	case DRM_FORMAT_ARGB8888:
-		/* For ARGB, use the pixel's alpha */
-		writel_bits_relaxed(OSD_REPLACE_EN, 0,
-				    priv->io_base + _REG(VIU_OSD1_CTRL_STAT2));
-		priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_32 |
-					      OSD_COLOR_MATRIX_32_ARGB;
-		break;
 	case DRM_FORMAT_ABGR8888:
 		/* For ARGB, use the pixel's alpha */
-		writel_bits_relaxed(OSD_REPLACE_EN, 0,
-				    priv->io_base + _REG(VIU_OSD1_CTRL_STAT2));
-		priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_32 |
-					      OSD_COLOR_MATRIX_32_ABGR;
-		break;
-	case DRM_FORMAT_RGB888:
-		priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_24 |
-					      OSD_COLOR_MATRIX_24_RGB;
-		break;
-	case DRM_FORMAT_RGB565:
-		priv->viu.osd1_blk0_cfg[0] |= OSD_BLK_MODE_16 |
-					      OSD_COLOR_MATRIX_16_RGB565;
+		priv->viu.osd1_ctrl_stat2 &= ~OSD_REPLACE_EN;
 		break;
 	};
 
···
 	priv->viu.osd1_height = fb->height;
 	priv->viu.osd1_width = fb->width;
 
+	if (priv->viu.osd1_afbcd) {
+		priv->afbcd.modifier = fb->modifier;
+		priv->afbcd.format = fb->format->format;
+
+		/* Calculate decoder write stride */
+		if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A))
+			priv->viu.osd1_blk2_cfg4 =
+				meson_g12a_afbcd_line_stride(priv);
+	}
+
 	if (!meson_plane->enabled) {
 		/* Reset OSD1 before enabling it on GXL+ SoCs */
 		if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXM) ||
···
 	struct meson_plane *meson_plane = to_meson_plane(plane);
 	struct meson_drm *priv = meson_plane->priv;
 
+	if (priv->afbcd.ops) {
+		priv->afbcd.ops->reset(priv);
+		priv->afbcd.ops->disable(priv);
+	}
+
 	/* Disable OSD1 */
 	if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A))
 		writel_bits_relaxed(VIU_OSD1_POSTBLD_SRC_OSD1, 0,
···
 	.prepare_fb	= drm_gem_fb_prepare_fb,
 };
 
+static bool meson_plane_format_mod_supported(struct drm_plane *plane,
+					     u32 format, u64 modifier)
+{
+	struct meson_plane *meson_plane = to_meson_plane(plane);
+	struct meson_drm *priv = meson_plane->priv;
+	int i;
+
+	if (modifier == DRM_FORMAT_MOD_INVALID)
+		return false;
+
+	if (modifier == DRM_FORMAT_MOD_LINEAR)
+		return true;
+
+	if (!meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXM) &&
+	    !meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A))
+		return false;
+
+	if (modifier & ~DRM_FORMAT_MOD_ARM_AFBC(MESON_MOD_AFBC_VALID_BITS))
+		return false;
+
+	for (i = 0 ; i < plane->modifier_count ; ++i)
+		if (plane->modifiers[i] == modifier)
+			break;
+
+	if (i == plane->modifier_count) {
+		DRM_DEBUG_KMS("Unsupported modifier\n");
+		return false;
+	}
+
+	if (priv->afbcd.ops && priv->afbcd.ops->supported_fmt)
+		return priv->afbcd.ops->supported_fmt(modifier, format);
+
+	DRM_DEBUG_KMS("AFBC Unsupported\n");
+	return false;
+}
+
 static const struct drm_plane_funcs meson_plane_funcs = {
 	.update_plane		= drm_atomic_helper_update_plane,
 	.disable_plane		= drm_atomic_helper_disable_plane,
···
 	.reset			= drm_atomic_helper_plane_reset,
 	.atomic_duplicate_state = drm_atomic_helper_plane_duplicate_state,
 	.atomic_destroy_state	= drm_atomic_helper_plane_destroy_state,
+	.format_mod_supported   = meson_plane_format_mod_supported,
 };
 
 static const uint32_t supported_drm_formats[] = {
···
 	DRM_FORMAT_RGB565,
 };
 
+static const uint64_t format_modifiers_afbc_gxm[] = {
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_16x16 |
+				AFBC_FORMAT_MOD_SPARSE |
+				AFBC_FORMAT_MOD_YTR),
+	/* SPLIT mandates SPARSE, RGB modes mandates YTR */
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_16x16 |
+				AFBC_FORMAT_MOD_YTR |
+				AFBC_FORMAT_MOD_SPARSE |
+				AFBC_FORMAT_MOD_SPLIT),
+	DRM_FORMAT_MOD_LINEAR,
+	DRM_FORMAT_MOD_INVALID,
+};
+
+static const uint64_t format_modifiers_afbc_g12a[] = {
+	/*
+	 * - TOFIX Support AFBC modifiers for YUV formats (16x16 + TILED)
+	 * - SPLIT is mandatory for performances reasons when in 16x16
+	 *   block size
+	 * - 32x8 block size + SPLIT is mandatory with 4K frame size
+	 *   for performances reasons
+	 */
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_16x16 |
+				AFBC_FORMAT_MOD_SPARSE |
+				AFBC_FORMAT_MOD_SPLIT),
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_16x16 |
+				AFBC_FORMAT_MOD_YTR |
+				AFBC_FORMAT_MOD_SPARSE |
+				AFBC_FORMAT_MOD_SPLIT),
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_32x8 |
+				AFBC_FORMAT_MOD_SPARSE),
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_32x8 |
+				AFBC_FORMAT_MOD_YTR |
+				AFBC_FORMAT_MOD_SPARSE),
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_32x8 |
+				AFBC_FORMAT_MOD_SPARSE |
+				AFBC_FORMAT_MOD_SPLIT),
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_32x8 |
+				AFBC_FORMAT_MOD_YTR |
+				AFBC_FORMAT_MOD_SPARSE |
+				AFBC_FORMAT_MOD_SPLIT),
+	DRM_FORMAT_MOD_LINEAR,
+	DRM_FORMAT_MOD_INVALID,
+};
+
+static const uint64_t format_modifiers_default[] = {
+	DRM_FORMAT_MOD_LINEAR,
+	DRM_FORMAT_MOD_INVALID,
+};
+
 int meson_plane_create(struct meson_drm *priv)
 {
 	struct meson_plane *meson_plane;
 	struct drm_plane *plane;
+	const uint64_t *format_modifiers = format_modifiers_default;
 
 	meson_plane = devm_kzalloc(priv->drm->dev, sizeof(*meson_plane),
 				   GFP_KERNEL);
···
 	meson_plane->priv = priv;
 	plane = &meson_plane->base;
 
+	if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXM))
+		format_modifiers = format_modifiers_afbc_gxm;
+	else if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A))
+		format_modifiers = format_modifiers_afbc_g12a;
+
 	drm_universal_plane_init(priv->drm, plane, 0xFF,
 				 &meson_plane_funcs,
 				 supported_drm_formats,
 				 ARRAY_SIZE(supported_drm_formats),
-				 NULL,
+				 format_modifiers,
 				 DRM_PLANE_TYPE_PRIMARY, "meson_primary_plane");
 
 	drm_plane_helper_add(plane, &meson_plane_helper_funcs);