}
}
-static uint32_t r600_translate_colorswap(enum pipe_format format)
-{
- switch (format) {
- /* 8-bit buffers. */
- case PIPE_FORMAT_A8_UNORM:
- case PIPE_FORMAT_A8_SNORM:
- case PIPE_FORMAT_A8_UINT:
- case PIPE_FORMAT_A8_SINT:
- case PIPE_FORMAT_A16_UNORM:
- case PIPE_FORMAT_A16_SNORM:
- case PIPE_FORMAT_A16_UINT:
- case PIPE_FORMAT_A16_SINT:
- case PIPE_FORMAT_A16_FLOAT:
- case PIPE_FORMAT_A32_UINT:
- case PIPE_FORMAT_A32_SINT:
- case PIPE_FORMAT_A32_FLOAT:
- case PIPE_FORMAT_R4A4_UNORM:
- return V_0280A0_SWAP_ALT_REV;
- case PIPE_FORMAT_I8_UNORM:
- case PIPE_FORMAT_I8_SNORM:
- case PIPE_FORMAT_I8_UINT:
- case PIPE_FORMAT_I8_SINT:
- case PIPE_FORMAT_L8_UNORM:
- case PIPE_FORMAT_L8_SNORM:
- case PIPE_FORMAT_L8_UINT:
- case PIPE_FORMAT_L8_SINT:
- case PIPE_FORMAT_L8_SRGB:
- case PIPE_FORMAT_L16_UNORM:
- case PIPE_FORMAT_L16_SNORM:
- case PIPE_FORMAT_L16_UINT:
- case PIPE_FORMAT_L16_SINT:
- case PIPE_FORMAT_L16_FLOAT:
- case PIPE_FORMAT_L32_UINT:
- case PIPE_FORMAT_L32_SINT:
- case PIPE_FORMAT_L32_FLOAT:
- case PIPE_FORMAT_I16_UNORM:
- case PIPE_FORMAT_I16_SNORM:
- case PIPE_FORMAT_I16_UINT:
- case PIPE_FORMAT_I16_SINT:
- case PIPE_FORMAT_I16_FLOAT:
- case PIPE_FORMAT_I32_UINT:
- case PIPE_FORMAT_I32_SINT:
- case PIPE_FORMAT_I32_FLOAT:
- case PIPE_FORMAT_R8_UNORM:
- case PIPE_FORMAT_R8_SNORM:
- case PIPE_FORMAT_R8_UINT:
- case PIPE_FORMAT_R8_SINT:
- return V_0280A0_SWAP_STD;
-
- case PIPE_FORMAT_L4A4_UNORM:
- case PIPE_FORMAT_A4R4_UNORM:
- return V_0280A0_SWAP_ALT;
-
- /* 16-bit buffers. */
- case PIPE_FORMAT_B5G6R5_UNORM:
- return V_0280A0_SWAP_STD_REV;
-
- case PIPE_FORMAT_B5G5R5A1_UNORM:
- case PIPE_FORMAT_B5G5R5X1_UNORM:
- return V_0280A0_SWAP_ALT;
-
- case PIPE_FORMAT_B4G4R4A4_UNORM:
- case PIPE_FORMAT_B4G4R4X4_UNORM:
- return V_0280A0_SWAP_ALT;
-
- case PIPE_FORMAT_Z16_UNORM:
- return V_0280A0_SWAP_STD;
-
- case PIPE_FORMAT_L8A8_UNORM:
- case PIPE_FORMAT_L8A8_SNORM:
- case PIPE_FORMAT_L8A8_UINT:
- case PIPE_FORMAT_L8A8_SINT:
- case PIPE_FORMAT_L8A8_SRGB:
- case PIPE_FORMAT_L16A16_UNORM:
- case PIPE_FORMAT_L16A16_SNORM:
- case PIPE_FORMAT_L16A16_UINT:
- case PIPE_FORMAT_L16A16_SINT:
- case PIPE_FORMAT_L16A16_FLOAT:
- case PIPE_FORMAT_L32A32_UINT:
- case PIPE_FORMAT_L32A32_SINT:
- case PIPE_FORMAT_L32A32_FLOAT:
- case PIPE_FORMAT_R8A8_UNORM:
- case PIPE_FORMAT_R8A8_SNORM:
- case PIPE_FORMAT_R8A8_UINT:
- case PIPE_FORMAT_R8A8_SINT:
- case PIPE_FORMAT_R16A16_UNORM:
- case PIPE_FORMAT_R16A16_SNORM:
- case PIPE_FORMAT_R16A16_UINT:
- case PIPE_FORMAT_R16A16_SINT:
- case PIPE_FORMAT_R16A16_FLOAT:
- case PIPE_FORMAT_R32A32_UINT:
- case PIPE_FORMAT_R32A32_SINT:
- case PIPE_FORMAT_R32A32_FLOAT:
- return V_0280A0_SWAP_ALT;
- case PIPE_FORMAT_R8G8_UNORM:
- case PIPE_FORMAT_R8G8_SNORM:
- case PIPE_FORMAT_R8G8_UINT:
- case PIPE_FORMAT_R8G8_SINT:
- return V_0280A0_SWAP_STD;
-
- case PIPE_FORMAT_R16_UNORM:
- case PIPE_FORMAT_R16_SNORM:
- case PIPE_FORMAT_R16_UINT:
- case PIPE_FORMAT_R16_SINT:
- case PIPE_FORMAT_R16_FLOAT:
- return V_0280A0_SWAP_STD;
-
- /* 32-bit buffers. */
-
- case PIPE_FORMAT_A8B8G8R8_SRGB:
- return V_0280A0_SWAP_STD_REV;
- case PIPE_FORMAT_B8G8R8A8_SRGB:
- return V_0280A0_SWAP_ALT;
-
- case PIPE_FORMAT_B8G8R8A8_UNORM:
- case PIPE_FORMAT_B8G8R8X8_UNORM:
- return V_0280A0_SWAP_ALT;
-
- case PIPE_FORMAT_A8R8G8B8_UNORM:
- case PIPE_FORMAT_X8R8G8B8_UNORM:
- return V_0280A0_SWAP_ALT_REV;
- case PIPE_FORMAT_R8G8B8A8_SNORM:
- case PIPE_FORMAT_R8G8B8A8_UNORM:
- case PIPE_FORMAT_R8G8B8X8_UNORM:
- case PIPE_FORMAT_R8G8B8X8_SNORM:
- case PIPE_FORMAT_R8G8B8X8_SRGB:
- case PIPE_FORMAT_R8G8B8X8_UINT:
- case PIPE_FORMAT_R8G8B8X8_SINT:
- case PIPE_FORMAT_R8G8B8A8_SINT:
- case PIPE_FORMAT_R8G8B8A8_UINT:
- return V_0280A0_SWAP_STD;
-
- case PIPE_FORMAT_A8B8G8R8_UNORM:
- case PIPE_FORMAT_X8B8G8R8_UNORM:
- /* case PIPE_FORMAT_R8SG8SB8UX8U_NORM: */
- return V_0280A0_SWAP_STD_REV;
-
- case PIPE_FORMAT_Z24X8_UNORM:
- case PIPE_FORMAT_Z24_UNORM_S8_UINT:
- return V_0280A0_SWAP_STD;
-
- case PIPE_FORMAT_R10G10B10A2_UNORM:
- case PIPE_FORMAT_R10G10B10X2_SNORM:
- case PIPE_FORMAT_R10SG10SB10SA2U_NORM:
- return V_0280A0_SWAP_STD;
-
- case PIPE_FORMAT_B10G10R10A2_UNORM:
- case PIPE_FORMAT_B10G10R10A2_UINT:
- case PIPE_FORMAT_B10G10R10X2_UNORM:
- return V_0280A0_SWAP_ALT;
-
- case PIPE_FORMAT_R11G11B10_FLOAT:
- case PIPE_FORMAT_R16G16_UNORM:
- case PIPE_FORMAT_R16G16_SNORM:
- case PIPE_FORMAT_R16G16_FLOAT:
- case PIPE_FORMAT_R16G16_UINT:
- case PIPE_FORMAT_R16G16_SINT:
- case PIPE_FORMAT_R32_UINT:
- case PIPE_FORMAT_R32_SINT:
- case PIPE_FORMAT_R32_FLOAT:
- case PIPE_FORMAT_Z32_FLOAT:
- return V_0280A0_SWAP_STD;
-
- /* 64-bit buffers. */
- case PIPE_FORMAT_R32G32_FLOAT:
- case PIPE_FORMAT_R32G32_UINT:
- case PIPE_FORMAT_R32G32_SINT:
- case PIPE_FORMAT_R16G16B16A16_UNORM:
- case PIPE_FORMAT_R16G16B16A16_SNORM:
- case PIPE_FORMAT_R16G16B16A16_UINT:
- case PIPE_FORMAT_R16G16B16A16_SINT:
- case PIPE_FORMAT_R16G16B16A16_FLOAT:
- case PIPE_FORMAT_R16G16B16X16_UNORM:
- case PIPE_FORMAT_R16G16B16X16_SNORM:
- case PIPE_FORMAT_R16G16B16X16_FLOAT:
- case PIPE_FORMAT_R16G16B16X16_UINT:
- case PIPE_FORMAT_R16G16B16X16_SINT:
- case PIPE_FORMAT_Z32_FLOAT_S8X24_UINT:
-
- /* 128-bit buffers. */
- case PIPE_FORMAT_R32G32B32A32_FLOAT:
- case PIPE_FORMAT_R32G32B32A32_SNORM:
- case PIPE_FORMAT_R32G32B32A32_UNORM:
- case PIPE_FORMAT_R32G32B32A32_SINT:
- case PIPE_FORMAT_R32G32B32A32_UINT:
- case PIPE_FORMAT_R32G32B32X32_FLOAT:
- case PIPE_FORMAT_R32G32B32X32_UINT:
- case PIPE_FORMAT_R32G32B32X32_SINT:
- return V_0280A0_SWAP_STD;
- default:
- R600_ERR("unsupported colorswap format %d\n", format);
- return ~0U;
- }
- return ~0U;
-}
-
-static uint32_t r600_translate_colorformat(enum pipe_format format)
-{
- switch (format) {
- case PIPE_FORMAT_L4A4_UNORM:
- case PIPE_FORMAT_R4A4_UNORM:
- case PIPE_FORMAT_A4R4_UNORM:
- return V_0280A0_COLOR_4_4;
-
- /* 8-bit buffers. */
- case PIPE_FORMAT_A8_UNORM:
- case PIPE_FORMAT_A8_SNORM:
- case PIPE_FORMAT_A8_UINT:
- case PIPE_FORMAT_A8_SINT:
- case PIPE_FORMAT_I8_UNORM:
- case PIPE_FORMAT_I8_SNORM:
- case PIPE_FORMAT_I8_UINT:
- case PIPE_FORMAT_I8_SINT:
- case PIPE_FORMAT_L8_UNORM:
- case PIPE_FORMAT_L8_SNORM:
- case PIPE_FORMAT_L8_UINT:
- case PIPE_FORMAT_L8_SINT:
- case PIPE_FORMAT_L8_SRGB:
- case PIPE_FORMAT_R8_UNORM:
- case PIPE_FORMAT_R8_SNORM:
- case PIPE_FORMAT_R8_UINT:
- case PIPE_FORMAT_R8_SINT:
- return V_0280A0_COLOR_8;
-
- /* 16-bit buffers. */
- case PIPE_FORMAT_B5G6R5_UNORM:
- return V_0280A0_COLOR_5_6_5;
-
- case PIPE_FORMAT_B5G5R5A1_UNORM:
- case PIPE_FORMAT_B5G5R5X1_UNORM:
- return V_0280A0_COLOR_1_5_5_5;
-
- case PIPE_FORMAT_B4G4R4A4_UNORM:
- case PIPE_FORMAT_B4G4R4X4_UNORM:
- return V_0280A0_COLOR_4_4_4_4;
-
- case PIPE_FORMAT_Z16_UNORM:
- return V_0280A0_COLOR_16;
-
- case PIPE_FORMAT_L8A8_UNORM:
- case PIPE_FORMAT_L8A8_SNORM:
- case PIPE_FORMAT_L8A8_UINT:
- case PIPE_FORMAT_L8A8_SINT:
- case PIPE_FORMAT_L8A8_SRGB:
- case PIPE_FORMAT_R8G8_UNORM:
- case PIPE_FORMAT_R8G8_SNORM:
- case PIPE_FORMAT_R8G8_UINT:
- case PIPE_FORMAT_R8G8_SINT:
- case PIPE_FORMAT_R8A8_UNORM:
- case PIPE_FORMAT_R8A8_SNORM:
- case PIPE_FORMAT_R8A8_UINT:
- case PIPE_FORMAT_R8A8_SINT:
- return V_0280A0_COLOR_8_8;
-
- case PIPE_FORMAT_R16_UNORM:
- case PIPE_FORMAT_R16_SNORM:
- case PIPE_FORMAT_R16_UINT:
- case PIPE_FORMAT_R16_SINT:
- case PIPE_FORMAT_A16_UNORM:
- case PIPE_FORMAT_A16_SNORM:
- case PIPE_FORMAT_A16_UINT:
- case PIPE_FORMAT_A16_SINT:
- case PIPE_FORMAT_L16_UNORM:
- case PIPE_FORMAT_L16_SNORM:
- case PIPE_FORMAT_L16_UINT:
- case PIPE_FORMAT_L16_SINT:
- case PIPE_FORMAT_I16_UNORM:
- case PIPE_FORMAT_I16_SNORM:
- case PIPE_FORMAT_I16_UINT:
- case PIPE_FORMAT_I16_SINT:
- return V_0280A0_COLOR_16;
-
- case PIPE_FORMAT_R16_FLOAT:
- case PIPE_FORMAT_A16_FLOAT:
- case PIPE_FORMAT_L16_FLOAT:
- case PIPE_FORMAT_I16_FLOAT:
- return V_0280A0_COLOR_16_FLOAT;
-
- /* 32-bit buffers. */
- case PIPE_FORMAT_A8B8G8R8_SRGB:
- case PIPE_FORMAT_A8B8G8R8_UNORM:
- case PIPE_FORMAT_A8R8G8B8_UNORM:
- case PIPE_FORMAT_B8G8R8A8_SRGB:
- case PIPE_FORMAT_B8G8R8A8_UNORM:
- case PIPE_FORMAT_B8G8R8X8_UNORM:
- case PIPE_FORMAT_R8G8B8A8_SNORM:
- case PIPE_FORMAT_R8G8B8A8_UNORM:
- case PIPE_FORMAT_R8G8B8X8_UNORM:
- case PIPE_FORMAT_R8G8B8X8_SNORM:
- case PIPE_FORMAT_R8G8B8X8_SRGB:
- case PIPE_FORMAT_R8G8B8X8_UINT:
- case PIPE_FORMAT_R8G8B8X8_SINT:
- case PIPE_FORMAT_R8SG8SB8UX8U_NORM:
- case PIPE_FORMAT_X8B8G8R8_UNORM:
- case PIPE_FORMAT_X8R8G8B8_UNORM:
- case PIPE_FORMAT_R8G8B8A8_SINT:
- case PIPE_FORMAT_R8G8B8A8_UINT:
- return V_0280A0_COLOR_8_8_8_8;
-
- case PIPE_FORMAT_R10G10B10A2_UNORM:
- case PIPE_FORMAT_R10G10B10X2_SNORM:
- case PIPE_FORMAT_B10G10R10A2_UNORM:
- case PIPE_FORMAT_B10G10R10A2_UINT:
- case PIPE_FORMAT_B10G10R10X2_UNORM:
- case PIPE_FORMAT_R10SG10SB10SA2U_NORM:
- return V_0280A0_COLOR_2_10_10_10;
-
- case PIPE_FORMAT_Z24X8_UNORM:
- case PIPE_FORMAT_Z24_UNORM_S8_UINT:
- return V_0280A0_COLOR_8_24;
-
- case PIPE_FORMAT_Z32_FLOAT_S8X24_UINT:
- return V_0280A0_COLOR_X24_8_32_FLOAT;
-
- case PIPE_FORMAT_R32_UINT:
- case PIPE_FORMAT_R32_SINT:
- case PIPE_FORMAT_A32_UINT:
- case PIPE_FORMAT_A32_SINT:
- case PIPE_FORMAT_L32_UINT:
- case PIPE_FORMAT_L32_SINT:
- case PIPE_FORMAT_I32_UINT:
- case PIPE_FORMAT_I32_SINT:
- return V_0280A0_COLOR_32;
-
- case PIPE_FORMAT_R32_FLOAT:
- case PIPE_FORMAT_A32_FLOAT:
- case PIPE_FORMAT_L32_FLOAT:
- case PIPE_FORMAT_I32_FLOAT:
- case PIPE_FORMAT_Z32_FLOAT:
- return V_0280A0_COLOR_32_FLOAT;
-
- case PIPE_FORMAT_R16G16_FLOAT:
- case PIPE_FORMAT_L16A16_FLOAT:
- case PIPE_FORMAT_R16A16_FLOAT:
- return V_0280A0_COLOR_16_16_FLOAT;
-
- case PIPE_FORMAT_R16G16_UNORM:
- case PIPE_FORMAT_R16G16_SNORM:
- case PIPE_FORMAT_R16G16_UINT:
- case PIPE_FORMAT_R16G16_SINT:
- case PIPE_FORMAT_L16A16_UNORM:
- case PIPE_FORMAT_L16A16_SNORM:
- case PIPE_FORMAT_L16A16_UINT:
- case PIPE_FORMAT_L16A16_SINT:
- case PIPE_FORMAT_R16A16_UNORM:
- case PIPE_FORMAT_R16A16_SNORM:
- case PIPE_FORMAT_R16A16_UINT:
- case PIPE_FORMAT_R16A16_SINT:
- return V_0280A0_COLOR_16_16;
-
- case PIPE_FORMAT_R11G11B10_FLOAT:
- return V_0280A0_COLOR_10_11_11_FLOAT;
-
- /* 64-bit buffers. */
- case PIPE_FORMAT_R16G16B16A16_UINT:
- case PIPE_FORMAT_R16G16B16A16_SINT:
- case PIPE_FORMAT_R16G16B16A16_UNORM:
- case PIPE_FORMAT_R16G16B16A16_SNORM:
- case PIPE_FORMAT_R16G16B16X16_UNORM:
- case PIPE_FORMAT_R16G16B16X16_SNORM:
- case PIPE_FORMAT_R16G16B16X16_UINT:
- case PIPE_FORMAT_R16G16B16X16_SINT:
- return V_0280A0_COLOR_16_16_16_16;
-
- case PIPE_FORMAT_R16G16B16A16_FLOAT:
- case PIPE_FORMAT_R16G16B16X16_FLOAT:
- return V_0280A0_COLOR_16_16_16_16_FLOAT;
-
- case PIPE_FORMAT_R32G32_FLOAT:
- case PIPE_FORMAT_L32A32_FLOAT:
- case PIPE_FORMAT_R32A32_FLOAT:
- return V_0280A0_COLOR_32_32_FLOAT;
-
- case PIPE_FORMAT_R32G32_SINT:
- case PIPE_FORMAT_R32G32_UINT:
- case PIPE_FORMAT_L32A32_UINT:
- case PIPE_FORMAT_L32A32_SINT:
- return V_0280A0_COLOR_32_32;
-
- /* 128-bit buffers. */
- case PIPE_FORMAT_R32G32B32A32_FLOAT:
- case PIPE_FORMAT_R32G32B32X32_FLOAT:
- return V_0280A0_COLOR_32_32_32_32_FLOAT;
- case PIPE_FORMAT_R32G32B32A32_SNORM:
- case PIPE_FORMAT_R32G32B32A32_UNORM:
- case PIPE_FORMAT_R32G32B32A32_SINT:
- case PIPE_FORMAT_R32G32B32A32_UINT:
- case PIPE_FORMAT_R32G32B32X32_UINT:
- case PIPE_FORMAT_R32G32B32X32_SINT:
- return V_0280A0_COLOR_32_32_32_32;
-
- /* YUV buffers. */
- case PIPE_FORMAT_UYVY:
- case PIPE_FORMAT_YUYV:
- default:
- return ~0U; /* Unsupported. */
- }
-}
-
-static uint32_t r600_colorformat_endian_swap(uint32_t colorformat)
-{
- if (R600_BIG_ENDIAN) {
- switch(colorformat) {
- case V_0280A0_COLOR_4_4:
- return ENDIAN_NONE;
-
- /* 8-bit buffers. */
- case V_0280A0_COLOR_8:
- return ENDIAN_NONE;
-
- /* 16-bit buffers. */
- case V_0280A0_COLOR_5_6_5:
- case V_0280A0_COLOR_1_5_5_5:
- case V_0280A0_COLOR_4_4_4_4:
- case V_0280A0_COLOR_16:
- case V_0280A0_COLOR_8_8:
- return ENDIAN_8IN16;
-
- /* 32-bit buffers. */
- case V_0280A0_COLOR_8_8_8_8:
- case V_0280A0_COLOR_2_10_10_10:
- case V_0280A0_COLOR_8_24:
- case V_0280A0_COLOR_24_8:
- case V_0280A0_COLOR_32_FLOAT:
- case V_0280A0_COLOR_16_16_FLOAT:
- case V_0280A0_COLOR_16_16:
- return ENDIAN_8IN32;
-
- /* 64-bit buffers. */
- case V_0280A0_COLOR_16_16_16_16:
- case V_0280A0_COLOR_16_16_16_16_FLOAT:
- return ENDIAN_8IN16;
-
- case V_0280A0_COLOR_32_32_FLOAT:
- case V_0280A0_COLOR_32_32:
- case V_0280A0_COLOR_X24_8_32_FLOAT:
- return ENDIAN_8IN32;
-
- /* 128-bit buffers. */
- case V_0280A0_COLOR_32_32_32_FLOAT:
- case V_0280A0_COLOR_32_32_32_32_FLOAT:
- case V_0280A0_COLOR_32_32_32_32:
- return ENDIAN_8IN32;
- default:
- return ENDIAN_NONE; /* Unsupported. */
- }
- } else {
- return ENDIAN_NONE;
- }
-}
-
static bool r600_is_sampler_format_supported(struct pipe_screen *screen, enum pipe_format format)
{
return r600_translate_texformat(screen, format, NULL, NULL, NULL) != ~0U;
}
-static bool r600_is_colorbuffer_format_supported(enum pipe_format format)
+static bool r600_is_colorbuffer_format_supported(enum chip_class chip, enum pipe_format format)
{
- return r600_translate_colorformat(format) != ~0U &&
+ return r600_translate_colorformat(chip, format) != ~0U &&
r600_translate_colorswap(format) != ~0U;
}
}
}
- if ((usage & PIPE_BIND_SAMPLER_VIEW) &&
- r600_is_sampler_format_supported(screen, format)) {
- retval |= PIPE_BIND_SAMPLER_VIEW;
+ if (usage & PIPE_BIND_SAMPLER_VIEW) {
+ if (target == PIPE_BUFFER) {
+ if (r600_is_vertex_format_supported(format))
+ retval |= PIPE_BIND_SAMPLER_VIEW;
+ } else {
+ if (r600_is_sampler_format_supported(screen, format))
+ retval |= PIPE_BIND_SAMPLER_VIEW;
+ }
}
if ((usage & (PIPE_BIND_RENDER_TARGET |
PIPE_BIND_DISPLAY_TARGET |
PIPE_BIND_SCANOUT |
- PIPE_BIND_SHARED)) &&
- r600_is_colorbuffer_format_supported(format)) {
+ PIPE_BIND_SHARED |
+ PIPE_BIND_BLENDABLE)) &&
+ r600_is_colorbuffer_format_supported(rscreen->b.chip_class, format)) {
retval |= usage &
(PIPE_BIND_RENDER_TARGET |
PIPE_BIND_DISPLAY_TARGET |
PIPE_BIND_SCANOUT |
PIPE_BIND_SHARED);
+ if (!util_format_is_pure_integer(format) &&
+ !util_format_is_depth_or_stencil(format))
+ retval |= usage & PIPE_BIND_BLENDABLE;
}
if ((usage & PIPE_BIND_DEPTH_STENCIL) &&
default:;
}
- r600_write_context_reg_seq(cs, R_028E00_PA_SU_POLY_OFFSET_FRONT_SCALE, 4);
+ radeon_set_context_reg_seq(cs, R_028E00_PA_SU_POLY_OFFSET_FRONT_SCALE, 4);
radeon_emit(cs, fui(offset_scale));
radeon_emit(cs, fui(offset_units));
radeon_emit(cs, fui(offset_scale));
S_028A0C_REPEAT_COUNT(state->line_stipple_factor) : 0;
rs->pa_cl_clip_cntl =
S_028810_PS_UCP_MODE(3) |
+ S_028810_DX_CLIP_SPACE_DEF(state->clip_halfz) |
S_028810_ZCLIP_NEAR_DISABLE(!state->depth_clip) |
S_028810_ZCLIP_FAR_DISABLE(!state->depth_clip) |
S_028810_DX_LINEAR_ATTR_CLIP_ENA(1);
+ if (rctx->b.chip_class == R700) {
+ rs->pa_cl_clip_cntl |=
+ S_028810_DX_RASTERIZATION_KILL(state->rasterizer_discard);
+ }
rs->multisample_enable = state->multisample;
/* offset */
rs->offset_units = state->offset_units;
- rs->offset_scale = state->offset_scale * 12.0f;
+ rs->offset_scale = state->offset_scale * 16.0f;
rs->offset_enable = state->offset_point || state->offset_line || state->offset_tri;
if (state->point_size_per_vertex) {
sc_mode_cntl = S_028A4C_MSAA_ENABLE(state->multisample) |
S_028A4C_LINE_STIPPLE_ENABLE(state->line_stipple_enable) |
- S_028A4C_FORCE_EOV_CNTDWN_ENABLE(1);
+ S_028A4C_FORCE_EOV_CNTDWN_ENABLE(1) |
+ S_028A4C_PS_ITER_SAMPLE(state->multisample && rctx->ps_iter_samples > 1);
+ if (rctx->b.family == CHIP_RV770) {
+ /* workaround possible rendering corruption on RV770 with hyperz together with sample shading */
+ sc_mode_cntl |= S_028A4C_TILE_COVER_DISABLE(state->multisample && rctx->ps_iter_samples > 1);
+ }
if (rctx->b.chip_class >= R700) {
sc_mode_cntl |= S_028A4C_FORCE_EOV_REZ_ENABLE(1) |
S_028A4C_R700_ZMM_LINE_OFFSET(1) |
S_028C08_PIX_CENTER_HALF(state->half_pixel_center) |
S_028C08_QUANT_MODE(V_028C08_X_1_256TH));
r600_store_context_reg(&rs->buffer, R_028DFC_PA_SU_POLY_OFFSET_CLAMP, fui(state->offset_clamp));
- r600_store_context_reg(&rs->buffer, R_028814_PA_SU_SC_MODE_CNTL,
- S_028814_PROVOKING_VTX_LAST(!state->flatshade_first) |
- S_028814_CULL_FRONT(state->cull_face & PIPE_FACE_FRONT ? 1 : 0) |
- S_028814_CULL_BACK(state->cull_face & PIPE_FACE_BACK ? 1 : 0) |
- S_028814_FACE(!state->front_ccw) |
- S_028814_POLY_OFFSET_FRONT_ENABLE(state->offset_tri) |
- S_028814_POLY_OFFSET_BACK_ENABLE(state->offset_tri) |
- S_028814_POLY_OFFSET_PARA_ENABLE(state->offset_tri) |
- S_028814_POLY_MODE(state->fill_front != PIPE_POLYGON_MODE_FILL ||
- state->fill_back != PIPE_POLYGON_MODE_FILL) |
- S_028814_POLYMODE_FRONT_PTYPE(r600_translate_fill(state->fill_front)) |
- S_028814_POLYMODE_BACK_PTYPE(r600_translate_fill(state->fill_back)));
- r600_store_context_reg(&rs->buffer, R_028350_SX_MISC, S_028350_MULTIPASS(state->rasterizer_discard));
+
+ rs->pa_su_sc_mode_cntl = S_028814_PROVOKING_VTX_LAST(!state->flatshade_first) |
+ S_028814_CULL_FRONT(state->cull_face & PIPE_FACE_FRONT ? 1 : 0) |
+ S_028814_CULL_BACK(state->cull_face & PIPE_FACE_BACK ? 1 : 0) |
+ S_028814_FACE(!state->front_ccw) |
+ S_028814_POLY_OFFSET_FRONT_ENABLE(util_get_offset(state, state->fill_front)) |
+ S_028814_POLY_OFFSET_BACK_ENABLE(util_get_offset(state, state->fill_back)) |
+ S_028814_POLY_OFFSET_PARA_ENABLE(state->offset_point || state->offset_line) |
+ S_028814_POLY_MODE(state->fill_front != PIPE_POLYGON_MODE_FILL ||
+ state->fill_back != PIPE_POLYGON_MODE_FILL) |
+ S_028814_POLYMODE_FRONT_PTYPE(r600_translate_fill(state->fill_front)) |
+ S_028814_POLYMODE_BACK_PTYPE(r600_translate_fill(state->fill_back));
+ if (rctx->b.chip_class == R700) {
+ r600_store_context_reg(&rs->buffer, R_028814_PA_SU_SC_MODE_CNTL, rs->pa_su_sc_mode_cntl);
+ }
+ if (rctx->b.chip_class == R600) {
+ r600_store_context_reg(&rs->buffer, R_028350_SX_MISC,
+ S_028350_MULTIPASS(state->rasterizer_discard));
+ }
return rs;
}
unsigned width0, unsigned height0)
{
- struct pipe_context *ctx = view->base.context;
struct r600_texture *tmp = (struct r600_texture*)view->base.texture;
- uint64_t va;
int stride = util_format_get_blocksize(view->base.format);
unsigned format, num_format, format_comp, endian;
- unsigned offset = view->base.u.buf.first_element * stride;
+ uint64_t offset = view->base.u.buf.first_element * stride;
unsigned size = (view->base.u.buf.last_element - view->base.u.buf.first_element + 1) * stride;
r600_vertex_data_type(view->base.format,
&format, &num_format, &format_comp,
&endian);
- va = r600_resource_va(ctx->screen, view->base.texture) + offset;
view->tex_resource = &tmp->resource;
-
view->skip_mip_address_reloc = true;
- view->tex_resource_words[0] = va;
+
+ view->tex_resource_words[0] = offset;
view->tex_resource_words[1] = size - 1;
- view->tex_resource_words[2] = S_038008_BASE_ADDRESS_HI(va >> 32UL) |
+ view->tex_resource_words[2] = S_038008_BASE_ADDRESS_HI(offset >> 32UL) |
S_038008_STRIDE(stride) |
S_038008_DATA_FORMAT(format) |
S_038008_NUM_FORMAT_ALL(num_format) |
S_038008_FORMAT_COMP_ALL(format_comp) |
- S_038008_SRF_MODE_ALL(1) |
S_038008_ENDIAN_SWAP(endian);
view->tex_resource_words[3] = 0;
/*
view->tex_resource_words[3] = tmp->surface.level[offset_level + 1].offset >> 8;
}
view->tex_resource_words[4] = (word4 |
- S_038010_SRF_MODE_ALL(V_038010_SRF_MODE_ZERO_CLAMP_MINUS_ONE) |
S_038010_REQUEST_SIZE(1) |
S_038010_ENDIAN_SWAP(endian) |
S_038010_BASE_LEVEL(0));
struct radeon_winsys_cs *cs = rctx->b.rings.gfx.cs;
struct pipe_clip_state *state = &rctx->clip_state.state;
- r600_write_context_reg_seq(cs, R_028E20_PA_CL_UCP0_X, 6*4);
+ radeon_set_context_reg_seq(cs, R_028E20_PA_CL_UCP0_X, 6*4);
radeon_emit_array(cs, (unsigned*)state, 6*4);
}
static void r600_emit_scissor_state(struct r600_context *rctx, struct r600_atom *atom)
{
struct radeon_winsys_cs *cs = rctx->b.rings.gfx.cs;
- struct pipe_scissor_state *state = &rctx->scissor.scissor;
+ struct r600_scissor_state *rstate = (struct r600_scissor_state *)atom;
+ struct pipe_scissor_state *state = &rstate->scissor;
+ unsigned offset = rstate->idx * 4 * 2;
- if (rctx->b.chip_class != R600 || rctx->scissor.enable) {
- r600_write_context_reg_seq(cs, R_028250_PA_SC_VPORT_SCISSOR_0_TL, 2);
+ if (rctx->b.chip_class != R600 || rctx->scissor[0].enable) {
+ radeon_set_context_reg_seq(cs, R_028250_PA_SC_VPORT_SCISSOR_0_TL + offset, 2);
radeon_emit(cs, S_028240_TL_X(state->minx) | S_028240_TL_Y(state->miny) |
S_028240_WINDOW_OFFSET_DISABLE(1));
radeon_emit(cs, S_028244_BR_X(state->maxx) | S_028244_BR_Y(state->maxy));
} else {
- r600_write_context_reg_seq(cs, R_028250_PA_SC_VPORT_SCISSOR_0_TL, 2);
+ radeon_set_context_reg_seq(cs, R_028250_PA_SC_VPORT_SCISSOR_0_TL, 2);
radeon_emit(cs, S_028240_TL_X(0) | S_028240_TL_Y(0) |
S_028240_WINDOW_OFFSET_DISABLE(1));
radeon_emit(cs, S_028244_BR_X(8192) | S_028244_BR_Y(8192));
const struct pipe_scissor_state *state)
{
struct r600_context *rctx = (struct r600_context *)ctx;
+ int i;
- rctx->scissor.scissor = *state;
+ for (i = start_slot ; i < start_slot + num_scissors; i++) {
+ rctx->scissor[i].scissor = state[i - start_slot];
+ }
- if (rctx->b.chip_class == R600 && !rctx->scissor.enable)
+ if (rctx->b.chip_class == R600 && !rctx->scissor[0].enable)
return;
- rctx->scissor.atom.dirty = true;
+ for (i = start_slot ; i < start_slot + num_scissors; i++) {
+ r600_mark_atom_dirty(rctx, &rctx->scissor[i].atom);
+ }
}
static struct r600_resource *r600_buffer_create_helper(struct r600_screen *rscreen,
buffer.target = PIPE_BUFFER;
buffer.format = PIPE_FORMAT_R8_UNORM;
buffer.bind = PIPE_BIND_CUSTOM;
- buffer.usage = PIPE_USAGE_STATIC;
+ buffer.usage = PIPE_USAGE_DEFAULT;
buffer.flags = 0;
buffer.width0 = size;
buffer.height0 = 1;
unsigned level = surf->base.u.tex.level;
unsigned pitch, slice;
unsigned color_info;
+ unsigned color_view;
unsigned format, swap, ntype, endian;
unsigned offset;
const struct util_format_description *desc;
}
offset = rtex->surface.level[level].offset;
- if (rtex->surface.level[level].mode < RADEON_SURF_MODE_1D) {
+ if (rtex->surface.level[level].mode == RADEON_SURF_MODE_LINEAR) {
+ assert(surf->base.u.tex.first_layer == surf->base.u.tex.last_layer);
offset += rtex->surface.level[level].slice_size *
- surf->base.u.tex.first_layer;
- }
+ surf->base.u.tex.first_layer;
+ color_view = 0;
+ } else
+ color_view = S_028080_SLICE_START(surf->base.u.tex.first_layer) |
+ S_028080_SLICE_MAX(surf->base.u.tex.last_layer);
+
pitch = rtex->surface.level[level].nblk_x / 8 - 1;
slice = (rtex->surface.level[level].nblk_x * rtex->surface.level[level].nblk_y) / 64;
if (slice) {
ntype = V_0280A0_NUMBER_UINT;
}
- format = r600_translate_colorformat(surf->base.format);
+ format = r600_translate_colorformat(rctx->b.chip_class, surf->base.format);
assert(format != ~0);
swap = r600_translate_colorswap(surf->base.format);
struct r600_cmask_info cmask;
struct r600_fmask_info fmask;
- r600_texture_get_cmask_info(rscreen, rtex, &cmask);
- r600_texture_get_fmask_info(rscreen, rtex, 8, &fmask);
+ r600_texture_get_cmask_info(&rscreen->b, rtex, &cmask);
+ r600_texture_get_fmask_info(&rscreen->b, rtex, 8, &fmask);
/* CMASK. */
if (!rctx->dummy_cmask ||
}
surf->cb_color_info = color_info;
-
- if (rtex->surface.level[level].mode < RADEON_SURF_MODE_1D) {
- surf->cb_color_view = 0;
- } else {
- surf->cb_color_view = S_028080_SLICE_START(surf->base.u.tex.first_layer) |
- S_028080_SLICE_MAX(surf->base.u.tex.last_layer);
- }
-
+ surf->cb_color_view = color_view;
surf->color_initialized = true;
}
default:;
}
- surf->htile_enabled = 0;
/* use htile only for first level */
- if (rtex->htile && !level) {
- uint64_t va = r600_resource_va(&rctx->screen->b.b, &rtex->htile->b.b);
- surf->htile_enabled = 1;
- surf->db_htile_data_base = va >> 8;
+ if (rtex->htile_buffer && !level) {
+ surf->db_htile_data_base = 0;
surf->db_htile_surface = S_028D24_HTILE_WIDTH(1) |
- S_028D24_HTILE_HEIGHT(1) |
- S_028D24_FULL_CACHE(1) |
- S_028D24_LINEAR(1);
+ S_028D24_HTILE_HEIGHT(1) |
+ S_028D24_FULL_CACHE(1);
/* preload is not working properly on r6xx/r7xx */
surf->db_depth_info |= S_028010_TILE_SURFACE_ENABLE(1);
}
if (rctx->framebuffer.state.nr_cbufs) {
rctx->b.flags |= R600_CONTEXT_WAIT_3D_IDLE | R600_CONTEXT_FLUSH_AND_INV;
- rctx->b.flags |= R600_CONTEXT_FLUSH_AND_INV_CB;
-
- if (rctx->b.chip_class >= R700 &&
- rctx->framebuffer.state.cbufs[0]->texture->nr_samples > 1) {
- rctx->b.flags |= R600_CONTEXT_FLUSH_AND_INV_CB_META;
- }
+ rctx->b.flags |= R600_CONTEXT_FLUSH_AND_INV_CB |
+ R600_CONTEXT_FLUSH_AND_INV_CB_META;
}
if (rctx->framebuffer.state.zsbuf) {
rctx->b.flags |= R600_CONTEXT_WAIT_3D_IDLE | R600_CONTEXT_FLUSH_AND_INV;
rctx->b.flags |= R600_CONTEXT_FLUSH_AND_INV_DB;
rtex = (struct r600_texture*)rctx->framebuffer.state.zsbuf->texture;
- if (rctx->b.chip_class >= R700 && rtex->htile) {
+ if (rctx->b.chip_class >= R700 && rtex->htile_buffer) {
rctx->b.flags |= R600_CONTEXT_FLUSH_AND_INV_DB_META;
}
}
util_copy_framebuffer_state(&rctx->framebuffer.state, state);
rctx->framebuffer.export_16bpc = state->nr_cbufs != 0;
- rctx->framebuffer.cb0_is_integer = state->nr_cbufs &&
+ rctx->framebuffer.cb0_is_integer = state->nr_cbufs && state->cbufs[0] &&
util_format_is_pure_integer(state->cbufs[0]->format);
rctx->framebuffer.compressed_cb_mask = 0;
rctx->framebuffer.is_msaa_resolve = state->nr_cbufs == 2 &&
+ state->cbufs[0] && state->cbufs[1] &&
state->cbufs[0]->texture->nr_samples > 1 &&
state->cbufs[1]->texture->nr_samples <= 1;
-
- if (state->nr_cbufs)
- rctx->framebuffer.nr_samples = state->cbufs[0]->texture->nr_samples;
- else if (state->zsbuf)
- rctx->framebuffer.nr_samples = state->zsbuf->texture->nr_samples;
- else
- rctx->framebuffer.nr_samples = 0;
+ rctx->framebuffer.nr_samples = util_framebuffer_get_num_samples(state);
/* Colorbuffers. */
for (i = 0; i < state->nr_cbufs; i++) {
i == 1;
surf = (struct r600_surface*)state->cbufs[i];
+ if (!surf)
+ continue;
+
rtex = (struct r600_texture*)surf->base.texture;
r600_context_add_resource_size(ctx, state->cbufs[i]->texture);
/* Update alpha-test state dependencies.
* Alpha-test is done on the first colorbuffer only. */
if (state->nr_cbufs) {
+ bool alphatest_bypass = false;
+
surf = (struct r600_surface*)state->cbufs[0];
- if (rctx->alphatest_state.bypass != surf->alphatest_bypass) {
- rctx->alphatest_state.bypass = surf->alphatest_bypass;
- rctx->alphatest_state.atom.dirty = true;
+ if (surf) {
+ alphatest_bypass = surf->alphatest_bypass;
+ }
+
+ if (rctx->alphatest_state.bypass != alphatest_bypass) {
+ rctx->alphatest_state.bypass = alphatest_bypass;
+ r600_mark_atom_dirty(rctx, &rctx->alphatest_state.atom);
}
}
if (state->zsbuf->format != rctx->poly_offset_state.zs_format) {
rctx->poly_offset_state.zs_format = state->zsbuf->format;
- rctx->poly_offset_state.atom.dirty = true;
+ r600_mark_atom_dirty(rctx, &rctx->poly_offset_state.atom);
}
if (rctx->db_state.rsurf != surf) {
rctx->db_state.rsurf = surf;
- rctx->db_state.atom.dirty = true;
- rctx->db_misc_state.atom.dirty = true;
+ r600_mark_atom_dirty(rctx, &rctx->db_state.atom);
+ r600_mark_atom_dirty(rctx, &rctx->db_misc_state.atom);
}
} else if (rctx->db_state.rsurf) {
rctx->db_state.rsurf = NULL;
- rctx->db_state.atom.dirty = true;
- rctx->db_misc_state.atom.dirty = true;
+ r600_mark_atom_dirty(rctx, &rctx->db_state.atom);
+ r600_mark_atom_dirty(rctx, &rctx->db_misc_state.atom);
}
if (rctx->cb_misc_state.nr_cbufs != state->nr_cbufs) {
rctx->cb_misc_state.nr_cbufs = state->nr_cbufs;
- rctx->cb_misc_state.atom.dirty = true;
+ r600_mark_atom_dirty(rctx, &rctx->cb_misc_state.atom);
}
if (state->nr_cbufs == 0 && rctx->alphatest_state.bypass) {
rctx->alphatest_state.bypass = false;
- rctx->alphatest_state.atom.dirty = true;
+ r600_mark_atom_dirty(rctx, &rctx->alphatest_state.atom);
}
- r600_update_db_shader_control(rctx);
-
/* Calculate the CS size. */
rctx->framebuffer.atom.num_dw =
10 /*COLOR_INFO*/ + 4 /*SCISSOR*/ + 3 /*SHADER_CONTROL*/ + 8 /*MSAA*/;
if (rctx->framebuffer.state.nr_cbufs) {
- rctx->framebuffer.atom.num_dw += 6 * (2 + rctx->framebuffer.state.nr_cbufs);
- rctx->framebuffer.atom.num_dw += 6 * rctx->framebuffer.state.nr_cbufs; /* relocs */
-
+ rctx->framebuffer.atom.num_dw += 15 * rctx->framebuffer.state.nr_cbufs;
+ rctx->framebuffer.atom.num_dw += 3 * (2 + rctx->framebuffer.state.nr_cbufs);
}
if (rctx->framebuffer.state.zsbuf) {
- rctx->framebuffer.atom.num_dw += 18;
+ rctx->framebuffer.atom.num_dw += 16;
} else if (rctx->screen->b.info.drm_minor >= 18) {
rctx->framebuffer.atom.num_dw += 3;
}
rctx->framebuffer.atom.num_dw += 2;
}
- rctx->framebuffer.atom.dirty = true;
-}
-
-#define FILL_SREG(s0x, s0y, s1x, s1y, s2x, s2y, s3x, s3y) \
- (((s0x) & 0xf) | (((s0y) & 0xf) << 4) | \
- (((s1x) & 0xf) << 8) | (((s1y) & 0xf) << 12) | \
- (((s2x) & 0xf) << 16) | (((s2y) & 0xf) << 20) | \
- (((s3x) & 0xf) << 24) | (((s3y) & 0xf) << 28))
+ r600_mark_atom_dirty(rctx, &rctx->framebuffer.atom);
+ r600_set_sample_locations_constant_buffer(rctx);
+}
static uint32_t sample_locs_2x[] = {
FILL_SREG(-4, 4, 4, -4, -4, 4, 4, -4),
nr_samples = 0;
break;
case 2:
- r600_write_config_reg(cs, R_008B40_PA_SC_AA_SAMPLE_LOCS_2S, sample_locs_2x[0]);
+ radeon_set_config_reg(cs, R_008B40_PA_SC_AA_SAMPLE_LOCS_2S, sample_locs_2x[0]);
max_dist = max_dist_2x;
break;
case 4:
- r600_write_config_reg(cs, R_008B44_PA_SC_AA_SAMPLE_LOCS_4S, sample_locs_4x[0]);
+ radeon_set_config_reg(cs, R_008B44_PA_SC_AA_SAMPLE_LOCS_4S, sample_locs_4x[0]);
max_dist = max_dist_4x;
break;
case 8:
- r600_write_config_reg_seq(cs, R_008B48_PA_SC_AA_SAMPLE_LOCS_8S_WD0, 2);
+ radeon_set_config_reg_seq(cs, R_008B48_PA_SC_AA_SAMPLE_LOCS_8S_WD0, 2);
radeon_emit(cs, sample_locs_8x[0]); /* R_008B48_PA_SC_AA_SAMPLE_LOCS_8S_WD0 */
radeon_emit(cs, sample_locs_8x[1]); /* R_008B4C_PA_SC_AA_SAMPLE_LOCS_8S_WD1 */
max_dist = max_dist_8x;
} else {
switch (nr_samples) {
default:
- r600_write_context_reg_seq(cs, R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX, 2);
+ radeon_set_context_reg_seq(cs, R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX, 2);
radeon_emit(cs, 0); /* R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX */
radeon_emit(cs, 0); /* R_028C20_PA_SC_AA_SAMPLE_LOCS_8D_WD1_MCTX */
nr_samples = 0;
break;
case 2:
- r600_write_context_reg_seq(cs, R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX, 2);
+ radeon_set_context_reg_seq(cs, R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX, 2);
radeon_emit(cs, sample_locs_2x[0]); /* R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX */
radeon_emit(cs, sample_locs_2x[1]); /* R_028C20_PA_SC_AA_SAMPLE_LOCS_8D_WD1_MCTX */
max_dist = max_dist_2x;
break;
case 4:
- r600_write_context_reg_seq(cs, R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX, 2);
+ radeon_set_context_reg_seq(cs, R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX, 2);
radeon_emit(cs, sample_locs_4x[0]); /* R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX */
radeon_emit(cs, sample_locs_4x[1]); /* R_028C20_PA_SC_AA_SAMPLE_LOCS_8D_WD1_MCTX */
max_dist = max_dist_4x;
break;
case 8:
- r600_write_context_reg_seq(cs, R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX, 2);
+ radeon_set_context_reg_seq(cs, R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX, 2);
radeon_emit(cs, sample_locs_8x[0]); /* R_028C1C_PA_SC_AA_SAMPLE_LOCS_MCTX */
radeon_emit(cs, sample_locs_8x[1]); /* R_028C20_PA_SC_AA_SAMPLE_LOCS_8D_WD1_MCTX */
max_dist = max_dist_8x;
}
if (nr_samples > 1) {
- r600_write_context_reg_seq(cs, R_028C00_PA_SC_LINE_CNTL, 2);
+ radeon_set_context_reg_seq(cs, R_028C00_PA_SC_LINE_CNTL, 2);
radeon_emit(cs, S_028C00_LAST_PIXEL(1) |
S_028C00_EXPAND_LINE_WIDTH(1)); /* R_028C00_PA_SC_LINE_CNTL */
radeon_emit(cs, S_028C04_MSAA_NUM_SAMPLES(util_logbase2(nr_samples)) |
S_028C04_MAX_SAMPLE_DIST(max_dist)); /* R_028C04_PA_SC_AA_CONFIG */
} else {
- r600_write_context_reg_seq(cs, R_028C00_PA_SC_LINE_CNTL, 2);
+ radeon_set_context_reg_seq(cs, R_028C00_PA_SC_LINE_CNTL, 2);
radeon_emit(cs, S_028C00_LAST_PIXEL(1)); /* R_028C00_PA_SC_LINE_CNTL */
radeon_emit(cs, 0); /* R_028C04_PA_SC_AA_CONFIG */
}
unsigned i, sbu = 0;
/* Colorbuffers. */
- r600_write_context_reg_seq(cs, R_0280A0_CB_COLOR0_INFO, 8);
+ radeon_set_context_reg_seq(cs, R_0280A0_CB_COLOR0_INFO, 8);
for (i = 0; i < nr_cbufs; i++) {
- radeon_emit(cs, cb[i]->cb_color_info);
+ radeon_emit(cs, cb[i] ? cb[i]->cb_color_info : 0);
}
/* set CB_COLOR1_INFO for possible dual-src blending */
- if (i == 1) {
+ if (i == 1 && cb[0]) {
radeon_emit(cs, cb[0]->cb_color_info);
i++;
}
}
if (nr_cbufs) {
- /* COLOR_BASE */
- r600_write_context_reg_seq(cs, R_028040_CB_COLOR0_BASE, nr_cbufs);
for (i = 0; i < nr_cbufs; i++) {
- radeon_emit(cs, cb[i]->cb_color_base);
- }
+ unsigned reloc;
- /* relocations */
- for (i = 0; i < nr_cbufs; i++) {
- unsigned reloc = r600_context_bo_reloc(&rctx->b,
- &rctx->b.rings.gfx,
- (struct r600_resource*)cb[i]->base.texture,
- RADEON_USAGE_READWRITE);
+ if (!cb[i])
+ continue;
+
+ /* COLOR_BASE */
+ radeon_set_context_reg(cs, R_028040_CB_COLOR0_BASE + i*4, cb[i]->cb_color_base);
+
+ reloc = r600_context_bo_reloc(&rctx->b,
+ &rctx->b.rings.gfx,
+ (struct r600_resource*)cb[i]->base.texture,
+ RADEON_USAGE_READWRITE,
+ cb[i]->base.texture->nr_samples > 1 ?
+ RADEON_PRIO_COLOR_BUFFER_MSAA :
+ RADEON_PRIO_COLOR_BUFFER);
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
radeon_emit(cs, reloc);
- }
- r600_write_context_reg_seq(cs, R_028060_CB_COLOR0_SIZE, nr_cbufs);
- for (i = 0; i < nr_cbufs; i++) {
- radeon_emit(cs, cb[i]->cb_color_size);
- }
+ /* FMASK */
+ radeon_set_context_reg(cs, R_0280E0_CB_COLOR0_FRAG + i*4, cb[i]->cb_color_fmask);
- r600_write_context_reg_seq(cs, R_028080_CB_COLOR0_VIEW, nr_cbufs);
- for (i = 0; i < nr_cbufs; i++) {
- radeon_emit(cs, cb[i]->cb_color_view);
- }
+ reloc = r600_context_bo_reloc(&rctx->b,
+ &rctx->b.rings.gfx,
+ cb[i]->cb_buffer_fmask,
+ RADEON_USAGE_READWRITE,
+ cb[i]->base.texture->nr_samples > 1 ?
+ RADEON_PRIO_COLOR_BUFFER_MSAA :
+ RADEON_PRIO_COLOR_BUFFER);
+ radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+ radeon_emit(cs, reloc);
- r600_write_context_reg_seq(cs, R_028100_CB_COLOR0_MASK, nr_cbufs);
- for (i = 0; i < nr_cbufs; i++) {
- radeon_emit(cs, cb[i]->cb_color_mask);
- }
+ /* CMASK */
+ radeon_set_context_reg(cs, R_0280C0_CB_COLOR0_TILE + i*4, cb[i]->cb_color_cmask);
- /* FMASK. */
- r600_write_context_reg_seq(cs, R_0280E0_CB_COLOR0_FRAG, nr_cbufs);
- for (i = 0; i < nr_cbufs; i++) {
- radeon_emit(cs, cb[i]->cb_color_fmask);
- }
- /* relocations */
- for (i = 0; i < nr_cbufs; i++) {
- unsigned reloc = r600_context_bo_reloc(&rctx->b,
- &rctx->b.rings.gfx,
- cb[i]->cb_buffer_fmask,
- RADEON_USAGE_READWRITE);
+ reloc = r600_context_bo_reloc(&rctx->b,
+ &rctx->b.rings.gfx,
+ cb[i]->cb_buffer_cmask,
+ RADEON_USAGE_READWRITE,
+ cb[i]->base.texture->nr_samples > 1 ?
+ RADEON_PRIO_COLOR_BUFFER_MSAA :
+ RADEON_PRIO_COLOR_BUFFER);
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
radeon_emit(cs, reloc);
}
- /* CMASK. */
- r600_write_context_reg_seq(cs, R_0280C0_CB_COLOR0_TILE, nr_cbufs);
+ radeon_set_context_reg_seq(cs, R_028060_CB_COLOR0_SIZE, nr_cbufs);
for (i = 0; i < nr_cbufs; i++) {
- radeon_emit(cs, cb[i]->cb_color_cmask);
+ radeon_emit(cs, cb[i] ? cb[i]->cb_color_size : 0);
}
- /* relocations */
+
+ radeon_set_context_reg_seq(cs, R_028080_CB_COLOR0_VIEW, nr_cbufs);
for (i = 0; i < nr_cbufs; i++) {
- unsigned reloc = r600_context_bo_reloc(&rctx->b,
- &rctx->b.rings.gfx,
- cb[i]->cb_buffer_cmask,
- RADEON_USAGE_READWRITE);
- radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
- radeon_emit(cs, reloc);
+ radeon_emit(cs, cb[i] ? cb[i]->cb_color_view : 0);
+ }
+
+ radeon_set_context_reg_seq(cs, R_028100_CB_COLOR0_MASK, nr_cbufs);
+ for (i = 0; i < nr_cbufs; i++) {
+ radeon_emit(cs, cb[i] ? cb[i]->cb_color_mask : 0);
}
sbu |= SURFACE_BASE_UPDATE_COLOR_NUM(nr_cbufs);
unsigned reloc = r600_context_bo_reloc(&rctx->b,
&rctx->b.rings.gfx,
(struct r600_resource*)state->zsbuf->texture,
- RADEON_USAGE_READWRITE);
+ RADEON_USAGE_READWRITE,
+ surf->base.texture->nr_samples > 1 ?
+ RADEON_PRIO_DEPTH_BUFFER_MSAA :
+ RADEON_PRIO_DEPTH_BUFFER);
- r600_write_context_reg(cs, R_028DF8_PA_SU_POLY_OFFSET_DB_FMT_CNTL,
+ radeon_set_context_reg(cs, R_028DF8_PA_SU_POLY_OFFSET_DB_FMT_CNTL,
surf->pa_su_poly_offset_db_fmt_cntl);
- r600_write_context_reg_seq(cs, R_028000_DB_DEPTH_SIZE, 2);
+ radeon_set_context_reg_seq(cs, R_028000_DB_DEPTH_SIZE, 2);
radeon_emit(cs, surf->db_depth_size); /* R_028000_DB_DEPTH_SIZE */
radeon_emit(cs, surf->db_depth_view); /* R_028004_DB_DEPTH_VIEW */
- r600_write_context_reg_seq(cs, R_02800C_DB_DEPTH_BASE, 2);
+ radeon_set_context_reg_seq(cs, R_02800C_DB_DEPTH_BASE, 2);
radeon_emit(cs, surf->db_depth_base); /* R_02800C_DB_DEPTH_BASE */
radeon_emit(cs, surf->db_depth_info); /* R_028010_DB_DEPTH_INFO */
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
radeon_emit(cs, reloc);
- r600_write_context_reg(cs, R_028D34_DB_PREFETCH_LIMIT, surf->db_prefetch_limit);
+ radeon_set_context_reg(cs, R_028D34_DB_PREFETCH_LIMIT, surf->db_prefetch_limit);
sbu |= SURFACE_BASE_UPDATE_DEPTH;
} else if (rctx->screen->b.info.drm_minor >= 18) {
/* DRM 2.6.18 allows the INVALID format to disable depth/stencil.
* Older kernels are out of luck. */
- r600_write_context_reg(cs, R_028010_DB_DEPTH_INFO, S_028010_FORMAT(V_028010_DEPTH_INVALID));
+ radeon_set_context_reg(cs, R_028010_DB_DEPTH_INFO, S_028010_FORMAT(V_028010_DEPTH_INVALID));
}
/* SURFACE_BASE_UPDATE */
}
/* Framebuffer dimensions. */
- r600_write_context_reg_seq(cs, R_028204_PA_SC_WINDOW_SCISSOR_TL, 2);
+ radeon_set_context_reg_seq(cs, R_028204_PA_SC_WINDOW_SCISSOR_TL, 2);
radeon_emit(cs, S_028240_TL_X(0) | S_028240_TL_Y(0) |
S_028240_WINDOW_OFFSET_DISABLE(1)); /* R_028204_PA_SC_WINDOW_SCISSOR_TL */
radeon_emit(cs, S_028244_BR_X(state->width) |
S_028244_BR_Y(state->height)); /* R_028208_PA_SC_WINDOW_SCISSOR_BR */
if (rctx->framebuffer.is_msaa_resolve) {
- r600_write_context_reg(cs, R_0287A0_CB_SHADER_CONTROL, 1);
+ radeon_set_context_reg(cs, R_0287A0_CB_SHADER_CONTROL, 1);
} else {
/* Always enable the first colorbuffer in CB_SHADER_CONTROL. This
* will assure that the alpha-test will work even if there is
* no colorbuffer bound. */
- r600_write_context_reg(cs, R_0287A0_CB_SHADER_CONTROL,
+ radeon_set_context_reg(cs, R_0287A0_CB_SHADER_CONTROL,
(1ull << MAX2(nr_cbufs, 1)) - 1);
}
r600_emit_msaa_state(rctx, rctx->framebuffer.nr_samples);
}
+static void r600_set_min_samples(struct pipe_context *ctx, unsigned min_samples)
+{
+ struct r600_context *rctx = (struct r600_context *)ctx;
+
+ if (rctx->ps_iter_samples == min_samples)
+ return;
+
+ rctx->ps_iter_samples = min_samples;
+ if (rctx->framebuffer.nr_samples > 1) {
+ r600_mark_atom_dirty(rctx, &rctx->rasterizer_state.atom);
+ if (rctx->b.chip_class == R600)
+ r600_mark_atom_dirty(rctx, &rctx->db_misc_state.atom);
+ }
+}
+
static void r600_emit_cb_misc_state(struct r600_context *rctx, struct r600_atom *atom)
{
struct radeon_winsys_cs *cs = rctx->b.rings.gfx.cs;
struct r600_cb_misc_state *a = (struct r600_cb_misc_state*)atom;
if (G_028808_SPECIAL_OP(a->cb_color_control) == V_028808_SPECIAL_RESOLVE_BOX) {
- r600_write_context_reg_seq(cs, R_028238_CB_TARGET_MASK, 2);
+ radeon_set_context_reg_seq(cs, R_028238_CB_TARGET_MASK, 2);
if (rctx->b.chip_class == R600) {
radeon_emit(cs, 0xff); /* R_028238_CB_TARGET_MASK */
radeon_emit(cs, 0xff); /* R_02823C_CB_SHADER_MASK */
radeon_emit(cs, 0xf); /* R_028238_CB_TARGET_MASK */
radeon_emit(cs, 0xf); /* R_02823C_CB_SHADER_MASK */
}
- r600_write_context_reg(cs, R_028808_CB_COLOR_CONTROL, a->cb_color_control);
+ radeon_set_context_reg(cs, R_028808_CB_COLOR_CONTROL, a->cb_color_control);
} else {
unsigned fb_colormask = (1ULL << ((unsigned)a->nr_cbufs * 4)) - 1;
unsigned ps_colormask = (1ULL << ((unsigned)a->nr_ps_color_outputs * 4)) - 1;
unsigned multiwrite = a->multiwrite && a->nr_cbufs > 1;
- r600_write_context_reg_seq(cs, R_028238_CB_TARGET_MASK, 2);
+ radeon_set_context_reg_seq(cs, R_028238_CB_TARGET_MASK, 2);
radeon_emit(cs, a->blend_colormask & fb_colormask); /* R_028238_CB_TARGET_MASK */
/* Always enable the first color output to make sure alpha-test works even without one. */
radeon_emit(cs, 0xf | (multiwrite ? fb_colormask : ps_colormask)); /* R_02823C_CB_SHADER_MASK */
- r600_write_context_reg(cs, R_028808_CB_COLOR_CONTROL,
+ radeon_set_context_reg(cs, R_028808_CB_COLOR_CONTROL,
a->cb_color_control |
S_028808_MULTIWRITE_ENABLE(multiwrite));
}
struct radeon_winsys_cs *cs = rctx->b.rings.gfx.cs;
struct r600_db_state *a = (struct r600_db_state*)atom;
- if (a->rsurf && a->rsurf->htile_enabled) {
+ if (a->rsurf && a->rsurf->db_htile_surface) {
struct r600_texture *rtex = (struct r600_texture *)a->rsurf->base.texture;
unsigned reloc_idx;
- r600_write_context_reg(cs, R_02802C_DB_DEPTH_CLEAR, fui(rtex->depth_clear));
- r600_write_context_reg(cs, R_028D24_DB_HTILE_SURFACE, a->rsurf->db_htile_surface);
- r600_write_context_reg(cs, R_028014_DB_HTILE_DATA_BASE, a->rsurf->db_htile_data_base);
- reloc_idx = r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rtex->htile, RADEON_USAGE_READWRITE);
+ radeon_set_context_reg(cs, R_02802C_DB_DEPTH_CLEAR, fui(rtex->depth_clear_value));
+ radeon_set_context_reg(cs, R_028D24_DB_HTILE_SURFACE, a->rsurf->db_htile_surface);
+ radeon_set_context_reg(cs, R_028014_DB_HTILE_DATA_BASE, a->rsurf->db_htile_data_base);
+ reloc_idx = r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rtex->htile_buffer,
+ RADEON_USAGE_READWRITE, RADEON_PRIO_DEPTH_META);
cs->buf[cs->cdw++] = PKT3(PKT3_NOP, 0, 0);
cs->buf[cs->cdw++] = reloc_idx;
} else {
- r600_write_context_reg(cs, R_028D24_DB_HTILE_SURFACE, 0);
+ radeon_set_context_reg(cs, R_028D24_DB_HTILE_SURFACE, 0);
}
}
}
db_render_override |= S_028D10_NOOP_CULL_DISABLE(1);
}
- if (rctx->db_state.rsurf && rctx->db_state.rsurf->htile_enabled) {
+ if (rctx->db_state.rsurf && rctx->db_state.rsurf->db_htile_surface) {
/* FORCE_OFF means HiZ/HiS are determined by DB_SHADER_CONTROL */
db_render_override |= S_028D10_FORCE_HIZ_ENABLE(V_028D10_FORCE_OFF);
/* This is to fix a lockup when hyperz and alpha test are enabled at
} else {
db_render_override |= S_028D10_FORCE_HIZ_ENABLE(V_028D10_FORCE_DISABLE);
}
+ if (rctx->b.chip_class == R600 && rctx->framebuffer.nr_samples > 1 && rctx->ps_iter_samples > 0) {
+ /* sample shading and hyperz causes lockups on R6xx chips */
+ db_render_override |= S_028D10_FORCE_HIZ_ENABLE(V_028D10_FORCE_DISABLE);
+ }
if (a->flush_depthstencil_through_cb) {
assert(a->copy_depth || a->copy_stencil);
S_028D0C_STENCIL_COPY_ENABLE(a->copy_stencil) |
S_028D0C_COPY_CENTROID(1) |
S_028D0C_COPY_SAMPLE(a->copy_sample);
+
+ if (rctx->b.chip_class == R600)
+ db_render_override |= S_028D10_NOOP_CULL_DISABLE(1);
+
+ if (rctx->b.family == CHIP_RV610 || rctx->b.family == CHIP_RV630 ||
+ rctx->b.family == CHIP_RV620 || rctx->b.family == CHIP_RV635)
+ db_render_override |= S_028D10_FORCE_HIZ_ENABLE(V_028D10_FORCE_DISABLE);
} else if (a->flush_depthstencil_in_place) {
db_render_control |= S_028D0C_DEPTH_COMPRESS_DISABLE(1) |
S_028D0C_STENCIL_COMPRESS_DISABLE(1);
db_render_control |= S_028D0C_DEPTH_CLEAR_ENABLE(1);
}
- r600_write_context_reg_seq(cs, R_028D0C_DB_RENDER_CONTROL, 2);
+ /* RV770 workaround for a hang with 8x MSAA. */
+ if (rctx->b.family == CHIP_RV770 && a->log_samples == 3) {
+ db_render_override |= S_028D10_MAX_TILES_IN_DTT(6);
+ }
+
+ radeon_set_context_reg_seq(cs, R_028D0C_DB_RENDER_CONTROL, 2);
radeon_emit(cs, db_render_control); /* R_028D0C_DB_RENDER_CONTROL */
radeon_emit(cs, db_render_override); /* R_028D10_DB_RENDER_OVERRIDE */
- r600_write_context_reg(cs, R_02880C_DB_SHADER_CONTROL, a->db_shader_control);
+ radeon_set_context_reg(cs, R_02880C_DB_SHADER_CONTROL, a->db_shader_control);
}
static void r600_emit_config_state(struct r600_context *rctx, struct r600_atom *atom)
struct radeon_winsys_cs *cs = rctx->b.rings.gfx.cs;
struct r600_config_state *a = (struct r600_config_state*)atom;
- r600_write_config_reg(cs, R_008C04_SQ_GPR_RESOURCE_MGMT_1, a->sq_gpr_resource_mgmt_1);
+ radeon_set_config_reg(cs, R_008C04_SQ_GPR_RESOURCE_MGMT_1, a->sq_gpr_resource_mgmt_1);
+ radeon_set_config_reg(cs, R_008C08_SQ_GPR_RESOURCE_MGMT_2, a->sq_gpr_resource_mgmt_2);
}
static void r600_emit_vertex_buffers(struct r600_context *rctx, struct r600_atom *atom)
radeon_emit(cs, 0xc0000000); /* RESOURCEi_WORD6 */
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
- radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rbuffer, RADEON_USAGE_READ));
+ radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rbuffer,
+ RADEON_USAGE_READ, RADEON_PRIO_SHADER_BUFFER_RO));
}
}
struct r600_resource *rbuffer;
unsigned offset;
unsigned buffer_index = ffs(dirty_mask) - 1;
-
+ unsigned gs_ring_buffer = (buffer_index == R600_GS_RING_CONST_BUFFER);
cb = &state->cb[buffer_index];
rbuffer = (struct r600_resource*)cb->buffer;
assert(rbuffer);
offset = cb->buffer_offset;
- r600_write_context_reg(cs, reg_alu_constbuf_size + buffer_index * 4,
- ALIGN_DIVUP(cb->buffer_size >> 4, 16));
- r600_write_context_reg(cs, reg_alu_const_cache + buffer_index * 4, offset >> 8);
+ if (!gs_ring_buffer) {
+ radeon_set_context_reg(cs, reg_alu_constbuf_size + buffer_index * 4,
+ ALIGN_DIVUP(cb->buffer_size >> 4, 16));
+ radeon_set_context_reg(cs, reg_alu_const_cache + buffer_index * 4, offset >> 8);
+ }
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
- radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rbuffer, RADEON_USAGE_READ));
+ radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rbuffer,
+ RADEON_USAGE_READ, RADEON_PRIO_SHADER_BUFFER_RO));
radeon_emit(cs, PKT3(PKT3_SET_RESOURCE, 7, 0));
radeon_emit(cs, (buffer_id_base + buffer_index) * 7);
radeon_emit(cs, offset); /* RESOURCEi_WORD0 */
radeon_emit(cs, rbuffer->buf->size - offset - 1); /* RESOURCEi_WORD1 */
radeon_emit(cs, /* RESOURCEi_WORD2 */
- S_038008_ENDIAN_SWAP(r600_endian_swap(32)) |
- S_038008_STRIDE(16));
+ S_038008_ENDIAN_SWAP(gs_ring_buffer ? ENDIAN_NONE : r600_endian_swap(32)) |
+ S_038008_STRIDE(gs_ring_buffer ? 4 : 16));
radeon_emit(cs, 0); /* RESOURCEi_WORD3 */
radeon_emit(cs, 0); /* RESOURCEi_WORD4 */
radeon_emit(cs, 0); /* RESOURCEi_WORD5 */
radeon_emit(cs, 0xc0000000); /* RESOURCEi_WORD6 */
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
- radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rbuffer, RADEON_USAGE_READ));
+ radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rbuffer,
+ RADEON_USAGE_READ, RADEON_PRIO_SHADER_BUFFER_RO));
dirty_mask &= ~(1 << buffer_index);
}
radeon_emit_array(cs, rview->tex_resource_words, 7);
reloc = r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rview->tex_resource,
- RADEON_USAGE_READ);
+ RADEON_USAGE_READ,
+ rview->tex_resource->b.b.nr_samples > 1 ?
+ RADEON_PRIO_SHADER_TEXTURE_MSAA :
+ RADEON_PRIO_SHADER_TEXTURE_RO);
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
radeon_emit(cs, reloc);
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
offset = border_color_reg;
offset += i * 16;
- r600_write_config_reg_seq(cs, offset, 4);
+ radeon_set_config_reg_seq(cs, offset, 4);
radeon_emit_array(cs, rstate->border_color.ui, 4);
}
}
if (!rctx->seamless_cube_map.enabled) {
tmp |= S_009508_DISABLE_CUBE_WRAP(1);
}
- r600_write_config_reg(cs, R_009508_TA_CNTL_AUX, tmp);
+ radeon_set_config_reg(cs, R_009508_TA_CNTL_AUX, tmp);
}
static void r600_emit_sample_mask(struct r600_context *rctx, struct r600_atom *a)
struct r600_sample_mask *s = (struct r600_sample_mask*)a;
uint8_t mask = s->sample_mask;
- r600_write_context_reg(rctx->b.rings.gfx.cs, R_028C48_PA_SC_AA_MASK,
+ radeon_set_context_reg(rctx->b.rings.gfx.cs, R_028C48_PA_SC_AA_MASK,
mask | (mask << 8) | (mask << 16) | (mask << 24));
}
struct r600_cso_state *state = (struct r600_cso_state*)a;
struct r600_fetch_shader *shader = (struct r600_fetch_shader*)state->cso;
- r600_write_context_reg(cs, R_028894_SQ_PGM_START_FS, shader->offset >> 8);
+ radeon_set_context_reg(cs, R_028894_SQ_PGM_START_FS, shader->offset >> 8);
radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
- radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, shader->buffer, RADEON_USAGE_READ));
+ radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, shader->buffer,
+ RADEON_USAGE_READ, RADEON_PRIO_SHADER_DATA));
+}
+
+static void r600_emit_shader_stages(struct r600_context *rctx, struct r600_atom *a)
+{
+ struct radeon_winsys_cs *cs = rctx->b.rings.gfx.cs;
+ struct r600_shader_stages_state *state = (struct r600_shader_stages_state*)a;
+
+ uint32_t v2 = 0, primid = 0;
+
+ if (rctx->vs_shader->current->shader.vs_as_gs_a) {
+ v2 = S_028A40_MODE(V_028A40_GS_SCENARIO_A);
+ primid = 1;
+ }
+
+ if (state->geom_enable) {
+ uint32_t cut_val;
+
+ if (rctx->gs_shader->gs_max_out_vertices <= 128)
+ cut_val = V_028A40_GS_CUT_128;
+ else if (rctx->gs_shader->gs_max_out_vertices <= 256)
+ cut_val = V_028A40_GS_CUT_256;
+ else if (rctx->gs_shader->gs_max_out_vertices <= 512)
+ cut_val = V_028A40_GS_CUT_512;
+ else
+ cut_val = V_028A40_GS_CUT_1024;
+
+ v2 = S_028A40_MODE(V_028A40_GS_SCENARIO_G) |
+ S_028A40_CUT_MODE(cut_val);
+
+ if (rctx->gs_shader->current->shader.gs_prim_id_input)
+ primid = 1;
+ }
+
+ radeon_set_context_reg(cs, R_028A40_VGT_GS_MODE, v2);
+ radeon_set_context_reg(cs, R_028A84_VGT_PRIMITIVEID_EN, primid);
+}
+
+static void r600_emit_gs_rings(struct r600_context *rctx, struct r600_atom *a)
+{
+ struct radeon_winsys_cs *cs = rctx->b.rings.gfx.cs;
+ struct r600_gs_rings_state *state = (struct r600_gs_rings_state*)a;
+ struct r600_resource *rbuffer;
+
+ radeon_set_config_reg(cs, R_008040_WAIT_UNTIL, S_008040_WAIT_3D_IDLE(1));
+ radeon_emit(cs, PKT3(PKT3_EVENT_WRITE, 0, 0));
+ radeon_emit(cs, EVENT_TYPE(EVENT_TYPE_VGT_FLUSH));
+
+ if (state->enable) {
+ rbuffer =(struct r600_resource*)state->esgs_ring.buffer;
+ radeon_set_config_reg(cs, R_008C40_SQ_ESGS_RING_BASE, 0);
+ radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+ radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rbuffer,
+ RADEON_USAGE_READWRITE,
+ RADEON_PRIO_SHADER_RESOURCE_RW));
+ radeon_set_config_reg(cs, R_008C44_SQ_ESGS_RING_SIZE,
+ state->esgs_ring.buffer_size >> 8);
+
+ rbuffer =(struct r600_resource*)state->gsvs_ring.buffer;
+ radeon_set_config_reg(cs, R_008C48_SQ_GSVS_RING_BASE, 0);
+ radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+ radeon_emit(cs, r600_context_bo_reloc(&rctx->b, &rctx->b.rings.gfx, rbuffer,
+ RADEON_USAGE_READWRITE,
+ RADEON_PRIO_SHADER_RESOURCE_RW));
+ radeon_set_config_reg(cs, R_008C4C_SQ_GSVS_RING_SIZE,
+ state->gsvs_ring.buffer_size >> 8);
+ } else {
+ radeon_set_config_reg(cs, R_008C44_SQ_ESGS_RING_SIZE, 0);
+ radeon_set_config_reg(cs, R_008C4C_SQ_GSVS_RING_SIZE, 0);
+ }
+
+ radeon_set_config_reg(cs, R_008040_WAIT_UNTIL, S_008040_WAIT_3D_IDLE(1));
+ radeon_emit(cs, PKT3(PKT3_EVENT_WRITE, 0, 0));
+ radeon_emit(cs, EVENT_TYPE(EVENT_TYPE_VGT_FLUSH));
}
/* Adjust GPR allocation on R6xx/R7xx */
bool r600_adjust_gprs(struct r600_context *rctx)
{
unsigned num_ps_gprs = rctx->ps_shader->current->shader.bc.ngpr;
- unsigned num_vs_gprs = rctx->vs_shader->current->shader.bc.ngpr;
+ unsigned num_vs_gprs, num_es_gprs, num_gs_gprs;
unsigned new_num_ps_gprs = num_ps_gprs;
- unsigned new_num_vs_gprs = num_vs_gprs;
+ unsigned new_num_vs_gprs, new_num_es_gprs, new_num_gs_gprs;
unsigned cur_num_ps_gprs = G_008C04_NUM_PS_GPRS(rctx->config_state.sq_gpr_resource_mgmt_1);
unsigned cur_num_vs_gprs = G_008C04_NUM_VS_GPRS(rctx->config_state.sq_gpr_resource_mgmt_1);
+ unsigned cur_num_gs_gprs = G_008C08_NUM_GS_GPRS(rctx->config_state.sq_gpr_resource_mgmt_2);
+ unsigned cur_num_es_gprs = G_008C08_NUM_ES_GPRS(rctx->config_state.sq_gpr_resource_mgmt_2);
unsigned def_num_ps_gprs = rctx->default_ps_gprs;
unsigned def_num_vs_gprs = rctx->default_vs_gprs;
+ unsigned def_num_gs_gprs = 0;
+ unsigned def_num_es_gprs = 0;
unsigned def_num_clause_temp_gprs = rctx->r6xx_num_clause_temp_gprs;
/* hardware will reserve twice num_clause_temp_gprs */
- unsigned max_gprs = def_num_ps_gprs + def_num_vs_gprs + def_num_clause_temp_gprs * 2;
- unsigned tmp;
+ unsigned max_gprs = def_num_gs_gprs + def_num_es_gprs + def_num_ps_gprs + def_num_vs_gprs + def_num_clause_temp_gprs * 2;
+ unsigned tmp, tmp2;
+
+ if (rctx->gs_shader) {
+ num_es_gprs = rctx->vs_shader->current->shader.bc.ngpr;
+ num_gs_gprs = rctx->gs_shader->current->shader.bc.ngpr;
+ num_vs_gprs = rctx->gs_shader->current->gs_copy_shader->shader.bc.ngpr;
+ } else {
+ num_es_gprs = 0;
+ num_gs_gprs = 0;
+ num_vs_gprs = rctx->vs_shader->current->shader.bc.ngpr;
+ }
+ new_num_vs_gprs = num_vs_gprs;
+ new_num_es_gprs = num_es_gprs;
+ new_num_gs_gprs = num_gs_gprs;
/* the sum of all SQ_GPR_RESOURCE_MGMT*.NUM_*_GPRS must <= to max_gprs */
- if (new_num_ps_gprs > cur_num_ps_gprs || new_num_vs_gprs > cur_num_vs_gprs) {
+ if (new_num_ps_gprs > cur_num_ps_gprs || new_num_vs_gprs > cur_num_vs_gprs ||
+ new_num_es_gprs > cur_num_es_gprs || new_num_gs_gprs > cur_num_gs_gprs) {
/* try to use switch back to default */
- if (new_num_ps_gprs > def_num_ps_gprs || new_num_vs_gprs > def_num_vs_gprs) {
+ if (new_num_ps_gprs > def_num_ps_gprs || new_num_vs_gprs > def_num_vs_gprs ||
+ new_num_gs_gprs > def_num_gs_gprs || new_num_es_gprs > def_num_es_gprs) {
/* always privilege vs stage so that at worst we have the
* pixel stage producing wrong output (not the vertex
* stage) */
- new_num_ps_gprs = max_gprs - (new_num_vs_gprs + def_num_clause_temp_gprs * 2);
+ new_num_ps_gprs = max_gprs - ((new_num_vs_gprs + new_num_es_gprs + new_num_gs_gprs) + def_num_clause_temp_gprs * 2);
new_num_vs_gprs = num_vs_gprs;
+ new_num_gs_gprs = num_gs_gprs;
+ new_num_es_gprs = num_es_gprs;
} else {
new_num_ps_gprs = def_num_ps_gprs;
new_num_vs_gprs = def_num_vs_gprs;
+ new_num_es_gprs = def_num_es_gprs;
+ new_num_gs_gprs = def_num_gs_gprs;
}
} else {
return true;
* it will lockup. So in this case just discard the draw command
* and don't change the current gprs repartitions.
*/
- if (num_ps_gprs > new_num_ps_gprs || num_vs_gprs > new_num_vs_gprs) {
- R600_ERR("ps & vs shader require too many register (%d + %d) "
+ if (num_ps_gprs > new_num_ps_gprs || num_vs_gprs > new_num_vs_gprs ||
+ num_gs_gprs > new_num_gs_gprs || num_es_gprs > new_num_es_gprs) {
+ R600_ERR("shaders require too many register (%d + %d + %d + %d) "
"for a combined maximum of %d\n",
- num_ps_gprs, num_vs_gprs, max_gprs);
+ num_ps_gprs, num_vs_gprs, num_es_gprs, num_gs_gprs, max_gprs);
return false;
}
tmp = S_008C04_NUM_PS_GPRS(new_num_ps_gprs) |
S_008C04_NUM_VS_GPRS(new_num_vs_gprs) |
S_008C04_NUM_CLAUSE_TEMP_GPRS(def_num_clause_temp_gprs);
- if (rctx->config_state.sq_gpr_resource_mgmt_1 != tmp) {
+
+ tmp2 = S_008C08_NUM_ES_GPRS(new_num_es_gprs) |
+ S_008C08_NUM_GS_GPRS(new_num_gs_gprs);
+ if (rctx->config_state.sq_gpr_resource_mgmt_1 != tmp || rctx->config_state.sq_gpr_resource_mgmt_2 != tmp2) {
rctx->config_state.sq_gpr_resource_mgmt_1 = tmp;
- rctx->config_state.atom.dirty = true;
+ rctx->config_state.sq_gpr_resource_mgmt_2 = tmp2;
+ r600_mark_atom_dirty(rctx, &rctx->config_state.atom);
rctx->b.flags |= R600_CONTEXT_WAIT_3D_IDLE;
}
return true;
int num_es_stack_entries;
enum radeon_family family;
struct r600_command_buffer *cb = &rctx->start_cs_cmd;
- uint32_t tmp;
+ uint32_t tmp, i;
r600_init_command_buffer(cb, 256);
num_es_stack_entries = 16;
break;
case CHIP_RV770:
- num_ps_gprs = 192;
+ num_ps_gprs = 130;
num_vs_gprs = 56;
num_temp_gprs = 4;
- num_gs_gprs = 0;
- num_es_gprs = 0;
- num_ps_threads = 188;
+ num_gs_gprs = 31;
+ num_es_gprs = 31;
+ num_ps_threads = 180;
num_vs_threads = 60;
- num_gs_threads = 0;
- num_es_threads = 0;
- num_ps_stack_entries = 256;
- num_vs_stack_entries = 256;
- num_gs_stack_entries = 0;
- num_es_stack_entries = 0;
+ num_gs_threads = 4;
+ num_es_threads = 4;
+ num_ps_stack_entries = 128;
+ num_vs_stack_entries = 128;
+ num_gs_stack_entries = 128;
+ num_es_stack_entries = 128;
break;
case CHIP_RV730:
case CHIP_RV740:
num_temp_gprs = 4;
num_gs_gprs = 0;
num_es_gprs = 0;
- num_ps_threads = 188;
+ num_ps_threads = 180;
num_vs_threads = 60;
- num_gs_threads = 0;
- num_es_threads = 0;
+ num_gs_threads = 4;
+ num_es_threads = 4;
num_ps_stack_entries = 128;
num_vs_stack_entries = 128;
num_gs_stack_entries = 0;
num_temp_gprs = 4;
num_gs_gprs = 0;
num_es_gprs = 0;
- num_ps_threads = 144;
+ num_ps_threads = 136;
num_vs_threads = 48;
- num_gs_threads = 0;
- num_es_threads = 0;
+ num_gs_threads = 4;
+ num_es_threads = 4;
num_ps_stack_entries = 128;
num_vs_stack_entries = 128;
num_gs_stack_entries = 0;
r600_store_config_reg(cb, R_009714_VC_ENHANCE, 0);
if (rctx->b.chip_class >= R700) {
+ r600_store_context_reg(cb, R_028A50_VGT_ENHANCE, 4);
r600_store_config_reg(cb, R_008D8C_SQ_DYN_GPR_CNTL_PS_FLUSH_REQ, 0x00004000);
r600_store_config_reg(cb, R_009830_DB_DEBUG, 0);
r600_store_config_reg(cb, R_009838_DB_WATERMARKS, 0x00420204);
r600_store_value(cb, 0); /* R_0288C8_SQ_GS_VERT_ITEMSIZE */
/* to avoid GPU doing any preloading of constant from random address */
- r600_store_context_reg_seq(cb, R_028140_ALU_CONST_BUFFER_SIZE_PS_0, 8);
- r600_store_value(cb, 0); /* R_028140_ALU_CONST_BUFFER_SIZE_PS_0 */
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_context_reg_seq(cb, R_028180_ALU_CONST_BUFFER_SIZE_VS_0, 8);
- r600_store_value(cb, 0); /* R_028180_ALU_CONST_BUFFER_SIZE_VS_0 */
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
- r600_store_value(cb, 0);
+ r600_store_context_reg_seq(cb, R_028140_ALU_CONST_BUFFER_SIZE_PS_0, 16);
+ for (i = 0; i < 16; i++)
+ r600_store_value(cb, 0);
+
+ r600_store_context_reg_seq(cb, R_028180_ALU_CONST_BUFFER_SIZE_VS_0, 16);
+ for (i = 0; i < 16; i++)
+ r600_store_value(cb, 0);
+
+ r600_store_context_reg_seq(cb, R_0281C0_ALU_CONST_BUFFER_SIZE_GS_0, 16);
+ for (i = 0; i < 16; i++)
+ r600_store_value(cb, 0);
r600_store_context_reg_seq(cb, R_028A10_VGT_OUTPUT_PATH_CNTL, 13);
r600_store_value(cb, 0); /* R_028A10_VGT_OUTPUT_PATH_CNTL */
r600_store_context_reg(cb, R_028AA0_VGT_INSTANCE_STEP_RATE_0, 0);
r600_store_context_reg(cb, R_028AA4_VGT_INSTANCE_STEP_RATE_1, 0);
- r600_store_context_reg_seq(cb, R_028AB0_VGT_STRMOUT_EN, 3);
- r600_store_value(cb, 0); /* R_028AB0_VGT_STRMOUT_EN */
+ r600_store_context_reg_seq(cb, R_028AB4_VGT_REUSE_OFF, 2);
r600_store_value(cb, 1); /* R_028AB4_VGT_REUSE_OFF */
r600_store_value(cb, 0); /* R_028AB8_VGT_VTX_CNT_EN */
r600_store_context_reg(cb, R_028A48_PA_SC_MPASS_PS_CNTL, 0);
r600_store_context_reg_seq(cb, R_028C0C_PA_CL_GB_VERT_CLIP_ADJ, 4);
- r600_store_value(cb, 0x3F800000); /* R_028C0C_PA_CL_GB_VERT_CLIP_ADJ */
- r600_store_value(cb, 0x3F800000); /* R_028C10_PA_CL_GB_VERT_DISC_ADJ */
- r600_store_value(cb, 0x3F800000); /* R_028C14_PA_CL_GB_HORZ_CLIP_ADJ */
- r600_store_value(cb, 0x3F800000); /* R_028C18_PA_CL_GB_HORZ_DISC_ADJ */
-
- r600_store_context_reg_seq(cb, R_0282D0_PA_SC_VPORT_ZMIN_0, 2);
- r600_store_value(cb, 0); /* R_0282D0_PA_SC_VPORT_ZMIN_0 */
- r600_store_value(cb, 0x3F800000); /* R_0282D4_PA_SC_VPORT_ZMAX_0 */
+ r600_store_value(cb, fui(1.0)); /* R_028C0C_PA_CL_GB_VERT_CLIP_ADJ */
+ r600_store_value(cb, fui(1.0)); /* R_028C10_PA_CL_GB_VERT_DISC_ADJ */
+ r600_store_value(cb, fui(1.0)); /* R_028C14_PA_CL_GB_HORZ_CLIP_ADJ */
+ r600_store_value(cb, fui(1.0)); /* R_028C18_PA_CL_GB_HORZ_DISC_ADJ */
- r600_store_context_reg(cb, R_028818_PA_CL_VTE_CNTL, 0x43F);
+ r600_store_context_reg_seq(cb, R_0282D0_PA_SC_VPORT_ZMIN_0, 2 * R600_MAX_VIEWPORTS);
+ for (tmp = 0; tmp < R600_MAX_VIEWPORTS; tmp++) {
+ r600_store_value(cb, 0); /* R_0282D0_PA_SC_VPORT_ZMIN_0 */
+ r600_store_value(cb, fui(1.0)); /* R_0282D4_PA_SC_VPORT_ZMAX_0 */
+ }
r600_store_context_reg(cb, R_028200_PA_SC_WINDOW_OFFSET, 0);
r600_store_context_reg(cb, R_02820C_PA_SC_CLIPRECT_RULE, 0xFFFF);
r600_store_value(cb, 0); /* R_028240_PA_SC_GENERIC_SCISSOR_TL */
r600_store_value(cb, S_028244_BR_X(8192) | S_028244_BR_Y(8192)); /* R_028244_PA_SC_GENERIC_SCISSOR_BR */
- r600_store_context_reg_seq(cb, R_0288CC_SQ_PGM_CF_OFFSET_PS, 2);
+ r600_store_context_reg_seq(cb, R_0288CC_SQ_PGM_CF_OFFSET_PS, 5);
r600_store_value(cb, 0); /* R_0288CC_SQ_PGM_CF_OFFSET_PS */
r600_store_value(cb, 0); /* R_0288D0_SQ_PGM_CF_OFFSET_VS */
+ r600_store_value(cb, 0); /* R_0288D4_SQ_PGM_CF_OFFSET_GS */
+ r600_store_value(cb, 0); /* R_0288D8_SQ_PGM_CF_OFFSET_ES */
+ r600_store_value(cb, 0); /* R_0288DC_SQ_PGM_CF_OFFSET_FS */
r600_store_context_reg(cb, R_0288E0_SQ_VTX_SEMANTIC_CLEAR, ~0);
r600_store_value(cb, 0); /* R_028404_VGT_MIN_VTX_INDX */
r600_store_context_reg(cb, R_0288A4_SQ_PGM_RESOURCES_FS, 0);
- r600_store_context_reg(cb, R_0288DC_SQ_PGM_CF_OFFSET_FS, 0);
- if (rctx->b.chip_class == R700 && rctx->screen->has_streamout)
+ if (rctx->b.chip_class == R700)
+ r600_store_context_reg(cb, R_028350_SX_MISC, 0);
+ if (rctx->b.chip_class == R700 && rctx->screen->b.has_streamout)
r600_store_context_reg(cb, R_028354_SX_SURFACE_SYNC, S_028354_SURFACE_SYNC_MASK(0xf));
+
r600_store_context_reg(cb, R_028800_DB_DEPTH_CONTROL, 0);
- if (rctx->screen->has_streamout) {
+ if (rctx->screen->b.has_streamout) {
r600_store_context_reg(cb, R_028B28_VGT_STRMOUT_DRAW_OPAQUE_OFFSET, 0);
}
r600_store_loop_const(cb, R_03E200_SQ_LOOP_CONST_0, 0x1000FFF);
r600_store_loop_const(cb, R_03E200_SQ_LOOP_CONST_0 + (32 * 4), 0x1000FFF);
+ r600_store_loop_const(cb, R_03E200_SQ_LOOP_CONST_0 + (64 * 4), 0x1000FFF);
}
void r600_update_ps_state(struct pipe_context *ctx, struct r600_pipe_shader *shader)
struct r600_command_buffer *cb = &shader->command_buffer;
struct r600_shader *rshader = &shader->shader;
unsigned i, exports_ps, num_cout, spi_ps_in_control_0, spi_input_z, spi_ps_in_control_1, db_shader_control;
- int pos_index = -1, face_index = -1;
+ int pos_index = -1, face_index = -1, fixed_pt_position_index = -1;
unsigned tmp, sid, ufi = 0;
int need_linear = 0;
- unsigned z_export = 0, stencil_export = 0;
+ unsigned z_export = 0, stencil_export = 0, mask_export = 0;
unsigned sprite_coord_enable = rctx->rasterizer ? rctx->rasterizer->sprite_coord_enable : 0;
if (!cb->buf) {
for (i = 0; i < rshader->ninput; i++) {
if (rshader->input[i].name == TGSI_SEMANTIC_POSITION)
pos_index = i;
- if (rshader->input[i].name == TGSI_SEMANTIC_FACE)
+ if (rshader->input[i].name == TGSI_SEMANTIC_FACE && face_index == -1)
face_index = i;
+ if (rshader->input[i].name == TGSI_SEMANTIC_SAMPLEID)
+ fixed_pt_position_index = i;
sid = rshader->input[i].spi_sid;
tmp |= S_028644_PT_SPRITE_TEX(1);
}
- if (rshader->input[i].centroid)
+ if (rshader->input[i].interpolate_location == TGSI_INTERPOLATE_LOC_CENTROID)
tmp |= S_028644_SEL_CENTROID(1);
+ if (rshader->input[i].interpolate_location == TGSI_INTERPOLATE_LOC_SAMPLE)
+ tmp |= S_028644_SEL_SAMPLE(1);
+
if (rshader->input[i].interpolate == TGSI_INTERPOLATE_LINEAR) {
need_linear = 1;
tmp |= S_028644_SEL_LINEAR(1);
z_export = 1;
if (rshader->output[i].name == TGSI_SEMANTIC_STENCIL)
stencil_export = 1;
+ if (rshader->output[i].name == TGSI_SEMANTIC_SAMPLEMASK &&
+ rctx->framebuffer.nr_samples > 1 && rctx->ps_iter_samples > 0)
+ mask_export = 1;
}
db_shader_control |= S_02880C_Z_EXPORT_ENABLE(z_export);
db_shader_control |= S_02880C_STENCIL_REF_EXPORT_ENABLE(stencil_export);
+ db_shader_control |= S_02880C_MASK_EXPORT_ENABLE(mask_export);
if (rshader->uses_kill)
db_shader_control |= S_02880C_KILL_ENABLE(1);
exports_ps = 0;
for (i = 0; i < rshader->noutput; i++) {
if (rshader->output[i].name == TGSI_SEMANTIC_POSITION ||
- rshader->output[i].name == TGSI_SEMANTIC_STENCIL) {
+ rshader->output[i].name == TGSI_SEMANTIC_STENCIL ||
+ rshader->output[i].name == TGSI_SEMANTIC_SAMPLEMASK) {
exports_ps |= 1;
}
}
spi_input_z = 0;
if (pos_index != -1) {
spi_ps_in_control_0 |= (S_0286CC_POSITION_ENA(1) |
- S_0286CC_POSITION_CENTROID(rshader->input[pos_index].centroid) |
+ S_0286CC_POSITION_CENTROID(rshader->input[pos_index].interpolate_location == TGSI_INTERPOLATE_LOC_CENTROID) |
S_0286CC_POSITION_ADDR(rshader->input[pos_index].gpr) |
- S_0286CC_BARYC_SAMPLE_CNTL(1));
+ S_0286CC_BARYC_SAMPLE_CNTL(1)) |
+ S_0286CC_POSITION_SAMPLE(rshader->input[pos_index].interpolate_location == TGSI_INTERPOLATE_LOC_SAMPLE);
spi_input_z |= S_0286D8_PROVIDE_Z_TO_SPI(1);
}
spi_ps_in_control_1 |= S_0286D0_FRONT_FACE_ENA(1) |
S_0286D0_FRONT_FACE_ADDR(rshader->input[face_index].gpr);
}
+ if (fixed_pt_position_index != -1) {
+ spi_ps_in_control_1 |= S_0286D0_FIXED_PT_POSITION_ENA(1) |
+ S_0286D0_FIXED_PT_POSITION_ADDR(rshader->input[fixed_pt_position_index].gpr);
+ }
/* HW bug in original R600 */
if (rctx->b.family == CHIP_R600)
/* only set some bits here, the other bits are set in the dsa state */
shader->db_shader_control = db_shader_control;
- shader->ps_depth_export = z_export | stencil_export;
+ shader->ps_depth_export = z_export | stencil_export | mask_export;
shader->sprite_coord_enable = sprite_coord_enable;
if (rctx->rasterizer)
r600_store_context_reg(cb, R_028868_SQ_PGM_RESOURCES_VS,
S_028868_NUM_GPRS(rshader->bc.ngpr) |
S_028868_STACK_SIZE(rshader->bc.nstack));
+ if (rshader->vs_position_window_space) {
+ r600_store_context_reg(cb, R_028818_PA_CL_VTE_CNTL,
+ S_028818_VTX_XY_FMT(1) | S_028818_VTX_Z_FMT(1));
+ } else {
+ r600_store_context_reg(cb, R_028818_PA_CL_VTE_CNTL,
+ S_028818_VTX_W0_FMT(1) |
+ S_028818_VPORT_X_SCALE_ENA(1) | S_028818_VPORT_X_OFFSET_ENA(1) |
+ S_028818_VPORT_Y_SCALE_ENA(1) | S_028818_VPORT_Y_OFFSET_ENA(1) |
+ S_028818_VPORT_Z_SCALE_ENA(1) | S_028818_VPORT_Z_OFFSET_ENA(1));
+
+ }
r600_store_context_reg(cb, R_028858_SQ_PGM_START_VS, 0);
/* After that, the NOP relocation packet must be emitted (shader->bo, RADEON_USAGE_READ). */
S_02881C_VS_OUT_CCDIST0_VEC_ENA((rshader->clip_dist_write & 0x0F) != 0) |
S_02881C_VS_OUT_CCDIST1_VEC_ENA((rshader->clip_dist_write & 0xF0) != 0) |
S_02881C_VS_OUT_MISC_VEC_ENA(rshader->vs_out_misc_write) |
- S_02881C_USE_VTX_POINT_SIZE(rshader->vs_out_point_size);
+ S_02881C_USE_VTX_POINT_SIZE(rshader->vs_out_point_size) |
+ S_02881C_USE_VTX_EDGE_FLAG(rshader->vs_out_edgeflag) |
+ S_02881C_USE_VTX_RENDER_TARGET_INDX(rshader->vs_out_layer) |
+ S_02881C_USE_VTX_VIEWPORT_INDX(rshader->vs_out_viewport);
}
+void r600_update_gs_state(struct pipe_context *ctx, struct r600_pipe_shader *shader)
+{
+ struct r600_context *rctx = (struct r600_context *)ctx;
+ struct r600_command_buffer *cb = &shader->command_buffer;
+ struct r600_shader *rshader = &shader->shader;
+ struct r600_shader *cp_shader = &shader->gs_copy_shader->shader;
+ unsigned gsvs_itemsize =
+ (cp_shader->ring_item_size * shader->selector->gs_max_out_vertices) >> 2;
+
+ r600_init_command_buffer(cb, 64);
+
+ /* VGT_GS_MODE is written by r600_emit_shader_stages */
+ r600_store_context_reg(cb, R_028AB8_VGT_VTX_CNT_EN, 1);
+
+ if (rctx->b.chip_class >= R700) {
+ r600_store_context_reg(cb, R_028B38_VGT_GS_MAX_VERT_OUT,
+ S_028B38_MAX_VERT_OUT(shader->selector->gs_max_out_vertices));
+ }
+ r600_store_context_reg(cb, R_028A6C_VGT_GS_OUT_PRIM_TYPE,
+ r600_conv_prim_to_gs_out(shader->selector->gs_output_prim));
+
+ r600_store_context_reg(cb, R_0288C8_SQ_GS_VERT_ITEMSIZE,
+ cp_shader->ring_item_size >> 2);
+
+ r600_store_context_reg(cb, R_0288A8_SQ_ESGS_RING_ITEMSIZE,
+ (rshader->ring_item_size) >> 2);
+
+ r600_store_context_reg(cb, R_0288AC_SQ_GSVS_RING_ITEMSIZE,
+ gsvs_itemsize);
+
+ /* FIXME calculate these values somehow ??? */
+ r600_store_config_reg_seq(cb, R_0088C8_VGT_GS_PER_ES, 2);
+ r600_store_value(cb, 0x80); /* GS_PER_ES */
+ r600_store_value(cb, 0x100); /* ES_PER_GS */
+ r600_store_config_reg_seq(cb, R_0088E8_VGT_GS_PER_VS, 1);
+ r600_store_value(cb, 0x2); /* GS_PER_VS */
+
+ r600_store_context_reg(cb, R_02887C_SQ_PGM_RESOURCES_GS,
+ S_02887C_NUM_GPRS(rshader->bc.ngpr) |
+ S_02887C_STACK_SIZE(rshader->bc.nstack));
+ r600_store_context_reg(cb, R_02886C_SQ_PGM_START_GS, 0);
+ /* After that, the NOP relocation packet must be emitted (shader->bo, RADEON_USAGE_READ). */
+}
+
+void r600_update_es_state(struct pipe_context *ctx, struct r600_pipe_shader *shader)
+{
+ struct r600_command_buffer *cb = &shader->command_buffer;
+ struct r600_shader *rshader = &shader->shader;
+
+ r600_init_command_buffer(cb, 32);
+
+ r600_store_context_reg(cb, R_028890_SQ_PGM_RESOURCES_ES,
+ S_028890_NUM_GPRS(rshader->bc.ngpr) |
+ S_028890_STACK_SIZE(rshader->bc.nstack));
+ r600_store_context_reg(cb, R_028880_SQ_PGM_START_ES, 0);
+ /* After that, the NOP relocation packet must be emitted (shader->bo, RADEON_USAGE_READ). */
+}
+
+
void *r600_create_resolve_blend(struct r600_context *rctx)
{
struct pipe_blend_state blend;
void r600_update_db_shader_control(struct r600_context * rctx)
{
- bool dual_export = rctx->framebuffer.export_16bpc &&
- !rctx->ps_shader->current->ps_depth_export;
+ bool dual_export;
+ unsigned db_shader_control;
- unsigned db_shader_control = rctx->ps_shader->current->db_shader_control |
- S_02880C_DUAL_EXPORT_ENABLE(dual_export);
+ if (!rctx->ps_shader) {
+ return;
+ }
+
+ dual_export = rctx->framebuffer.export_16bpc &&
+ !rctx->ps_shader->current->ps_depth_export;
+
+ db_shader_control = rctx->ps_shader->current->db_shader_control |
+ S_02880C_DUAL_EXPORT_ENABLE(dual_export);
/* When alpha test is enabled we can't trust the hw to make the proper
* decision on the order in which ztest should be run related to fragment
if (db_shader_control != rctx->db_misc_state.db_shader_control) {
rctx->db_misc_state.db_shader_control = db_shader_control;
- rctx->db_misc_state.atom.dirty = true;
+ r600_mark_atom_dirty(rctx, &rctx->db_misc_state.atom);
}
}
-static INLINE unsigned r600_array_mode(unsigned mode)
+static inline unsigned r600_array_mode(unsigned mode)
{
switch (mode) {
case RADEON_SURF_MODE_LINEAR_ALIGNED: return V_0280A0_ARRAY_LINEAR_ALIGNED;
unsigned ncopy, height, cheight, detile, i, x, y, z, src_mode, dst_mode;
uint64_t base, addr;
- /* make sure that the dma ring is only one active */
- rctx->b.rings.gfx.flush(rctx, RADEON_FLUSH_ASYNC);
-
dst_mode = rdst->surface.level[dst_level].mode;
src_mode = rsrc->surface.level[src_level].mode;
/* downcast linear aligned to linear to simplify test */
y = 0;
lbpp = util_logbase2(bpp);
- pitch_tile_max = ((pitch / bpp) >> 3) - 1;
+ pitch_tile_max = ((pitch / bpp) / 8) - 1;
if (dst_mode == RADEON_SURF_MODE_LINEAR) {
/* T2L */
array_mode = r600_array_mode(src_mode);
- slice_tile_max = (rsrc->surface.level[src_level].nblk_x * rsrc->surface.level[src_level].nblk_y) >> 6;
+ slice_tile_max = (rsrc->surface.level[src_level].nblk_x * rsrc->surface.level[src_level].nblk_y) / (8*8);
slice_tile_max = slice_tile_max ? slice_tile_max - 1 : 0;
/* linear height must be the same as the slice tile max height, it's ok even
* if the linear destination/source have smaller heigh as the size of the
} else {
/* L2T */
array_mode = r600_array_mode(dst_mode);
- slice_tile_max = (rdst->surface.level[dst_level].nblk_x * rdst->surface.level[dst_level].nblk_y) >> 6;
+ slice_tile_max = (rdst->surface.level[dst_level].nblk_x * rdst->surface.level[dst_level].nblk_y) / (8*8);
slice_tile_max = slice_tile_max ? slice_tile_max - 1 : 0;
/* linear height must be the same as the slice tile max height, it's ok even
* if the linear destination/source have smaller heigh as the size of the
addr += src_y * pitch + src_x * bpp;
}
/* check that we are in dw/base alignment constraint */
- if ((addr & 0x3) || (base & 0xff)) {
+ if (addr % 4 || base % 256) {
return FALSE;
}
/* It's a r6xx/r7xx limitation, the blit must be on 8 boundary for number
* line in the blit. Compute max 8 line we can copy in the size limit
*/
- cheight = ((0x0000ffff << 2) / pitch) & 0xfffffff8;
+ cheight = ((R600_DMA_COPY_MAX_SIZE_DW * 4) / pitch) & 0xfffffff8;
ncopy = (copy_height / cheight) + !!(copy_height % cheight);
- r600_need_dma_space(rctx, ncopy * 7);
+ r600_need_dma_space(&rctx->b, ncopy * 7);
for (i = 0; i < ncopy; i++) {
cheight = cheight > copy_height ? copy_height : cheight;
- size = (cheight * pitch) >> 2;
- /* emit reloc before writting cs so that cs is always in consistent state */
- r600_context_bo_reloc(&rctx->b, &rctx->b.rings.dma, &rsrc->resource, RADEON_USAGE_READ);
- r600_context_bo_reloc(&rctx->b, &rctx->b.rings.dma, &rdst->resource, RADEON_USAGE_WRITE);
+ size = (cheight * pitch) / 4;
+ /* emit reloc before writing cs so that cs is always in consistent state */
+ r600_context_bo_reloc(&rctx->b, &rctx->b.rings.dma, &rsrc->resource, RADEON_USAGE_READ,
+ RADEON_PRIO_MIN);
+ r600_context_bo_reloc(&rctx->b, &rctx->b.rings.dma, &rdst->resource, RADEON_USAGE_WRITE,
+ RADEON_PRIO_MIN);
cs->buf[cs->cdw++] = DMA_PACKET(DMA_PACKET_COPY, 1, 0, size);
cs->buf[cs->cdw++] = base >> 8;
cs->buf[cs->cdw++] = (detile << 31) | (array_mode << 27) |
return TRUE;
}
-static boolean r600_dma_blit(struct pipe_context *ctx,
- struct pipe_resource *dst,
- unsigned dst_level,
- unsigned dst_x, unsigned dst_y, unsigned dst_z,
- struct pipe_resource *src,
- unsigned src_level,
- const struct pipe_box *src_box)
+static void r600_dma_copy(struct pipe_context *ctx,
+ struct pipe_resource *dst,
+ unsigned dst_level,
+ unsigned dstx, unsigned dsty, unsigned dstz,
+ struct pipe_resource *src,
+ unsigned src_level,
+ const struct pipe_box *src_box)
{
struct r600_context *rctx = (struct r600_context *)ctx;
struct r600_texture *rsrc = (struct r600_texture*)src;
unsigned dst_pitch, src_pitch, bpp, dst_mode, src_mode, copy_height;
unsigned src_w, dst_w;
unsigned src_x, src_y;
+ unsigned dst_x = dstx, dst_y = dsty, dst_z = dstz;
if (rctx->b.rings.dma.cs == NULL) {
- return FALSE;
+ goto fallback;
}
- if (src->format != dst->format) {
- return FALSE;
+
+ if (dst->target == PIPE_BUFFER && src->target == PIPE_BUFFER) {
+ if (dst_x % 4 || src_box->x % 4 || src_box->width % 4)
+ goto fallback;
+
+ r600_dma_copy_buffer(rctx, dst, src, dst_x, src_box->x, src_box->width);
+ return;
+ }
+
+ if (src->format != dst->format || src_box->depth > 1) {
+ goto fallback;
}
src_x = util_format_get_nblocksx(src->format, src_box->x);
dst_mode = dst_mode == RADEON_SURF_MODE_LINEAR_ALIGNED ? RADEON_SURF_MODE_LINEAR : dst_mode;
if (src_pitch != dst_pitch || src_box->x || dst_x || src_w != dst_w) {
- /* strick requirement on r6xx/r7xx */
- return FALSE;
+ /* strict requirement on r6xx/r7xx */
+ goto fallback;
}
/* lot of constraint on alignment this should capture them all */
- if ((src_pitch & 0x7) || (src_box->y & 0x7) || (dst_y & 0x7)) {
- return FALSE;
+ if (src_pitch % 8 || src_box->y % 8 || dst_y % 8) {
+ goto fallback;
}
if (src_mode == dst_mode) {
dst_offset += dst_y * dst_pitch + dst_x * bpp;
size = src_box->height * src_pitch;
/* must be dw aligned */
- if ((dst_offset & 0x3) || (src_offset & 0x3) || (size & 0x3)) {
- return FALSE;
+ if (dst_offset % 4 || src_offset % 4 || size % 4) {
+ goto fallback;
}
- r600_dma_copy(rctx, dst, src, dst_offset, src_offset, size);
+ r600_dma_copy_buffer(rctx, dst, src, dst_offset, src_offset, size);
} else {
- return r600_dma_copy_tile(rctx, dst, dst_level, dst_x, dst_y, dst_z,
+ if (!r600_dma_copy_tile(rctx, dst, dst_level, dst_x, dst_y, dst_z,
src, src_level, src_x, src_y, src_box->z,
- copy_height, dst_pitch, bpp);
+ copy_height, dst_pitch, bpp)) {
+ goto fallback;
+ }
}
- return TRUE;
+ return;
+
+fallback:
+ r600_resource_copy_region(ctx, dst, dst_level, dstx, dsty, dstz,
+ src, src_level, src_box);
}
void r600_init_state_functions(struct r600_context *rctx)
{
unsigned id = 4;
+ int i;
/* !!!
* To avoid GPU lockup registers must be emited in a specific order
r600_init_atom(rctx, &rctx->samplers[PIPE_SHADER_FRAGMENT].views.atom, id++, r600_emit_ps_sampler_views, 0);
r600_init_atom(rctx, &rctx->vertex_buffer_state.atom, id++, r600_emit_vertex_buffers, 0);
- r600_init_atom(rctx, &rctx->vgt_state.atom, id++, r600_emit_vgt_state, 7);
+ r600_init_atom(rctx, &rctx->vgt_state.atom, id++, r600_emit_vgt_state, 10);
r600_init_atom(rctx, &rctx->seamless_cube_map.atom, id++, r600_emit_seamless_cube_map, 3);
r600_init_atom(rctx, &rctx->sample_mask.atom, id++, r600_emit_sample_mask, 3);
r600_init_atom(rctx, &rctx->dsa_state.atom, id++, r600_emit_cso_state, 0);
r600_init_atom(rctx, &rctx->poly_offset_state.atom, id++, r600_emit_polygon_offset, 6);
r600_init_atom(rctx, &rctx->rasterizer_state.atom, id++, r600_emit_cso_state, 0);
- r600_init_atom(rctx, &rctx->scissor.atom, id++, r600_emit_scissor_state, 4);
+ for (i = 0;i < R600_MAX_VIEWPORTS; i++) {
+ r600_init_atom(rctx, &rctx->scissor[i].atom, id++, r600_emit_scissor_state, 4);
+ r600_init_atom(rctx, &rctx->viewport[i].atom, id++, r600_emit_viewport_state, 8);
+ rctx->scissor[i].idx = i;
+ rctx->viewport[i].idx = i;
+ }
r600_init_atom(rctx, &rctx->config_state.atom, id++, r600_emit_config_state, 3);
r600_init_atom(rctx, &rctx->stencil_ref.atom, id++, r600_emit_stencil_ref, 4);
- r600_init_atom(rctx, &rctx->viewport.atom, id++, r600_emit_viewport_state, 8);
r600_init_atom(rctx, &rctx->vertex_fetch_shader.atom, id++, r600_emit_vertex_fetch_shader, 5);
- rctx->atoms[id++] = &rctx->b.streamout.begin_atom;
+ r600_add_atom(rctx, &rctx->b.streamout.begin_atom, id++);
+ r600_add_atom(rctx, &rctx->b.streamout.enable_atom, id++);
r600_init_atom(rctx, &rctx->vertex_shader.atom, id++, r600_emit_shader, 23);
r600_init_atom(rctx, &rctx->pixel_shader.atom, id++, r600_emit_shader, 0);
+ r600_init_atom(rctx, &rctx->geometry_shader.atom, id++, r600_emit_shader, 0);
+ r600_init_atom(rctx, &rctx->export_shader.atom, id++, r600_emit_shader, 0);
+ r600_init_atom(rctx, &rctx->shader_stages.atom, id++, r600_emit_shader_stages, 0);
+ r600_init_atom(rctx, &rctx->gs_rings.atom, id++, r600_emit_gs_rings, 0);
rctx->b.b.create_blend_state = r600_create_blend_state;
rctx->b.b.create_depth_stencil_alpha_state = r600_create_dsa_state;
rctx->b.b.create_sampler_view = r600_create_sampler_view;
rctx->b.b.set_framebuffer_state = r600_set_framebuffer_state;
rctx->b.b.set_polygon_stipple = r600_set_polygon_stipple;
+ rctx->b.b.set_min_samples = r600_set_min_samples;
rctx->b.b.set_scissor_states = r600_set_scissor_states;
rctx->b.b.get_sample_position = r600_get_sample_position;
- rctx->b.dma_copy = r600_dma_blit;
+ rctx->b.dma_copy = r600_dma_copy;
}
/* this function must be last */
projects / mesa.git / blobdiff