June 2022 - Wine-Devel - list.winehq.org

[PATCH 1/5] winegstreamer: Introduce new wg_transform_(push|read)_mf helpers.

by Rémi Bernon

From: Rémi Bernon <rbernon(a)codeweavers.com> To read MF sample properties before pushing, and update them after sucessfully reading a sample. Signed-off-by: Rémi Bernon <rbernon(a)codeweavers.com> --- dlls/winegstreamer/gst_private.h | 4 +++ dlls/winegstreamer/h264_decoder.c | 6 ++-- dlls/winegstreamer/mfplat.c | 55 +++++++++++++++++++++---------- dlls/winegstreamer/wma_decoder.c | 4 +-- 4 files changed, 45 insertions(+), 24 deletions(-) diff --git a/dlls/winegstreamer/gst_private.h b/dlls/winegstreamer/gst_private.h index 159143d7e54..ca7396ad5cd 100644 --- a/dlls/winegstreamer/gst_private.h +++ b/dlls/winegstreamer/gst_private.h @@ -123,6 +123,10 @@ void mf_media_type_to_wg_format(IMFMediaType *type, struct wg_format *format); HRESULT mf_create_wg_sample(IMFSample *sample, struct wg_sample **out); void mf_destroy_wg_sample(struct wg_sample *wg_sample); +HRESULT wg_transform_push_mf(struct wg_transform *transform, struct wg_sample *sample); +HRESULT wg_transform_read_mf(struct wg_transform *transform, struct wg_sample *sample, + struct wg_format *format); + HRESULT winegstreamer_stream_handler_create(REFIID riid, void **obj); HRESULT h264_decoder_create(REFIID riid, void **ret); diff --git a/dlls/winegstreamer/h264_decoder.c b/dlls/winegstreamer/h264_decoder.c index 19a36a9a77a..0ee9084e12e 100644 --- a/dlls/winegstreamer/h264_decoder.c +++ b/dlls/winegstreamer/h264_decoder.c @@ -543,8 +543,7 @@ static HRESULT WINAPI transform_ProcessInput(IMFTransform *iface, DWORD id, IMFS if (FAILED(hr = mf_create_wg_sample(sample, &wg_sample))) return hr; - hr = wg_transform_push_data(decoder->wg_transform, wg_sample); - + hr = wg_transform_push_mf(decoder->wg_transform, wg_sample); mf_destroy_wg_sample(wg_sample); return hr; } @@ -583,8 +582,7 @@ static HRESULT WINAPI transform_ProcessOutput(IMFTransform *iface, DWORD flags, return MF_E_BUFFERTOOSMALL; } - hr = wg_transform_read_data(decoder->wg_transform, wg_sample, - &wg_format); + hr = wg_transform_read_mf(decoder->wg_transform, wg_sample, &wg_format); mf_destroy_wg_sample(wg_sample); if (hr == MF_E_TRANSFORM_STREAM_CHANGE) diff --git a/dlls/winegstreamer/mfplat.c b/dlls/winegstreamer/mfplat.c index 0226e7a2e45..026237bdf5d 100644 --- a/dlls/winegstreamer/mfplat.c +++ b/dlls/winegstreamer/mfplat.c @@ -969,8 +969,6 @@ HRESULT mf_create_wg_sample(IMFSample *sample, struct wg_sample **out) { DWORD current_length, max_length; struct mf_sample *mf_sample; - LONGLONG time, duration; - UINT32 value; BYTE *buffer; HRESULT hr; @@ -981,19 +979,6 @@ HRESULT mf_create_wg_sample(IMFSample *sample, struct wg_sample **out) if (FAILED(hr = IMFMediaBuffer_Lock(mf_sample->media_buffer, &buffer, &max_length, &current_length))) goto out; - if (SUCCEEDED(IMFSample_GetSampleTime(sample, &time))) - { - mf_sample->wg_sample.flags |= WG_SAMPLE_FLAG_HAS_PTS; - mf_sample->wg_sample.pts = time; - } - if (SUCCEEDED(IMFSample_GetSampleDuration(sample, &duration))) - { - mf_sample->wg_sample.flags |= WG_SAMPLE_FLAG_HAS_DURATION; - mf_sample->wg_sample.duration = duration; - } - if (SUCCEEDED(IMFSample_GetUINT32(sample, &MFSampleExtension_CleanPoint, &value)) && value) - mf_sample->wg_sample.flags |= WG_SAMPLE_FLAG_SYNC_POINT; - IMFSample_AddRef((mf_sample->sample = sample)); mf_sample->wg_sample.data = buffer; mf_sample->wg_sample.size = current_length; @@ -1015,8 +1000,43 @@ void mf_destroy_wg_sample(struct wg_sample *wg_sample) struct mf_sample *mf_sample = CONTAINING_RECORD(wg_sample, struct mf_sample, wg_sample); IMFMediaBuffer_Unlock(mf_sample->media_buffer); - IMFMediaBuffer_SetCurrentLength(mf_sample->media_buffer, wg_sample->size); IMFMediaBuffer_Release(mf_sample->media_buffer); + IMFSample_Release(mf_sample->sample); + free(mf_sample); +} + +HRESULT wg_transform_push_mf(struct wg_transform *transform, struct wg_sample *sample) +{ + struct mf_sample *mf_sample = CONTAINING_RECORD(sample, struct mf_sample, wg_sample); + LONGLONG time, duration; + UINT32 value; + + if (SUCCEEDED(IMFSample_GetSampleTime(mf_sample->sample, &time))) + { + mf_sample->wg_sample.flags |= WG_SAMPLE_FLAG_HAS_PTS; + mf_sample->wg_sample.pts = time; + } + if (SUCCEEDED(IMFSample_GetSampleDuration(mf_sample->sample, &duration))) + { + mf_sample->wg_sample.flags |= WG_SAMPLE_FLAG_HAS_DURATION; + mf_sample->wg_sample.duration = duration; + } + if (SUCCEEDED(IMFSample_GetUINT32(mf_sample->sample, &MFSampleExtension_CleanPoint, &value)) && value) + mf_sample->wg_sample.flags |= WG_SAMPLE_FLAG_SYNC_POINT; + + return wg_transform_push_data(transform, sample); +} + +HRESULT wg_transform_read_mf(struct wg_transform *transform, struct wg_sample *wg_sample, + struct wg_format *format) +{ + struct mf_sample *mf_sample = CONTAINING_RECORD(wg_sample, struct mf_sample, wg_sample); + HRESULT hr; + + if (FAILED(hr = wg_transform_read_data(transform, wg_sample, format))) + return hr; + + IMFMediaBuffer_SetCurrentLength(mf_sample->media_buffer, wg_sample->size); if (wg_sample->flags & WG_SAMPLE_FLAG_HAS_PTS) IMFSample_SetSampleTime(mf_sample->sample, wg_sample->pts); @@ -1025,6 +1045,5 @@ void mf_destroy_wg_sample(struct wg_sample *wg_sample) if (wg_sample->flags & WG_SAMPLE_FLAG_SYNC_POINT) IMFSample_SetUINT32(mf_sample->sample, &MFSampleExtension_CleanPoint, 1); - IMFSample_Release(mf_sample->sample); - free(mf_sample); + return S_OK; } diff --git a/dlls/winegstreamer/wma_decoder.c b/dlls/winegstreamer/wma_decoder.c index 106d32adce9..fdb68328c3a 100644 --- a/dlls/winegstreamer/wma_decoder.c +++ b/dlls/winegstreamer/wma_decoder.c @@ -544,7 +544,7 @@ static HRESULT WINAPI transform_ProcessInput(IMFTransform *iface, DWORD id, IMFS return S_OK; } - hr = wg_transform_push_data(decoder->wg_transform, wg_sample); + hr = wg_transform_push_mf(decoder->wg_transform, wg_sample); mf_destroy_wg_sample(wg_sample); return hr; } @@ -586,7 +586,7 @@ static HRESULT WINAPI transform_ProcessOutput(IMFTransform *iface, DWORD flags, return MF_E_BUFFERTOOSMALL; } - if (SUCCEEDED(hr = wg_transform_read_data(decoder->wg_transform, wg_sample, NULL))) + if (SUCCEEDED(hr = wg_transform_read_mf(decoder->wg_transform, wg_sample, NULL))) { if (wg_sample->flags & WG_SAMPLE_FLAG_INCOMPLETE) samples[0].dwStatus |= MFT_OUTPUT_DATA_BUFFER_INCOMPLETE; -- GitLab https://gitlab.winehq.org/wine/wine/-/merge_requests/197

June 7, 2022

[PATCH 0/5] MR197: winegstreamer: Add support for zero-copy in wg_transform.

by Rémi Bernon

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/197

June 7, 2022

[PATCH 1/1] winegstreamer: Return void from align_video_info_planes.

by Rémi Bernon

From: Rémi Bernon <rbernon(a)codeweavers.com> Some older GStreamer versions return void from gst_video_info_align. Signed-off-by: Rémi Bernon <rbernon(a)codeweavers.com> --- dlls/winegstreamer/wg_transform.c | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/dlls/winegstreamer/wg_transform.c b/dlls/winegstreamer/wg_transform.c index fb852b4cf3d..9a5c16afe74 100644 --- a/dlls/winegstreamer/wg_transform.c +++ b/dlls/winegstreamer/wg_transform.c @@ -70,7 +70,7 @@ static bool is_caps_video(GstCaps *caps) return g_str_has_prefix(media_type, "video/"); } -static bool align_video_info_planes(gsize plane_align, GstVideoInfo *info, GstVideoAlignment *align) +static void align_video_info_planes(gsize plane_align, GstVideoInfo *info, GstVideoAlignment *align) { gst_video_alignment_reset(align); @@ -81,7 +81,7 @@ static bool align_video_info_planes(gsize plane_align, GstVideoInfo *info, GstVi align->stride_align[2] = plane_align; align->stride_align[3] = plane_align; - return gst_video_info_align(info, align); + gst_video_info_align(info, align); } static GstFlowReturn transform_sink_chain_cb(GstPad *pad, GstObject *parent, GstBuffer *buffer) @@ -130,10 +130,11 @@ static gboolean transform_sink_query_cb(GstPad *pad, GstObject *parent, GstQuery break; if (!gst_video_info_from_caps(&info, caps) - || !align_video_info_planes(plane_align, &info, &align) || !(pool = gst_video_buffer_pool_new())) break; + align_video_info_planes(plane_align, &info, &align); + if ((params = gst_structure_new("video-meta", "padding-top", G_TYPE_UINT, align.padding_top, "padding-bottom", G_TYPE_UINT, align.padding_bottom, @@ -546,11 +547,8 @@ static bool copy_video_buffer(GstBuffer *buffer, GstCaps *caps, gsize plane_alig } dst_info = src_info; - if (!align_video_info_planes(plane_align, &dst_info, &align)) - { - GST_ERROR("Failed to align video info."); - return false; - } + align_video_info_planes(plane_align, &dst_info, &align); + if (sample->max_size < dst_info.size) { GST_ERROR("Output buffer is too small."); @@ -703,9 +701,10 @@ NTSTATUS wg_transform_read_data(void *args) wg_format_from_caps(format, output_caps); if (format->major_type == WG_MAJOR_TYPE_VIDEO - && gst_video_info_from_caps(&info, output_caps) - && align_video_info_planes(plane_align, &info, &align)) + && gst_video_info_from_caps(&info, output_caps)) { + align_video_info_planes(plane_align, &info, &align); + GST_INFO("Returning video alignment left %u, top %u, right %u, bottom %u.", align.padding_left, align.padding_top, align.padding_right, align.padding_bottom); -- GitLab https://gitlab.winehq.org/wine/wine/-/merge_requests/196

June 7, 2022

[PATCH 0/1] MR196: winegstreamer: Return void from align_video_info_planes.

by Rémi Bernon

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/196

June 7, 2022

Re: [PATCH 0/11] MR193: wineandroid: PE conversion - approved

by Huw Davies (＠huw)

This merge request was approved by Huw Davies. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/193

June 7, 2022

[PATCH] msvcp90: Fix segfault in bad_alloc exception default constructor.

by Torge Matthies

Signed-off-by: Torge Matthies <tmatthies(a)codeweavers.com> --- EXCEPTION_NAME needs a pointer variable passed in, not an array. I removed the cast from the macro so that in the future a warning is displayed if this happens again. dlls/msvcp90/exception.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/dlls/msvcp90/exception.c b/dlls/msvcp90/exception.c index 7f1dfd2188a1..effe57063a14 100644 --- a/dlls/msvcp90/exception.c +++ b/dlls/msvcp90/exception.c @@ -36,7 +36,7 @@ int* __cdecl __processing_throw(void); #if _MSVCP_VER >= 70 || defined(_MSVCIRT) typedef const char **exception_name; #define EXCEPTION_STR(name) (*name) -#define EXCEPTION_NAME(str) ((exception_name)&str) +#define EXCEPTION_NAME(str) (&str) #else typedef const char *exception_name; #define EXCEPTION_STR(name) (name) @@ -96,7 +96,7 @@ exception* __thiscall exception_copy_ctor(exception *this, const exception *rhs) this->name = rhs->name; this->do_free = FALSE; } else - MSVCP_exception_ctor(this, EXCEPTION_NAME(rhs->name)); + MSVCP_exception_ctor(this, (exception_name)EXCEPTION_NAME(rhs->name)); TRACE("name = %s\n", this->name); return this; } @@ -214,7 +214,7 @@ bad_alloc* __thiscall MSVCP_bad_alloc_ctor(bad_alloc *this, exception_name name) DEFINE_THISCALL_WRAPPER(MSVCP_bad_alloc_default_ctor, 4) bad_alloc* __thiscall MSVCP_bad_alloc_default_ctor(bad_alloc *this) { - static const char name[] = "bad allocation"; + static const char *name = "bad allocation"; return MSVCP_bad_alloc_ctor(this, EXCEPTION_NAME(name)); } -- 2.36.1

June 7, 2022

[PATCH vkd3d v4 9/9] vkd3d-shader/hlsl: Parse the mul() intrinsic.

by Giovanni Mascellani

Signed-off-by: Giovanni Mascellani <gmascellani(a)codeweavers.com> --- libs/vkd3d-shader/hlsl.y | 118 ++++++++++++++++++++++++ tests/hlsl-majority-pragma.shader_test | 2 +- tests/hlsl-majority-typedef.shader_test | 2 +- tests/hlsl-mul.shader_test | 36 ++++---- 4 files changed, 138 insertions(+), 20 deletions(-) diff --git a/libs/vkd3d-shader/hlsl.y b/libs/vkd3d-shader/hlsl.y index 2aee5153..df5fda47 100644 --- a/libs/vkd3d-shader/hlsl.y +++ b/libs/vkd3d-shader/hlsl.y @@ -2046,6 +2046,123 @@ static bool intrinsic_min(struct hlsl_ctx *ctx, return !!add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_MIN, params->args[0], params->args[1], loc); } +static bool intrinsic_mul(struct hlsl_ctx *ctx, + const struct parse_initializer *params, const struct vkd3d_shader_location *loc) +{ + struct hlsl_ir_node *arg1 = params->args[0], *arg2 = params->args[1], *cast1, *cast2; + enum hlsl_base_type base = expr_common_base_type(arg1->data_type->base_type, arg2->data_type->base_type); + struct hlsl_type *cast_type1 = arg1->data_type, *cast_type2 = arg2->data_type, *matrix_type, *ret_type; + unsigned int i, j, k, vect_count = 0; + struct vkd3d_string_buffer *name; + static unsigned int counter = 0; + struct hlsl_ir_load *load; + struct hlsl_ir_var *var; + + if (arg1->data_type->type == HLSL_CLASS_SCALAR || arg2->data_type->type == HLSL_CLASS_SCALAR) + return !!add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_MUL, arg1, arg2, loc); + + if (arg1->data_type->type == HLSL_CLASS_VECTOR) + { + vect_count++; + cast_type1 = hlsl_get_matrix_type(ctx, base, arg1->data_type->dimx, 1); + } + if (arg2->data_type->type == HLSL_CLASS_VECTOR) + { + vect_count++; + cast_type2 = hlsl_get_matrix_type(ctx, base, 1, arg2->data_type->dimx); + } + + matrix_type = hlsl_get_matrix_type(ctx, base, cast_type2->dimx, cast_type1->dimy); + + if (vect_count == 0) + { + ret_type = matrix_type; + } + else if (vect_count == 1) + { + assert(matrix_type->dimx == 1 || matrix_type->dimy == 1); + ret_type = hlsl_get_vector_type(ctx, base, matrix_type->dimx * matrix_type->dimy); + } + else + { + assert(matrix_type->dimx == 1 && matrix_type->dimy == 1); + ret_type = hlsl_get_scalar_type(ctx, base); + } + + if (!(cast1 = add_implicit_conversion(ctx, params->instrs, arg1, cast_type1, loc))) + return false; + + if (!(cast2 = add_implicit_conversion(ctx, params->instrs, arg2, cast_type2, loc))) + return false; + + name = vkd3d_string_buffer_get(&ctx->string_buffers); + vkd3d_string_buffer_printf(name, "<mul-%u>", counter++); + var = hlsl_new_synthetic_var(ctx, name->buffer, matrix_type, *loc); + vkd3d_string_buffer_release(&ctx->string_buffers, name); + if (!var) + return false; + + for (i = 0; i < matrix_type->dimx; ++i) + for (j = 0; j < matrix_type->dimy; ++j) + { + struct hlsl_ir_node *node = NULL; + struct hlsl_type *scalar_type; + struct hlsl_ir_store *store; + struct hlsl_ir_constant *c; + unsigned int offset; + + for (k = 0; k < cast_type1->dimx && k < cast_type2->dimy; ++k) + { + struct hlsl_ir_load *value1, *value2; + struct hlsl_ir_node *mul; + + offset = hlsl_compute_component_offset(ctx, cast_type1, j * cast_type1->dimx + k, &scalar_type); + if (!(c = hlsl_new_uint_constant(ctx, offset, loc))) + return false; + list_add_tail(params->instrs, &c->node.entry); + + if (!(value1 = add_load(ctx, params->instrs, cast1, &c->node, scalar_type, *loc))) + return false; + + offset = hlsl_compute_component_offset(ctx, cast_type2, k * cast_type2->dimx + i, &scalar_type); + if (!(c = hlsl_new_uint_constant(ctx, offset, loc))) + return false; + list_add_tail(params->instrs, &c->node.entry); + + if (!(value2 = add_load(ctx, params->instrs, cast2, &c->node, scalar_type, *loc))) + return false; + + if (!(mul = add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_MUL, &value1->node, &value2->node, loc))) + return false; + + if (node) + { + if (!(node = add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_ADD, node, mul, loc))) + return false; + } + else + { + node = mul; + } + } + + offset = hlsl_compute_component_offset(ctx, matrix_type, j * matrix_type->dimx + i, &scalar_type); + if (!(c = hlsl_new_uint_constant(ctx, offset, loc))) + return false; + list_add_tail(params->instrs, &c->node.entry); + + if (!(store = hlsl_new_store(ctx, var, &c->node, node, 0, *loc))) + return false; + list_add_tail(params->instrs, &store->node.entry); + } + + if (!(load = hlsl_new_load(ctx, var, NULL, matrix_type, *loc))) + return false; + list_add_tail(params->instrs, &load->node.entry); + + return !!add_implicit_conversion(ctx, params->instrs, &load->node, ret_type, loc); +} + static bool intrinsic_pow(struct hlsl_ctx *ctx, const struct parse_initializer *params, const struct vkd3d_shader_location *loc) { @@ -2106,6 +2223,7 @@ intrinsic_functions[] = {"floor", 1, true, intrinsic_floor}, {"max", 2, true, intrinsic_max}, {"min", 2, true, intrinsic_min}, + {"mul", 2, true, intrinsic_mul}, {"pow", 2, true, intrinsic_pow}, {"round", 1, true, intrinsic_round}, {"saturate", 1, true, intrinsic_saturate}, diff --git a/tests/hlsl-majority-pragma.shader_test b/tests/hlsl-majority-pragma.shader_test index a9f917ff..e7fc75cd 100644 --- a/tests/hlsl-majority-pragma.shader_test +++ b/tests/hlsl-majority-pragma.shader_test @@ -17,5 +17,5 @@ uniform 0 float4 0.1 0.2 0.0 0.0 uniform 4 float4 0.3 0.4 0.0 0.0 uniform 8 float4 0.1 0.3 0.0 0.0 uniform 12 float4 0.2 0.4 0.0 0.0 -todo draw quad +draw quad probe all rgba (0.17, 0.39, 0.17, 0.39) 1 diff --git a/tests/hlsl-majority-typedef.shader_test b/tests/hlsl-majority-typedef.shader_test index 192c96db..1460e9a0 100644 --- a/tests/hlsl-majority-typedef.shader_test +++ b/tests/hlsl-majority-typedef.shader_test @@ -18,5 +18,5 @@ uniform 0 float4 0.1 0.2 0.0 0.0 uniform 4 float4 0.3 0.4 0.0 0.0 uniform 8 float4 0.1 0.3 0.0 0.0 uniform 12 float4 0.2 0.4 0.0 0.0 -todo draw quad +draw quad probe all rgba (0.17, 0.39, 0.17, 0.39) 1 diff --git a/tests/hlsl-mul.shader_test b/tests/hlsl-mul.shader_test index 1d137e69..7b453187 100644 --- a/tests/hlsl-mul.shader_test +++ b/tests/hlsl-mul.shader_test @@ -12,7 +12,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (30.0, 70.0, 110.0, 150.0) [pixel shader] @@ -28,7 +28,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (90.0, 100.0, 110.0, 120.0) [pixel shader] @@ -44,7 +44,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (14.0, 38.0, 62.0, 86.0) [pixel shader] @@ -60,7 +60,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (38.0, 44.0, 50.0, 56.0) [pixel shader] @@ -75,7 +75,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (14.0, 32.0, 50.0, 0.0) [pixel shader] @@ -90,7 +90,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (30.0, 36.0, 42.0, 0.0) [pixel shader] @@ -106,7 +106,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (50.0, 60.0, 70.0, 80.0) [pixel shader] @@ -122,7 +122,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (50.0, 60.0, 70.0, 80.0) [pixel shader] @@ -138,7 +138,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (10.0, 20.0, 30.0, 40.0) [pixel shader] @@ -154,7 +154,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (10.0, 50.0, 90.0, 130.0) [pixel shader] @@ -170,7 +170,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (10.0, 20.0, 30.0, 40.0) [pixel shader] @@ -186,7 +186,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (10.0, 50.0, 90.0, 130.0) [pixel shader] @@ -202,7 +202,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (90.0, 100.0, 110.0, 120.0) [pixel shader] @@ -218,7 +218,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (5.0, 10.0, 15.0, 20.0) [pixel shader] @@ -234,7 +234,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -250,7 +250,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (30.0, 70.0, 110.0, 150.0) [pixel shader] @@ -268,7 +268,7 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (83.0, 98.0, 113.0, 128.0) [pixel shader] @@ -286,5 +286,5 @@ float4 main(float4 pos : sv_position) : sv_target } [test] -todo draw quad +draw quad probe all rgba (78.0, 96.0, 114.0, 0.0) -- 2.36.1

June 7, 2022

[PATCH vkd3d v4 8/9] vkd3d-shader/hlsl: Split matrix operations.

by Giovanni Mascellani

Signed-off-by: Giovanni Mascellani <gmascellani(a)codeweavers.com> --- libs/vkd3d-shader/hlsl.y | 74 ++++++++++++++++++++++++++++++++++++ tests/hlsl-shape.shader_test | 30 +++++++-------- 2 files changed, 89 insertions(+), 15 deletions(-) diff --git a/libs/vkd3d-shader/hlsl.y b/libs/vkd3d-shader/hlsl.y index 687569e8..2aee5153 100644 --- a/libs/vkd3d-shader/hlsl.y +++ b/libs/vkd3d-shader/hlsl.y @@ -1201,6 +1201,22 @@ static bool expr_common_shape(struct hlsl_ctx *ctx, struct hlsl_type *t1, struct return true; } +static unsigned int minor_size(const struct hlsl_type *type) +{ + if (type->modifiers & HLSL_MODIFIER_ROW_MAJOR) + return type->dimx; + else + return type->dimy; +} + +static unsigned int major_size(const struct hlsl_type *type) +{ + if (type->modifiers & HLSL_MODIFIER_ROW_MAJOR) + return type->dimy; + else + return type->dimx; +} + static struct hlsl_ir_node *add_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *operands[HLSL_MAX_OPERANDS], struct hlsl_type *type, const struct vkd3d_shader_location *loc) @@ -1208,6 +1224,64 @@ static struct hlsl_ir_node *add_expr(struct hlsl_ctx *ctx, struct list *instrs, struct hlsl_ir_expr *expr; unsigned int i; + if (type->type == HLSL_CLASS_MATRIX) + { + struct vkd3d_string_buffer *name; + static unsigned int counter = 0; + struct hlsl_type *vector_type; + struct hlsl_ir_load *load; + struct hlsl_ir_var *var; + + vector_type = hlsl_get_vector_type(ctx, type->base_type, minor_size(type)); + + name = vkd3d_string_buffer_get(&ctx->string_buffers); + vkd3d_string_buffer_printf(name, "<split_op-%u>", counter++); + var = hlsl_new_synthetic_var(ctx, name->buffer, type, *loc); + vkd3d_string_buffer_release(&ctx->string_buffers, name); + if (!var) + return NULL; + + for (i = 0; i < major_size(type); i++) + { + struct hlsl_ir_node *value, *vector_operands[HLSL_MAX_OPERANDS] = { NULL }; + struct hlsl_ir_store *store; + struct hlsl_ir_constant *c; + unsigned int j; + + if (!(c = hlsl_new_uint_constant(ctx, 4 * i, loc))) + return NULL; + list_add_tail(instrs, &c->node.entry); + + for (j = 0; j < HLSL_MAX_OPERANDS; j++) + { + if (operands[j]) + { + struct hlsl_type *vector_arg_type; + struct hlsl_ir_load *load; + + vector_arg_type = hlsl_get_vector_type(ctx, operands[j]->data_type->base_type, minor_size(type)); + + if (!(load = add_load(ctx, instrs, operands[j], &c->node, vector_arg_type, *loc))) + return NULL; + vector_operands[j] = &load->node; + } + } + + if (!(value = add_expr(ctx, instrs, op, vector_operands, vector_type, loc))) + return NULL; + + if (!(store = hlsl_new_store(ctx, var, &c->node, value, 0, *loc))) + return NULL; + list_add_tail(instrs, &store->node.entry); + } + + if (!(load = hlsl_new_load(ctx, var, NULL, type, *loc))) + return NULL; + list_add_tail(instrs, &load->node.entry); + + return &load->node; + } + if (!(expr = hlsl_alloc(ctx, sizeof(*expr)))) return NULL; init_node(&expr->node, HLSL_IR_EXPR, type, *loc); diff --git a/tests/hlsl-shape.shader_test b/tests/hlsl-shape.shader_test index 65cc322c..b96f0fd2 100644 --- a/tests/hlsl-shape.shader_test +++ b/tests/hlsl-shape.shader_test @@ -93,7 +93,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 7.0, 9.0) [pixel shader] @@ -107,7 +107,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 7.0, 9.0) [pixel shader] @@ -122,7 +122,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -137,7 +137,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -152,7 +152,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 7.0, 12.0, 17.0) [pixel shader] @@ -167,7 +167,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 7.0, 12.0, 17.0) [pixel shader] @@ -183,7 +183,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 0.0) [pixel shader] @@ -199,7 +199,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (9.0, 11.0, 13.0, 0.0) [pixel shader] @@ -223,7 +223,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -247,7 +247,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -273,7 +273,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -285,7 +285,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 3.0, 4.0, 5.0) [pixel shader] @@ -297,7 +297,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 3.0, 4.0, 5.0) [pixel shader] @@ -336,7 +336,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 3.0, 4.0, 5.0) [pixel shader] @@ -351,5 +351,5 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (6.0, 7.0, 8.0, 9.0) -- 2.36.1

June 7, 2022

[PATCH vkd3d v4 7/9] vkd3d-shader/hlsl: Have add_expr() return hlsl_ir_node*.

by Giovanni Mascellani

Signed-off-by: Giovanni Mascellani <gmascellani(a)codeweavers.com> Signed-off-by: Zebediah Figura <zfigura(a)codeweavers.com> --- libs/vkd3d-shader/hlsl.y | 54 +++++++++++++++++----------------------- 1 file changed, 23 insertions(+), 31 deletions(-) diff --git a/libs/vkd3d-shader/hlsl.y b/libs/vkd3d-shader/hlsl.y index f85d4252..687569e8 100644 --- a/libs/vkd3d-shader/hlsl.y +++ b/libs/vkd3d-shader/hlsl.y @@ -674,7 +674,7 @@ static bool add_record_load(struct hlsl_ctx *ctx, struct list *instrs, struct hl return !!add_load(ctx, instrs, record, &c->node, field->type, loc); } -static struct hlsl_ir_expr *add_binary_arithmetic_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_binary_arithmetic_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg1, struct hlsl_ir_node *arg2, const struct vkd3d_shader_location *loc); @@ -682,8 +682,7 @@ static struct hlsl_ir_node *add_matrix_scalar_load(struct hlsl_ctx *ctx, struct struct hlsl_ir_node *matrix, struct hlsl_ir_node *x, struct hlsl_ir_node *y, const struct vkd3d_shader_location *loc) { - struct hlsl_ir_node *major, *minor; - struct hlsl_ir_expr *mul, *add; + struct hlsl_ir_node *major, *minor, *mul, *add; struct hlsl_ir_constant *four; struct hlsl_ir_load *load; struct hlsl_type *type = matrix->data_type, *scalar_type; @@ -708,10 +707,10 @@ static struct hlsl_ir_node *add_matrix_scalar_load(struct hlsl_ctx *ctx, struct if (!(mul = add_binary_arithmetic_expr(ctx, instrs, HLSL_OP2_MUL, &four->node, major, loc))) return NULL; - if (!(add = add_binary_arithmetic_expr(ctx, instrs, HLSL_OP2_ADD, &mul->node, minor, loc))) + if (!(add = add_binary_arithmetic_expr(ctx, instrs, HLSL_OP2_ADD, mul, minor, loc))) return NULL; - if (!(load = add_load(ctx, instrs, matrix, &add->node, scalar_type, *loc))) + if (!(load = add_load(ctx, instrs, matrix, add, scalar_type, *loc))) return NULL; return &load->node; @@ -1202,7 +1201,7 @@ static bool expr_common_shape(struct hlsl_ctx *ctx, struct hlsl_type *t1, struct return true; } -static struct hlsl_ir_expr *add_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *operands[HLSL_MAX_OPERANDS], struct hlsl_type *type, const struct vkd3d_shader_location *loc) { @@ -1217,7 +1216,7 @@ static struct hlsl_ir_expr *add_expr(struct hlsl_ctx *ctx, struct list *instrs, hlsl_src_from_node(&expr->operands[i], operands[i]); list_add_tail(instrs, &expr->node.entry); - return expr; + return &expr->node; } static void check_integer_type(struct hlsl_ctx *ctx, const struct hlsl_ir_node *instr) @@ -1241,7 +1240,7 @@ static void check_integer_type(struct hlsl_ctx *ctx, const struct hlsl_ir_node * } } -static struct hlsl_ir_expr *add_unary_arithmetic_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_unary_arithmetic_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg, const struct vkd3d_shader_location *loc) { struct hlsl_ir_node *args[HLSL_MAX_OPERANDS] = {arg}; @@ -1249,7 +1248,7 @@ static struct hlsl_ir_expr *add_unary_arithmetic_expr(struct hlsl_ctx *ctx, stru return add_expr(ctx, instrs, op, args, arg->data_type, loc); } -static struct hlsl_ir_expr *add_unary_bitwise_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_unary_bitwise_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg, const struct vkd3d_shader_location *loc) { check_integer_type(ctx, arg); @@ -1257,7 +1256,7 @@ static struct hlsl_ir_expr *add_unary_bitwise_expr(struct hlsl_ctx *ctx, struct return add_unary_arithmetic_expr(ctx, instrs, op, arg, loc); } -static struct hlsl_ir_expr *add_unary_logical_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_unary_logical_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg, const struct vkd3d_shader_location *loc) { struct hlsl_ir_node *args[HLSL_MAX_OPERANDS] = {0}; @@ -1272,7 +1271,7 @@ static struct hlsl_ir_expr *add_unary_logical_expr(struct hlsl_ctx *ctx, struct return add_expr(ctx, instrs, op, args, bool_type, loc); } -static struct hlsl_ir_expr *add_binary_arithmetic_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_binary_arithmetic_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg1, struct hlsl_ir_node *arg2, const struct vkd3d_shader_location *loc) { @@ -1307,7 +1306,7 @@ static struct list *add_binary_arithmetic_expr_merge(struct hlsl_ctx *ctx, struc return list1; } -static struct hlsl_ir_expr *add_binary_bitwise_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_binary_bitwise_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg1, struct hlsl_ir_node *arg2, const struct vkd3d_shader_location *loc) { @@ -1329,7 +1328,7 @@ static struct list *add_binary_bitwise_expr_merge(struct hlsl_ctx *ctx, struct l return list1; } -static struct hlsl_ir_expr *add_binary_comparison_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_binary_comparison_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg1, struct hlsl_ir_node *arg2, struct vkd3d_shader_location *loc) { @@ -1365,7 +1364,7 @@ static struct list *add_binary_comparison_expr_merge(struct hlsl_ctx *ctx, struc return list1; } -static struct hlsl_ir_expr *add_binary_logical_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_binary_logical_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg1, struct hlsl_ir_node *arg2, const struct vkd3d_shader_location *loc) { @@ -1400,7 +1399,7 @@ static struct list *add_binary_logical_expr_merge(struct hlsl_ctx *ctx, struct l return list1; } -static struct hlsl_ir_expr *add_binary_shift_expr(struct hlsl_ctx *ctx, struct list *instrs, +static struct hlsl_ir_node *add_binary_shift_expr(struct hlsl_ctx *ctx, struct list *instrs, enum hlsl_ir_expr_op op, struct hlsl_ir_node *arg1, struct hlsl_ir_node *arg2, const struct vkd3d_shader_location *loc) { @@ -1509,22 +1508,17 @@ static struct hlsl_ir_node *add_assignment(struct hlsl_ctx *ctx, struct list *in if (assign_op == ASSIGN_OP_SUB) { - struct hlsl_ir_expr *expr; - - if (!(expr = add_unary_arithmetic_expr(ctx, instrs, HLSL_OP1_NEG, rhs, &rhs->loc))) + if (!(rhs = add_unary_arithmetic_expr(ctx, instrs, HLSL_OP1_NEG, rhs, &rhs->loc))) return NULL; - rhs = &expr->node; assign_op = ASSIGN_OP_ADD; } if (assign_op != ASSIGN_OP_ASSIGN) { enum hlsl_ir_expr_op op = op_from_assignment(assign_op); - struct hlsl_ir_expr *expr; assert(op); - if (!(expr = add_binary_arithmetic_expr(ctx, instrs, op, lhs, rhs, &rhs->loc))) + if (!(rhs = add_binary_arithmetic_expr(ctx, instrs, op, lhs, rhs, &rhs->loc))) return NULL; - rhs = &expr->node; } if (lhs_type->type <= HLSL_CLASS_LAST_NUMERIC) @@ -1894,12 +1888,12 @@ static bool intrinsic_abs(struct hlsl_ctx *ctx, static bool intrinsic_clamp(struct hlsl_ctx *ctx, const struct parse_initializer *params, const struct vkd3d_shader_location *loc) { - struct hlsl_ir_expr *max; + struct hlsl_ir_node *max; if (!(max = add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_MAX, params->args[0], params->args[1], loc))) return false; - return !!add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_MIN, &max->node, params->args[2], loc); + return !!add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_MIN, max, params->args[2], loc); } static bool intrinsic_cross(struct hlsl_ctx *ctx, @@ -1907,8 +1901,7 @@ static bool intrinsic_cross(struct hlsl_ctx *ctx, { struct hlsl_ir_swizzle *arg1_swzl1, *arg1_swzl2, *arg2_swzl1, *arg2_swzl2; struct hlsl_ir_node *arg1 = params->args[0], *arg2 = params->args[1]; - struct hlsl_ir_node *arg1_cast, *arg2_cast, *mul1_neg; - struct hlsl_ir_expr *mul1, *mul2; + struct hlsl_ir_node *arg1_cast, *arg2_cast, *mul1_neg, *mul1, *mul2; struct hlsl_type *cast_type; enum hlsl_base_type base; @@ -1937,7 +1930,7 @@ static bool intrinsic_cross(struct hlsl_ctx *ctx, &arg1_swzl1->node, &arg2_swzl1->node, loc))) return false; - if (!(mul1_neg = hlsl_new_unary_expr(ctx, HLSL_OP1_NEG, &mul1->node, *loc))) + if (!(mul1_neg = hlsl_new_unary_expr(ctx, HLSL_OP1_NEG, mul1, *loc))) return false; list_add_tail(params->instrs, &mul1_neg->entry); @@ -1953,7 +1946,7 @@ static bool intrinsic_cross(struct hlsl_ctx *ctx, &arg1_swzl2->node, &arg2_swzl2->node, loc))) return false; - return !!add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_ADD, &mul2->node, mul1_neg, loc); + return !!add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_ADD, mul2, mul1_neg, loc); } static bool intrinsic_floor(struct hlsl_ctx *ctx, @@ -1982,8 +1975,7 @@ static bool intrinsic_min(struct hlsl_ctx *ctx, static bool intrinsic_pow(struct hlsl_ctx *ctx, const struct parse_initializer *params, const struct vkd3d_shader_location *loc) { - struct hlsl_ir_node *log, *exp, *arg; - struct hlsl_ir_expr *mul; + struct hlsl_ir_node *log, *exp, *arg, *mul; if (!(arg = intrinsic_float_convert_arg(ctx, params, params->args[0], loc))) return false; @@ -1995,7 +1987,7 @@ static bool intrinsic_pow(struct hlsl_ctx *ctx, if (!(mul = add_binary_arithmetic_expr(ctx, params->instrs, HLSL_OP2_MUL, params->args[1], log, loc))) return false; - if (!(exp = hlsl_new_unary_expr(ctx, HLSL_OP1_EXP2, &mul->node, *loc))) + if (!(exp = hlsl_new_unary_expr(ctx, HLSL_OP1_EXP2, mul, *loc))) return false; list_add_tail(params->instrs, &exp->entry); return true; -- 2.36.1

June 7, 2022

[PATCH vkd3d v4 6/9] vkd3d-shader/hlsl: Lower matrix casts.

by Giovanni Mascellani

Signed-off-by: Giovanni Mascellani <gmascellani(a)codeweavers.com> --- I had to add a forward declaration for add_load(). I don't know if in cases like that it is preferred to add a forward declaration (which adds cruft) or move the definition backwards (which break the vague logical order functions have). v4: * Mark a couple of tests as passing, thanks to Zeb's commits for splitting matrix variables. --- libs/vkd3d-shader/hlsl.y | 93 ++++++++++++++++++- tests/hlsl-duplicate-modifiers.shader_test | 2 +- tests/hlsl-initializer-matrix.shader_test | 2 +- ...lsl-return-implicit-conversion.shader_test | 8 +- tests/hlsl-shape.shader_test | 10 +- tests/matrix-semantics.shader_test | 18 ++-- 6 files changed, 109 insertions(+), 24 deletions(-) diff --git a/libs/vkd3d-shader/hlsl.y b/libs/vkd3d-shader/hlsl.y index 23c48cdc..f85d4252 100644 --- a/libs/vkd3d-shader/hlsl.y +++ b/libs/vkd3d-shader/hlsl.y @@ -266,6 +266,9 @@ static bool implicit_compatible_data_types(struct hlsl_type *t1, struct hlsl_typ return false; } +static struct hlsl_ir_load *add_load(struct hlsl_ctx *ctx, struct list *instrs, struct hlsl_ir_node *var_node, + struct hlsl_ir_node *offset, struct hlsl_type *data_type, const struct vkd3d_shader_location loc); + static struct hlsl_ir_node *add_cast(struct hlsl_ctx *ctx, struct list *instrs, struct hlsl_ir_node *node, struct hlsl_type *dst_type, const struct vkd3d_shader_location *loc) { @@ -275,10 +278,92 @@ static struct hlsl_ir_node *add_cast(struct hlsl_ctx *ctx, struct list *instrs, if (hlsl_types_are_equal(src_type, dst_type)) return node; - if (!(cast = hlsl_new_cast(ctx, node, dst_type, loc))) - return NULL; - list_add_tail(instrs, &cast->node.entry); - return &cast->node; + if ((src_type->type == HLSL_CLASS_MATRIX || dst_type->type == HLSL_CLASS_MATRIX) + && src_type->type <= HLSL_CLASS_LAST_NUMERIC && dst_type->type <= HLSL_CLASS_LAST_NUMERIC) + { + struct vkd3d_string_buffer *name; + static unsigned int counter = 0; + struct hlsl_ir_load *load; + struct hlsl_ir_var *var; + unsigned int dst_idx; + bool broadcast; + + broadcast = src_type->dimx == 1 && src_type->dimy == 1; + assert(dst_type->dimx * dst_type->dimy <= src_type->dimx * src_type->dimy || broadcast); + if (src_type->type == HLSL_CLASS_MATRIX && dst_type->type == HLSL_CLASS_MATRIX && !broadcast) + { + assert(dst_type->dimx <= src_type->dimx); + assert(dst_type->dimy <= src_type->dimy); + } + + name = vkd3d_string_buffer_get(&ctx->string_buffers); + vkd3d_string_buffer_printf(name, "<cast-%u>", counter++); + var = hlsl_new_synthetic_var(ctx, name->buffer, dst_type, *loc); + vkd3d_string_buffer_release(&ctx->string_buffers, name); + if (!var) + return NULL; + + for (dst_idx = 0; dst_idx < dst_type->dimx * dst_type->dimy; ++dst_idx) + { + struct hlsl_type *src_scalar_type, *dst_scalar_type; + unsigned int src_idx, src_offset, dst_offset; + struct hlsl_ir_store *store; + struct hlsl_ir_constant *c; + + if (broadcast) + { + src_idx = 0; + } + else + { + if (src_type->type == HLSL_CLASS_MATRIX && dst_type->type == HLSL_CLASS_MATRIX) + { + unsigned int x = dst_idx % dst_type->dimx, y = dst_idx / dst_type->dimx; + + src_idx = y * src_type->dimx + x; + } + else + { + src_idx = dst_idx; + } + } + + dst_offset = hlsl_compute_component_offset(ctx, dst_type, dst_idx, &dst_scalar_type); + src_offset = hlsl_compute_component_offset(ctx, src_type, src_idx, &src_scalar_type); + + if (!(c = hlsl_new_uint_constant(ctx, src_offset, loc))) + return NULL; + list_add_tail(instrs, &c->node.entry); + + if (!(load = add_load(ctx, instrs, node, &c->node, src_scalar_type, *loc))) + return NULL; + + if (!(cast = hlsl_new_cast(ctx, &load->node, dst_scalar_type, loc))) + return NULL; + list_add_tail(instrs, &cast->node.entry); + + if (!(c = hlsl_new_uint_constant(ctx, dst_offset, loc))) + return NULL; + list_add_tail(instrs, &c->node.entry); + + if (!(store = hlsl_new_store(ctx, var, &c->node, &cast->node, 0, *loc))) + return NULL; + list_add_tail(instrs, &store->node.entry); + } + + if (!(load = hlsl_new_load(ctx, var, NULL, dst_type, *loc))) + return NULL; + list_add_tail(instrs, &load->node.entry); + + return &load->node; + } + else + { + if (!(cast = hlsl_new_cast(ctx, node, dst_type, loc))) + return NULL; + list_add_tail(instrs, &cast->node.entry); + return &cast->node; + } } static struct hlsl_ir_node *add_implicit_conversion(struct hlsl_ctx *ctx, struct list *instrs, diff --git a/tests/hlsl-duplicate-modifiers.shader_test b/tests/hlsl-duplicate-modifiers.shader_test index fcae12da..6491701a 100644 --- a/tests/hlsl-duplicate-modifiers.shader_test +++ b/tests/hlsl-duplicate-modifiers.shader_test @@ -7,5 +7,5 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (0.1, 0.2, 0.3, 0.4) diff --git a/tests/hlsl-initializer-matrix.shader_test b/tests/hlsl-initializer-matrix.shader_test index ea9de9c0..7e12b0a0 100644 --- a/tests/hlsl-initializer-matrix.shader_test +++ b/tests/hlsl-initializer-matrix.shader_test @@ -55,7 +55,7 @@ float4 main() : SV_TARGET } [test] -todo draw quad +draw quad probe all rgba (21, 22, 31, 32) diff --git a/tests/hlsl-return-implicit-conversion.shader_test b/tests/hlsl-return-implicit-conversion.shader_test index bf99d9cb..545340eb 100644 --- a/tests/hlsl-return-implicit-conversion.shader_test +++ b/tests/hlsl-return-implicit-conversion.shader_test @@ -5,7 +5,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (0.4, 0.3, 0.2, 0.1) [pixel shader] @@ -15,7 +15,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (0.4, 0.3, 0.2, 0.1) [pixel shader] @@ -25,7 +25,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (0.4, 0.3, 0.2, 0.1) [pixel shader] @@ -35,7 +35,7 @@ float4x1 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (0.4, 0.3, 0.2, 0.1) [pixel shader] diff --git a/tests/hlsl-shape.shader_test b/tests/hlsl-shape.shader_test index 57d59534..65cc322c 100644 --- a/tests/hlsl-shape.shader_test +++ b/tests/hlsl-shape.shader_test @@ -211,7 +211,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -235,7 +235,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -260,7 +260,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 6.0, 8.0) [pixel shader] @@ -309,7 +309,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 0.0, 0.0) [pixel shader] @@ -321,7 +321,7 @@ float4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (2.0, 4.0, 0.0, 0.0) [pixel shader] diff --git a/tests/matrix-semantics.shader_test b/tests/matrix-semantics.shader_test index 6a089683..2d40dac9 100644 --- a/tests/matrix-semantics.shader_test +++ b/tests/matrix-semantics.shader_test @@ -5,7 +5,7 @@ float4x1 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (1.0, 2.0, 3.0, 4.0) [pixel shader] @@ -15,7 +15,7 @@ row_major float1x4 main() : sv_target } [test] -todo draw quad +draw quad probe all rgba (1.0, 2.0, 3.0, 4.0) [require] @@ -28,8 +28,8 @@ row_major float4x1 main() : sv_target } [test] -todo draw quad -probe all r (1.0) +draw quad +todo probe all r (1.0) [pixel shader] float1x4 main() : sv_target @@ -38,8 +38,8 @@ float1x4 main() : sv_target } [test] -todo draw quad -probe all r (1.0) +draw quad +todo probe all r (1.0) [pixel shader] void main(out row_major float1x4 x : sv_target0, out float1x4 y : sv_target1) @@ -49,7 +49,7 @@ void main(out row_major float1x4 x : sv_target0, out float1x4 y : sv_target1) } [test] -todo draw quad +draw quad probe all rgba (1.0, 2.0, 3.0, 4.0) [pixel shader fail todo] @@ -67,5 +67,5 @@ void main(out float1x4 x : sv_target0, out float1x4 y : sv_target4) } [test] -todo draw quad -probe all r (1.0) +draw quad +todo probe all r (1.0) -- 2.36.1

June 7, 2022