i965: Move the entire compiler API into a single file

At this point, the compiler API has been substantially simplified.  In the
spirit of Kristian's making a compiler library, this commit makes a single
header file that contains, more-or-less, the entire compiler API.

There's still a bit of cleanup to do particularly in the area of geometry
shaders.  However, this gets us much closer to having a separate compiler.

Reviewed-by: Topi Pohjolainen <[email protected]>

1 files changed, 1 insertions, 355 deletions

diff --git a/src/mesa/drivers/dri/i965/brw_context.h b/src/mesa/drivers/dri/i965/brw_context.h
index 4aba7b814c6..4f503ae4869 100644
--- a/src/mesa/drivers/dri/i965/brw_context.h
+++ b/src/mesa/drivers/dri/i965/brw_context.h
@@ -40,6 +40,7 @@
 #include "main/mm.h"
 #include "main/mtypes.h"
 #include "brw_structs.h"
+#include "brw_compiler.h"
 #include "intel_aub.h"
 #include "program/prog_parameter.h"
 
@@ -340,261 +341,6 @@ struct brw_shader {
    bool compiled_once;
 };
 
-struct brw_stage_prog_data {
-   struct {
-      /** size of our binding table. */
-      uint32_t size_bytes;
-
-      /** @{
-       * surface indices for the various groups of surfaces
-       */
-      uint32_t pull_constants_start;
-      uint32_t texture_start;
-      uint32_t gather_texture_start;
-      uint32_t ubo_start;
-      uint32_t ssbo_start;
-      uint32_t abo_start;
-      uint32_t image_start;
-      uint32_t shader_time_start;
-      /** @} */
-   } binding_table;
-
-   GLuint nr_params;       /**< number of float params/constants */
-   GLuint nr_pull_params;
-   unsigned nr_image_params;
-
-   unsigned curb_read_length;
-   unsigned total_scratch;
-
-   /**
-    * Register where the thread expects to find input data from the URB
-    * (typically uniforms, followed by vertex or fragment attributes).
-    */
-   unsigned dispatch_grf_start_reg;
-
-   bool use_alt_mode; /**< Use ALT floating point mode?  Otherwise, IEEE. */
-
-   /* Pointers to tracked values (only valid once
-    * _mesa_load_state_parameters has been called at runtime).
-    */
-   const gl_constant_value **param;
-   const gl_constant_value **pull_param;
-
-   /** Image metadata passed to the shader as uniforms. */
-   struct brw_image_param *image_param;
-};
-
-/*
- * Image metadata structure as laid out in the shader parameter
- * buffer.  Entries have to be 16B-aligned for the vec4 back-end to be
- * able to use them.  That's okay because the padding and any unused
- * entries [most of them except when we're doing untyped surface
- * access] will be removed by the uniform packing pass.
- */
-#define BRW_IMAGE_PARAM_SURFACE_IDX_OFFSET      0
-#define BRW_IMAGE_PARAM_OFFSET_OFFSET           4
-#define BRW_IMAGE_PARAM_SIZE_OFFSET             8
-#define BRW_IMAGE_PARAM_STRIDE_OFFSET           12
-#define BRW_IMAGE_PARAM_TILING_OFFSET           16
-#define BRW_IMAGE_PARAM_SWIZZLING_OFFSET        20
-#define BRW_IMAGE_PARAM_SIZE                    24
-
-struct brw_image_param {
-   /** Surface binding table index. */
-   uint32_t surface_idx;
-
-   /** Offset applied to the X and Y surface coordinates. */
-   uint32_t offset[2];
-
-   /** Surface X, Y and Z dimensions. */
-   uint32_t size[3];
-
-   /** X-stride in bytes, Y-stride in pixels, horizontal slice stride in
-    * pixels, vertical slice stride in pixels.
-    */
-   uint32_t stride[4];
-
-   /** Log2 of the tiling modulus in the X, Y and Z dimension. */
-   uint32_t tiling[3];
-
-   /**
-    * Right shift to apply for bit 6 address swizzling.  Two different
-    * swizzles can be specified and will be applied one after the other.  The
-    * resulting address will be:
-    *
-    *  addr' = addr ^ ((1 << 6) & ((addr >> swizzling[0]) ^
-    *                              (addr >> swizzling[1])))
-    *
-    * Use \c 0xff if any of the swizzles is not required.
-    */
-   uint32_t swizzling[2];
-};
-
-/* Data about a particular attempt to compile a program.  Note that
- * there can be many of these, each in a different GL state
- * corresponding to a different brw_wm_prog_key struct, with different
- * compiled programs.
- */
-struct brw_wm_prog_data {
-   struct brw_stage_prog_data base;
-
-   GLuint num_varying_inputs;
-
-   GLuint dispatch_grf_start_reg_16;
-   GLuint reg_blocks;
-   GLuint reg_blocks_16;
-
-   struct {
-      /** @{
-       * surface indices the WM-specific surfaces
-       */
-      uint32_t render_target_start;
-      /** @} */
-   } binding_table;
-
-   uint8_t computed_depth_mode;
-
-   bool early_fragment_tests;
-   bool no_8;
-   bool dual_src_blend;
-   bool uses_pos_offset;
-   bool uses_omask;
-   bool uses_kill;
-   bool pulls_bary;
-   uint32_t prog_offset_16;
-
-   /**
-    * Mask of which interpolation modes are required by the fragment shader.
-    * Used in hardware setup on gen6+.
-    */
-   uint32_t barycentric_interp_modes;
-
-   /**
-    * Map from gl_varying_slot to the position within the FS setup data
-    * payload where the varying's attribute vertex deltas should be delivered.
-    * For varying slots that are not used by the FS, the value is -1.
-    */
-   int urb_setup[VARYING_SLOT_MAX];
-};
-
-struct brw_cs_prog_data {
-   struct brw_stage_prog_data base;
-
-   GLuint dispatch_grf_start_reg_16;
-   unsigned local_size[3];
-   unsigned simd_size;
-   bool uses_barrier;
-   bool uses_num_work_groups;
-   unsigned local_invocation_id_regs;
-
-   struct {
-      /** @{
-       * surface indices the CS-specific surfaces
-       */
-      uint32_t work_groups_start;
-      /** @} */
-   } binding_table;
-};
-
-/**
- * Enum representing the i965-specific vertex results that don't correspond
- * exactly to any element of gl_varying_slot.  The values of this enum are
- * assigned such that they don't conflict with gl_varying_slot.
- */
-typedef enum
-{
-   BRW_VARYING_SLOT_NDC = VARYING_SLOT_MAX,
-   BRW_VARYING_SLOT_PAD,
-   /**
-    * Technically this is not a varying but just a placeholder that
-    * compile_sf_prog() inserts into its VUE map to cause the gl_PointCoord
-    * builtin variable to be compiled correctly. see compile_sf_prog() for
-    * more info.
-    */
-   BRW_VARYING_SLOT_PNTC,
-   BRW_VARYING_SLOT_COUNT
-} brw_varying_slot;
-
-
-/**
- * Data structure recording the relationship between the gl_varying_slot enum
- * and "slots" within the vertex URB entry (VUE).  A "slot" is defined as a
- * single octaword within the VUE (128 bits).
- *
- * Note that each BRW register contains 256 bits (2 octawords), so when
- * accessing the VUE in URB_NOSWIZZLE mode, each register corresponds to two
- * consecutive VUE slots.  When accessing the VUE in URB_INTERLEAVED mode (as
- * in a vertex shader), each register corresponds to a single VUE slot, since
- * it contains data for two separate vertices.
- */
-struct brw_vue_map {
-   /**
-    * Bitfield representing all varying slots that are (a) stored in this VUE
-    * map, and (b) actually written by the shader.  Does not include any of
-    * the additional varying slots defined in brw_varying_slot.
-    */
-   GLbitfield64 slots_valid;
-
-   /**
-    * Is this VUE map for a separate shader pipeline?
-    *
-    * Separable programs (GL_ARB_separate_shader_objects) can be mixed and matched
-    * without the linker having a chance to dead code eliminate unused varyings.
-    *
-    * This means that we have to use a fixed slot layout, based on the output's
-    * location field, rather than assigning slots in a compact contiguous block.
-    */
-   bool separate;
-
-   /**
-    * Map from gl_varying_slot value to VUE slot.  For gl_varying_slots that are
-    * not stored in a slot (because they are not written, or because
-    * additional processing is applied before storing them in the VUE), the
-    * value is -1.
-    */
-   signed char varying_to_slot[BRW_VARYING_SLOT_COUNT];
-
-   /**
-    * Map from VUE slot to gl_varying_slot value.  For slots that do not
-    * directly correspond to a gl_varying_slot, the value comes from
-    * brw_varying_slot.
-    *
-    * For slots that are not in use, the value is BRW_VARYING_SLOT_COUNT (this
-    * simplifies code that uses the value stored in slot_to_varying to
-    * create a bit mask).
-    */
-   signed char slot_to_varying[BRW_VARYING_SLOT_COUNT];
-
-   /**
-    * Total number of VUE slots in use
-    */
-   int num_slots;
-};
-
-/**
- * Convert a VUE slot number into a byte offset within the VUE.
- */
-static inline GLuint brw_vue_slot_to_offset(GLuint slot)
-{
-   return 16*slot;
-}
-
-/**
- * Convert a vertex output (brw_varying_slot) into a byte offset within the
- * VUE.
- */
-static inline GLuint brw_varying_to_offset(struct brw_vue_map *vue_map,
-                                           GLuint varying)
-{
-   return brw_vue_slot_to_offset(vue_map->varying_to_slot[varying]);
-}
-
-void brw_compute_vue_map(const struct brw_device_info *devinfo,
-                         struct brw_vue_map *vue_map,
-                         GLbitfield64 slots_valid,
-                         bool separate_shader);
-
-
 /**
  * Bitmask indicating which fragment shader inputs represent varyings (and
  * hence have to be delivered to the fragment shader by the SF/SBE stage).
@@ -671,41 +417,6 @@ struct brw_ff_gs_prog_data {
    unsigned svbi_postincrement_value;
 };
 
-enum shader_dispatch_mode {
-   DISPATCH_MODE_4X1_SINGLE = 0,
-   DISPATCH_MODE_4X2_DUAL_INSTANCE = 1,
-   DISPATCH_MODE_4X2_DUAL_OBJECT = 2,
-   DISPATCH_MODE_SIMD8 = 3,
-};
-
-struct brw_vue_prog_data {
-   struct brw_stage_prog_data base;
-   struct brw_vue_map vue_map;
-
-   GLuint urb_read_length;
-   GLuint total_grf;
-
-   /* Used for calculating urb partitions.  In the VS, this is the size of the
-    * URB entry used for both input and output to the thread.  In the GS, this
-    * is the size of the URB entry used for output.
-    */
-   GLuint urb_entry_size;
-
-   enum shader_dispatch_mode dispatch_mode;
-};
-
-
-struct brw_vs_prog_data {
-   struct brw_vue_prog_data base;
-
-   GLbitfield64 inputs_read;
-
-   unsigned nr_attributes;
-
-   bool uses_vertexid;
-   bool uses_instanceid;
-};
-
 /** Number of texture sampler units */
 #define BRW_MAX_TEX_UNIT 32
 
@@ -763,71 +474,6 @@ struct brw_vs_prog_data {
 
 #define SURF_INDEX_GEN6_SOL_BINDING(t) (t)
 
-struct brw_gs_prog_data
-{
-   struct brw_vue_prog_data base;
-
-   /**
-    * Size of an output vertex, measured in HWORDS (32 bytes).
-    */
-   unsigned output_vertex_size_hwords;
-
-   unsigned output_topology;
-
-   /**
-    * Size of the control data (cut bits or StreamID bits), in hwords (32
-    * bytes).  0 if there is no control data.
-    */
-   unsigned control_data_header_size_hwords;
-
-   /**
-    * Format of the control data (either GEN7_GS_CONTROL_DATA_FORMAT_GSCTL_SID
-    * if the control data is StreamID bits, or
-    * GEN7_GS_CONTROL_DATA_FORMAT_GSCTL_CUT if the control data is cut bits).
-    * Ignored if control_data_header_size is 0.
-    */
-   unsigned control_data_format;
-
-   bool include_primitive_id;
-
-   /**
-    * The number of vertices emitted, if constant - otherwise -1.
-    */
-   int static_vertex_count;
-
-   int invocations;
-
-   /**
-    * Gen6 transform feedback enabled flag.
-    */
-   bool gen6_xfb_enabled;
-
-   /**
-    * Gen6: Provoking vertex convention for odd-numbered triangles
-    * in tristrips.
-    */
-   GLuint pv_first:1;
-
-   /**
-    * Gen6: Number of varyings that are output to transform feedback.
-    */
-   GLuint num_transform_feedback_bindings:7; /* 0-BRW_MAX_SOL_BINDINGS */
-
-   /**
-    * Gen6: Map from the index of a transform feedback binding table entry to the
-    * gl_varying_slot that should be streamed out through that binding table
-    * entry.
-    */
-   unsigned char transform_feedback_bindings[BRW_MAX_SOL_BINDINGS];
-
-   /**
-    * Gen6: Map from the index of a transform feedback binding table entry to the
-    * swizzles that should be used when streaming out data through that
-    * binding table entry.
-    */
-   unsigned char transform_feedback_swizzles[BRW_MAX_SOL_BINDINGS];
-};
-
 /**
  * Stride in bytes between shader_time entries.
  *