CELL: changes to generate SPU code for stenciling

This set of code changes are for stencil code generation
support.  Both one-sided and two-sided stenciling are supported.
In addition to the raw code generation changes, these changes had
to be made elsewhere in the system:

- Added new "register set" feature to the SPE assembly generation.
  A "register set" is a way to allocate multiple registers and free
  them all at the same time, delegating register allocation management
  to the spe_function unit.  It's quite useful in complex register
  allocation schemes (like stenciling).

- Added and improved SPE macro calculations.
  These are operations between registers and unsigned integer
  immediates.  In many cases, the calculation can be performed
  with a single instruction; the macros will generate the
  single instruction if possible, or generate a register load
  and register-to-register operation if not.  These macro
  functions are: spe_load_uint() (which has new ways to
  load a value in a single instruction), spe_and_uint(),
  spe_xor_uint(), spe_compare_equal_uint(), and spe_compare_greater_uint().

- Added facing to fragment generation.  While rendering, the rasterizer
  needs to be able to determine front- and back-facing fragments, in order
  to correctly apply two-sided stencil.  That requires these changes:
  - Added front_winding field to the cell_command_render block, so that
    the state tracker could communicate to the rasterizer what it
    considered to be the front-facing direction.
  - Added fragment facing as an input to the fragment function.
  - Calculated facing is passed during emit_quad().

1 files changed, 34 insertions, 7 deletions

diff --git a/src/gallium/auxiliary/rtasm/rtasm_ppc_spe.h b/src/gallium/auxiliary/rtasm/rtasm_ppc_spe.h
index 61c7edeb60..cd2e245409 100644
--- a/src/gallium/auxiliary/rtasm/rtasm_ppc_spe.h
+++ b/src/gallium/auxiliary/rtasm/rtasm_ppc_spe.h
@@ -53,17 +53,26 @@ struct spe_function
    uint num_inst;
    uint max_inst;
 
-    /**
-     * Mask of used / unused registers
-     *
-     * Each set bit corresponds to an available register.  Each cleared bit
-     * corresponds to an allocated register.
+   /**
+    * The "set count" reflects the number of nested register sets
+    * are allowed.  In the unlikely case that we exceed the set count,
+    * register allocation will start to be confused, which is critical
+    * enough that we check for it.
+    */
+   unsigned char set_count;
+
+   /** 
+    * Flags for used and unused registers.  Each byte corresponds to a
+    * register; a 0 in that byte means that the register is available.
+    * A value of 1 means that the register was allocated in the current
+    * register set.  Any other value N means that the register was allocated
+    * N register sets ago.
      *
      * \sa
      * spe_allocate_register, spe_allocate_available_register,
-     * spe_release_register
+     * spe_allocate_register_set, spe_release_register_set, spe_release_register, 
      */
-    uint64_t regs[SPE_NUM_REGS / 64];
+    unsigned char regs[SPE_NUM_REGS];
 
     boolean print; /**< print/dump instructions as they're emitted? */
     int indent;    /**< number of spaces to indent */
@@ -77,6 +86,8 @@ extern unsigned spe_code_size(const struct spe_function *p);
 extern int spe_allocate_available_register(struct spe_function *p);
 extern int spe_allocate_register(struct spe_function *p, int reg);
 extern void spe_release_register(struct spe_function *p, int reg);
+extern void spe_allocate_register_set(struct spe_function *p);
+extern void spe_release_register_set(struct spe_function *p);
 
 extern void spe_print_code(struct spe_function *p, boolean enable);
 extern void spe_indent(struct spe_function *p, int spaces);
@@ -307,6 +318,22 @@ spe_load_int(struct spe_function *p, unsigned rT, int i);
 extern void
 spe_load_uint(struct spe_function *p, unsigned rT, unsigned int ui);
 
+/** And immediate value into rT. */
+extern void
+spe_and_uint(struct spe_function *p, unsigned rT, unsigned rA, unsigned int ui);
+
+/** Xor immediate value into rT. */
+extern void
+spe_xor_uint(struct spe_function *p, unsigned rT, unsigned rA, unsigned int ui);
+
+/** Compare equal with immediate value. */
+extern void
+spe_compare_equal_uint(struct spe_function *p, unsigned rT, unsigned rA, unsigned int ui);
+
+/** Compare greater with immediate value. */
+extern void
+spe_compare_greater_uint(struct spe_function *p, unsigned rT, unsigned rA, unsigned int ui);
+
 /** Replicate word 0 of rA across rT. */
 extern void
 spe_splat(struct spe_function *p, unsigned rT, unsigned rA);