r600g: support up to 64 shader constants

From the r600 ISA:
Each ALU clause can lock up to four sets of constants
into the constant cache.  Each set (one cache line) is
16 128-bit constants. These are split into two groups.
Each group can be from a different constant buffer
(out of 16 buffers). Each group of two constants consists
of either [Line] and [Line+1] or [line + loop_ctr]
and [line + loop_ctr +1].

For supporting more than 64 constants, we need to
break the code into multiple ALU clauses based
on what sets of constants are needed in that clause.

Note: This is a candidate for the 7.10 branch.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>

2 files changed, 20 insertions, 1 deletions

diff --git a/src/gallium/drivers/r600/r600_asm.c b/src/gallium/drivers/r600/r600_asm.c
index 1f41269534..894d0d2fcd 100644
--- a/src/gallium/drivers/r600/r600_asm.c
+++ b/src/gallium/drivers/r600/r600_asm.c
@@ -470,7 +470,22 @@ int r600_bc_add_alu_type(struct r600_bc *bc, const struct r600_bc_alu *alu, int
 	bc->cf_last->ndw += 2;
 	bc->ndw += 2;
 
-	bc->cf_last->kcache0_mode = 2;
+	/* The following configuration provides 64 128-bit constants.
+	 * Each cacheline holds 16 128-bit constants and each
+	 * kcache can lock 2 cachelines and there are 2 kcaches per
+	 * ALU clause for a max of 64 constants.
+	 * For supporting more than 64 constants, the code needs
+	 * to be broken down into multiple ALU clauses.
+	 */
+	/* select the constant buffer (0-15) for each kcache */
+	bc->cf_last->kcache0_bank = 0;
+	bc->cf_last->kcache1_bank = 0;
+	/* lock 2 cachelines per kcache; 4 total */
+	bc->cf_last->kcache0_mode = V_SQ_CF_KCACHE_LOCK_2;
+	bc->cf_last->kcache1_mode = V_SQ_CF_KCACHE_LOCK_2;
+	/* set the cacheline offsets for each kcache */
+	bc->cf_last->kcache0_addr = 0;
+	bc->cf_last->kcache1_addr = 2;
 
 	/* process cur ALU instructions for bank swizzle */
 	if (alu->last) {
diff --git a/src/gallium/drivers/r600/r600_sq.h b/src/gallium/drivers/r600/r600_sq.h
index 0573e63dc8..d812bfd1fe 100644
--- a/src/gallium/drivers/r600/r600_sq.h
+++ b/src/gallium/drivers/r600/r600_sq.h
@@ -74,6 +74,10 @@
 #define   S_SQ_CF_ALU_WORD0_KCACHE_MODE0(x)                          (((x) & 0x3) << 30)
 #define   G_SQ_CF_ALU_WORD0_KCACHE_MODE0(x)                          (((x) >> 30) & 0x3)
 #define   C_SQ_CF_ALU_WORD0_KCACHE_MODE0                             0x3FFFFFFF
+#define     V_SQ_CF_KCACHE_NOP                                       0x00000000
+#define     V_SQ_CF_KCACHE_LOCK_1                                    0x00000001
+#define     V_SQ_CF_KCACHE_LOCK_2                                    0x00000002
+#define     V_SQ_CF_KCACHE_LOCK_LOOP_INDEX                           0x00000003
 #define P_SQ_CF_ALU_WORD1
 #define   S_SQ_CF_ALU_WORD1_KCACHE_MODE1(x)                          (((x) & 0x3) << 0)
 #define   G_SQ_CF_ALU_WORD1_KCACHE_MODE1(x)                          (((x) >> 0) & 0x3)