diff options
| -rw-r--r-- | src/mesa/drivers/dri/i965/brw_eu_emit.c | 11 | ||||
| -rw-r--r-- | src/mesa/drivers/dri/i965/brw_vec4.cpp | 13 |
2 files changed, 18 insertions, 6 deletions
diff --git a/src/mesa/drivers/dri/i965/brw_eu_emit.c b/src/mesa/drivers/dri/i965/brw_eu_emit.c index 5fb96626649..35d8039ed13 100644 --- a/src/mesa/drivers/dri/i965/brw_eu_emit.c +++ b/src/mesa/drivers/dri/i965/brw_eu_emit.c @@ -847,12 +847,11 @@ brw_alu2(struct brw_codegen *p, unsigned opcode, static int get_3src_subreg_nr(struct brw_reg reg) { - if (reg.vstride == BRW_VERTICAL_STRIDE_0) { - assert(brw_is_single_value_swizzle(reg.swizzle)); - return reg.subnr / 4 + BRW_GET_SWZ(reg.swizzle, 0); - } else { - return reg.subnr / 4; - } + /* Normally, SubRegNum is in bytes (0..31). However, 3-src instructions + * use 32-bit units (components 0..7). Since they only support F/D/UD + * types, this doesn't lose any flexibility, but uses fewer bits. + */ + return reg.subnr / 4; } static brw_inst * diff --git a/src/mesa/drivers/dri/i965/brw_vec4.cpp b/src/mesa/drivers/dri/i965/brw_vec4.cpp index dd223985d1c..c6a52c5d183 100644 --- a/src/mesa/drivers/dri/i965/brw_vec4.cpp +++ b/src/mesa/drivers/dri/i965/brw_vec4.cpp @@ -1784,9 +1784,22 @@ vec4_visitor::convert_to_hw_regs() case ATTR: unreachable("not reached"); } + src = reg; } + if (inst->is_3src()) { + /* 3-src instructions with scalar sources support arbitrary subnr, + * but don't actually use swizzles. Convert swizzle into subnr. + */ + for (int i = 0; i < 3; i++) { + if (inst->src[i].vstride == BRW_VERTICAL_STRIDE_0) { + assert(brw_is_single_value_swizzle(inst->src[i].swizzle)); + inst->src[i].subnr += 4 * BRW_GET_SWZ(inst->src[i].swizzle, 0); + } + } + } + dst_reg &dst = inst->dst; struct brw_reg reg; |