Revert "aco: only emit waitcnt on loop continues if we there was some load or export"

We don't properly pass on ctx.lgkm_cnt/ctx.barrier_imm/etc, so this
waitcnt was necessary for barriers and correctly waiting for SMEM before
s_dcache_wb on GFX10.

Totals from affected shaders:
SGPRS: 33200 -> 33200 (0.00 %)
VGPRS: 31376 -> 31376 (0.00 %)
Spilled SGPRs: 0 -> 0 (0.00 %)
Spilled VGPRs: 0 -> 0 (0.00 %)
Private memory VGPRs: 0 -> 0 (0.00 %)
Scratch size: 0 -> 0 (0.00 %) dwords per thread
Code Size: 2431804 -> 2433956 (0.09 %) bytes
LDS: 316 -> 316 (0.00 %) blocks
Max Waves: 1609 -> 1609 (0.00 %)

This reverts commit 2c050b49b3d776f054f1265d5523cabb61f22fc3.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>

diff --git a/src/amd/compiler/aco_insert_waitcnt.cpp b/src/amd/compiler/aco_insert_waitcnt.cpp
index af609132deed64473516875f947420a7c0532240..1f4733ed356cecdb9659a204d9624de3981efb9b 100644 (file)
--- a/src/amd/compiler/aco_insert_waitcnt.cpp
+++ b/src/amd/compiler/aco_insert_waitcnt.cpp
@@ -680,7 +680,7 @@ void handle_block(Program *program, Block& block, wait_ctx& ctx)
    /* check if this block is at the end of a loop */
    for (unsigned succ_idx : block.linear_succs) {
       /* eliminate any remaining counters */
-      if (succ_idx <= block.index && (ctx.vm_cnt || ctx.exp_cnt || ctx.lgkm_cnt || ctx.vs_cnt) && !ctx.gpr_map.empty()) {
+      if (succ_idx <= block.index && (ctx.vm_cnt || ctx.exp_cnt || ctx.lgkm_cnt || ctx.vs_cnt)) {
          // TODO: we could do better if we only wait if the regs between the block and other predecessors differ
 
          aco_ptr<Instruction> branch = std::move(new_instructions.back());