aco: add pending_lds_access option for insert waitcnt

For tcs epilog to add p_barrier at the beginning to sync
main shader part tess factor LDS write.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24442>

src/amd/compiler/aco_insert_waitcnt.cpp

@@ -1049,6 +1049,11 @@ insert_wait_states(Program* program)
    std::stack<unsigned, std::vector<unsigned>> loop_header_indices;
    unsigned loop_progress = 0;
 
+   if (program->pending_lds_access) {
+      update_barrier_imm(in_ctx[0], get_counters_for_event(event_lds), event_lds,
+                         memory_sync_info(storage_shared));
+   }
+
    for (Definition def : program->args_pending_vmem) {
       update_counters(in_ctx[0], event_vmem);
       insert_wait_entry(in_ctx[0], def, event_vmem);

src/amd/compiler/aco_ir.h

@@ -2145,6 +2145,9 @@ public:
 
    std::vector<Definition> args_pending_vmem;
 
+   /* For shader part with previous shader part that has lds access. */
+   bool pending_lds_access = false;
+
    struct {
       FILE* output = stderr;
       bool shorten_messages = false;