2003-06-18 Stephen Clarke <stephen.clarke@superh.com>

            J"orn Rennecke <joern.rennecke@superh.com>

	* bt-load.c: New file.
	* Makefile.in (OBJS): Include bt-load.o
	(bt-load.o): Add dependencies.
	* flags.h (flag_branch_target_load_optimize): Declare.
	(flag_branch_target_load_optimize2): Likewise.
	* hooks.c (hook_reg_class_void_no_regs): New function.
	(hook_bool_bool_false): Likewise.
	* hooks.h (hook_reg_class_void_no_regs, hook_bool_bool_false): Declare.
	* rtl.h (branch_target_load_optimize): Declare.
	* target-def.h (TARGET_BRANCH_TARGET_REGISTER_CLASS): Define.
	(TARGET_BRANCH_TARGET_REGISTER_CALLEE_SAVED): Likewise.
	(TARGET_INITIALIZER): Include these.
	* target.h (struct gcc_target): Add branch_target_register_class
	and branch_target_register_callee_saved members.
	* toplev.c (enum dump_file_index): Add DFI_branch_target_load
	(dump_file) Add "tars" entry.
	(flag_branch_target_load_optimize): New variable.
	(flag_branch_target_load_optimize2): Likewise.
	(lang_independent_options): Add entries for new options.
	(rest_of_compilation): Call branch_target_load_optimize.
	* doc/tm.texi (TARGET_BRANCH_TARGET_REGISTER_CLASS): Document.
	(TARGET_BRANCH_TARGET_REGISTER_CALLEE_SAVED): Likewise.
	* doc/invoke.texi: Document -fbranch-target-load-optimize and
	-fbranch-target-load-optimize2.
	* rtl.h (epilogue_completed): Declare.
	* recog.c (epilogue_completed): New variable.
	* toplev.c (rest_of_compilation): Set it.
	* flow.c (mark_regs_live_at_end): Use it.
	* config/ia64/ia64.c (ia64_output_mi_thunk): Set it.
	* config/rs6000/rs6000.c (rs6000_output_mi_thunk): Likewise.
	* config/sh/sh.c (sh_output_mi_thunk): Likewise.
	* config/sparc/sparc.c (sparc_output_mi_thunk): Likewise.

	* sh.c (shmedia_space_reserved_for_target_registers): New variable.
	(sh_target_reg_class): New function.
	(sh_optimize_target_register_callee_saved): Likwise.
	(shmedia_target_regs_stack_space): Likewise.
	(shmedia_reserve_space_for_target_registers_p): Likewise.
	(shmedia_target_regs_stack_adjust): Likewise.
	(TARGET_BRANCH_TARGET_REGISTER_CLASS): Override.
	(TARGET_BRANCH_TARGET_REGISTER_CALLEE_SAVED): Likewise.
	(calc_live_regs): If flag_branch_target_load_optimize2 and
	TARGET_SAVE_ALL_TARGET_REGS is enabled, and we have space reserved
	for target registers, make sure that we save all target registers.
	(sh_expand_prologue, sh_expand_epilogue): Take target register
	optimizations into account.  Collapse stack adjustments if that
	is beneficial.
	(initial_elimination_offset): Reserve space for target registers
	if necessary.
	* sh.h (SAVE_ALL_TR_BIT, TARGET_SAVE_ALL_TARGET_REGS): Define.
	(OPTIMIZATION_OPTIONS): Enable flag_branch_target_load_optimize.


git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@68165 138bc75d-0d04-0410-961f-82ee72b054a4

1 files changed, 47 insertions, 0 deletions

diff --git a/gcc/toplev.c b/gcc/toplev.c
index 87892a56fdc..b4d8a388a8c 100644
--- a/gcc/toplev.c
+++ b/gcc/toplev.c
@@ -287,6 +287,7 @@ enum dump_file_index
   DFI_rnreg,
   DFI_bbro,
   DFI_ce3,
+  DFI_branch_target_load,
   DFI_sched2,
   DFI_stack,
   DFI_mach,
@@ -338,6 +339,7 @@ static struct dump_file_info dump_file[DFI_MAX] =
   { "rnreg",	'n', 1, 0, 0 },
   { "bbro",	'B', 1, 0, 0 },
   { "ce3",	'E', 1, 0, 0 },
+  { "btl",	'd', 1, 0, 0 }, /* Yes, duplicate enable switch.  */
   { "sched2",	'R', 1, 0, 0 },
   { "stack",	'k', 1, 0, 0 },
   { "mach",	'M', 1, 0, 0 },
@@ -689,6 +691,16 @@ int flag_gcse_lm = 1;
 
 int flag_gcse_sm = 1;
 
+/* Perform target register optimization before prologue / epilogue
+   threading.  */
+
+int flag_branch_target_load_optimize = 0;
+
+/* Perform target register optimization after prologue / epilogue
+   threading and jump2.  */
+
+int flag_branch_target_load_optimize2 = 0;
+
 /* Nonzero means to rerun cse after loop optimization.  This increases
    compilation time about 20% and picks up a few more common expressions.  */
 
@@ -1118,6 +1130,10 @@ static const lang_independent_options f_options[] =
    N_("Perform enhanced load motion during global subexpression elimination") },
   {"gcse-sm", &flag_gcse_sm, 1,
    N_("Perform store motion after global subexpression elimination") },
+  {"branch-target-load-optimize", &flag_branch_target_load_optimize, 1,
+   N_("Perform branch target load optimization before prologue / epilogue threading") },
+  {"branch-target-load-optimize2", &flag_branch_target_load_optimize2, 1,
+   N_("Perform branch target load optimization after prologue / epilogue threading") },
   {"loop-optimize", &flag_loop_optimize, 1,
    N_("Perform the loop optimizations") },
   {"crossjumping", &flag_crossjumping, 1,
@@ -3761,6 +3777,17 @@ rest_of_compilation (tree decl)
 #endif
     split_all_insns (0);
 
+    if (flag_branch_target_load_optimize)
+      {
+	open_dump_file (DFI_branch_target_load, decl);
+
+	branch_target_load_optimize (insns, false);
+
+	close_dump_file (DFI_branch_target_load, print_rtl_with_bb, insns);
+
+	ggc_collect ();
+      }
+
   if (optimize)
     cleanup_cfg (CLEANUP_EXPENSIVE);
 
@@ -3769,6 +3796,7 @@ rest_of_compilation (tree decl)
      it and the rest of the code and also allows delayed branch
      scheduling to operate in the epilogue.  */
   thread_prologue_and_epilogue_insns (insns);
+  epilogue_completed = 1;
 
   if (optimize)
     {
@@ -3825,6 +3853,24 @@ rest_of_compilation (tree decl)
       timevar_pop (TV_IFCVT2);
     }
 
+    if (flag_branch_target_load_optimize2)
+      {
+	/* Leave this a warning for now so that it is possible to experiment
+	   with running this pass twice.  In 3.6, we should either make this
+	   an error, or use separate dump files.  */
+	if (flag_branch_target_load_optimize)
+	  warning ("branch target register load optimization is not intended "
+		   "to be run twice");
+
+	open_dump_file (DFI_branch_target_load, decl);
+
+	branch_target_load_optimize (insns, true);
+
+	close_dump_file (DFI_branch_target_load, print_rtl_with_bb, insns);
+
+	ggc_collect ();
+      }
+
 #ifdef INSN_SCHEDULING
   if (optimize > 0 && flag_schedule_insns_after_reload)
     rest_of_handle_sched2 (decl, insns);
@@ -3904,6 +3950,7 @@ rest_of_compilation (tree decl)
 #endif
 
   reload_completed = 0;
+  epilogue_completed = 0;
   flow2_completed = 0;
   no_new_pseudos = 0;