path: root/0124-Backport-SME-aarch64-Tweak-handling-of-mgeneral-regs.patch

From 91f7471cbc7dec42673b58a1896330d64eb6be2a Mon Sep 17 00:00:00 2001
From: Richard Sandiford <richard.sandiford@arm.com>
Date: Thu, 29 Sep 2022 11:32:57 +0100
Subject: [PATCH 025/157] [Backport][SME] aarch64: Tweak handling of
 -mgeneral-regs-only

Reference: https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=2a269bda9e7b8f9353699d0c965e7e9246500aa0

-mgeneral-regs-only is effectively "+nofp for the compiler without
changing the assembler's ISA flags".  Currently that's implemented
by making TARGET_FLOAT, TARGET_SIMD and TARGET_SVE depend on
!TARGET_GENERAL_REGS_ONLY and then making any feature that needs FP
registers depend (directly or indirectly) on one of those three TARGET
macros.  The problem is that it's easy to forgot to do the last bit.

This patch instead represents the distinction between "assemnbler
ISA flags" and "compiler ISA flags" more directly, funnelling
all updates through a new function that sets both sets of flags
together.

gcc/
	* config/aarch64/aarch64.opt (aarch64_asm_isa_flags): New variable.
	* config/aarch64/aarch64.h (aarch64_asm_isa_flags)
	(aarch64_isa_flags): Redefine as read-only macros.
	(TARGET_SIMD, TARGET_FLOAT, TARGET_SVE): Don't depend on
	!TARGET_GENERAL_REGS_ONLY.
	* common/config/aarch64/aarch64-common.cc
	(aarch64_set_asm_isa_flags): New function.
	(aarch64_handle_option): Call it when updating -mgeneral-regs.
	* config/aarch64/aarch64-protos.h (aarch64_simd_switcher): Replace
	m_old_isa_flags with m_old_asm_isa_flags.
	(aarch64_set_asm_isa_flags): Declare.
	* config/aarch64/aarch64-builtins.cc
	(aarch64_simd_switcher::aarch64_simd_switcher)
	(aarch64_simd_switcher::~aarch64_simd_switcher): Save and restore
	aarch64_asm_isa_flags instead of aarch64_isa_flags.
	* config/aarch64/aarch64-sve-builtins.cc
	(check_required_extensions): Use aarch64_asm_isa_flags instead
	of aarch64_isa_flags.
	* config/aarch64/aarch64.cc (aarch64_set_asm_isa_flags): New function.
	(aarch64_override_options, aarch64_handle_attr_arch)
	(aarch64_handle_attr_cpu, aarch64_handle_attr_isa_flags): Use
	aarch64_set_asm_isa_flags to set the ISA flags.
	(aarch64_option_print, aarch64_declare_function_name)
	(aarch64_start_file): Use aarch64_asm_isa_flags instead
	of aarch64_isa_flags.
	(aarch64_can_inline_p): Check aarch64_asm_isa_flags as well as
	aarch64_isa_flags.
---
 gcc/common/config/aarch64/aarch64-common.cc | 12 ++++++
 gcc/config/aarch64/aarch64-builtins.cc      |  6 +--
 gcc/config/aarch64/aarch64-protos.h         |  5 ++-
 gcc/config/aarch64/aarch64-sve-builtins.cc  |  2 +-
 gcc/config/aarch64/aarch64.cc               | 45 ++++++++++++++-------
 gcc/config/aarch64/aarch64.h                | 17 ++++++--
 gcc/config/aarch64/aarch64.opt              |  3 ++
 7 files changed, 68 insertions(+), 22 deletions(-)

diff --git a/gcc/common/config/aarch64/aarch64-common.cc b/gcc/common/config/aarch64/aarch64-common.cc
index 752ba5632..c64b4987e 100644
--- a/gcc/common/config/aarch64/aarch64-common.cc
+++ b/gcc/common/config/aarch64/aarch64-common.cc
@@ -137,6 +137,17 @@ reset_tsv110_option ()
     }
 }
 
+/* Set OPTS->x_aarch64_asm_isa_flags to FLAGS and update
+   OPTS->x_aarch64_isa_flags accordingly.  */
+void
+aarch64_set_asm_isa_flags (gcc_options *opts, aarch64_feature_flags flags)
+{
+  opts->x_aarch64_asm_isa_flags = flags;
+  opts->x_aarch64_isa_flags = flags;
+  if (opts->x_target_flags & MASK_GENERAL_REGS_ONLY)
+    opts->x_aarch64_isa_flags &= ~feature_deps::get_flags_off (AARCH64_FL_FP);
+}
+
 /* Implement TARGET_HANDLE_OPTION.
    This function handles the target specific options for CPU/target selection.
 
@@ -174,6 +185,7 @@ aarch64_handle_option (struct gcc_options *opts,
 
     case OPT_mgeneral_regs_only:
       opts->x_target_flags |= MASK_GENERAL_REGS_ONLY;
+      aarch64_set_asm_isa_flags (opts, opts->x_aarch64_asm_isa_flags);
       return true;
 
     case OPT_mfix_cortex_a53_835769:
diff --git a/gcc/config/aarch64/aarch64-builtins.cc b/gcc/config/aarch64/aarch64-builtins.cc
index 42276e7ca..015e9d975 100644
--- a/gcc/config/aarch64/aarch64-builtins.cc
+++ b/gcc/config/aarch64/aarch64-builtins.cc
@@ -1336,20 +1336,20 @@ aarch64_scalar_builtin_type_p (aarch64_simd_type t)
 /* Enable AARCH64_FL_* flags EXTRA_FLAGS on top of the base Advanced SIMD
    set.  */
 aarch64_simd_switcher::aarch64_simd_switcher (unsigned int extra_flags)
-  : m_old_isa_flags (aarch64_isa_flags),
+  : m_old_asm_isa_flags (aarch64_asm_isa_flags),
     m_old_general_regs_only (TARGET_GENERAL_REGS_ONLY)
 {
   /* Changing the ISA flags should be enough here.  We shouldn't need to
      pay the compile-time cost of a full target switch.  */
-  aarch64_isa_flags = AARCH64_FL_FP | AARCH64_FL_SIMD | extra_flags;
   global_options.x_target_flags &= ~MASK_GENERAL_REGS_ONLY;
+  aarch64_set_asm_isa_flags (AARCH64_FL_FP | AARCH64_FL_SIMD | extra_flags);
 }
 
 aarch64_simd_switcher::~aarch64_simd_switcher ()
 {
   if (m_old_general_regs_only)
     global_options.x_target_flags |= MASK_GENERAL_REGS_ONLY;
-  aarch64_isa_flags = m_old_isa_flags;
+  aarch64_set_asm_isa_flags (m_old_asm_isa_flags);
 }
 
 /* Implement #pragma GCC aarch64 "arm_neon.h".  */
diff --git a/gcc/config/aarch64/aarch64-protos.h b/gcc/config/aarch64/aarch64-protos.h
index ef84df731..86e444a60 100644
--- a/gcc/config/aarch64/aarch64-protos.h
+++ b/gcc/config/aarch64/aarch64-protos.h
@@ -747,7 +747,7 @@ public:
   ~aarch64_simd_switcher ();
 
 private:
-  unsigned long m_old_isa_flags;
+  unsigned long m_old_asm_isa_flags;
   bool m_old_general_regs_only;
 };
 
@@ -1032,7 +1032,10 @@ extern bool aarch64_classify_address (struct aarch64_address_info *, rtx,
 				      machine_mode, bool,
 				      aarch64_addr_query_type = ADDR_QUERY_M);
 
+void aarch64_set_asm_isa_flags (aarch64_feature_flags);
+
 /* Defined in common/config/aarch64-common.cc.  */
+void aarch64_set_asm_isa_flags (gcc_options *, aarch64_feature_flags);
 bool aarch64_handle_option (struct gcc_options *, struct gcc_options *,
 			     const struct cl_decoded_option *, location_t);
 const char *aarch64_rewrite_selected_cpu (const char *name);
diff --git a/gcc/config/aarch64/aarch64-sve-builtins.cc b/gcc/config/aarch64/aarch64-sve-builtins.cc
index b927a886e..a70e3a6b4 100644
--- a/gcc/config/aarch64/aarch64-sve-builtins.cc
+++ b/gcc/config/aarch64/aarch64-sve-builtins.cc
@@ -696,7 +696,7 @@ static bool
 check_required_extensions (location_t location, tree fndecl,
 			   aarch64_feature_flags required_extensions)
 {
-  auto missing_extensions = required_extensions & ~aarch64_isa_flags;
+  auto missing_extensions = required_extensions & ~aarch64_asm_isa_flags;
   if (missing_extensions == 0)
     return check_required_registers (location, fndecl);
 
diff --git a/gcc/config/aarch64/aarch64.cc b/gcc/config/aarch64/aarch64.cc
index 8cb820767..3e83e48ec 100644
--- a/gcc/config/aarch64/aarch64.cc
+++ b/gcc/config/aarch64/aarch64.cc
@@ -18432,10 +18432,19 @@ aarch64_convert_sve_vector_bits (aarch64_sve_vector_bits_enum value)
     return (int) value / 64;
 }
 
+/* Set the global aarch64_asm_isa_flags to FLAGS and update
+   aarch64_isa_flags accordingly.  */
+
+void
+aarch64_set_asm_isa_flags (aarch64_feature_flags flags)
+{
+  aarch64_set_asm_isa_flags (&global_options, flags);
+}
+
 /* Implement TARGET_OPTION_OVERRIDE.  This is called once in the beginning
    and is used to parse the -m{cpu,tune,arch} strings and setup the initial
    tuning structs.  In particular it must set selected_tune and
-   aarch64_isa_flags that define the available ISA features and tuning
+   aarch64_asm_isa_flags that define the available ISA features and tuning
    decisions.  It must also set selected_arch as this will be used to
    output the .arch asm tags for each function.  */
 
@@ -18444,7 +18453,7 @@ aarch64_override_options (void)
 {
   aarch64_feature_flags cpu_isa = 0;
   aarch64_feature_flags arch_isa = 0;
-  aarch64_isa_flags = 0;
+  aarch64_set_asm_isa_flags (0);
 
   const struct processor *cpu = NULL;
   const struct processor *arch = NULL;
@@ -18484,25 +18493,25 @@ aarch64_override_options (void)
 	}
 
       selected_arch = arch->arch;
-      aarch64_isa_flags = arch_isa;
+      aarch64_set_asm_isa_flags (arch_isa);
     }
   else if (cpu)
     {
       selected_arch = cpu->arch;
-      aarch64_isa_flags = cpu_isa;
+      aarch64_set_asm_isa_flags (cpu_isa);
     }
   else if (arch)
     {
       cpu = &all_cores[arch->ident];
       selected_arch = arch->arch;
-      aarch64_isa_flags = arch_isa;
+      aarch64_set_asm_isa_flags (arch_isa);
     }
   else
     {
       /* No -mcpu or -march specified, so use the default CPU.  */
       cpu = &all_cores[TARGET_CPU_DEFAULT];
       selected_arch = cpu->arch;
-      aarch64_isa_flags = cpu->flags;
+      aarch64_set_asm_isa_flags (cpu->flags);
     }
 
   selected_tune = tune ? tune->ident : cpu->ident;
@@ -18644,7 +18653,7 @@ aarch64_option_print (FILE *file, int indent, struct cl_target_option *ptr)
     = aarch64_get_tune_cpu (ptr->x_selected_tune);
   const struct processor *arch = aarch64_get_arch (ptr->x_selected_arch);
   std::string extension
-    = aarch64_get_extension_string_for_isa_flags (ptr->x_aarch64_isa_flags,
+    = aarch64_get_extension_string_for_isa_flags (ptr->x_aarch64_asm_isa_flags,
 						  arch->flags);
 
   fprintf (file, "%*sselected tune = %s\n", indent, "", cpu->name);
@@ -18752,13 +18761,15 @@ aarch64_handle_attr_arch (const char *str)
 {
   const struct processor *tmp_arch = NULL;
   std::string invalid_extension;
+  aarch64_feature_flags tmp_flags;
   enum aarch64_parse_opt_result parse_res
-    = aarch64_parse_arch (str, &tmp_arch, &aarch64_isa_flags, &invalid_extension);
+    = aarch64_parse_arch (str, &tmp_arch, &tmp_flags, &invalid_extension);
 
   if (parse_res == AARCH64_PARSE_OK)
     {
       gcc_assert (tmp_arch);
       selected_arch = tmp_arch->arch;
+      aarch64_set_asm_isa_flags (tmp_flags);
       return true;
     }
 
@@ -18790,14 +18801,16 @@ aarch64_handle_attr_cpu (const char *str)
 {
   const struct processor *tmp_cpu = NULL;
   std::string invalid_extension;
+  aarch64_feature_flags tmp_flags;
   enum aarch64_parse_opt_result parse_res
-    = aarch64_parse_cpu (str, &tmp_cpu, &aarch64_isa_flags, &invalid_extension);
+    = aarch64_parse_cpu (str, &tmp_cpu, &tmp_flags, &invalid_extension);
 
   if (parse_res == AARCH64_PARSE_OK)
     {
       gcc_assert (tmp_cpu);
       selected_tune = tmp_cpu->ident;
       selected_arch = tmp_cpu->arch;
+      aarch64_set_asm_isa_flags (tmp_flags);
       return true;
     }
 
@@ -18891,7 +18904,7 @@ static bool
 aarch64_handle_attr_isa_flags (char *str)
 {
   enum aarch64_parse_opt_result parse_res;
-  auto isa_flags = aarch64_isa_flags;
+  auto isa_flags = aarch64_asm_isa_flags;
 
   /* We allow "+nothing" in the beginning to clear out all architectural
      features if the user wants to handpick specific features.  */
@@ -18906,7 +18919,7 @@ aarch64_handle_attr_isa_flags (char *str)
 
   if (parse_res == AARCH64_PARSE_OK)
     {
-      aarch64_isa_flags = isa_flags;
+      aarch64_set_asm_isa_flags (isa_flags);
       return true;
     }
 
@@ -19328,8 +19341,12 @@ aarch64_can_inline_p (tree caller, tree callee)
 					   : target_option_default_node);
 
   /* Callee's ISA flags should be a subset of the caller's.  */
+  if ((caller_opts->x_aarch64_asm_isa_flags
+       & callee_opts->x_aarch64_asm_isa_flags)
+      != callee_opts->x_aarch64_asm_isa_flags)
+    return false;
   if ((caller_opts->x_aarch64_isa_flags & callee_opts->x_aarch64_isa_flags)
-       != callee_opts->x_aarch64_isa_flags)
+      != callee_opts->x_aarch64_isa_flags)
     return false;
 
   /* Allow non-strict aligned functions inlining into strict
@@ -22772,7 +22789,7 @@ aarch64_declare_function_name (FILE *stream, const char* name,
   const struct processor *this_arch
     = aarch64_get_arch (targ_options->x_selected_arch);
 
-  auto isa_flags = targ_options->x_aarch64_isa_flags;
+  auto isa_flags = targ_options->x_aarch64_asm_isa_flags;
   std::string extension
     = aarch64_get_extension_string_for_isa_flags (isa_flags,
 						  this_arch->flags);
@@ -22902,7 +22919,7 @@ aarch64_start_file (void)
 
   const struct processor *default_arch
     = aarch64_get_arch (default_options->x_selected_arch);
-  auto default_isa_flags = default_options->x_aarch64_isa_flags;
+  auto default_isa_flags = default_options->x_aarch64_asm_isa_flags;
   std::string extension
     = aarch64_get_extension_string_for_isa_flags (default_isa_flags,
 						  default_arch->flags);
diff --git a/gcc/config/aarch64/aarch64.h b/gcc/config/aarch64/aarch64.h
index 50a2ef444..521031efe 100644
--- a/gcc/config/aarch64/aarch64.h
+++ b/gcc/config/aarch64/aarch64.h
@@ -22,6 +22,17 @@
 #ifndef GCC_AARCH64_H
 #define GCC_AARCH64_H
 
+/* Make these flags read-only so that all uses go via
+   aarch64_set_asm_isa_flags.  */
+#ifndef GENERATOR_FILE
+#undef aarch64_asm_isa_flags
+#define aarch64_asm_isa_flags \
+  ((aarch64_feature_flags) global_options.x_aarch64_asm_isa_flags)
+#undef aarch64_isa_flags
+#define aarch64_isa_flags \
+  ((aarch64_feature_flags) global_options.x_aarch64_isa_flags)
+#endif
+
 /* Target CPU builtins.  */
 #define TARGET_CPU_CPP_BUILTINS()	\
   aarch64_cpu_cpp_builtins (pfile)
@@ -51,8 +62,8 @@
 
 /* AdvSIMD is supported in the default configuration, unless disabled by
    -mgeneral-regs-only or by the +nosimd extension.  */
-#define TARGET_SIMD (!TARGET_GENERAL_REGS_ONLY && AARCH64_ISA_SIMD)
-#define TARGET_FLOAT (!TARGET_GENERAL_REGS_ONLY && AARCH64_ISA_FP)
+#define TARGET_SIMD (AARCH64_ISA_SIMD)
+#define TARGET_FLOAT (AARCH64_ISA_FP)
 
 #define UNITS_PER_WORD		8
 
@@ -242,7 +253,7 @@ enum class aarch64_feature : unsigned char {
 #define TARGET_DOTPROD (TARGET_SIMD && AARCH64_ISA_DOTPROD)
 
 /* SVE instructions, enabled through +sve.  */
-#define TARGET_SVE (!TARGET_GENERAL_REGS_ONLY && AARCH64_ISA_SVE)
+#define TARGET_SVE (AARCH64_ISA_SVE)
 
 /* SVE2 instructions, enabled through +sve2.  */
 #define TARGET_SVE2 (TARGET_SVE && AARCH64_ISA_SVE2)
diff --git a/gcc/config/aarch64/aarch64.opt b/gcc/config/aarch64/aarch64.opt
index 47ec7824f..5f507abd4 100644
--- a/gcc/config/aarch64/aarch64.opt
+++ b/gcc/config/aarch64/aarch64.opt
@@ -27,6 +27,9 @@ enum aarch64_processor selected_tune = aarch64_none
 TargetVariable
 enum aarch64_arch selected_arch = aarch64_no_arch
 
+TargetVariable
+aarch64_feature_flags aarch64_asm_isa_flags = 0
+
 TargetVariable
 aarch64_feature_flags aarch64_isa_flags = 0
 
-- 
2.33.0