x86: Don't save callee-saved registers in noreturn functions

There is no need to save callee-saved registers in noreturn functions
if they don't throw nor support exceptions.  We can treat them the same
as functions with no_callee_saved_registers attribute.

Adjust stack-check-17.c for noreturn function which no longer saves any
registers.

With this change, __libc_start_main in glibc 2.39, which is a noreturn
function, is changed from

__libc_start_main:
	endbr64
	push   %r15
	push   %r14
	mov    %rcx,%r14
	push   %r13
	push   %r12
	push   %rbp
	mov    %esi,%ebp
	push   %rbx
	mov    %rdx,%rbx
	sub    $0x28,%rsp
	mov    %rdi,(%rsp)
	mov    %fs:0x28,%rax
	mov    %rax,0x18(%rsp)
	xor    %eax,%eax
	test   %r9,%r9

to

__libc_start_main:
	endbr64
        sub    $0x28,%rsp
        mov    %esi,%ebp
        mov    %rdx,%rbx
        mov    %rcx,%r14
        mov    %rdi,(%rsp)
        mov    %fs:0x28,%rax
        mov    %rax,0x18(%rsp)
        xor    %eax,%eax
        test   %r9,%r9

In Linux kernel 6.7.0 on x86-64, do_exit is changed from

do_exit:
        endbr64
        call   <do_exit+0x9>
        push   %r15
        push   %r14
        push   %r13
        push   %r12
        mov    %rdi,%r12
        push   %rbp
        push   %rbx
        mov    %gs:0x0,%rbx
        sub    $0x28,%rsp
        mov    %gs:0x28,%rax
        mov    %rax,0x20(%rsp)
        xor    %eax,%eax
        call   *0x0(%rip)        # <do_exit+0x39>
        test   $0x2,%ah
        je     <do_exit+0x8d3>

to

do_exit:
        endbr64
        call   <do_exit+0x9>
        sub    $0x28,%rsp
        mov    %rdi,%r12
        mov    %gs:0x28,%rax
        mov    %rax,0x20(%rsp)
        xor    %eax,%eax
        mov    %gs:0x0,%rbx
        call   *0x0(%rip)        # <do_exit+0x2f>
        test   $0x2,%ah
        je     <do_exit+0x8c9>

I compared GCC master branch bootstrap and test times on a slow machine
with 6.6 Linux kernels compiled with the original GCC 13 and the GCC 13
with the backported patch.  The performance data isn't precise since the
measurements were done on different days with different GCC sources under
different 6.6 kernel versions.

GCC master branch build time in seconds:

before                after                  improvement
30043.75user          30013.16user           0%
1274.85system         1243.72system          2.4%

GCC master branch test time in seconds (new tests added):

before                after                  improvement
216035.90user         216547.51user          0
27365.51system        26658.54system         2.6%

gcc/

	PR target/38534
	* config/i386/i386-options.cc (ix86_set_func_type): Don't
	save and restore callee saved registers for a noreturn function
	with nothrow or compiled with -fno-exceptions.

gcc/testsuite/

	PR target/38534
	* gcc.target/i386/pr38534-1.c: New file.
	* gcc.target/i386/pr38534-2.c: Likewise.
	* gcc.target/i386/pr38534-3.c: Likewise.
	* gcc.target/i386/pr38534-4.c: Likewise.
	* gcc.target/i386/stack-check-17.c: Updated.

gcc/config/i386/i386-options.cc

@@ -3380,9 +3380,21 @@ ix86_simd_clone_adjust (struct cgraph_node *node)
 static void
 ix86_set_func_type (tree fndecl)
 {
+  /* No need to save and restore callee-saved registers for a noreturn
+     function with nothrow or compiled with -fno-exceptions.
+
+     NB: Don't use TREE_THIS_VOLATILE to check if this is a noreturn
+     function.  The local-pure-const pass turns an interrupt function
+     into a noreturn function by setting TREE_THIS_VOLATILE.  Normally
+     the local-pure-const pass is run after ix86_set_func_type is called.
+     When the local-pure-const pass is enabled for LTO, the interrupt
+     function is marked as noreturn in the IR output, which leads the
+     incompatible attribute error in LTO1.  */
   bool has_no_callee_saved_registers
-    = lookup_attribute ("no_callee_saved_registers",
-			TYPE_ATTRIBUTES (TREE_TYPE (fndecl)));
+    = (((TREE_NOTHROW (fndecl) || !flag_exceptions)
+	&& lookup_attribute ("noreturn", DECL_ATTRIBUTES (fndecl)))
+       || lookup_attribute ("no_callee_saved_registers",
+			    TYPE_ATTRIBUTES (TREE_TYPE (fndecl))));
 
   if (cfun->machine->func_type == TYPE_UNKNOWN)
     {

gcc/testsuite/gcc.target/i386/pr38534-1.c

@@ -0,0 +1,26 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mtune-ctrl=^prologue_using_move,^epilogue_using_move" } */
+
+#define ARRAY_SIZE 256
+
+extern int array[ARRAY_SIZE][ARRAY_SIZE][ARRAY_SIZE];
+extern int value (int, int, int)
+#ifndef __x86_64__
+__attribute__ ((regparm(3)))
+#endif
+;
+
+void
+__attribute__((noreturn))
+no_return_to_caller (void)
+{
+  unsigned i, j, k;
+  for (i = ARRAY_SIZE; i > 0; --i)
+    for (j = ARRAY_SIZE; j > 0; --j)
+      for (k = ARRAY_SIZE; k > 0; --k)
+	array[i - 1][j - 1][k - 1] = value (i, j, k);
+  while (1);
+}
+
+/* { dg-final { scan-assembler-not "push" } } */
+/* { dg-final { scan-assembler-not "pop" } } */

gcc/testsuite/gcc.target/i386/pr38534-2.c

@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mtune-ctrl=^prologue_using_move,^epilogue_using_move" } */
+
+extern void bar (void) __attribute__ ((no_callee_saved_registers));
+extern void fn (void) __attribute__ ((noreturn));
+
+__attribute__ ((noreturn))
+void
+foo (void)
+{
+  bar ();
+  fn ();
+}
+
+/* { dg-final { scan-assembler-not "push" } } */
+/* { dg-final { scan-assembler-not "pop" } } */
+/* { dg-final { scan-assembler-not "jmp\[\\t \]+_?bar" } } */
+/* { dg-final { scan-assembler "call\[\\t \]+_?bar" } } */

gcc/testsuite/gcc.target/i386/pr38534-3.c

@@ -0,0 +1,19 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mtune-ctrl=^prologue_using_move,^epilogue_using_move" } */
+
+typedef void (*fn_t) (void) __attribute__ ((no_callee_saved_registers));
+extern fn_t bar;
+extern void fn (void) __attribute__ ((noreturn));
+
+__attribute__ ((noreturn))
+void
+foo (void)
+{
+  bar ();
+  fn ();
+}
+
+/* { dg-final { scan-assembler-not "push" } } */
+/* { dg-final { scan-assembler-not "pop" } } */
+/* { dg-final { scan-assembler-not "jmp" } } */
+/* { dg-final { scan-assembler "call\[\\t \]+" } } */

gcc/testsuite/gcc.target/i386/pr38534-4.c

@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mtune-ctrl=^prologue_using_move,^epilogue_using_move" } */
+
+typedef void (*fn_t) (void) __attribute__ ((no_callee_saved_registers));
+extern void fn (void) __attribute__ ((noreturn));
+
+__attribute__ ((noreturn))
+void
+foo (fn_t bar)
+{
+  bar ();
+  fn ();
+}
+
+/* { dg-final { scan-assembler-not "push" } } */
+/* { dg-final { scan-assembler-not "pop" } } */
+/* { dg-final { scan-assembler-not "jmp" } } */
+/* { dg-final { scan-assembler "call\[\\t \]+" } } */

gcc/testsuite/gcc.target/i386/stack-check-17.c

@@ -23,19 +23,14 @@ f3 (void)
 /* Verify no explicit probes.  */
 /* { dg-final { scan-assembler-not "or\[ql\]" } } */
 
-/* We also want to verify we did not use a push/pop sequence
-   to probe *sp as the callee register saves are sufficient
-   to probe *sp.
-
-   y0/y1 are live across the call and thus must be allocated
+/* y0/y1 are live across the call and thus must be allocated
    into either a stack slot or callee saved register.  The former
    would be rather dumb.  So assume it does not happen.
 
-   So search for two/four pushes for the callee register saves/argument pushes
-   (plus one for the PIC register if needed on ia32) and no pops (since the
-   function has no reachable epilogue).  */
-/* { dg-final { scan-assembler-times "push\[ql\]" 2 { target { ! ia32 } } } }  */
-/* { dg-final { scan-assembler-times "push\[ql\]" 4 { target { ia32 && nonpic } } } }  */
-/* { dg-final { scan-assembler-times "push\[ql\]" 5 { target { ia32 && { ! nonpic } } } } }  */
-/* { dg-final { scan-assembler-not "pop" } } */
+   So search for a push/pop sequence for stack probe and 2 argument
+   pushes on ia32.  There is no need to save and restore the PIC
+   register on ia32 for a noreturn function.  */
+/* { dg-final { scan-assembler-times "push\[ql\]" 1 { target { ! ia32 } } } }  */
+/* { dg-final { scan-assembler-times "push\[ql\]" 3 { target ia32 } } }  */
+/* { dg-final { scan-assembler-times "pop" 1 } } */