Hi!

The following code is rejected e.g. on mips64el-linux (but I think many
other targets which don't support target attribute or pragma).
The problem is that the change to decl_attributes below is done
unconditionally and with just #pragma GCC push_options/pop_options pair
we have target_option_default_node NULL, but after popping options
target_option_current_node becomes non-NULL and this decl_attribute
spot fills in DECL_FUNCTION_SPECIFIC_TARGET of a subset of a functions.
Those appearing before push_options/pop_options will have it NULL and
as target_option_default_node is also NULL on those targets, the default
can_inline_p will refuse to inline any functions defined with NULL
DECL_FUNCTION_SPECIFIC_TARGET into any function with non-NULL
DECL_FUNCTION_SPECIFIC_TARGET (even when nothing in the options really
changed).

The following patch restricts that snippet to targets that care (initialize
target_option_default_node to non-NULL to the command line options early)
which include all targets that actually implement target attribute and/or
pragma.

Bootstrapped/regtested on x86_64-linux and i686-linux, tested on the
testcase with cross-compiler to mips64el-linux, ok for trunk?

2022-04-13  Jakub Jelinek  <ja...@redhat.com>

        PR target/105234
        * attribs.cc (decl_attributes): Don't set
        DECL_FUNCTION_SPECIFIC_TARGET if target_option_default_node is
        NULL.

        * gcc.c-torture/compile/pr105234.c: New test.

--- gcc/attribs.cc.jj   2022-02-04 14:36:53.947620076 +0100
+++ gcc/attribs.cc      2022-04-12 12:52:36.630967329 +0200
@@ -636,15 +636,20 @@ decl_attributes (tree *node, tree attrib
       && !DECL_FUNCTION_SPECIFIC_OPTIMIZATION (*node))
     {
       DECL_FUNCTION_SPECIFIC_OPTIMIZATION (*node) = optimization_current_node;
-      tree cur_tree
-       = build_target_option_node (&global_options, &global_options_set);
-      tree old_tree = DECL_FUNCTION_SPECIFIC_TARGET (*node);
-      if (!old_tree)
-       old_tree = target_option_default_node;
-      /* The changes on optimization options can cause the changes in
-        target options, update it accordingly if it's changed.  */
-      if (old_tree != cur_tree)
-       DECL_FUNCTION_SPECIFIC_TARGET (*node) = cur_tree;
+      /* Don't set DECL_FUNCTION_SPECIFIC_TARGET for targets that don't
+        support #pragma GCC target or target attribute.  */
+      if (target_option_default_node)
+       {
+         tree cur_tree
+           = build_target_option_node (&global_options, &global_options_set);
+         tree old_tree = DECL_FUNCTION_SPECIFIC_TARGET (*node);
+         if (!old_tree)
+           old_tree = target_option_default_node;
+         /* The changes on optimization options can cause the changes in
+            target options, update it accordingly if it's changed.  */
+         if (old_tree != cur_tree)
+           DECL_FUNCTION_SPECIFIC_TARGET (*node) = cur_tree;
+       }
     }
 
   /* If this is a function and the user used #pragma GCC target, add the
--- gcc/testsuite/gcc.c-torture/compile/pr105234.c.jj   2022-04-12 
13:07:55.021142913 +0200
+++ gcc/testsuite/gcc.c-torture/compile/pr105234.c      2022-04-12 
13:07:43.596302442 +0200
@@ -0,0 +1,14 @@
+/* PR target/105234 */
+/* { dg-do compile } */
+
+static inline __attribute__((always_inline)) int foo (int x) { return x + 1; }
+#pragma GCC push_options
+static inline __attribute__((always_inline)) int bar (int x) { return x + 2; }
+#pragma GCC pop_options
+static inline __attribute__((always_inline)) int baz (int x) { return x + 3; }
+
+int
+qux (void)
+{
+  return foo (bar (baz (42)));
+}

        Jakub

Reply via email to