public inbox for gcc-cvs@sourceware.org help / color / mirror / Atom feed
From: Richard Biener <rguenth@gcc.gnu.org> To: gcc-cvs@gcc.gnu.org Subject: [gcc r13-1106] tree-optimization/105971 - less surprising refs_may_alias_p_2 Date: Wed, 15 Jun 2022 11:15:27 +0000 (GMT) [thread overview] Message-ID: <20220615111527.96669385C32F@sourceware.org> (raw) https://gcc.gnu.org/g:8c2733e16ec1c0cdda3db4cdc5ad158a96a658e8 commit r13-1106-g8c2733e16ec1c0cdda3db4cdc5ad158a96a658e8 Author: Richard Biener <rguenther@suse.de> Date: Wed Jun 15 11:27:31 2022 +0200 tree-optimization/105971 - less surprising refs_may_alias_p_2 When DSE asks whether __real a is using __imag a it gets a surprising result when a is a FUNCTION_DECL. The following makes sure this case is less surprising to callers but keeping the bail-out for the non-decl case where it is true that PTA doesn't track aliases to code correctly. 2022-06-15 Richard Biener <rguenther@suse.de> PR tree-optimization/105971 * tree-ssa-alias.cc (refs_may_alias_p_2): Put bail-out for FUNCTION_DECL and LABEL_DECL refs after decl-decl disambiguation to leak less surprising alias results. * gcc.dg/torture/pr106971.c: New testcase. Diff: --- gcc/testsuite/gcc.dg/torture/pr106971.c | 12 ++++++++++++ gcc/tree-ssa-alias.cc | 18 +++++++++--------- 2 files changed, 21 insertions(+), 9 deletions(-) diff --git a/gcc/testsuite/gcc.dg/torture/pr106971.c b/gcc/testsuite/gcc.dg/torture/pr106971.c new file mode 100644 index 00000000000..33f21056e62 --- /dev/null +++ b/gcc/testsuite/gcc.dg/torture/pr106971.c @@ -0,0 +1,12 @@ +/* { dg-do compile } */ + +void a() +{ + int b; + int c; + int d = (__INTPTR_TYPE__)a; + _Complex float *e = (_Complex float *)a; + for (;;) { + (*e += d) / b ?: 0; + } +} diff --git a/gcc/tree-ssa-alias.cc b/gcc/tree-ssa-alias.cc index 063f1893851..b1e7a2d5afc 100644 --- a/gcc/tree-ssa-alias.cc +++ b/gcc/tree-ssa-alias.cc @@ -2397,15 +2397,6 @@ refs_may_alias_p_2 (ao_ref *ref1, ao_ref *ref2, bool tbaa_p) || CONSTANT_CLASS_P (base2)) return false; - /* We can end up referring to code via function and label decls. - As we likely do not properly track code aliases conservatively - bail out. */ - if (TREE_CODE (base1) == FUNCTION_DECL - || TREE_CODE (base1) == LABEL_DECL - || TREE_CODE (base2) == FUNCTION_DECL - || TREE_CODE (base2) == LABEL_DECL) - return true; - /* Two volatile accesses always conflict. */ if (ref1->volatile_p && ref2->volatile_p) @@ -2432,6 +2423,15 @@ refs_may_alias_p_2 (ao_ref *ref1, ao_ref *ref2, bool tbaa_p) ref2ref, base2, offset2, max_size2, ref2->size); + /* We can end up referring to code via function and label decls. + As we likely do not properly track code aliases conservatively + bail out. */ + if (TREE_CODE (base1) == FUNCTION_DECL + || TREE_CODE (base1) == LABEL_DECL + || TREE_CODE (base2) == FUNCTION_DECL + || TREE_CODE (base2) == LABEL_DECL) + return true; + /* Handle restrict based accesses. ??? ao_ref_base strips inner MEM_REF [&decl], recover from that here. */
reply other threads:[~2022-06-15 11:15 UTC|newest] Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20220615111527.96669385C32F@sourceware.org \ --to=rguenth@gcc.gnu.org \ --cc=gcc-cvs@gcc.gnu.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).