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 r9-9558] c++: 'this' adjustment for devirtualized call Date: Fri, 28 May 2021 09:23:10 +0000 (GMT) [thread overview] Message-ID: <20210528092310.B6D053857001@sourceware.org> (raw) https://gcc.gnu.org/g:ebfe8b28d40746ff33724bd5b9ade2552e619213 commit r9-9558-gebfe8b28d40746ff33724bd5b9ade2552e619213 Author: Jason Merrill <jason@redhat.com> Date: Thu May 27 23:54:52 2021 -0400 c++: 'this' adjustment for devirtualized call My patch for 95719 made us do a better job of finding the actual virtual function we want to call, but didn't update the 'this' pointer adjustment to match. This backport also incorporates a bit of the r11-1638 reorganization. PR c++/100797 PR c++/95719 gcc/cp/ChangeLog: * call.c (build_over_call): Adjust base_binfo in resolves_to_fixed_type_p case. gcc/testsuite/ChangeLog: * g++.dg/inherit/virtual15.C: New test. * g++.dg/inherit/virtual15a.C: New test. Diff: --- gcc/cp/call.c | 44 ++++++++++++++++++------------- gcc/testsuite/g++.dg/inherit/virtual15.C | 18 +++++++++++++ gcc/testsuite/g++.dg/inherit/virtual15a.C | 19 +++++++++++++ 3 files changed, 62 insertions(+), 19 deletions(-) diff --git a/gcc/cp/call.c b/gcc/cp/call.c index 7f76bd905ca..304c01619da 100644 --- a/gcc/cp/call.c +++ b/gcc/cp/call.c @@ -8309,19 +8309,6 @@ build_over_call (struct z_candidate *cand, int flags, tsubst_flags_t complain) || CLASSTYPE_FINAL (TYPE_METHOD_BASETYPE (TREE_TYPE (fn)))) flags |= LOOKUP_NONVIRTUAL; - /* If we know the dynamic type of the object, look up the final overrider - in the BINFO. */ - if (DECL_VINDEX (fn) && (flags & LOOKUP_NONVIRTUAL) == 0 - && resolves_to_fixed_type_p (arg)) - { - tree binfo = cand->conversion_path; - if (BINFO_TYPE (binfo) != DECL_CONTEXT (fn)) - binfo = lookup_base (binfo, DECL_CONTEXT (fn), ba_unique, - NULL, complain); - fn = lookup_vfn_in_binfo (DECL_VINDEX (fn), binfo); - flags |= LOOKUP_NONVIRTUAL; - } - /* [class.mfct.nonstatic]: If a nonstatic member function of a class X is called for an object that is not of type X, or of a type derived from X, the behavior is undefined. @@ -8331,10 +8318,6 @@ build_over_call (struct z_candidate *cand, int flags, tsubst_flags_t complain) gcc_assert (TYPE_PTR_P (parmtype)); /* Convert to the base in which the function was declared. */ gcc_assert (cand->conversion_path != NULL_TREE); - converted_arg = build_base_path (PLUS_EXPR, - arg, - cand->conversion_path, - 1, complain); /* Check that the base class is accessible. */ if (!accessible_base_p (TREE_TYPE (argtype), BINFO_TYPE (cand->conversion_path), true)) @@ -8349,10 +8332,33 @@ build_over_call (struct z_candidate *cand, int flags, tsubst_flags_t complain) /* If fn was found by a using declaration, the conversion path will be to the derived class, not the base declaring fn. We must convert from derived to base. */ - base_binfo = lookup_base (TREE_TYPE (TREE_TYPE (converted_arg)), + base_binfo = lookup_base (cand->conversion_path, TREE_TYPE (parmtype), ba_unique, NULL, complain); - converted_arg = build_base_path (PLUS_EXPR, converted_arg, + + /* If we know the dynamic type of the object, look up the final overrider + in the BINFO. */ + if (DECL_VINDEX (fn) && (flags & LOOKUP_NONVIRTUAL) == 0 + && resolves_to_fixed_type_p (arg)) + { + tree ov = lookup_vfn_in_binfo (DECL_VINDEX (fn), base_binfo); + + /* And unwind base_binfo to match. If we don't find the type we're + looking for in BINFO_INHERITANCE_CHAIN, we're looking at diamond + inheritance; for now do a normal virtual call in that case. */ + tree octx = DECL_CONTEXT (ov); + tree obinfo = base_binfo; + while (obinfo && !SAME_BINFO_TYPE_P (BINFO_TYPE (obinfo), octx)) + obinfo = BINFO_INHERITANCE_CHAIN (obinfo); + if (obinfo) + { + fn = ov; + base_binfo = obinfo; + flags |= LOOKUP_NONVIRTUAL; + } + } + + converted_arg = build_base_path (PLUS_EXPR, arg, base_binfo, 1, complain); argarray[j++] = converted_arg; diff --git a/gcc/testsuite/g++.dg/inherit/virtual15.C b/gcc/testsuite/g++.dg/inherit/virtual15.C new file mode 100644 index 00000000000..ebd8e3ad29b --- /dev/null +++ b/gcc/testsuite/g++.dg/inherit/virtual15.C @@ -0,0 +1,18 @@ +// PR c++/100797 +// { dg-do run } + +bool ok = false; +struct S1 { virtual ~S1() {} }; +struct S2 { virtual void f1() = 0; }; +struct S3: S1, S2 { + void f1() { f2(); } + virtual void f2() = 0; +}; +struct S4: S3 { + void f2() { ok = true; } + using S2::f1; +}; +int main() { + S4().f1(); + if (!ok) __builtin_abort (); +} diff --git a/gcc/testsuite/g++.dg/inherit/virtual15a.C b/gcc/testsuite/g++.dg/inherit/virtual15a.C new file mode 100644 index 00000000000..6139385192d --- /dev/null +++ b/gcc/testsuite/g++.dg/inherit/virtual15a.C @@ -0,0 +1,19 @@ +// PR c++/100797 plus diamond inheritance +// { dg-do run } + +bool ok = false; +struct S1 { virtual ~S1() {} }; +struct S2 { virtual void f1() = 0; }; +struct S3: S1, virtual S2 { + void f1() { f2(); } + virtual void f2() = 0; +}; +struct SX: virtual S2 { }; +struct S4: SX, S3 { + void f2() { ok = true; } + using S2::f1; +}; +int main() { + S4().f1(); + if (!ok) __builtin_abort (); +}
reply other threads:[~2021-05-28 9:23 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=20210528092310.B6D053857001@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).