This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: [PATCH, PR 46053] Devirtualization and thunks revisited

From: Andi Kleen <andi at firstfloor dot org>
To: Richard Guenther <rguenther at suse dot de>
Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>, Jan Hubicka <hubicka at ucw dot cz>
Date: Thu, 21 Oct 2010 19:00:21 +0200
Subject: Re: [PATCH, PR 46053] Devirtualization and thunks revisited
References: <20101021134336.GB21785@virgil.arch.suse.de> <alpine.LNX.2.00.1010211606320.23074@zhemvz.fhfr.qr> <20101021160319.GC21785@virgil.arch.suse.de>

Martin Jambor <mjambor@suse.cz> writes:
>
> In fact, IPA passes get direct calls to thunks horribly wrong.  I have
> just added a loop to main in testsuite/g++.dg/torture/pr45699.C so
> that gcc attempts to inline the direct call to thunk and I got aborts
> caused by ipa-cp and ipa-inline (when enabling either or both).

At least I found devirtualization doesn't even work for very simple C 
pointers, see PR45631
-Andi

-- 
ak@linux.intel.com -- Speaking for myself only.

References:
- [PATCH, PR 46053] Devirtualization and thunks revisited
  - From: Martin Jambor
- Re: [PATCH, PR 46053] Devirtualization and thunks revisited
  - From: Richard Guenther
- Re: [PATCH, PR 46053] Devirtualization and thunks revisited
  - From: Martin Jambor

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]