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] Fix COMDAT group of thunks


On Sun, Feb 27, 2005 at 12:17:53PM +0000, Nathan Sidwell wrote:
> Jakub Jelinek wrote:
> >On Fri, Feb 25, 2005 at 01:50:42PM +0000, Nathan Sidwell wrote:
> 
> >>With covariancy THUNK_TARGET can itself be a thunk.  Do you need an 
> >>additional
> >>	if (THUNK_P (target))
> >>	target = THUNK_TARGET (target)
> >>here?  (and in use_thunk?)
> >
> >
> >How do you think should a testcase for that look like?
> >I have tried g++.dg/abi/covariant*.C and on none of the these
> 
> abi/covariant4.C contains such a case.  In use_thunk it encounters such a
> condition (I think use_thunk handles it incorrectly, btw).  I don't have
> a comdat-group supporting assembler - binutils 2.15 segfaults during the
> configure test.  do you know what version is good on i86?

You'll have to try HEAD since 2.16 isn't branched yet.

-- 
Daniel Jacobowitz
CodeSourcery, LLC


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