[lto]: merge from trunk

Rafael Espindola espindola@google.com
Thu Jul 10 08:28:00 GMT 2008


2008/7/10 Bill Maddox <maddox@google.com>:
> On Wed, Jul 9, 2008 at 4:05 PM, Kenneth Zadeck <zadeck@naturalbridge.com> wrote:
>
>> seems to assert in a number of tests.    I assume that this is related to no
>> quite getting everything in the va type correct.   This needs to be looked
>> into further.
>
> I attempted to build the branch, and found that it failed due to 1)
> the use of an uninitialized variable
> in input_binfo (see my remarks on Rafael's last patch)

I was able to reproduce this one. Sorry for the breakage.

> and 2) zillions
> of warnings, treated as errors,
> about implicit casts and casting away of qualifiers presumably due to
> the -Wcast-qual and -Wc++-compat
> flags.  My configuration is very vanilla, with default compiler
> options.  It's not clear why I have seen
> these problems and Kenny did not, unless my host compiler interprets
> these options differently.
> In any case, here is a patch for all these issues that compiles and
> appears to work -- no change log or
> regression test run yet.  I suppose it could be argued that we should
> be using these warnings and make
> our code clean, but was it our intent to throw the switch now?

I was able to bootstrap with 137664 + you patch. Has the merge changed
the warning flags?

> --Bill
>


Cheers,
-- 
Rafael Avila de Espindola

Google Ireland Ltd.
Gordon House
Barrow Street
Dublin 4
Ireland

Registered in Dublin, Ireland
Registration Number: 368047



More information about the Gcc-patches mailing list