This is the mail archive of the
mailing list for the GCC project.
Re: [MinGW] Allow Building GCC with --disable-sjlj-exceptions
- From: Danny Smith <dannysmith at clear dot net dot nz>
- To: GCC-patches <gcc-patches at gcc dot gnu dot org>
- Cc: Ranjit Mathew <rmathew at gmail dot com>
- Date: Fri, 16 Jun 2006 09:25:42 +1200
- Subject: Re: [MinGW] Allow Building GCC with --disable-sjlj-exceptions
Hello Ranjit, Brian
Brian Dessent wrote:
> Ranjit Mathew wrote:
>> -#define DWARF2_UNWIND_INFO 0
>> +#define DWARF2_UNWIND_INFO 1
> This makes no sense at all. DW2 EH is seriously broken under
> cygwin/mingw, and it's not clear that this is even fundamentally
> (e.g. the situation of throwing an exception in a callback from third
> party code that doesn't use DW2 like the win32 API.) Someone posted
> the mingw list recently with a testcase showing even further breakage
> DW2 EH when throwing across a DLL boundary. Flipping this switch
> be the *last* step that happens after all these other problems are
> fixed, not the first step. SJLJ is still the only sane default for
I would disagree about the "seriously broken". I have been using it in
my own gcc toolset on mingw for three years and I am convinced that it
has more benefits than deficiencies.
The report of "even further breakage" was actually exciting to me
because the reporter cared enough to also submit a tidy little testcase
Now we can fix the bug. (FWIW, a very similar initialization bug was
reported on the mingw list when the experimental EH-acoss-DLL-boundaries
-- using SjLj -- was first unleashed on mingw users for testing.)
I agree with Ranjit that the option to enable DW2 EH should be made
available on mingw32 target.
But I think it should wait until stage 1.