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 PR target/6476


On Mon, Apr 29, 2002 at 07:37:30AM -0700, Mark Mitchell wrote:
> > The following patch adds the needed configury checks for .eh_frame
> > optimization bugs recently discovered in binutils.
> 
> Let's back up a second.
> 
> Does it make more sense just to require some version of binutils with
> a version number greater than x?

Unfortunately 2.12.1 with this fix won't be out until mid-May I'm afraid,
plus H.J's binutils are there too, so test for ld version >= 2.12.1 doesn't
work very well, when e.g. 2.12.90.0.{1,3,4} has the same problems as 2.12
or 2.11.93.0.2 and 2.12.90.0.7 has it fixed as well as 2.12.1 will.

> It seems like it would be simpler to me to just to this:
> 
> 1) If the --enable switch is explicitly present, honor it.
> 
> 2) Otherwise, if the version number is large enough that we know it will
>    work, use the magic.
> 
> 3) Otherwise, don't use the magic.
> 
> Frankly, I think this will be more reliable than the kind of configury
> test you've written -- which, while I'm sure it's correct, must have
> taken you a lot of brain power to write down.
> 
> Would you reimplement the patch to work like that, unless you object?
> 
> --
> Mark Mitchell                   mark@codesourcery.com
> CodeSourcery, LLC               http://www.codesourcery.com

	Jakub


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