Re: correct -fno-guess-branch-probabillity docs

["Mark D. Baushke" <mdb at juniper dot net>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

James E Wilson <wilson@specifixinc.com> writes:

> A customer was confused by the wording of the
> -fno-guess-branch-probability documentation.  The docs say gcc by
> default uses a randomized model, and has non-deterministic behaviour. 
> This is not true.  If it was, then the object file comparisons which are
> part of the bootstrap process would not work.
> 
> Here is an attempt to rewrite the doc entry to be more accurate.
> 
> The option and the docs were added here:
>     http://gcc.gnu.org/ml/gcc-patches/2001-01/msg01275.html
>     http://gcc.gnu.org/ml/gcc-patches/2001-01/msg01380.html
> And my original explanation of what was going on is here:
>     http://gcc.gnu.org/ml/gcc-patches/2001-01/msg01349.html
> 
> Comments anyone?  If I don't get any comments, I will go ahead and add
> this in a few days.

I like the rewritten text. It is much more understandable and stops
giving bad advise to real-time system people.

	Thanks!
	-- Mark
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (FreeBSD)

iD8DBQFBiCq8DsmuPPFOO2cRAiUPAJ0fyx5eGLrV6y1bDT93gCvUGDVhjgCfYJK6
ux+7C4TSSFv8umnxlgNCaNU=
=9x08
-----END PGP SIGNATURE-----