This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug rtl-optimization/53176] [4.8 Regression] gcc.target/i386/movbe-2.c and gcc.dg/lower-subreg-1.c
- From: "zadeck at naturalbridge dot com" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Tue, 01 May 2012 18:32:32 +0000
- Subject: [Bug rtl-optimization/53176] [4.8 Regression] gcc.target/i386/movbe-2.c and gcc.dg/lower-subreg-1.c
- Auto-submitted: auto-generated
- References: <bug-53176-4@http.gcc.gnu.org/bugzilla/>
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53176
Kenneth Zadeck <zadeck at naturalbridge dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |zadeck at naturalbridge dot
| |com
--- Comment #1 from Kenneth Zadeck <zadeck at naturalbridge dot com> 2012-05-01 18:32:32 UTC ---
hj
this is exactly the kind of regression that we expect!!! The difference
between this version and the previous versions of lower-subreg, is that this
version takes into account the rtl costs of the operations transformed. If
the rtl costs do not indicate that this would be profitable, then the
transformations are not preformed.
it is up to the port maintainer to set the rtl costs correctly for their port.
If the port maintainer believes that his/her costs are correct and we are
still missing optimizations, then we will look into it.
Lower-subreg is most likely a terrible transformation to do on later x86s but
is likely to have been useful on the earlier ones. So this regression is
likely a good thing.
kenny