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: Default Solaris 2/x86 to -xarch/-xtune=pentium4


Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE> writes:

> While config.guess still detects Solaris 2/x86 as i386-pc-solaris2.*, it
> seems to make sense to default -xarch/-xtune to pentium4, as is done on
> most other i386 targets.  The following patch implements this.  It only
> sets with_arch_32 since setting with_arch breaks the bi-arch
> i386-pc-solaris2.1[0-9] ports.  I'm setting with_tune_32 at the same
> time, since otherwise -xtune still defaults to i386, which doesn't seem
> to make sense.
>
> For comparionds, I've looked at what recent versions of the Sun Studio
> compilers do:
>
> Release  		OS support	-xarch default
>
> Studio 12 Update 1	S10 1/06+	generic (pentium_pro @ -m32,
>                                         sse2 @ -m64)
> Studio 12		S9+	   	generic (most likely pentium_pro, 386
> 					isn't listed)
> Studio 11		S8, 9, 10	generic (386)
>
> Since we still support Solaris 7 (although obsoleted) and 8, one could
> argue to leave the -xarch at 386, but this seems counterproductive
> compared to what other platforms do.  I'll probably add a hint to the
> installation guide.
>
> Bootstrapped without regressions on i386-pc-solaris2.11.
>
> In fact, this patch has most likely fixed a couple of testsuite
> failures:
>
> -FAIL: gcc.dg/tree-ssa/ltrans-1.c scan-tree-dump-times ltrans "converted loop nest to perfect loop nest" 1
> -FAIL: gcc.dg/tree-ssa/ltrans-1.c scan-tree-dump-times ltrans "transformed loop" 1
> -FAIL: gcc.dg/tree-ssa/ltrans-3.c scan-tree-dump-times ltrans "transformed loop" 1
> -FAIL: gcc.dg/tree-ssa/ltrans-4.c scan-tree-dump-times ltrans "transformed loop" 1
> -FAIL: gcc.dg/tree-ssa/ltrans-5.c scan-tree-dump-times ltrans "transformed loop" 1
> -FAIL: gcc.dg/tree-ssa/ltrans-6.c scan-tree-dump-times ltrans "transformed loop" 1
> -FAIL: gcc.dg/tree-ssa/ltrans-8.c scan-tree-dump-times ltrans "transformed loop" 1
> -FAIL: gcc.dg/tree-ssa/vrp47.c scan-tree-dump-times vrp1 "[xy][^ ]* !=" 0
> -FAIL: gcc.dg/tree-ssa/vrp47.c scan-tree-dump-times dom1 "x[^ ]* & y" 1
> -FAIL: gcc.dg/tree-ssa/vrp47.c scan-tree-dump-times vrp1 "x[^ ]* [|] y" 1
>
> If this turns out to be true, these testcases silently depending on
> -xarch=i686 would be testsuite bugs by themselves.
>
> While I don't need approval for this change, I'd like to hear if I've
> overlooked any caveats here.

Given the imminent freeze and no comments whatsoever, I've checked this in.

	Rainer

-- 
-----------------------------------------------------------------------------
Rainer Orth, Center for Biotechnology, Bielefeld University


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