Re: Reload bug

[Jan Hubicka <jh@suse.cz>]

> > > Concerning this testcase, the reload used to manage to get around this
> > > problem.  It didn't get around intentionally and just because the
> > > instruction contrain does not accept SSE register so reload is needed
> > > anyway.
> > 
> > I think you're completely right: the reload pass has no specific 
> > infrastructure for dealing with invalid subregs. It may "fix" these subregs, 
> > but only if it happens that the operand needs reloading because of the insn 
> > constraints.
> > 
> > I now have an example: PR target/10286
> > 
> > We pass
> > 
> > (insn 33 30 35 (set (subreg:SI (reg/v:DI 67) 4)
> >         (mem/f:SI (plus:SI (reg/f:SI 16 argp)
> >                 (const_int 4 [0x4])) [0 hi+0 S4 A32])) 45 {*movsi_1} 
> > (insn_list 30 (nil))
> > 
> > to reload, which produces
> > 
> > (insn 33 30 35 (set (reg:SI 29 emm0)
> >         (mem/f:SI (plus:SI (reg/f:SI 6 ebp)
> >                 (const_int 12 [0xc])) [0 hi+0 S4 A32])) 45 {*movsi_1} 
> > (insn_list 30 (nil))
> >     (nil))
> > 
> > that is, the only reload for insn 33 is
> > 
> > Reloads for insn # 33
> > Reload 0: reload_in (SI) = (mem/f:SI (plus:SI (reg/f:SI 6 ebp)
> >                                                         (const_int 12 [0xc])) 
> > [0 hi+0 S4 A32])
> > 	GENERAL_REGS, RELOAD_FOR_INPUT (opnum = 1), optional
> > 	reload_in_reg: (mem/f:SI (plus:SI (reg/f:SI 6 ebp)
> >                                                         (const_int 12 [0xc])) 
> > [0 hi+0 S4 A32])
> > 
> > 
> > Incidentally, PR target/10286 is a regression from GCC 3.2 present in
> > GCC 3.2.1 and later. The culprit is this patch:
> > 
> > Sat Oct 19 15:49:14 CEST 2002  Jan Hubicka  <jh@suse.cz>
> > 
> > 	* mmintrin.h (__m64): typedef it to v2si.
> > 	(_mm_cvtsi32_si64, _mm_cvtsi32_si64_mm_sll_pi16,
> > 	_mm_sll_pi32, _mm_sll_pi64, _mm_slli_pi64, _mm_sra_pi16,
> > 	_mm_sra_pi32, _mm_srl_pi16, _mm_srl_pi32, _mm_srl_pi64,
> > 	_mm_srli_pi64, _mm_and_si64, _mm_andnot_si64,
> > 	_mm_or_si64, _mm_xor_si64): Add neccesary casts.
> > 	* xmmintrin.h (_mm_setzero_si64): Likewise.
> > 
> > See http://gcc.gnu.org/ml/gcc-patches/2002-10/msg01091.html
> > 
> > Before the patch, __m64 was defined as 'unsigned long long' so (%eax, %edx) 
> > was used instead of %rmm0 and we didn't produce invalid subregs. I think 
> > this patch has opened a can of worms and the MMX intrinsics are seriously 
> > broken on the 3.2 branch now.
> 
> They always have been - I've fixed 50+ different bugs exposed by simple
> testsuite executing each individual intrincisc with few diferent args,
> so I would not brother about this on the 3.2.  3.3 is different issue
> and should be the first GCC where intrincisc are generaly usable.  I
> hope.
> 
> Will look into this in the detail this evening, now I have about -10
> minutes of time.  Thanks!

Perhaps I would vote for simply disabling the xmm intrincisc support in
the 3.2.x compiler by adding #error into the headers.  It is better than
confusing user by completely broken interface especially one so dificult
to use as xmm/mmx intrincisc are.

Honza
> 
> Honza
> > 
> > -- 
> > Eric Botcazou