Re: [PATCH] Fix SFmode subreg of DImode and TImode

[Hongtao Liu <crazylht@gmail.com>]

On Fri, Sep 10, 2021 at 7:25 PM Hongtao Liu <crazylht@gmail.com> wrote:
>
> On Fri, Sep 10, 2021 at 6:54 PM Richard Biener
> <richard.guenther@gmail.com> wrote:
> >
> > On Fri, Sep 10, 2021 at 5:03 AM Hongtao Liu <crazylht@gmail.com> wrote:
> > >
> > > On Fri, Sep 10, 2021 at 7:49 AM Segher Boessenkool
> > > <segher@kernel.crashing.org> wrote:
> > > >
> > > > On Thu, Sep 09, 2021 at 08:16:16AM +0200, Richard Biener wrote:
> > > > > > I think we should (longer term) get rid of the overloaded meanings and
> > > > > > uses of subregs.  One fairly simple thing is to make a new rtx code
> > > > > > "bit_cast" (or is there a nice short more traditional name for it?)
> > > > >
> > > > > But subreg _is_ bit_cast.
> > > >
> > > > It is not.  (subreg:M (reg:N) O) for O>0, little-endian, is not a
> > > > bit_cast.  It is taking a part of a register, or a single register from
> > > > a multi-register thing.  Paradoxicals are not bit-casts either.
> > > >
> > > > Subregs from or to (but not both) integer modes are generally bit_cast,
> > > > yeah.
> > > >
> > > > > What is odd to me is that a "disallowed" subreg
> > > > > like (subreg:SF (reg:TI ..) 0) magically becomes valid (in terms of
> > > > > validate_subreg) if you rewrite it as (subreg:SF (subreg:SI (reg:TI ..) 0) 0).
> > > > > Of course that's nested and invalid but just push the inner subreg to a
> > > > > new pseudo and the thing becomes valid.
> > > >
> > > > Bingo.
> > > >
> > > > And many targets have strange rules for bit-strings in which modes can
> > > > be used as bit-strings in which other modes, and at what offsets in
> > > > which registers.  Now perhaps none of that is optimal (I bet it isn't),
> > > > but changing this without a transition plan simply does not work.
> > > >
> > > > > > But that is not the core problem we had here.  The behaviour of the
> > > > > > generic parts of the compiler was changed, without testing if that
> > > > > > works on other targets but x86.  That is an understandable mistake, it
> > > > > > takes some experience to know where the morasses are.  But this change
> > > > > > should have been accompanied by testcases exercising the changed code.
> > > > > > We would have clearly seen there are issues then, simply by watching
> > > > > > gcc-testresults@ (and/or maintainers are on top of the test results
> > > > > > anyway).  Also, if there were testcases for this, we could have some
> > > > > > confidence that a change in this area is robust.
> > > > >
> > > > > Well, that only works if some maintainers that are familiar enough
> > > > > with all this chime in ;)
> > > >
> > > > Not really.  It works always.  And it works way better than the
> > > > pandemonium we now have with broken targets left and right.
> > > >
> > > > With testcases anyone can see if any specific target is broken here.
> > > >
> > > > > It's stage1 so it's understandable that some
> > > > > people (like me ...) are tyring to help people making progress even
> > > > > if that involves trying to decipher 30 years of GCC history in this
> > > > > area (without much success in the end as we see) ;)
> > > >
> > > > Yeah :-)  And my thanks to you and everyone involved for tackling this
> > > > problematic part of GCC, which has been neglected and patched over for
> > > > way too long.  But from that same history it follows that anything you
> > > > do not super carefully (with testing everywhere) will cause some serious
> > > Frankly, testing everywhere is too heavy a burden for developers,
> > > after all, everyone has a limited variety of machines, and may not be
> > > familiar with using  other targets' simulators.
> > > And back to the problem we were trying to solve at the beginning
> > > (subreg:HF(reg:SI)), I guess this is not just a problem in x86
> > > backend, any backend can encounter similar problems, that's why we
> > > remove all the weird cases in validate_subreg.
> >
> > So can you please revert the change for now?  I think we need to go
> > back to the issue in extract_bit_field - does it somehow work to use
> > validate_subreg to avoid creating the subreg we ICE on in the first
> > place and what happens then to code quality?
> Sure, let me test the patch.
Survive regtest, code quality seems to be ok.
But lose some performance.
        .cfi_startproc
-       pinsrw  $0, %edi, %xmm0
+       movw    %di, -2(%rsp)
+       pinsrw  $0, -2(%rsp), %xmm0
        ret
        .cfi_endproc

Anyway I'll post the patches first and see if I can fix the
performance issue in the x86 backend.

> >
> > Thanks,
> > Richard.
> >
> > > > problems.  And nonse of these are easy to fix at all -- there is a
> > > > *reason* targets did this nastiness.
> > > >
> > > > > > p.s. Very unrelated...  Should we have __builtin_bit_cast for C as well?
> > > > > > Is there any reason this could not work?
> > > >
> > > > Still interested in this btw :-)  (And still very unrelated.)
> > > >
> > > >
> > > > Segher
> > >
> > >
> > >
> > > --
> > > BR,
> > > Hongtao
>
>
>
> --
> BR,
> Hongtao



-- 
BR,
Hongtao