On Sun, Jun 25, 2023 at 7:39 AM Roger Sayle <ro...@nextmovesoftware.com> wrote:
>
>
> On Tue, 13 June 2023 12:02, Richard Biener wrote:
> > On Mon, Jun 12, 2023 at 4:04 PM Roger Sayle <ro...@nextmovesoftware.com>
> > wrote:
> > > The following simple test case, from PR 104610, shows that memcmp ()
> > > == 0 can result in some bizarre code sequences on x86.
> > >
> > > int foo(char *a)
> > > {
> > > static const char t[] = "0123456789012345678901234567890";
> > > return __builtin_memcmp(a, &t[0], sizeof(t)) == 0; }
> > >
> > > with -O2 currently contains both:
> > > xorl %eax, %eax
> > > xorl $1, %eax
> > > and also
> > > movl $1, %eax
> > > xorl $1, %eax
> > >
> > > Changing the return type of foo to _Bool results in the equally
> > > bizarre:
> > > xorl %eax, %eax
> > > testl %eax, %eax
> > > sete %al
> > > and also
> > > movl $1, %eax
> > > testl %eax, %eax
> > > sete %al
> > >
> > > All these sequences set the result to a constant, but this
> > > optimization opportunity only occurs very late during compilation, by
> > > basic block duplication in the 322r.bbro pass, too late for CSE or
> > > peephole2 to do anything about it. The problem is that the idiom
> > > expanded by compare_by_pieces for __builtin_memcmp_eq contains basic
> > > blocks that can't easily be optimized by if-conversion due to the
> > > multiple incoming edges on the fail block.
> > >
> > > In summary, compare_by_pieces generates code that looks like:
> > >
> > > if (x[0] != y[0]) goto fail_label;
> > > if (x[1] != y[1]) goto fail_label;
> > > ...
> > > if (x[n] != y[n]) goto fail_label;
> > > result = 1;
> > > goto end_label;
> > > fail_label:
> > > result = 0;
> > > end_label:
> > >
> > > In theory, the RTL if-conversion pass could be enhanced to tackle
> > > arbitrarily complex if-then-else graphs, but the solution proposed
> > > here is to allow suitable targets to perform if-conversion during
> > > compare_by_pieces. The x86, for example, can take advantage that all
> > > of the above comparisons set and test the zero flag (ZF), which can
> > > then be used in combination with sete. Hence compare_by_pieces could
> > > instead generate:
> > >
> > > if (x[0] != y[0]) goto fail_label;
> > > if (x[1] != y[1]) goto fail_label;
> > > ...
> > > if (x[n] != y[n]) goto fail_label;
> > > fail_label:
> > > sete result
> > >
> > > which requires one less basic block, and the redundant conditional
> > > branch to a label immediately after is cleaned up by GCC's existing
> > > RTL optimizations.
> > >
> > > For the test case above, where -O2 -msse4 previously generated:
> > >
> > > foo: movdqu (%rdi), %xmm0
> > > pxor .LC0(%rip), %xmm0
> > > ptest %xmm0, %xmm0
> > > je .L5
> > > .L2: movl $1, %eax
> > > xorl $1, %eax
> > > ret
> > > .L5: movdqu 16(%rdi), %xmm0
> > > pxor .LC1(%rip), %xmm0
> > > ptest %xmm0, %xmm0
> > > jne .L2
> > > xorl %eax, %eax
> > > xorl $1, %eax
> > > ret
> > >
> > > we now generate:
> > >
> > > foo: movdqu (%rdi), %xmm0
> > > pxor .LC0(%rip), %xmm0
> > > ptest %xmm0, %xmm0
> > > jne .L2
> > > movdqu 16(%rdi), %xmm0
> > > pxor .LC1(%rip), %xmm0
> > > ptest %xmm0, %xmm0
> > > .L2: sete %al
> > > movzbl %al, %eax
> > > ret
> > >
> > > Using a target hook allows the large amount of intelligence already in
> > > compare_by_pieces to be re-used by the i386 backend, but this can also
> > > help other backends with condition flags where the equality result can
> > > be materialized.
> > >
> > > This patch has been tested on x86_64-pc-linux-gnu with make bootstrap
> > > and make -k check, both with and without --target_board=unix{-m32}
> > > with no new failures. Ok for mainline?
> >
> > What's the guarantee that the zero flag is appropriately set on all edges
> > incoming
> > now and forever?
>
> Is there any reason why this target hook can't be removed (in future) should
> it stop
> being useful? It's completely optional and not required for the correct
> functioning
> of the compiler.
>
> > Does this require target specific knowledge on how do_compare_rtx_and_jump
> > is emitting RTL?
>
> Yes. Each backend can decide how best to implement finish_compare_by_pieces
> given its internal knowledge of how do_compare_rtx_and_jump works. It's not
> important to the middle-end how the underlying invariants are guaranteed, just
> that they are and the backend produces correct code. A backend may store
> flags
> on the target label, or maintain state in cfun. Future changes to the i386
> backend
> might cause it to revert to the default finish_compare_by_pieces, or provide
> an
> alternate implementation, but at the moment this patch improves the code that
> GCC generates. Very little (in software like GCC) is forever.
>
> > Do you see matching this in ifcvt to be unreasonable? I'm thinking of
> > "reducing"
> > the incoming edges pairwise without actually looking at the ifcvt code.
>
> There's nothing about the proposed patch that prevents or blocks improvements
> to ifcvt, which I agree completely could be improved. But even (in future)
> if later
> RTL passes could clean things up, that's no reason for RTL expansion to
> initially
> generate poor/inefficient code. I'm not sure that a (hypothetical) ifcvt
> improvement
> would be sufficient reason to revert/remove enhancements to compare_by_pieces.
>
> Is it that there's not enough (bootstrap and) testsuite coverage of
> compare_by_pieces
> to make you feel comfortable with this change? The proposed implementation
> should
> be "obvious enough" to future generations what the intended behaviour should
> be.
> And the x86 target hook implementation (i.e. the change) is only four lines
> long, a
> fraction of the size of the new documentation and comments.
My main concern was that we are communicating "implicit" dependences between
the target hook and expand RTL code generation - we don't seem to pass down
enough info for example to have the target verify constraints and
excuse itself if
they do not hold. They also seem to be poorly documented and the
compare_by_pieces (and all _by_pieces) stuff has been extended to be quite
"configurable" and thus is probably prone to emit vastly different RTL in
some special cases?
That said, I'm not familiar enough with the compare_by_pices logic to say that
this looks obviously safe, but maybe it is.
Yes - all target hooks have this kind of "implicit" dependence on how
the calling
code works, so I guess the main thing is that it's not obvious how the calling
code works in this case?
Anyway, I'm not standing in the way of somebody approving this change - the code
generation improvements are definitely worth it.
Maybe it's the abstraction boundary that makes me worry - can we make the
target simply indicate whether a do_compare_rtx_and_jump sets a flag
(maybe query the CC mode used?) and have the generic code then emit
the setcc? The calling code would then know it emitted suitable CC producing
compare and branch and it would emit the corresponding setcc with the proper
CCmode?
Richard.
> Thanks in advance.
> Roger
>
>
> > > 2023-06-12 Roger Sayle <ro...@nextmovesoftware.com>
> > >
> > > gcc/ChangeLog
> > > * config/i386/i386.cc (ix86_finish_compare_by_pieces): New
> > > function to provide a backend specific implementation.
> > > (TARGET_FINISH_COMPARE_BY_PIECES): Use the above function.
> > >
> > > * doc/tm.texi.in (TARGET_FINISH_COMPARE_BY_PIECES): New @hook.
> > > * doc/tm.texi: Regenerate.
> > >
> > > * expr.cc (compare_by_pieces): Call finish_compare_by_pieces in
> > > targetm to finalize the RTL expansion. Move the current
> > > implementation to a default target hook.
> > > * target.def (finish_compare_by_pieces): New target hook to allow
> > > compare_by_pieces to be customized by the target.
> > > * targhooks.cc (default_finish_compare_by_pieces): Default
> > > implementation moved here from expr.cc's compare_by_pieces.
> > > * targhooks.h (default_finish_compare_by_pieces): Prototype.
> > >
> > > gcc/testsuite/ChangeLog
> > > * gcc.target/i386/pieces-memcmp-1.c: New test case.
> > >
>
>
