[llvm-dev] funnel shift, select, and poison

[Nuno Lopes via llvm-dev]

You are very right! Transformation to rotate is correct.

So I guess the remaining case is if you want to be able to transform funnel 
shifts into other arithmetic operations when %x != %y. I think I saw some 
optimizations where fshl was being transformed into shl. This wouldn't be OK 
because shl doesn't stop poison. Unless these are only being done for 
guaranteed non-zero %sh? Then it's ok because fshl can't possibly block 
poison in that case.

Nuno


-----Original Message-----
From: Sanjay Patel
Sent: Monday, February 25, 2019 10:30 PM
Subject: Re: [llvm-dev] funnel shift, select, and poison


Don't we need to distinguish funnel shift from the more specific rotate?
I'm not seeing how rotate (a single input op shifted by some amount) gets 
into trouble like funnel shift (two variables concatenated and shifted by 
some amount).
Eg, if in pseudo IR we have:
%funnel_shift = fshl %x, %y, %sh ; this is problematic because either x or y 
can be poison, but we may not touch the poison when sh==0
%rotate = fshl %x, %x, %sh ; if x is poison, the op is unquestionably 
producing poison; there's no sh==0 loophole here



On Mon, Feb 25, 2019 at 1:12 PM Nuno Lopes <nunoplopes at sapo.pt> wrote:
Thanks Sanjay!

I did a quick study of these funnel shifts:
The generic lowering to SDAG is correct for the optimization below. It
actually stops poison if shift amount is zero:
    SDValue ShAmt = DAG.getNode(ISD::UREM, sdl, VT, Z, BitWidthC);
(...)
    SDValue IsZeroShift = DAG.getSetCC(sdl, CCVT, ShAmt, Zero, ISD::SETEQ);
    setValue(&I, DAG.getSelect(sdl, VT, IsZeroShift, IsFSHL ? X : Y, Or));

This is assuming select in SDAG stops poison in the same way we've proposed
for the IR.

However, the lowering has 2 optimizations. It can lower funnel shifts to:
1) A special funnel shift instruction if the backend supports it
2) Rotate

At least lowering to rotate would be incorrect if rotate didn't stop poison
as well when shift amount == 0. Most likely rotate doesn't block poison
though. So this doesn't seem correct.

Blocking poison, while tempting, is usually not a good idea because it
blocks many optimizations. It becomes hard to remove instructions that block
poison. Exactly as you see here: if select blocks poison (and we claim it
does), then it cannot be removed.

I have 2 separate proposals:
1) Keep select blocking poison, and remove the transformation below because
it doesn't play well with 1) lowering to rotate, and 2) because it blocks
optimizations like converting funnel shifts to plain shifts
2) Introduce a flag in select, like we have nsw/nuw today that changes the
semantics regarding poison. Essentially a select that doesn't stop poison.
This can be safely emitted by most frontends of the languages we support
today, but wouldn't be used by compiler-introduced selects. The optimization
below would only kick in when this flag is present. Of course then we can
have an analysis that inserts these flags like we have for nsw.

I like 2), but 1) is simpler. I don't know if 2) is worth it, but is
appealing :)

Nuno


-----Original Message-----
From: Sanjay Patel via llvm-dev
Sent: Monday, February 25, 2019 4:29 PM
Subject: [llvm-dev] funnel shift, select, and poison


There's a question about the behavior of funnel shift [1] + select and
poison here that reminds me of previous discussions about select and poison
[2]:
https://github.com/AliveToolkit/alive2/pull/32#discussion_r257528880

Example:
define i8 @fshl_zero_shift_guard(i8 %x, i8 %y, i8 %sh) {
%c = icmp eq i8 %sh, 0
%f = fshl i8 %x, i8 %y, i8 %sh
%s = select i1 %c, i8 %x, i8 %f ; shift amount is 0 returns x (same as fshl)
ret i8 %s
}
=>
define i8 @fshl_zero_shift_guard(i8 %x, i8 %y, i8 %sh) {
%f = fshl i8 %x, i8 %y, i8 %sh
ret i8 %f
}
Transformation doesn't verify!
ERROR: Target is more poisonous than source

----------------------------------------------------------------------------

The problem is that if %y is poison and we assume that funnel shift uses all
of its operands unconditionally, the reduced code sees poison while the
original code is protected by the "conditional poison" (terminology?)
property of a select op and is safe.

If we treat funnel shift more like a select based on its operation (when the
shift amount is 0, we know that the output is exactly 1 of the inputs), then
the transform should be allowed?

This transform was implemented in instcombine [3] with the motivation of
reducing UB-safe rotate code in C to the LLVM intrinsic [4]. So a potential
sidestep of the problem would be to limit that transform to a rotate pattern
(single value is being shifted) rather than the more general funnel pattern
(two values are being shifted).

[1] https://llvm.org/docs/LangRef.html#llvm-fshl-intrinsic
[2] http://llvm.1065342.n5.nabble.com/poison-and-select-td72262.html
[3] https://reviews.llvm.org/D54552
[4] https://bugs.llvm.org/show_bug.cgi?id=34924



_______________________________________________
LLVM Developers mailing list
llvm-dev at lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev