similar to: Unsafe floating point operation (FDiv & FRem) in LoopVectorizer

I came across this comment in SelectionDAG.cpp: case ISD::FADD: case ISD::FSUB: case ISD::FMUL: case ISD::FDIV: case ISD::FREM: // If both operands are undef, the result is undef. If 1 operand is undef, // the result is NaN. This should match the behavior of the IR optimizer. That isn't intuitive to me. I would have expected a binary FP operation with one undef operand to

%y = fadd float %x, undef Can we simplify this? Currently in IR, we do nothing for fadd/fsub/fmul. For fdiv/frem, we propagate undef. The code comment for fdiv/frem says: "the undef could be a snan" If that's correct, then shouldn't it be the same for fadd/fsub/fmul? But this can't be correct because we support targets that don't raise exceptions...and even targets

Here are some clarifications for the reference manual. Please verify that my assumptions are correct. Shall I post a patch? Floating-point Constants: Add "The assembler requires the exact decimal value of a floating-point constant. For example, the assembler accepts '1.25' but rejects '1.3' because '1.3' is a repeating decimal in binary." Binary

Yes, if %x is a NaN, we should expect that NaN is propagated. I'm still not sure what to do here. We can take comfort in knowing that whatever we do is likely an improvement over the current situation though. :) That's because the code in InstSimplify is inconsistent with the LangRef: http://llvm.org/docs/LangRef.html#undefined-values (UB for fdiv by 0?) ...and both of those are

I’m not sure the transformation happening with fdiv is correct. If we have “%y = fdiv float %x, undef” and %x is a NaN then the result will be NaN for any value of the undef, right? So if I understand the undef rules correctly (never a certainty) then we can’t safely replace the expression with undef. We could, I think, replace it with “%y = %x” though. I think the same is true for fadd, fsub,

Correct - NaN is not undef in IR. But we don't have a NaN in this example. We have its moral equivalent in LLVM - an uninitialized value, undef. So we're not introducing any extra uncertainty by propagating the undef. The backend can choose whatever encoding of undef makes sense when lowering? And yes, I don't know why FP-div-by-zero would ever be UB. I think that text in the LangRef

Why is NaN “just ‘undef’ in IR”? NaN is a specific value with well-defined behavior. I would think that unless the no-NaNs flag is used we need to preserve the behavior of NaNs. From: Sanjay Patel [mailto:spatel at rotateright.com] Sent: Wednesday, February 28, 2018 12:08 PM To: Kaylor, Andrew <andrew.kaylor at intel.com> Cc: llvm-dev <llvm-dev at lists.llvm.org>; Nuno Lopes

Ah, thanks for explaining. So given that any of these ops will return NaN with a NaN operand, let's choose the undef operand value to be NaN. That means we can fold all of these to a NaN constant in the general case. But if we have 'nnan' FMF, then we can fold harder to undef? nnan - Allow optimizations to assume the arguments and result are not NaN. Such optimizations are required to

For the first part of Sanjay’s question, I think the answer is, “Yes, we can fold all of these to NaN in the general case.” For the second part, which the nnan FMF is present, I’m not sure. The particulars of the semantics of nnan are unclear to me. But let me explore what Eli is saying. It sounds reasonable, but I have a question about it. Suppose we have the nnan FMF set, and we encounter

What I’m saying is that if we have one operand that is not an undef value then that operand might be NaN and if it is then the result must be NaN. So while it may be true that we don’t have a NaN, it is not true that we definitely do not have a NaN in the example. This is analogous to the example in the language reference where it says “%A = or %X, undef” -> “%A = undef” is unsafe because any

I'm pretty sure that isn't what nnan is supposed to mean. If the result of nnan math were undefined in the sense of "undef", programs using nnan could have undefined behavior if the result is used in certain ways which would not be undefined for any actual float value (e.g. converting the result to a string), which seems like a surprising result.  And I don't think we

We can do "add %x, undef" => "undef" because for any value of %x, we can always find a value that when added to %x produces any value in the domain of integers. This is not the case with floats since with some inputs, e.g., NaNs, we are not able to produce some values in the domain (e.g., there's no value of %x that makes "fadd NaN, %x" return 42.0). In

So you don’t think sNaNs can just be treated as if they were qNaNs? I understand why we would want to ignore the signaling part of things, but the rules for operating on NaNs are pretty clear and reasonable to implement. The signaling aspect can, I think, be safely ignored when we are in the mode of assuming the default FP environment. As for the distinction between IEEE and LLVM IR, I would

On Mar 1, 2018, at 10:07 AM, Kaylor, Andrew <andrew.kaylor at intel.com> wrote: > So you don’t think sNaNs can just be treated as if they were qNaNs? I understand why we would want to ignore the signaling part of things, but the rules for operating on NaNs are pretty clear and reasonable to implement. The signaling aspect can, I think, be safely ignored when we are in the mode of assuming

Hi all, I came across an issue caused by the Call-Graph Simplify Pass. Here is a a small repro: ``` define double @foo(double %a1, double %a2, double %a3) #0 { entry: %a_mul = fmul double %a1, %a2 %a_cmp = fcmp ogt double %a3, %a_mul br i1 %a_cmp, label %a.then, label %a.end a.then: %a_div = fdiv double %a_mul, %a3 br label %a.end a.end: %a_factor = phi double [ %a_div, %a.then ],

[+current llvm-dev address] On 03/15/2017 09:23 AM, Hal Finkel wrote: > Hi Samuel, > > What are you taking about? ;) > > The only way to get a SIGFPE from a floating-point division by zero is > to modify the floating-point environment to enable those exceptions. > We don't support that (*). In the default (supported) environment, > floating point division is well

> > Do you have plans to add other intrinsics? I'm curious as to why there > > is an llvm.sin intrinsic and an llvm.cos intrinsic, but no llvm.atan > > intrinsic. Why is there an llvm.pow intrinsic but no llvm.log > > intrinsic? > > Intrinsics get added on demand. Generally there has to be a good reason > to add them. llvm.sin was implemented (for

Hello, I have a question about the constant folding for fdiv instructions. For the instruction "fdiv double 0.0, 0.0", the folded result is inf. I think this should be nan. Can anyone tell me why it is not nan? Thanks. Leo _________________________________________________________________ Exercise your brain! Try Flexicon.

Hey guys, I am wondering what the semantics for fdiv is, if the denominator is zero. For sdiv, the language reference specifies that this is undefined behavior. For fdiv, the language reference says nothing (hence, I assume that it is defined somehow). So how is it defined? Does it follow IEEE-754 definition, i.e., +Inf if nominator is > 0, -Inf if nominator is < 0, otherwise NaN?

On 08.08.2013, at 18:25, Chad Rosier <chad.rosier at gmail.com> wrote: > I would like to transform X/Y -> X*1/Y. Specifically, I would like to convert: > > define void @t1a(double %a, double %b, double %d) { > entry: > %div = fdiv fast double %a, %d > %div1 = fdiv fast double %b, %d > %call = tail call i32 @foo(double %div, double %div1) > ret void >