GHC fails to terminate while compiling with optimization level 2

When I try to compile random-fu-0.2.4.0 from Hackage, using the command:

cabal install --ghc --ghc-options="-fllvm -O2 -dcore-lint" --enable-optimization=2 random-fu-0.2.4.0

I get the following output:

Resolving dependencies...
Configuring random-fu-0.2.4.0...
Building random-fu-0.2.4.0...
Preprocessing library random-fu-0.2.4.0...
[ 1 of 27] Compiling Data.Random.Internal.Find ( src/Data/Random/Internal/Find.hs, dist/build/Data/Random/Internal/Find.o )
[ 2 of 27] Compiling Data.Random.Internal.Fixed ( src/Data/Random/Internal/Fixed.hs, dist/build/Data/Random/Internal/Fixed.o )
[ 3 of 27] Compiling Data.Random.Internal.TH ( src/Data/Random/Internal/TH.hs, dist/build/Data/Random/Internal/TH.o )
[ 4 of 27] Compiling Data.Random.Lift ( src/Data/Random/Lift.hs, dist/build/Data/Random/Lift.o )
[ 5 of 27] Compiling Data.Random.RVar ( src/Data/Random/RVar.hs, dist/build/Data/Random/RVar.o )
[ 6 of 27] Compiling Data.Random.Distribution ( src/Data/Random/Distribution.hs, dist/build/Data/Random/Distribution.o )
[ 7 of 27] Compiling Data.Random.Distribution.Uniform ( src/Data/Random/Distribution/Uniform.hs, dist/build/Data/Random/Distribution/Uniform.o )
Loading package ghc-prim ... linking ... done.
Loading package integer-gmp ... linking ... done.
Loading package base ... linking ... done.
Loading package transformers-0.3.0.0 ... linking ... done.
Loading package mtl-2.1.2 ... linking ... done.
Loading package MonadPrompt-1.0.0.3 ... linking ... done.
Loading package array-0.4.0.1 ... linking ... done.
Loading package deepseq-1.3.0.1 ... linking ... done.
Loading package containers-0.5.0.0 ... linking ... done.
Loading package pretty-1.1.1.0 ... linking ... done.
Loading package template-haskell ... linking ... done.
Loading package syb-0.4.0 ... linking ... done.
Loading package th-extras-0.0.0.2 ... linking ... done.
Loading package flexible-defaults-0.0.1.1 ... linking ... done.
Loading package old-locale-1.0.0.5 ... linking ... done.
Loading package old-time-1.1.0.1 ... linking ... done.
Loading package time-1.4.0.1 ... linking ... done.
Loading package random-1.0.1.1 ... linking ... done.
Loading package mersenne-random-pure64-0.2.0.3 ... linking ... done.
Loading package primitive-0.5.0.1 ... linking ... done.
Loading package vector-0.10.0.1 ... linking ... done.
Loading package mwc-random-0.12.0.1 ... linking ... done.
Loading package stm-2.4.2 ... linking ... done.
Loading package stateref-0.3 ... linking ... done.
Loading package random-source-0.3.0.4 ... linking ... done.
Loading package rvar-0.2.0.1 ... linking ... done.
Loading package MonadRandom-0.1.9 ... linking ... done.
Loading package random-shuffle-0.0.4 ... linking ... done.
Loading package monad-loops-0.4.2 ... linking ... done.
Loading package continued-fractions-0.9.1.1 ... linking ... done.
Loading package converge-0.1.0.1 ... linking ... done.
Loading package gamma-0.9.0.2 ... linking ... done.
Loading package erf-2.0.0.0 ... linking ... done.
[ 8 of 27] Compiling Data.Random.List ( src/Data/Random/List.hs, dist/build/Data/Random/List.o )
[ 9 of 27] Compiling Data.Random.Distribution.Bernoulli ( src/Data/Random/Distribution/Bernoulli.hs, dist/build/Data/Random/Distribution/Bernoulli.o )
[10 of 27] Compiling Data.Random.Distribution.Categorical ( src/Data/Random/Distribution/Categorical.hs, dist/build/Data/Random/Distribution/Categorical.o )
*** Core Lint warnings : in result of Desugar (after optimization) ***
<no location info>: Warning:
    [RHS of $c>>_aMYc :: forall p_aMMn.
                         GHC.Num.Num p_aMMn =>
                         forall a_a3Jr b_a3Js.
                         Data.Random.Distribution.Categorical.Categorical p_aMMn a_a3Jr
                         -> Data.Random.Distribution.Categorical.Categorical p_aMMn b_a3Js
                         -> Data.Random.Distribution.Categorical.Categorical p_aMMn b_a3Js]
    INLINE binder is (non-rule) loop breaker: $c>>_aMYc

It then freezes for an as-far-as-I-known indefinite amount of time; I have tried it several times, and it never moves beyond that. Once, I left it for two whole hours without any change. Every time, I have had to manually interrupt it with ctrl-c.

On the other hand, if I instead use the command:

cabal install --ghc --ghc-options="-fllvm -O1 -dcore-lint" --enable-optimization=1 random-fu-0.2.4.0

It produces literally identitical output up to the point that the first command stopped at, but it continues as normal afterwards and finishes installing random-fu as expected.

My terminal confirms that the ghc executable (as opposed to, say, opt or llc) is the one that the first command gets stuck on.

Trac metadata

Trac field	Value
Version	7.6.3
Type	Bug
TypeOfFailure	OtherFailure
Priority	normal
Resolution	Unresolved
Component	Compiler
Test case
Differential revisions
BlockedBy
Related
Blocking
CC
Operating system
Architecture