Cannot build benchmarks

[Mark Karpov]

I wanted to run cabal bench on your library, but I'm getting this:

Preprocessing benchmark 'lift-benchmarks' for pipes-4.1.5...
[1 of 2] Compiling Common           ( benchmarks/Common.hs, dist/build/lift-benchmarks/lift-benchmarks-tmp/Common.o )
[2 of 2] Compiling Main             ( benchmarks/LiftBench.hs, dist/build/lift-benchmarks/lift-benchmarks-tmp/Main.o )

benchmarks/LiftBench.hs:17:10:
    Duplicate instance declarations:
      instance NFData a => NFData (Sum a)
        -- Defined at benchmarks/LiftBench.hs:17:10
      instance [safe] NFData a => NFData (Sum a)
        -- Defined in ‘Control.DeepSeq’

Compilation exited abnormally with code 1 at Sun Aug 16 02:05:37

GHC 7.10, Cabal 1.22.

[Gabriella Gonzalez]

It looks like this is due to a newly added NFData instance for Sum in deepseq-1.4. I'll update the benchmark suite to remove the orphan instance and increase the lower bound on deepseq to 1.4