|
RE: profiling in Cabal in 6.4.1-aug-1: msg#00026lang.haskell.glasgow.bugs
On 04 August 2005 12:54, Serge D. Mechveliani wrote: > I have `made' ghc-6.4.1.20050801 > > by its binary, on Linux Debian, and tested on DoCon and Dumatel > application. > The first `make' crashed (I do not know, maybe, indeed, a hardware > fail). The second (by new) succeeded. > > It looks all right -- except `making' with profiling. > > Without profiling, it `makes' and installs. > For profiling, dm.cabal adds > -prof -auto-all -hisuf p_hi -osuf p_o > to its options. > The commands > runhaskell Setup configure --ghc --prefix=$dmSource/inst > runhaskell Setup build > > seem to work all right. Then, runhaskell Setup install --user > reports > > Installing: /home/mechvel/dm/1.03/dm/source/inst/lib/dumatel-1.3 & > /home/mechvel/dm/1.03/dm/source/inst/bin dumatel-1.3... > Error: Could not find module: Dumatel with any suffix: ["hi"] > > And the directory $dmSource/dist/build/ contains > Dumatel.p_hi, Dumatel.p_o > among its files. Newer versions of Cabal have support for profiling. I suggest you upgrade Cabal independently: there's a downloadable pre-release version available from http://www.haskell.org/cabal/. Cheers, Simon
|
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| Previous by Date: | [ ghc-Bugs-1251699 ] ghc-6.4.1.20050801: panic!, SourceForge.net |
|---|---|
| Next by Date: | RE: "Error: symbol `__stg_split_marker' is already defined", Simon Marlow |
| Previous by Thread: | "Error: symbol `__stg_split_marker' is already defined", Frederik Eaton |
| Next by Thread: | profiling in Cabal in 6.4.1-aug-1, Serge D. Mechveliani |
| Indexes: | [Date] [Thread] [Top] [All Lists] |
| News | FAQ | advertise |