- Glasgow-haskell-users

Closure elimination transformation (continuation style passing code)
by Tyson Whitehead 20 May '09

20 May '09

I was wondering about GHC's ability to optimize the following code module Test (test) where newtype Step a b = Step { unStep :: (a -> Step a b -> b) -> b -> b } iter :: [a] -> (a -> Step a b -> b) -> b -> b iter [] next done = done iter (x:xs) next done = next x (Step $ iter xs) count :: Int -> Char -> Step Char Int -> Int count i x step = unStep step (count $ i+1) (i+1) test :: String -> Int test xs = iter xs (count 0) 0 (test implements the string length function). The transformations steps that reduce this to the optimized version are 1- avoid forming the (iter xs) and (count i+1) closures by passing the function and the arguments instead of the function bound to the argument iter [] next i done = done iter (x:xs) next i done = next i x iter xs count i x step xs = step xs count (i+1) (i+1) test xs = iter xs count 0 0 2- specialize count for step = iter iter [] next i done = done iter (x:xs) next i done = next i x iter xs count i x xs = iter xs count (i+1) (i+1) test xs = iter xs count 0 0 3- specializing iter for next = count iter [] i done = done iter (x:xs) i done = count i x iter xs count i x xs = iter xs (i+1) (i+1) test xs = iter xs 0 0 4- inline count iter [] i done = done iter (x:xs) i done = iter xs (i+1) (i+1) test xs = iter xs 0 0 5- eliminate the done argument as it is always the same as the i argument iter [] i = i iter (x:xs) i = iter xs (i+1) test xs = iter xs 0 Currently 6.10.1 with -O2 seems stuck with regard to step 1 (eliminating the closures). Is there any hope of getting it to these transformations? Thanks! -Tyson

4 7

baffling manual sections
by Jason Dusek 20 May '09

20 May '09

I'm a little puzzled by these seeming duplicate pages in the manual: http://www.haskell.org/ghc/dist/current/docs/users_guide/type-extensions.ht… http://www.haskell.org/ghc/dist/current/docs/users_guide/data-type-extensio… They give slightly different section numbers to similar stuff. The former page also has a curious discussion of standalone deriving with a `for` keyword: http://www.haskell.org/ghc/dist/current/docs/users_guide/type-extensions.ht… -- Jason Dusek

4 7

6.8.1 on ubuntu jaunty
by Martin Matusiak 20 May '09

20 May '09

Hello, I need ghc 6.8.1 for a codebase that depends on it and I'm having trouble building it on the current ubuntu. Error output attached. Martin configure: creating ./config.status config.status: creating network.buildinfo config.status: creating include/HsNetworkConfig.h Configuring network-2.1.0.0... rm -f network/GNUmakefile cp Makefile.local network if ifBuildable/ifBuildable network; then \ cd network && setup/Setup makefile -f GNUmakefile; \ fi Socket.hsc: In function ‘main’: Socket.hsc:1143: error: invalid application of ‘sizeof’ to incomplete type ‘struct ucred’ Socket.hsc:1143: error: invalid application of ‘sizeof’ to incomplete type ‘struct ucred’ Socket.hsc:1143: error: invalid application of ‘sizeof’ to incomplete type ‘struct ucred’ Socket.hsc:1149: error: invalid use of undefined type ‘struct ucred’ Socket.hsc:1150: error: invalid use of undefined type ‘struct ucred’ Socket.hsc:1151: error: invalid use of undefined type ‘struct ucred’ compiling dist/build/Network/Socket_hsc_make.c failed command was: gcc -c -D__GLASGOW_HASKELL__=608 -I/tmp/ghc-6.8.1/includes -I/tmp/ghc-6.8.1/gmp/gmpbuild -D__GLASGOW_HASKELL__=608 -DCALLCONV=ccall -Iinclude dist/build/Network/Socket_hsc_make.c -o dist/build/Network/Socket_hsc_make.o Preprocessing library network-2.1.0.0... make[1]: *** [network/GNUmakefile] Error 1 make[1]: Leaving directory `/tmp/ghc-6.8.1/libraries' make: *** [stage1] Error 2

1 0

Re: strictness of interpreted haskell implementations
by Geraint Jones 18 May '09

18 May '09

Sorry to revive a year-old thread, but... On Fri, 25 Apr 2008 at 20:17:53 +0100 Duncan Coutts wrote: > On Fri, 2008-04-25 at 09:08 -0700, Don Stewart wrote: > > Geraint.Jones: > > > Are there well-known differences in the implementations of Haskell in > > > ghci and hugs? I've got some moderately intricate code (simulations > > > of pipelined processors) that behave differently - apparently because > > > ghci Haskell is stricter than hugs Haskell, and I cannot find any > > > obviously relevant claims about strictness in the documentation. > > I think they should give the same answer. It sounds like a bug in one > implementation or the other. > > > Hugs does no optimisations, while GHC does a truckload, including > > strictness analysis. Some of these optimisations prevent space leaks. > > Though none should change the static semantics. > > Post the code. Even if you don't have time to track down the difference, > someone might. At the time I was reluctant to impose all the code on anyone and I found it hard to cut the example down to a manageable size. I've just got it down to a one-liner: it's the implementation of what I think ought to be strict fields in records: data S = S { a :: Int, b :: ! Int } I think ghci is correct: *Main> a (S { a = 0, b = 1 }) 0 *Main> a (S { a = 0, b = undefined }) *** Exception: Prelude.undefined and that hugs had been concealing a bug in my program by not demanding one of the fields of S when it ought to: Main> a (S { a = 0, b = 1 }) 0 Main> a (S { a = 0, b = undefined }) 0 Ho hum. Is this a "known difference"? (What makes you think I'm teaching the same course again this year?)

4 3

possible alternative to libFFI
by Bulat Ziganshin 18 May '09

18 May '09

Hello glasgow-haskell-users, http://www.nongnu.org/cinvoke/faq.html -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

3 3

defaultsHook isn't documented
by Bulat Ziganshin 18 May '09

18 May '09

Hello glasgow-haskell-users, it seems that defaultsHook isn't documented on http://www.haskell.org/ghc/docs/latest/html/users_guide/runtime-control.htm… neither anywhere else in user manual -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

2 2

Re: [Haskell-cafe] ghc ./configure stalls on docbook DTD
by Donnie Jones 18 May '09

18 May '09

Hello Dan, Best place to ask is glasgow-haskell-users(a)haskell.org since that is the GHC users list. I have CC'd your email to the GHC user list. Cheers. -- Donnie Jones On Wed, May 13, 2009 at 1:35 PM, Dan <danielkcook(a)gmail.com> wrote: > Hi, > > Not sure if this is the right place to ask. > GHC 6.10.3 source dist: ./configure takes about 10 minutes to look for > DocBook DTD and another 10 to look for DocBook XSL directory. I was writing > this e-mail thinking it had completely crashed. Any reason why it's so > ridiculously slow? A look at 'top' says the CPU is mostly idle during this > time, but xmllint is running. > > Mac OS X > > uname -a: Darwin 9.6.0 Darwin Kernel Version 9.6.0: Mon Nov 24 17:37:00 PST > 2008; root:xnu-1228.9.59~1/RELEASE_I386 i386 > > CookBook:ghc-6.10.3 dan$ ./configure > checking build system type... i386-apple-darwin9.6.0 > checking host system type... i386-apple-darwin9.6.0 > checking target system type... i386-apple-darwin9.6.0 > Canonicalised to: i386-apple-darwin > [blah blah blah] > checking for xmllint... /usr/bin/xmllint > checking for DocBook DTD... > (blocks for ages) > > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > >

2 1

comprehension vs `map'
by Serge D. Mechveliani 15 May '09

15 May '09

Dear GHC team, I would like to write [rl {ruleMode = AlwaysApply} | rl <- rules calc] (I) instead of map (\ rl -> rl {ruleMode = AlwaysApply}) $ rules calc (II) and instead of let rs = rules calc in [rl {ruleMode = AlwaysApply} | rl <- rs] (III). But is this reliable in GHC that the compiler converts (I) into something which is not not worse than III ? What about other implementations? Regards, ----------- Mechveliani mechvel(a)botik.ru

2 1

ANN: GhciExt 0.6 for GHC 6.10.3
by Peter Hercek 14 May '09

14 May '09

Hi, I know there were 4 unique IP addresses which checked it out. Since I do not know who they are I just spam this list again :) You can get it here: http://www.hck.sk/users/peter/pub/ If you decide to give it a try then read the README file before installing. It should work with the stock GHC 6.10.2 and higher, but I only tried briefly with GHC 6.10.3 (as it was released) and I use it without any problem with my customized GHC 6.10.3 (dirty support for ansi escape sequences in ':set prompt' for color highlighting of the prompt, patch for ticket http://hackage.haskell.org/trac/ghc/ticket/3084 (allow macros to redefine builtin GHCi commands), and few more details). Here is what it does: Prelude> :defs long :. <file> -- source commands from <file> :* <count> <cmd>... -- run <cmd> <count> times :x <cmd>... -- run <cmd> with stdout suppressed :out -- redirect ghci stdout back to console :redir <var> <cmd>... -- execute <cmd> redirecting stdout to <var> :grep <args> < <cmd>... -- run grep <args> on <cmd> output Runs grep from your OS on the output of <cmd>. This means all the options of your OS grep are available. '<' separates grep options from the command. :find <cmd> [-]<var> -- step with <cmd> until <var> is found Prepend variable name with '-' to disable printing of its value. :locate <cmd> <bpArgs> -- step with <cmd> until location <bpArgs> is hit :bp <bpArgs> -- put breakpoint at <bpArgs> (adds hit count) :inject <cc> <c> <sc> -- at location run <c> if <cc> and stop if <sc> There are two special identifiers which can be used in the breakpoint code (<c>) and the breakpoint stop condition (<sc>); but which are not usable in the breakpoint code condition (<cc>). These are: - ghciExt_BpId - macro name; translates to breakpoint id - ghciExt_HitCnt - macro name; translates to breakpoint hit count :strobe ["<c>"] -- at location show hit count if <c> :monitor ["<c>"] <vs> -- show comma-sep. variables at location if <c> :watch <cond> <bpArgs> -- break at location <bpArgs> when <cond> is True :count <N> [<bpArgs>] -- count/stop execution at <bpArgs> or query bp hits There are three ways how to use this command: - :count 0 <bpArgs> - never stops; counts hits and extends trace history - :count N <bpArgs> - stops when location <bpArgs> is hit N-th time - :count N - shows hit count of a ghciExt breakpoint with number N. :hl <cmd>... -- highlight the output of <cmd> :defs [long] -- list user commands (possibly with long help) General information for all GhciExt commands: - Most arguments may be in quotation marks (to allow spaces in them). - Instead of a regular quotation mark, @x can be used to start a string argument too. In such a case the argument is finished when character x appears again after `@x' string. Use any character in place of x. This was added so that one does not need to escape quotation marks and back-slashes in arguments (especially those representing code). Prelude> Not so much obsolete description of how it works (and how to use it) is here: http://permalink.gmane.org/gmane.comp.lang.haskell.glasgow.user/16270 Notable changes are: * much better :grep command (you can even get color highlighted output) * you do not need to prefix commands :main, :continute, :step, and :trace with :x any more * :findex replaced with much better :locate It works on linux. The only reason I recall it would not work on windows is the use of '/dev/null'. But I never tried on windows. Enjoy, Peter.

1 0

[web] distro page update for Fedora
by Jens Petersen 13 May '09

13 May '09

Hi, I can't remember if I actually reported this before - I certainly meant too... :) Could you please update the Fedora entry on the distribution packages page: http://haskell.org/ghc/distribution_packages.html Please replace `RPMs are available from Fedora Extras' with `<a href="http://fedoraproject.org">Fedora users can install with "yum install ghc"'. "Fedora Extras" also appears under the openSUSE entry! Could you please remove that. Thank you, Jens on behalf of the Fedora Haskell SIG

3 4