- Cvs-nhc98 - Haskell.org

Modified combinators and also slightly the transformation.
by olaf＠sparud.net 29 Mar '01

29 Mar '01

olaf: Wed Mar 28 20:33:56 CEST 2001 Update of /usr/src/master/nhc/include In directory hinken:/tmp/cvs-serv11925/include Modified Files: Prelude.T.hi Log Message: Modified combinators and also slightly the transformation. Moved introduction of Sat's from functions side (fun?, prim?) to application side (ap?, rap?). - Removes previous bug that transformed program was too eager, e.g. in (True || h True) the application of h was evaluated. - Creation of Sat and App? trace nodes now next to each other. Integration of writing a Sat and an App? now straightforward. - Subsequent RHS-optimisation possible. Corrected bug in transformation of let. Expression after in is only in projective context if whole let is in projective context. This bug could lead to missing Hidden node in trusted code and too eager evaluation when the RHS-optimisation is used. Introduced separate combinator trif for if in projective context. The use of original tif could lead to too eager evaluation (never observed) and this would have become worse with the RHS-optimisation. RHS-optimisation: If an application (or if) is in projective context, then it doesn't need a Sat, because the application node only serves as a parent, but is never pointed to from an App? as an argument. So modified rap? and trif to not produce Sats. This halfs the number of SatCs and reduces runtime by ~10%.

1 0

No log message
by colin＠sparud.net 28 Mar '01

28 Mar '01

colin: Wed Mar 28 18:58:07 CEST 2001 Update of /usr/src/master/nhc/src/tracer/hat In directory hinken:/tmp/cvs-serv11797 Modified Files: hat-check.c Log Message: Added -r option to report statistics about reachable nodes.

1 0

Make arg of hat-check consistent with other hat programs.
by malcolm＠sparud.net 28 Mar '01

28 Mar '01

malcolm: Wed Mar 28 17:23:54 CEST 2001 Update of /usr/src/master/nhc/src/tracer/hat In directory hinken:/tmp/cvs-serv11650 Modified Files: hat-check.c Log Message: Make arg of hat-check consistent with other hat programs. The main argument of hat-check should be a program name without a ".hat" extension, just like the other hat programs.

1 0

Oops in Makefile. Need to remove separate tracer targets
by malcolm＠sparud.net 27 Mar '01

27 Mar '01

malcolm: Tue Mar 27 14:57:54 CEST 2001 Update of /usr/src/master/nhc In directory hinken:/tmp/cvs-serv7434 Modified Files: Makefile Log Message: Oops in Makefile. Need to remove separate tracer targets for each build compiler, leaving just one target for all.

1 0

Introduced new combinator conCons to make transformed stri
by olaf＠sparud.net 27 Mar '01

27 Mar '01

olaf: Tue Mar 27 13:22:17 CEST 2001 Update of /usr/src/master/nhc/src/prelude/PreludeDebug In directory hinken:/tmp/cvs-serv7089/src/prelude/PreludeDebug Modified Files: PreludeDebug.hs Log Message: Introduced new combinator conCons to make transformed string literals shorter. The new tracing code for literal strings is a bit smaller. It also shares all uses of the (:) in a string. To improve it further, we would need to be able to use the tranformed list type and the untransformed one together in a combinator.

1 0

Introduced new combinator conCons to make transformed stri
by olaf＠sparud.net 27 Mar '01

27 Mar '01

olaf: Tue Mar 27 13:22:17 CEST 2001 Update of /usr/src/master/nhc/src/compiler98 In directory hinken:/tmp/cvs-serv7089/src/compiler98 Modified Files: DbgId.hs DbgTrans.hs Log Message: Introduced new combinator conCons to make transformed string literals shorter. The new tracing code for literal strings is a bit smaller. It also shares all uses of the (:) in a string. To improve it further, we would need to be able to use the tranformed list type and the untransformed one together in a combinator.

1 0

Introduced new combinator conCons to make transformed stri
by olaf＠sparud.net 27 Mar '01

27 Mar '01

olaf: Tue Mar 27 13:22:16 CEST 2001 Update of /usr/src/master/nhc/include In directory hinken:/tmp/cvs-serv7089/include Modified Files: Prelude.T.hi Log Message: Introduced new combinator conCons to make transformed string literals shorter. The new tracing code for literal strings is a bit smaller. It also shares all uses of the (:) in a string. To improve it further, we would need to be able to use the tranformed list type and the untransformed one together in a combinator.

1 0

Mode for createDirectory should be octal not hexadecimal.
by malcolm＠sparud.net 27 Mar '01

27 Mar '01

malcolm: Tue Mar 27 13:04:18 CEST 2001 Update of /usr/src/master/nhc/src/prelude/Directory In directory hinken:/tmp/cvs-serv7007 Modified Files: CreateDirectory.gc Log Message: Mode for createDirectory should be octal not hexadecimal.

1 0

'make tracer' now ensures the unified compiler is built.
by malcolm＠sparud.net 27 Mar '01

27 Mar '01

malcolm: Tue Mar 27 11:59:36 CEST 2001 Update of /usr/src/master/nhc In directory hinken:/tmp/cvs-serv6693 Modified Files: Makefile Log Message: 'make tracer' now ensures the unified compiler is built.

1 0

Fixed transformation of default methods.
by olaf＠sparud.net 26 Mar '01

26 Mar '01

olaf: Mon Mar 26 18:27:22 CEST 2001 Update of /usr/src/master/nhc/src/prelude/PreludeDebug In directory hinken:/tmp/cvs-serv4381/src/prelude/PreludeDebug Modified Files: PreludeDebug.hs Log Message: Fixed transformation of default methods. A missing default method lead to a corrupted trace file and sometimes even a segmentation fault. The problem was that the call to `fatal' was not wrapped within an R. Furthermore the error message "No match in pattern." was rather misleading. Now transform default method like any other function definition. The error message which has been introduced by the renamer is transformed like a normal expression. Changed renamer so that this compiler introduced expression has no source positions. So the explicit call of error is visible in the trace, but the browser user cannot ask for the source position.

1 0