- Gpc - gnu.de

Usage of commandline path-options ?
by Samoth 02 Aug '03

02 Aug '03

Hi ! I'm not very familar with unix-based systems and gcc, but I need to change from borland pascal to gnu-pascal on Windows (and later on other os', too). But as soon as I try to configure the system I got the first problems and the docs aren't very helpful at all. My problem is that most of my old projects need units from several paths on different drives, so I need to add more than one searchpath, but I don't know what is meant by dir[:dir...]. (The ':' follows after the drive-char in windows, so I wanted to now the usage of this seperator.) Perhabs someone could do an example with two unit-paths ( 'a:\Units1' and 'b:\Units2' ) ? Another problem is to use the cfg-file with gpcc.exe The lines with the paths seemed to be ignored, because the output of the arguments passed to gpc.exe does not include them. My config file looks like -U a:\unit1\;b:\Unit1\ -E g:\exe\ Many thanks for answering ... Thomas -- COMPUTERBILD 15/03: Premium-e-mail-Dienste im Test -------------------------------------------------- 1. GMX TopMail - Platz 1 und Testsieger! 2. GMX ProMail - Platz 2 und Preis-QualitÃ¤tssieger! 3. Arcor - 4. web.de - 5. T-Online - 6. freenet.de - 7. daybyday - 8. e-Post

2 1

Dynamic Array Size
by Silvio a Beccara 29 Jul '03

29 Jul '03

Hi, how do I get the dimensions of a dynamic array? For instance, like the following: type PMatr2Dob10 = ^Tmatr2Dob10; Tmatr2Dob10 (Size1, Size2: Integer) = array [ 1 .. Size1, 0 .. Size2 ] of double; var kets: Pmatr2Dob10 Thanks very much Silvio a Beccara - Cornell University, USA

4 3

Conformant array problem
by Mirsad Todorovac 29 Jul '03

29 Jul '03

Hi, Here is the variant of Francois' test fsc28.pas: ----------------------- {$gnu-pascal} program mir037c; {slice array write, what Francois has done for Integer here is for Char} uses GPC; var v : array['A'..'Z'] of Integer; i,j : Char; procedure conformantArray(a: array[l..u : Char] of Integer); begin end; procedure ExpectError; begin if ExitCode = 0 then WriteLn ('failed') else begin WriteLn ('OK'); Halt (0) {!} end end; begin AtExit(ExpectError); i:='W'; j:=Succ('Z', 5); {j out of bounds so slice will be out of bounds too (write)} conformantArray(v[i..j]); { line 26 } end. ------------------------ here is what I get: bash-2.05b$ gpc mir037c.pas mir037c.pas: In main program: mir037c.pas:26: invalid operands to binary `+' mir037c.pas:26: incompatible type for argument 1 of `conformantArray' mir037c.pas:8: routine declaration bash-2.05b$ Now, what is done in fsc28.pas looks apparently the same: --------------------------- {$gnu-pascal} program fsc28; {slice array write} uses GPC; var v : array[1..100] of Integer; i,j : Integer; procedure conformantArray(a: array[l..u : Integer] of Integer); begin end; procedure ExpectError; begin if ExitCode = 0 then WriteLn ('failed') else begin WriteLn ('OK'); Halt (0) {!} end end; begin AtExit(ExpectError); i:=99; j:=110; {j out of bounds so slice will be out of bounds too (write)} conformantArray(v[i..j]); end. ----------------------------- And it passes. Thanks, Mirsad -- "I have a dream!" -- Martin Luther King Jr.

1 0

Bug report/feature request: overly aggressive warning on use of for-loop variable
by Joseph Oswald 29 Jul '03

29 Jul '03

The attached program demonstrates an issue with Gnu Pascal, version 20030507. The program cannot be compiled because of a non-existent "threat" on the global loop variable used in a subroutine. The main program has a loop which uses the variable, but calls no subroutines within the loop, and therefore is not vulnerable to any threat from use of that variable by other subroutines. Attempting to compile this program results in the message gpc -o forwarning forwarning.pas forwarning.pas: In main program: forwarning.pas:13: error: `for' loop counter is threatened in a subroutine This usage is present in Donald E. Knuth's WEB and TeX programs, and requires changing these loops to use unique loop variables. If there is a way to disable this check, I cannot find it in the documentation. gpc --version results in: Reading specs from /Developer/Pascal/gpc33d6/lib/gcc-lib/powerpc-apple-darwin/3.3/specs Configured with: ../gpc-3.3d6/configure --enable-languages=pascal,c --enable-threads=posix --prefix=/Developer/Pascal/gpc33d6 --target=powerpc-apple-darwin Thread model: posix gpc version 20030507, based on gcc-3.3 Thank you for your help. --Joe Oswald

4 4

Range tests
by Mirsad Todorovac 29 Jul '03

29 Jul '03

I have been rather busy in all this rush before holidays. I've seen that some of those range tests had problem compiling within testsuite, mostly because of rather harmless problems such as capitalisation. I've fixed that (hopefully), and henceforth I'll take as a standard that I don't submit a test unless it's successfully passed (or in this case - successfully failed) in the testsuite, under testsuite's compile options. Please find what's been done so far at http://www.alu.hr/~mtodorov/pascal/rngv0.02.tar.gz At my side they are of course all "failed". It would be reassuring if I get a confirmation that on range-checking compiler they all pass "OK". Mirsad -- "I have a dream!" -- Martin Luther King Jr.

1 0

Redelaration problems.
by nick 28 Jul '03

28 Jul '03

Hi! As I told I have some problems in the last GPC version (20030507) with repeated declaration in implementation part. consider the following: module test1 interface; export test1=all; procedure mt1 (str:string); function mt2 (str:string):string; end. module test1 implementation; procedure mt1 (str:string); begin writeln (str); end; function mt2 (str:string):string; begin mt2:=str; end; end. this file doesn't compile. GPC tells that mt2 redeclaration doesn't match the first one. The funny thing is that if I add type mystr=string after the "export" line and then change the function to use mystr - everything compile. Nick

2 2

Get still an error under FreeBSD (for Mr. Heckenbach)
by Karim Forsthofer 28 Jul '03

28 Jul '03

Hello Sorry, for the delay.. I was in holiday. At the 15. of july, I wrote several Emails to the gpc mailing list, because I got same "make" error outputs, when I tried to compile gpc and gcc on my FreeBSD 5.1 OS. Frank Heckenback told me, to change a simple command in the ../gcc.3.2.1/gcc/intl/plural.y file. I changed it. But now , I get an other error msg from make (after running a long time). The last output is : gcc: No input files specified ***Error code 1 Stop in /usr/home/karim/gpc-build/gcc. etc. The msg is very short and I don`t know how to handle that or to fix it. Need some feedback Greetings __________________________________________________________________ Gesendet von Yahoo! Mail - http://mail.yahoo.de Logos und KlingeltÃ¶ne fÃ¼rs Handy bei http://sms.yahoo.de

2 2

Re: overly aggressive warning message
by Joseph Oswald 27 Jul '03

27 Jul '03

Frank & Gale-- Thanks for your quick response to my e-mail. Most of the errors you refer to are well-localized in Knuth's source, and are part of the program that is meant to be modified to suit the particular platform that is being ported to. Once the WEB system is bootstrapped, each of the issues except for the for-loop one is a one-line change in the original source. At least in this respect, Knuth had the foresight to allow his programs to survive for what is a very long time in computer terms! Knuth was not programming in dec-pascal; as I recall it was in Pascal-H on the SAIL system. So I am not surprised that compiler directives have changed. However, on the for loop issue, I can see no practical advantage to enforcing this point of the standard to the point of not compiling the program. Can this aggressive behavior possibly be restricted to a --pedantic mode? I.e., can the error be demoted to a warning unless a --pedantic mode is enabled? On a larger scale, I think it is more useful for GPC to allow existing Pascal programs, with all their flaws, to be ported to GNU and other modern systems than for it to act as an enforcement mechanism to ensure that any Pascal programs that have harmless violations of the standard get weeded out. gcc as a whole takes a similar approach, I think. Thanks again to all the GPC contributors for an excellent tool! --Joe Oswald >I tried to compile Web and TeX and noticed a few more problems, >though apparently all quite minor: >- a lot of warnings which all seem to be harmless, but point to > unniceties in the code > >Time for `--dec-pascal' ...? > > > In > > the area of for-statements, there are latent errors in quite a few > > programs since there are quite a few compilers around which don't > > enforce all the for-statement requirements. > >Seems quite so. (GPC does one or two more now, but still not all.) _________________________________________________________________ The new MSN 8: smart spam protection and 2 months FREE* http://join.msn.com/?page=features/junkmail

2 1

BUG: GPC 20030507 on Mac OS X: does not recognize Macintosh line endings
by Joseph Oswald 27 Jul '03

27 Jul '03

The attached program and gzipped .txt files demonstrate a bug in GPC 20030507 compiled for Mac OS X. The EoLn(f) function does not recognize Macintosh line endings (character code decimal 13), only UNIX (decimal 10) line endings. Ideally, the EoLn behavior would recognize multiple possible line endings (including DOS CRLF), or at least be configurable at run-time. Running the macosx_eoln.pas on a Macintosh will result in the following output: file mac_eoln.txt had 0 lines. file unix_eoln.txt had 3 lines. Please be careful to preserve the line endings of the .txt.gz files in order to make the test a valid one. File transfer programs will often change the line endings of files believed to contain text, invalidating the test. gpc -v returns Reading specs from /Developer/Pascal/gpc33d6/lib/gcc-lib/powerpc-apple-darwin/3.3/specs Configured with: ../gpc-3.3d6/configure --enable-languages=pascal,c --enable-threads=posix --prefix=/Developer/Pascal/gpc33d6 --target=powerpc-apple-darwin Thread model: posix gpc version 20030507, based on gcc-3.3 Thank you for your help. --Joe oswald

1 0

Re: implicit object pointer dereference
by Adriaan van Os 26 Jul '03

26 Jul '03

Prof A Olowofoyeku (The African Chief) wrote: > What you are describing is almost equal to the "new" style Delphi > Classes (actually, this is already over 7 years old, so it is not so > new - but I use "new" here to distinguish it from the original BP > object model). That type of Delphi compatibility, is, I believe, on the > to-do list - but perhaps not a high priority. There is also the "Object-Oriented Extensions to Pascal" proposal, of the Technical Committee X3J9, see <http://www.pascal-central.com/OOE-stds.html>. In passing, I have to mention that I personally prefer the BP object model, see section "E.6.5.1 The Object Model. ... The reference model has the following disadvantages: ...". Still, it is a given fact that traditional Macintosh Pascal compilers implemented the reference model. Regards, Adriaan van Os

7 16