Well, just an update on my attempts to build GPC 20020410 on Mandrake (based on the SuSE GCC 2.95.3 source archive) ...
I followed the instructions from the web pages verbatim, and as the Chief said, it's "quite straightforward". As far as I could tell, there were no errors in building gpc-20020410, just a warning or two, which AFAIK is "normal" with C programs.
I then ran the test script, expecting this would produce a summary at the end. It didn't, so I ran this again but this time I redirected the output to a file. After this, I searched the file for the word "failed" and extracted the following :
TEST bill.pas: ./bill.pas: In main program: ./bill.pas:7: undeclared identifier `Dectobin' (first use in this routine) ./bill.pas:7: (Each undeclared identifier is reported only once ./bill.pas:7: for each routine it appears in.) failed
TEST confarr9.pas: ./confarr9.pas: In main program: ./confarr9.pas:7: undeclared identifier `Test' (first use in this routine) ./confarr9.pas:7: (Each undeclared identifier is reported only once ./confarr9.pas:7: for each routine it appears in.) failed
TEST fjf623.pas: failed: failed 1 2
I had a quick look at the first two "failures", and I am not sure why these would have failed. The Pascal code looked OK, so I suspect the problem might be in the script file, which I will try to decipher a bit later ...
I also invoked GPC to compile a small "hello" program in both Pascal and in C. Both worked, however, the "a.out" file from the Pascal compilation was over 500K, whereas that from the C compilation was about 13K. Is this normal???
Joe.
-----Original Message----- From: Frank Heckenbach [SMTP:frank@g-n-u.de] Sent: Thursday, April 11, 2002 9:31 PM To: gpc@gnu.de Subject: GPC 2.1 RC 6 released
Hi everyone,
GPC version 20020410, the release candidate #6 for GPC 2.1, has been uploaded to http://www.gnu-pascal.de/current/gpc-20020410.tar.gz
A diff against RC 5 can be found at http://www.gnu-pascal.de/current/gpc-20020402-20020410.diff.gz
According to a request by the Chief, the install target `pascal.install' and the Pascal only binary distributions (`pascal.bindist') do not include libgcc.a anymore. It is now only included with `pascal.install-with-gcc' and Pascal binaries with gcc (`pascal.bindist-with-gcc'), since the former should be used if (and only if) a matching GCC version is already installed which includes libgcc.a.
A module name conflict (which prohibited certain module names, e.g., `Files') was resolved, thanks to a report by Carel Fellinger. In the course of fixing this problem, the directive `asmname' was enabled for programs, units and module interfaces.
Some small bugs were fixed, including a problem when building GPC with `-pipe' (thanks to a report by George Shapovalov), the problem that ParamStr was not accepted as the first argument of `ReadStr' and `Val' (thanks to a report by Maurice Lombardi), a bug with differences of integer constans and a bug with a special case of nested schema types (thanks to reports by Emil Jerabek) and some other small bugs.
Frank
-- Frank Heckenbach, frank@g-n-u.de, http://fjf.gnu.de/, 7977168E GPC To-Do list, latest features, fixed bugs: http://www.gnu-pascal.de/todo.html
da Silva, Joe wrote:
Well, just an update on my attempts to build GPC 20020410 on Mandrake (based on the SuSE GCC 2.95.3 source archive) ...
I followed the instructions from the web pages verbatim, and as the Chief said, it's "quite straightforward". As far as I could tell, there were no errors in building gpc-20020410, just a warning or two, which AFAIK is "normal" with C programs.
Maybe, maybe not. Without seeing the text of the warnings, it could be anything.
I then ran the test script, expecting this would produce a summary at the end.
What exactly did you do? Simply make, or make pascal.check-long or whatever?
It didn't, so I ran this again but this time I redirected the output to a file. After this, I searched the file for the word "failed" and extracted the following :
TEST bill.pas: ./bill.pas: In main program: ./bill.pas:7: undeclared identifier `Dectobin' (first use in this routine) ./bill.pas:7: (Each undeclared identifier is reported only once ./bill.pas:7: for each routine it appears in.) failed
TEST confarr9.pas: ./confarr9.pas: In main program: ./confarr9.pas:7: undeclared identifier `Test' (first use in this routine) ./confarr9.pas:7: (Each undeclared identifier is reported only once ./confarr9.pas:7: for each routine it appears in.) failed
TEST fjf623.pas: failed: failed 1 2
I had a quick look at the first two "failures", and I am not sure why these would have failed. The Pascal code looked OK, so I suspect the problem might be in the script file, which I will try to decipher a bit later ...
It might be easier to try compiling the programs manually (with `gpc --automake'). If they still fail, the problem is not in the script.
I also invoked GPC to compile a small "hello" program in both Pascal and in C. Both worked, however, the "a.out" file from the Pascal compilation was over 500K, whereas that from the C compilation was about 13K. Is this normal???
Yes, FAQ.
Frank
Yuri Prokushev wrote:
http://www.gnu-pascal.de/current/gpc-20020402-20020410.diff.gz
Sorry, my mistake. It's back now.
Frank