Gpc

gpc@gnu.de

5400 discussions

gpc bootstrapping fails on makeinfo
by Treutwein Bernhard 13 Mar '16

13 Mar '16

Hi all, when I try to bootstrap gpc, on a TinyCore 7.0 Linux with the necessary extensions installed. I downloaded the source distribution from http://alignment.hep.brandeis.edu/Software/Pascal/gpc-20070904-build.tar.gz and followed the instructions given in http://alignment.hep.brandeis.edu/Software/Pascal/Pascal.html But make aborts with the following error (after quite some time workin and some warnings) during the attempt to create the info files: [...] make[2]: Entering directory '/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/rts' make[2]: Nothing to be done for 'generated-files'. make[2]: Leaving directory '/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/rts' cd /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/rts && make -f Makefile.in srcdir=. GCC_VERSION="3.4.6" SHELL="/bin/sh" CFLAGS="-g -O2" PFLAGS="" AR_FLAGS="rc" RTSFLAGS="" DESTDIR="../.." AR="ar" RANLIB="ranlib" RANLIB_TEST="" generated-files make[2]: Entering directory '/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/rts' make[2]: Nothing to be done for 'generated-files'. make[2]: Leaving directory '/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/rts' [ -d "/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info" ] || /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/script/mkdir-p "/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info" LANG=en_GB LC_ALL=en_GB makeinfo --split-size=5000000 -I /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc -I /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/en -I /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/generated -o /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info/gpc.info gpc.texi could not open gpc.texi: No such file or directory /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/Make-lang.in:808: recipe for target '/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info/gpc.info' failed make[1]: *** [/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info/gpc.info] Error 1 make[1]: Leaving directory '/mnt/sda5/TeX-GPC/work/gcc' Makefile:23375: recipe for target 'all-gcc' failed make: *** [all-gcc] Error 2 tc@box:/mnt/sda5/TeX-GPC/work$ I'm neither an autoconf nor a make guru so I am lost. I have no idea how I could tell make/configure to skip the step for creation of the documentation. I guess that it is almost done, but I am not sure about that, I tried to install with sudo make pascal.install-with -gcc which did something, but also ended with errors, which look very similar to the errors I hadbefore: [...] make[1]: Nothing to be done for 'generated-files'. make[1]: Leaving directory '/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/rts' [ -d "/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info" ] || /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/script/mkdir-p "/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info" LANG=en_GB LC_ALL=en_GB makeinfo -I /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc -I /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/en -I /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/generated -o /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info/gpc.info gpc.texi could not open gpc.texi: No such file or directory /mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/Make-lang.in:808: recipe for target '/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info/gpc.info' failed make: *** [/mnt/sda5/TeX-GPC/work/gcc-3.4.6/gcc/p/doc/info/gpc.info] Error 1 Although gpc --version now works. But I cannot compile anything. Going to demos and trying gpc hello.pas -o hello yields gpc: installation problem, cannot exec gpc1: No such file or directory Here are the versions of the different tools used: $ bash --version GNU bash, version 4.3.39(1)-release (i486-pc-linux-gnu) ... $ bzip2 --version bzip2, a block-sorting file compressor. Version 1.0.6, 6-Sept-2010. ... $ sed --version sed (GNU sed) 4.2.2 ... $ awk --version GNU Awk 4.1.3, API: 1.1 (GNU MPFR 3.1.3-p4, GNU MP 6.0.0) ... $ m4 --version m4 (GNU M4) 1.4.17 ... $ bison --version bison (GNU Bison) 3.0.4 ... $ flex --version flex 2.5.39 ... $ autoconf --version autoconf (GNU Autoconf) 2.69 ... $ makeinfo --version texi2any (GNU texinfo) 6.0 ... $ help2man --version GNU help2man 1.46.1 Regards and thanks in advance for any help -- Bernhard Treutwein

4 4

Re: GPC source code errors
by Waldek Hebisch 10 Mar '16

10 Mar '16

Gale Paeper wrote: > > While trying to get a successful compilation of GPC on Mac OS X 10.11 using AppleĂ˘ÂÂs LLVM based tool chain, I discovered what looks to be a couple of source code errors in p/statements.c in the expand_pascal_assignment2 functionĂ˘ÂÂs code. That function is declared with a void return type so there shouldnĂ˘ÂÂt be any return statements returning a value anywhere within the body of the function but there are two uses of the CHK_EM macro which does expand to a return statement that potentially returns a value (a tree type). Trying to potentially return something into the void of a non-existant return isnĂ˘ÂÂt technically correct code and I havenĂ˘ÂÂt found a way to get the LLVM based clang compiler to accept it. > > IĂ˘ÂÂll note that you can get a gnu gcc version 5.2.0 compiler to accept the code when compiling with the -std=gnu89 command line option. > > The GPC source code version IĂ˘ÂÂm trying to get compiling is the Adriaan van OsĂ˘ÂÂs Mac OS X patched gcc 3.4.6 with the p directory replaced with Waldek HebischĂ˘ÂÂs git repositoryĂ˘ÂÂs (<https://github.com/hebisch/gpc>) p directory source code. The code fragment containing the error producing code starting around line 1231 in file p/statements.c (in expand_pascal_assignment2 functionĂ˘ÂÂs body) is: > > schema_source = undo_schema_dereference (source); > schema_target = undo_schema_dereference (target); > DEREFERENCE_SCHEMA (source); > DEREFERENCE_SCHEMA (target); > CHK_EM (source); > CHK_EM (target); The patch below should fix this: diff --git a/p/statements.c b/p/statements.c index 4904834..e0026b7 100644 --- a/p/statements.c +++ b/p/statements.c @@ -1232,8 +1232,8 @@ expand_pascal_assignment2 (tree target, tree source, int is_init) schema_target = undo_schema_dereference (target); DEREFERENCE_SCHEMA (source); DEREFERENCE_SCHEMA (target); - CHK_EM (source); - CHK_EM (target); + if (EM (source)||EM (target)) + return; /* Restricted types. @@@@ Maybe this needs further checking */ if (TREE_CODE (source) == CALL_EXPR && PASCAL_TYPE_RESTRICTED (TREE_TYPE (source))) -- Waldek Hebisch

5 9

Inverse ord
by Paul Isaacs 07 Mar '16

07 Mar '16

Hello Waldek, Thanks, I should have looked longer and more carefully. I am retired now but spent my working years with micro/embedded processors. I want to keep my mind alive so I am setting forth to write a 10206 compliant compiler for embedded backends. Pascal should be a good fit for Arduino and Raspberry-Pi experimentalists. It is not as obscure as C and not as complex and picky as ADA. I am currently writing a 10206 compliant parser to produce a stream of lexically correct terminals from Pascal source. If I ever complete the lexical analysis, I will move on to semantically checking the results and producing a stream of intermediate opcodes for a backend. I know that I will need to feed extra information through the parser ( e.g. an "interrupt" directive for an interrupt procedure and "peek" and "poke" procedures for hardware I/O ) so I am looking for ways to do that and still be 10206 compliant. Thus the language lawyer interest. Thanks for your time. Maybe someday I will be able to make a contribution to the contributions section of the gpc web site. Regards, Paul Isaacs

1 0

Thanks
by Paul Isaacs 07 Mar '16

07 Mar '16

Hello Gale and Waldek, Thank you for your extensive replies. I will change over to using "gp". I notice that "inverse ord" is mentioned in the spec as an addition. However, other than a single mention, the spec is silent on inverse ord. Was such a thing implemented? The Gnu Pascal manual does not include the word "inverse". I would like to do a binary search on a set of keywords whose value is specified in a "TokenTypes" enumeration. The range of the array that stores the keywords and their value is the enumeration. Is it possible to compute the enumeration that is midway between two other enumerations? In other words is it possible to find an enumeration from its "ord" value? Regards, Paul Isaacs

2 1

Importing
by Paul Isaacs 07 Mar '16

07 Mar '16

Hello, I noticed that imports done outside the interface-specification-part are not dependency checked by --automake. This creates problems for first time gnu pascal uses such as myself. I started placing imports outside the interface-specification-part to avoid circular dependencies and it took a while to understand that --automake was no longer finding the import dependencies. I also assumed that the reason for the, rather awkward, multiple interface capability was to get around the module circular dependency problem. Ada resorted to the kludgy 'limited with' workaround. Gnu pascal seems to have resorted to the import 'in' workaround. However, having filenames in the source code does not seem to me to be a good idea. Would "import modulename_interfacename" work? The first '_' character would take on special meaning but the change would be 10602 specification compliant and having the module name and interface name combined into one identifier would eliminate the need for file names in the source code. Regards, Paul Isaacs

3 2

GPC source code errors
by Gale Paeper 25 Feb '16

25 Feb '16

While trying to get a successful compilation of GPC on Mac OS X 10.11 using Apple’s LLVM based tool chain, I discovered what looks to be a couple of source code errors in p/statements.c in the expand_pascal_assignment2 function’s code. That function is declared with a void return type so there shouldn’t be any return statements returning a value anywhere within the body of the function but there are two uses of the CHK_EM macro which does expand to a return statement that potentially returns a value (a tree type). Trying to potentially return something into the void of a non-existant return isn’t technically correct code and I haven’t found a way to get the LLVM based clang compiler to accept it. I’ll note that you can get a gnu gcc version 5.2.0 compiler to accept the code when compiling with the -std=gnu89 command line option. The GPC source code version I’m trying to get compiling is the Adriaan van Os’s Mac OS X patched gcc 3.4.6 with the p directory replaced with Waldek Hebisch’s git repository’s (<https://github.com/hebisch/gpc>) p directory source code. The code fragment containing the error producing code starting around line 1231 in file p/statements.c (in expand_pascal_assignment2 function’s body) is: schema_source = undo_schema_dereference (source); schema_target = undo_schema_dereference (target); DEREFERENCE_SCHEMA (source); DEREFERENCE_SCHEMA (target); CHK_EM (source); CHK_EM (target); I’m not sure how long the code “error” has been around. I checked the gpc-20070904 sources and it is also in that version. I don’t have a working GPC at the moment so I don’t know what sort of ill formed Pascal source code would be needed to exercise the CHK_EM macro return statement’s code and what sort of ills would result if the code was exercised. The compiler version information on the compiler being used for the compiling stopping errors is: gcc --version Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1 Apple LLVM version 7.0.2 (clang-700.1.81) Target: x86_64-apple-darwin15.3.0 Thread model: posix Note: Apple has got their tool chain wired up so that “gcc” invokes LLVM based clang. In case you want to see it I’ve included the gcc [really clang] and make diagnostic output below my signature. P.S. I think I’m getting close to seeing the light at the end of the tunnel of building a gcc-3.4.6 based GPC on the latest version of Mac OS X. The path to success looks to be using a real gnu based gcc and as for compiling and assembling to object code. At least for the fairly old code base of gcc-3.4.6, trying to use a LLVM based tool chain build path seems to be a never ending series of “close but no cigar” dead ends, issues, and headaches. Gale Paeper gpaeper(a)empirenet.com gcc -std=gnu99 -arch i386 -o p/statements.o -c -g -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -DHAVE_CONFIG_H -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -Wno-traditional -I. -Ip -I../../gcc-3.4.6/gcc -I../../gcc-3.4.6/gcc/p -I../../gcc-3.4.6/gcc/../include -DGPC -DGPC_UNITS_DIR=\"/usr/local/gpc346u4/gpc346u4/lib/gcc/i386-apple-darwin15/3.4.6/units\" -I../../gcc-3.4.6/gcc/p -DRTS_VERSION_CHECK="\"_p_GPC_RTS_VERSION_`cat ../../gcc-3.4.6/gcc/p/rts/rts-version`\"" ../../gcc-3.4.6/gcc/p/statements.c 4 warnings generated. warning: unknown warning option '-Wno-traditional' [-Wunknown-warning-option] In file included from ../../gcc-3.4.6/gcc/p/statements.c:29: In file included from ../../gcc-3.4.6/gcc/p/gpc.h:46: In file included from ../../gcc-3.4.6/gcc/p/gbe.h:151: ../../gcc-3.4.6/gcc/output.h:122:6: warning: '__format__' attribute argument not supported: asm_fprintf [-Wignored-attributes] ATTRIBUTE_ASM_FPRINTF(2, 3); ^ ../../gcc-3.4.6/gcc/output.h:112:53: note: expanded from macro 'ATTRIBUTE_ASM_FPRINTF' #define ATTRIBUTE_ASM_FPRINTF(m, n) __attribute__ ((__format__ (__asm_fprintf__, m, n))) ATTRIBUTE_NONNULL(m) ^ ../../gcc-3.4.6/gcc/p/statements.c:1235:3: error: void function 'expand_pascal_assignment2' should not return a value [-Wreturn-type] CHK_EM (source); ^~~~~~~~~~~~~~~ ../../gcc-3.4.6/gcc/p/gpc.h:155:36: note: expanded from macro 'CHK_EM' #define CHK_EM(t) do { if (EM (t)) return error_mark_node; } while (0) ^ ~~~~~~~~~~~~~~~ ../../gcc-3.4.6/gcc/p/statements.c:1236:3: error: void function 'expand_pascal_assignment2' should not return a value [-Wreturn-type] CHK_EM (target); ^~~~~~~~~~~~~~~ ../../gcc-3.4.6/gcc/p/gpc.h:155:36: note: expanded from macro 'CHK_EM' #define CHK_EM(t) do { if (EM (t)) return error_mark_node; } while (0) ^ ~~~~~~~~~~~~~~~ ../../gcc-3.4.6/gcc/p/statements.c:1342:1: warning: no previous prototype for function 'assign_string2' [-Wmissing-prototypes] assign_string2 (tree target, tree source, int is_init) ^ ../../gcc-3.4.6/gcc/p/statements.c:1342:47: warning: unused parameter 'is_init' [-Wunused-parameter] assign_string2 (tree target, tree source, int is_init) ^ 4 warnings and 2 errors generated. make[2]: *** [p/statements.o] Error 1 make[1]: *** [stage1_build] Error 2 make: *** [bootstrap] Error 2

1 0

RE: GPC compile on Mac OS X 10.10.5?
by samiam＠moorecad.com 28 Jan '16

28 Jan '16

2 1

RE: GPC compile on Mac OS X 10.10.5?
by samiam＠moorecad.com 27 Jan '16

27 Jan '16

2 1

RE: GPC compile on Mac OS X 10.10.5?
by samiam＠moorecad.com 27 Jan '16

27 Jan '16

1 0

Re: GPC compile on Mac OS X 10.10.5?
by Schneider 27 Jan '16

27 Jan '16

Scott: > On another subject entirely, Do you publish your programs, ie., are > they available somewhere? https://alum.mit.edu/www/toms/delila/delilaprograms.html Note that alum.mit.edu/www/toms points to the current location of my website, so if you use that it should always work even if my website moves again. > I'd be interested in large ISO 7185 compliant programs for compiler > testing. Strict ISO 7185 programs are kinda rare. The only exception is that some programs need the date and time. They all use a procedure called getdatetime which is compiler specific. I have several programs named time*.p for different compilers. They contain 'modules' for the time functions. Long ago I recognized that insertion methods were not standard in Pascal, so I wrote a program - 'module' https://alum.mit.edu/www/toms/delila/module.html - which can transfer chunks of text between Pascal programs. So I use module to transfer the time modules into a program and then they compile. Other than the time modules, I can't guarantee that they are all "strict ISO 7185 programs" but I try to stick to the original Wirth definition. So if you set aside the programs that contain 'getdatetime' it would be a pretty clean set. You could also use the timenul.p program to neuter any program. (That is, use module to transfer the modules of timenul.p into Delila programs that use time). Since you would like large programs, the biggest is probably delila.p - but it uses time modules. If you find bugs, I'd appreciate knowing about them. Tom Thomas D. Schneider, Ph.D. Senior Investigator National Institutes of Health National Cancer Institute Center for Cancer Research Gene Regulation and Chromosome Biology Laboratory Molecular Information Theory Group Frederick, Maryland 21702-1201 schneidt(a)mail.nih.gov https://schneider.ncifcrf.gov (current link) https://alum.mit.edu/www/toms (permanent link)

1 0

← Newer
1
...
5
6
7
8
9
10
11
...
540
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

Gpc