Pike-devel January 2004

pike-devel@lists.lysator.liu.se

11 participants
77 discussions

Re: Pike @ Debian
by Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum 26 Jan '04

26 Jan '04

pike -x module is _my_ friend, and it works everywhere except in Debian. / Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) Previous text: >2004-01-26 18:18: >Subject: Re: Pike @ Debian >-------------------------------------------------------------------- >On Mon, Jan 26, 2004 at 06:05:03PM +0100, Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) @ Pike (-) developers forum scribbled: >> I don't mind if Debian mucks around with the files, as long as they >> make sure that everything _works_ afterwards. ;-) >Caudium compiles with pike on debian, so it works. pike7.4-config --cflags >is your friend. I cannot and will not try to change the debian standards to >fit the Pike standards file location wise. The thing to be fixed and made >more flexible is the build system which should be able to find the files >anywhere they are given certain tips as to their locations. A simple config >file in a well-known location would do just fine. > >marek > > > > / Brevbäraren >

1 0

Re: Pike @ Debian
by Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum 26 Jan '04

26 Jan '04

It is told, by the master. The path is passed through the variable master()->include_prefix, which is initialized to the value of ¤include_prefix¤ when master.pike is generated from master.pike.in. So as you can see, it's fully configurable. / Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) Previous text: >2004-01-26 18:12: >Subject: Re: Pike @ Debian >-------------------------------------------------------------------- >On Mon, Jan 26, 2004 at 06:00:03PM +0100, Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) @ Pike (-) developers forum scribbled: >> Pike looks for the includes where it installs them. If Debian wants >> them installed somewhere else, it's their job to make sure Pike looks >> there instead. Pike can't be clairvoyant about these things. >Fix the build system then. It should be flexible enough to be told where the >includes are. > >marek > > > > / Brevbäraren >

1 0

Pike @ Debian
by Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum 26 Jan '04

26 Jan '04

I'm trying to use the Debian Pike packages together with some Pike software, but it looks like the packages are broken somehow. Although I have pike7.4-dev installed, /usr/lib/pike/7.4.35/include/pike/ only contains the specs file. There are no headers, no makefiles, no aclocal files. Hence modules can not be build. What gives?

2 2

Re: Pike @ Debian
by Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum 26 Jan '04

26 Jan '04

>all the includes are in /usr/include/pike/PIKEVER/ That's not where they're supposed to be. The module build system doesn't look there, and so putting them there will not work. / Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) Previous text: >2004-01-26 17:37: >Subject: Re: Pike @ Debian >-------------------------------------------------------------------- >On Mon, Jan 26, 2004 at 03:40:01PM +0100, Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) @ Pike (-) developers forum scribbled: >> I'm trying to use the Debian Pike packages together with some Pike >> software, but it looks like the packages are broken somehow. Although >> I have pike7.4-dev installed, /usr/lib/pike/7.4.35/include/pike/ only >> contains the specs file. There are no headers, no makefiles, no >> aclocal files. Hence modules can not be build. What gives? >all the includes are in /usr/include/pike/PIKEVER/ > >marek > > > > / Brevbäraren >

2 2

Re: nasm and non x86 platforms
by Mirar ＠ Pike developers forum 26 Jan '04

26 Jan '04

>> Possibly the test should be if it can compile *and* link the compiled >> (assembled) code with something else. > >Should be it can compile *and* if the link is compatible with target >linker/code ? Nah? Linking is what fails today. Compile/assembler -> .o file and then link .o files to a .so (or an executable). Sometimes the .o files are collected in a ".a" library too, but that part is just archiving. / Mirar Previous text: >2004-01-26 13:14: >Subject: Re: nasm and non x86 platforms >-------------------------------------------------------------------- > >Le 26 janv. 04, à 11:05, Mirar @ Pike developers forum a écrit : > >> --without-machine-code should do it. Please find out why it doesn't. >> >> But I guess you could always expand the test in Image/configure.in: >> >> elif test "`uname -s`" = "OpenBSD"; then >> #nasm on OpenBSD is braindamaged >> AC_MSG_RESULT(no) >> >> add another three rows and s/OpenBSD/FreeBSD...? :) > >Works like a charm on FreeBSD x86. > >> Possibly the test should be if it can compile *and* link the compiled >> (assembled) code with something else. > >Should be it can compile *and* if the link is compatible with target >linker/code ? > >/Xavier > > > > / Brevbäraren >

1 0

nasm and non x86 platforms
by Xavier Beaudouin 26 Jan '04

26 Jan '04

Hello there, I have several non x86 platforms to maintain on FreeBSD. Seems that even if I send --without-machine-code to configure of pike 7.4 (7.5 and 7.2 as well), the auto-configure script still wants to use nasm if it can be found on his path. This give some "interressing" errors like this (on amd64 for example) : gmake[3]: Entering directory `/tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src' Compiling las_t.c Compiling builtin_functions_t.c Compiling peep_t.c Compiling module_t.c /tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src/smartlink cc -DPIKE_SRC_ROOT="/tmp/a/ports/lang/pike74/work/Pike-v7.4.31" -I. -I/tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src -I/usr/local/include -pthread -R/usr/local/lib -L/usr/local/lib -rdynamic language.o security.o bignum.o pike_cpulib.o interpret.o constants.o cpp.o fdlib.o cyclic.o array.o backend.o callback.o encode.o docode.o dynamic_buffer.o dynamic_load.o error.o fd_control.o fsort.o gc.o hashtable.o lex.o multiset.o signal_handler.o pike_search.o pike_types.o main.o mapping.o pike_memory.o module_support.o pikecode.o object.o opcodes.o operators.o port.o program.o rbtree.o rusage.o stralloc.o stuff.o threads.o version.o queue.o builtin.o iterators.o svalue.o las_t.o builtin_functions_t.o peep_t.o module_t.o `cat modules/linker_options` -lm -lcrypt -o tpike /usr/bin/ld: warning: i386 architecture of input file `/tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src/modules/Image/ module.a(assembly.o)' is incompatible with i386:x86-64 output /tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src/modules/Image/ module.a(operator.o): In function `image_operator_plus': operator.o(.text+0xdf7): undefined reference to `image_add_buffers_mmx_x86asm' operator.o(.text+0x10ec): undefined reference to `image_add_buffer_mmx_x86asm' operator.o(.text+0x11fd): undefined reference to `image_sub_buffer_mmx_x86asm' /tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src/modules/Image/ module.a(operator.o): In function `image_operator_multiply': operator.o(.text+0x1a39): undefined reference to `image_mult_buffers_mmx_x86asm' operator.o(.text+0x1cc1): undefined reference to `image_mult_buffer_mmx_x86asm' gmake[3]: *** [tpike-real] Error 1 gmake[3]: Leaving directory `/tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src' gmake[2]: *** [/tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src/tpike] Error 2 gmake[2]: Leaving directory `/tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src' gmake[1]: *** [pike] Error 2 gmake[1]: Leaving directory `/tmp/a/ports/lang/pike74/work/Pike-v7.4.31/src' gmake: *** [all] Error 2 *** Error code 2 (Full make log here : http://bento.freebsd.org/errorlogs/amd64-5-latest/pike74-7.4.31_1.log) Is there any "official" / good way to make configure script to avoid using nasm on non x86 platforms (I have likewise problems in : alpha, ppc, sparc64....) ????? Thanks... /Xavier

2 2

rntcl warning
by Peter Bortas ＠ Pike developers forum 25 Jan '04

25 Jan '04

Exactly why does rntcl give a warning? Is there some new way of building that I should know about? checking if /export/aphraelraid/home/peter/hack/Pike/7.5/build/windows_xp-5.1.2600-x86/smartlink exists is up to date and works... no configure: WARNING: rntcc/rntcl/rnticl/rntecl detected. checking for gcc... /usr/bin/gcc Using /usr/bin/gcc to compile smartlink. checking whether smartlink compiles... no - use the sh script

1 0

Warnings
by Martin Nilsson (saturator) ＠ Pike (-) developers forum 25 Jan '04

25 Jan '04

The following code generates the warning: Attempt to override a non local variable "foo" with a non variable. class A { function foo; } class B { inherit A; void foo() { return; } } How can that warning be avoided?

2 3

Pike 7.4.31 is broken on FreeBSD 4.x (was: caudium-devel-1.3.21.20031129,2 broken on 4.x)
by Xavier Beaudouin 23 Jan '04

23 Jan '04

Hello there, Seems that for some obscure reasons that pike 7.4.31 is broken on FreeBSD 4.x... Why does it need String* when ./configuring it ???? /Xavier Début du message réexpédié : > De: Kris Kennaway <kris(a)obsecurity.org> > Date: 23 janvier 2004 17:24:23 GMT+01:00 > À: Xavier Beaudouin <kiwi(a)oav.net>, pav(a)FreeBSD.org > Objet: caudium-devel-1.3.21.20031129,2 broken on 4.x > > http://bento.freebsd.org/errorlogs/i386-4-latest/caudium-devel > -1.3.21.20031129,2.log > > Can you please investigate and/or report to the developers? > > Kris

2 2

Sql interface
by Peter Bortas ＠ Pike developers forum 22 Jan '04

22 Jan '04

Is there anything deterministic about what Sql.Sql->create() throws? I just added this to AIDOs startup to autocreate a database, and I'm not overly pleased with it: mixed err = catch { sql = objectp(url)?url:Sql.Sql( url ); }; //FIXME: Very crude, but works for the normal sqlurl case. if(err && !objectp(url)) { string baseurl = reverse(reverse(url/"/")[1..])*"/"; string dbname = (url/"/")[-1]; werror( "NOTE: Trying to create database %O\n", dbname ); Sql.Sql( baseurl ) ->query( "CREATE DATABASE IF NOT EXISTS "+ dbname ); } else if(err) { werror("FATAL: Could not connect to local media database!\n"); //FIXME: Just letting this through results in a unfriendly backtrace. } sql = objectp(url)?url:Sql.Sql( url ); To keep things on topic: What was the result of the latest discusion about how to write ranges ending before the last index? [..-1]...

6 9

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Pike-devel January 2004