pike issueshttps://git.lysator.liu.se/pikelang/pike/-/issues2009-04-16T14:11:39Zhttps://git.lysator.liu.se/pikelang/pike/-/issues/369class { ; } bugs out.2009-04-16T14:11:39ZPeter Bortasclass { ; } bugs out.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=369
Reported by Fredrik Noring, Roxen Internet Software <noring@roxen.com>
All examples below should compile nicely.
> class { }
Ok.
> class { ; }
Parse error.
...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=369
Reported by Fredrik Noring, Roxen Internet Software <noring@roxen.com>
All examples below should compile nicely.
> class { }
Ok.
> class { ; }
Parse error.
> class { class { }; }
Ok.
> class { class { }; ; }
Ok.
> class { class { }; ; ; }
Ok.
> class { ; class { }; ; ; }
Parse error.Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/355Unsorted charset list in _Charset module2009-04-16T14:11:39ZPeter BortasUnsorted charset list in _Charset moduleImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=355
Reported by @marcus
```
The charsets iso-8859-10 through iso-8859-15 violate the lexiographic sort
order in the list "charset_map" in tables.c. All iso-8859-1* must ...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=355
Reported by @marcus
```
The charsets iso-8859-10 through iso-8859-15 violate the lexiographic sort
order in the list "charset_map" in tables.c. All iso-8859-1* must come
before iso-8859-2, or the binary search won't work. This goes for
iso_8859-1x:199x as well.
```Pike 7.2https://git.lysator.liu.se/pikelang/pike/-/issues/340Roxen uses all available CPU on AIX.2009-04-16T14:11:39ZPeter BortasRoxen uses all available CPU on AIX.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=340
Reported by @grubba
```
From: Ulf Lundstrom <ulf@softwarerun.com>
Organization: SoftwareRun AB
To: Henrik Grubbström <grubba@roxen.com>
Date: Tue, 19 Sep 2000 11:26:5...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=340
Reported by @grubba
```
From: Ulf Lundstrom <ulf@softwarerun.com>
Organization: SoftwareRun AB
To: Henrik Grubbström <grubba@roxen.com>
Date: Tue, 19 Sep 2000 11:26:50 +0200
Subject: Re: Roxen och AIX Bug Id 82
```
Hej igen...
Lyckades inte kompilera Pike från CVS'en...
Laddade hem senaste binären av Pike från Autobuild istället.
Kör nu Pike 2.0.227...
Kunde inte labba mer med 4 propps maskinen, så denna output är från en
singelcpu maskin... Iövrigt identiska vad gäller konfiguration, både
hårdvara
och mjukvara....
Får veta detta av roxen:
Server uptime : 4:42 m:s
CPU-Time used : 5:05 m:s (107.0%)
Page faults (non I/O) : 7863
Number of signals received : 30543
Hmm... Ser inge vidare ut om man säger så... ;-)
Alla tips är välkommna...
Mvh
Ulf Lundström
SoftwareRun ABPike 7.0Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/339Purely void variables2009-04-16T14:11:39ZPeter BortasPurely void variablesImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=339
Reported by Fredrik Noring, Roxen Internet Software <noring@roxen.com>
Maybe purely void variables ough to give an error?
```
string f(void s)
{ // ^^^^
void x...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=339
Reported by Fredrik Noring, Roxen Internet Software <noring@roxen.com>
Maybe purely void variables ough to give an error?
```
string f(void s)
{ // ^^^^
void x = "baz";
//^^^^
return s + x;
}
void main()
{
write(f("zot")+"\n");
}
```Pike 7.2https://git.lysator.liu.se/pikelang/pike/-/issues/327T_UNKNOWN defined in both svalue.h and <sys/conf.h>2009-04-16T14:11:39ZPeter BortasT_UNKNOWN defined in both svalue.h and <sys/conf.h>Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=327
Reported by @grubba
```
Compiling
/tmp/autobuild/pike7.1-20000918181831.tar/src/modules/system/system.c
In file included from
/tmp/autobuild/pike7.1-20000918181831.ta...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=327
Reported by @grubba
```
Compiling
/tmp/autobuild/pike7.1-20000918181831.tar/src/modules/system/system.c
In file included from
/tmp/autobuild/pike7.1-20000918181831.tar/src/modules/system/system.c:58:
/usr/include/sys/conf.h:199: warning: `T_UNKNOWN' redefined
/tmp/autobuild/pike7.1-20000918181831.tar/src/modules/system/../../svalue.h:103:
warning: this is the location of the previous definition
```
Suggested fix: Rename T_UNKNOWN to PIKE_T_UNKNOWN.Pike 7.2https://git.lysator.liu.se/pikelang/pike/-/issues/262Pike hangs on checking if signal handlers reset automatically2009-04-16T14:11:39ZPeter BortasPike hangs on checking if signal handlers reset automaticallyImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=262
Reported by Honza Petrous <hop@unibase.cz>
Solaris SPARC 2.6, Sparcstation 1
Configure hangs on "checking if signal handlers reset automatically...".
```
sunny[/]# ...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=262
Reported by Honza Petrous <hop@unibase.cz>
Solaris SPARC 2.6, Sparcstation 1
Configure hangs on "checking if signal handlers reset automatically...".
```
sunny[/]# uname -a
SunOS sunny 5.6 Generic_105181-12 sun4c sparc SUNW,Sun_4_60
sunny[/]# ps -ft pts/5
UID PID PPID C STIME TTY TIME CMD
root 18523 18522 0 10:28:09 pts/5 0:02 ./conftest
root 15004 11234 0 10:01:14 pts/5 0:51 /bin/sh ./configure
root 18522 15004 0 10:28:09 pts/5 0:00 /bin/sh ./configure
root 11234 11231 0 14:18:45 pts/5 0:04 tcsh
sunny[/]# truss -p 18523
Received signal #14, SIGALRM, in lwp_sema_wait() [caught]
lwp_sema_wait(0xEF6984D8) Err#91 ERESTART
sigprocmask(SIG_SETMASK, 0xEF39CE2C, 0x00000000) = 0
lwp_sema_post(0x00021D70) = 0
lwp_sema_wait(0x00021D70) = 0
lwp_mutex_unlock(0xEF69B1C8) = 0
lwp_mutex_lock(0xEF69B1C8) = 0
setitimer(ITIMER_REAL, 0xEF39C888, 0x00000000) = 0
setcontext(0xEF39C820)
sigprocmask(SIG_BLOCK, 0xEF6984C8, 0x00000000) = 0
setitimer(ITIMER_REAL, 0xEF39CCD0, 0x00000000) = 0
sigprocmask(SIG_UNBLOCK, 0xEF6984C8, 0x00000000) = 0
Received signal #14, SIGALRM, in lwp_sema_wait() [caught]
lwp_sema_wait(0xEF6984D8) Err#91 ERESTART
sigprocmask(SIG_SETMASK, 0xEF39CE2C, 0x00000000) = 0
lwp_sema_post(0x00021D70) = 0
lwp_sema_wait(0x00021D70) = 0
setcontext(0xEF39C820)
lwp_sema_post(0xEF6984D8) = 0
sigprocmask(SIG_BLOCK, 0xEF6984C8, 0x00000000) = 0
setitimer(ITIMER_REAL, 0xEF39CCD0, 0x00000000) = 0
sigprocmask(SIG_UNBLOCK, 0xEF6984C8, 0x00000000) = 0
lwp_sema_wait(0xEF6984D8) = 0
sigprocmask(SIG_BLOCK, 0xEF6984C8, 0x00000000) = 0
setitimer(ITIMER_REAL, 0xEF39CCD0, 0x00000000) = 0
sigprocmask(SIG_UNBLOCK, 0xEF6984C8, 0x00000000) = 0
Received signal #14, SIGALRM, in lwp_sema_wait() [caught]
lwp_sema_wait(0xEF6984D8) Err#91 ERESTART
sigprocmask(SIG_SETMASK, 0xEF39CE2C, 0x00000000) = 0
```
[ ... and so on ...]
/HonzaPike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/223Still some Pike error output in the "add modules list".2009-04-16T14:11:39ZPeter BortasStill some Pike error output in the "add modules list".Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=223
Reported by Hans-Göran Wendelhof, Roxen Internet Software <hg@roxen.com>
```
In the "add modules" list:
pike/lib/pike/modules/process.pmod:157 Error: Index 'fork' no...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=223
Reported by Hans-Göran Wendelhof, Roxen Internet Software <hg@roxen.com>
```
In the "add modules" list:
pike/lib/pike/modules/process.pmod:157 Error: Index 'fork' not present in module 'predef'.
pike/lib/pike/modules/process.pmod:276 Error: Class definition failed.
```Pike 7.0https://git.lysator.liu.se/pikelang/pike/-/issues/202Prototypes without argument names causes "Duplicate local variable ''".2009-04-16T14:11:39ZPeter BortasPrototypes without argument names causes "Duplicate local variable ''".Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=202
Reported by @grubba
```
Pike v7.1 release 6 running Hilfe v2.0 (Incremental Pike Frontend)
> class { void foo(int, string); };
-:1:Duplicate local variable '', previo...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=202
Reported by @grubba
```
Pike v7.1 release 6 running Hilfe v2.0 (Incremental Pike Frontend)
> class { void foo(int, string); };
-:1:Duplicate local variable '', previous declaration on line 1
-:1:Class definition failed.
```
See also lyskom://kom.idonex.se/216307/.Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/199Image.Colortable doesn't find all colors on IA64.2020-02-29T17:57:14ZPeter BortasImage.Colortable doesn't find all colors on IA64.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=199
Reported by @grubba
```
Pike v7.1 release 6 running Hilfe v2.0 (Incremental Pike Frontend)
> object img=Image.Image(100,100)->test(99)->scale(400,100);
Result: object...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=199
Reported by @grubba
```
Pike v7.1 release 6 running Hilfe v2.0 (Incremental Pike Frontend)
> object img=Image.Image(100,100)->test(99)->scale(400,100);
Result: object
> object c=Image.Colortable(img,256);
Result: object
> img=c*img;
Result: object
> sizeof(c);
Result: 256
> object c2=Image.Colortable(img,256);
Result: object
> sizeof(c2);
Result: 246
> c2*img == img;
Result: 0
```
This behaviour causes the BMP testsuite to fail on Digital Unix.Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/193backend.c has a select which doesn't work on NetBSD-1.52009-04-16T14:11:39ZPeter Bortasbackend.c has a select which doesn't work on NetBSD-1.5Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=193
Reported by Hanno Wagner, Xlink Internet Service GmbH <wagner@fitug.de>
I tried to use Roxen-2.0r3 on NetBSD, but it fails as soon as I try
to connect to the Administ...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=193
Reported by Hanno Wagner, Xlink Internet Service GmbH <wagner@fitug.de>
I tried to use Roxen-2.0r3 on NetBSD, but it fails as soon as I try
to connect to the Administration Interface, after the password the
process (more or less) freezes.
After looking with gdb, I got the following:
```
Program received signal SIGINT, Interrupt.
0x483bf747 in select ()
(gdb) bt
#0 0x483bf747 in select ()
#1 0x80807c4 in backend ()
at /usr/local/src/roxen/roxen2.0r3/pike/src/backend.c:688
#2 0x80a0d5d in main (argc=9, argv=0xbfbfda14)
at /usr/local/src/roxen/roxen2.0r3/pike/src/main.c:527
#3 0x8055c61 in ___start ()
```
in backend.c I find:
```
#ifdef HAVE_POLL
{
int msec = (next_timeout.tv_sec*1000) + next_timeout.tv_usec/1000;
i = poll(active_poll_fds, active_num_in_poll, msec);
}
#else
/* FIXME: OOB? */
i = fd_select(max_fd+1, &rset, &wset,
#ifdef WITH_OOB
&eset,
#else /* !WITH_OOB */
0,
endif /* WITH_OOB */
&next_timeout);
#endif
```
and so on.
Does anyone know why there is a problem on NetBSD?Pike 7.0https://git.lysator.liu.se/pikelang/pike/-/issues/185Core dump2009-04-16T14:11:39ZPeter BortasCore dumpImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=185
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
gdb dump in attachment (if it still works).Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=185
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
gdb dump in attachment (if it still works).Pike 7.0Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/184Windows efuns/constants/functions not documented.2009-04-16T14:11:39ZPeter BortasWindows efuns/constants/functions not documented.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=184
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
Windows specific efuns (and functions) like RegGetValue is not present in
the manual.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=184
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
Windows specific efuns (and functions) like RegGetValue is not present in
the manual.Pike 7.4Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/182Pike fails to create backend wakup pike errno 10014 (WSAEFAULT)2009-04-16T14:11:39ZPeter BortasPike fails to create backend wakup pike errno 10014 (WSAEFAULT)Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=182
Reported by @grubba
This seems to occurr with Pike 7.1 on both NT4/IA32 and Windows 2000/IA64.
Pike 7.0 works fine on NT4/IA32.
errno 10014 is WSAEFAULT.
Pike compil...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=182
Reported by @grubba
This seems to occurr with Pike 7.1 on both NT4/IA32 and Windows 2000/IA64.
Pike 7.0 works fine on NT4/IA32.
errno 10014 is WSAEFAULT.
Pike compiled with FD_DEBUG and SOCKETPAIR_DEBUG:
```
H:\grubba\src\Pike7.1\build\ia64>pike
Using WinSock 2.0
New socket: 3 (1920)
bind on 3 (1920)
bind returned 0 (0)
getsockname on 3 (1920)
getsockname returned 0 (0)
listen on 3 (1920)
listen returned 0 (0)
ioctl(3 (1920,-2147195266,000000000012FB88)
ioctlsocket returned 0 (0)
New socket: 4 (1912)
connect on 4 (1912)
02 00 04 9d 7f 00 00 01 00 00 00 00 00 00 00 00
connect returned 0 (0)
Accept on 3 (1920)..
Accept failed with errno 10014
my_socketpair:fd_accept() failed, errno:10014 (2)
Closing 4 (1912)
4 (1912) closed
\grubba\src\Pike7.1\src\backend.c:250: Fatal error:
Couldn't create backend wakup pipe! errno=10014.
No stack - no backtrace.
```
This application has requested the Runtime to terminate it in an unusual
way.
Please contact the application's support team for more information.
H:\grubba\src\Pike7.1\build\ia64>Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/177sprshd sometimes doesn't detect termination correctly2009-04-16T14:11:39ZPeter Bortassprshd sometimes doesn't detect termination correctlyImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=177
Reported by @grubba
When the spawned tpike dies with:
```
/home/build/autobuild/nt4_build/pike7.1-20000831142409.tar/build/windows_nt-x86/tpike
-DNOT_INSTALLED
-m/ho...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=177
Reported by @grubba
When the spawned tpike dies with:
```
/home/build/autobuild/nt4_build/pike7.1-20000831142409.tar/build/windows_nt-x86/tpike
-DNOT_INSTALLED
-m/home/build/autobuild/nt4_build/pike7.1-20000831142409.tar/build/windows_nt-x86/master.pike
/home/build/autobuild/nt4_build/pike7.1-20000831142409.tar/bin/mktreeopt.pike
/home/build/autobuild/nt4_build/pike7.1-20000831142409.tar/src/treeopt.in
|| ( mv
/home/build/autobuild/nt4_build/pike7.1-20000831142409.tar/build/windows_nt-x86/tpike
/home/build/autobuild/nt4_build/pike7.1-20000831142409.tar/build/windows_nt-x86/tpike.bug
; exit 1 )
\build\autobuild\nt4_build\pike7.1-20000831142409.tar\src\backend.c:250:
Fatal error:
Couldn't create backend wakup pipe! errno=10014.
No stack - no backtrace.
```
sprshd doesn't signal it correctly to the client, so the client hangs in a
read(4):
```
# /usr/proc/bin/ptree 3872
3872 /bin/bash ./megalon_nt4.sh_
3873 pike nt4.pike
9631 /bin/sh -c make > output_make 2>&1
9632 make
10274 /bin/sh -c builddir="build/`uname -srm|sed -e 's/ /-/g'|tr
'[A-Z]
10309 make MAKE=make MAKE_PARALLEL= all
23121 /bin/sh -c make pike
23122 make pike
23323 /bin/sh -c make las.o builtin_functions.o peep.o
23324 make las.o builtin_functions.o peep.o
23484 /bin/sh -c
/home/build/autobuild/nt4_build/pike7.1-20
23485 /usr/local/bin/pike
/home/build/autobuild/nt4_build
# truss -p 23485
read(3, 0x002FCCEC, 4) (sleeping...)
signotifywait() (sleeping...)
lwp_sema_wait(0xFEF0DE30) (sleeping...)
lwp_cond_wait(0xFF3A55B0, 0xFF3A55C0, 0xFF045BF0) (sleeping...)
door_return(0x00000000, 0, 0x00000000, 0) (sleeping...)
^C
```Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/172sqltable "ascii" shows wrong number of columns.2009-04-16T14:11:39ZPeter Bortassqltable "ascii" shows wrong number of columns.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=172
Reported by Joachim Nyman Silverstig, Idonex AB <jocke@roxen.com>
When I put ascii="ascii" attribute in the sqltable tag, it only returns 1
column instead of three, a...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=172
Reported by Joachim Nyman Silverstig, Idonex AB <jocke@roxen.com>
When I put ascii="ascii" attribute in the sqltable tag, it only returns 1
column instead of three, although it returns three ok in HTML table mode
Here are the examples:
==========================================================================
<sqltable query="select empno, ename, floor(sal) as INTSAL from scott.emp
order by ename" />
works OK, returning a titled HTML table of all three columns
==========================================================================
<sqltable ascii="ascii" query="select empno, ename, floor(sal) as INTSAL
from scott.emp order by ename" />
returns just the ename column.
Both of these give a </sqltable> in the output.
scott.emp is the standard Oracle example table owned by scott with all the
employees in, installed by default with Oracle.
Here is the description -
SQL> desc scott.emp
Name Null? Type
----------------------------------------- -------- --------------------
EMPNO NOT NULL NUMBER(4)
ENAME VARCHAR2(10)
JOB VARCHAR2(9)
MGR NUMBER(4)
HIREDATE DATE
SAL NUMBER(7,2)
COMM NUMBER(7,2)
DEPTNO NUMBER(2)Pike 7.0https://git.lysator.liu.se/pikelang/pike/-/issues/168Filesystem()->stat()->islnk() doesn't detect links.2009-04-16T14:11:39ZPeter BortasFilesystem()->stat()->islnk() doesn't detect links.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=168
Reported by Mikael Brandström <mikael@brandstrom.org>
```
Current behavior:
Pike v7.1 release 6 running Hilfe v2.0 (Incremental Pike Frontend)
> Filesystem.System("/"...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=168
Reported by Mikael Brandström <mikael@brandstrom.org>
```
Current behavior:
Pike v7.1 release 6 running Hilfe v2.0 (Incremental Pike Frontend)
> Filesystem.System("/")->stat("opt")->islnk();
Result: 0
>
After applying patch below:
Pike v7.1 release 6 running Hilfe v2.0 (Incremental Pike Frontend)
> Filesystem.System("/")->stat("opt")->islnk();
Result: 1
>
```
I'm not sure if this patch breakes something that's dependant on
Filesystem()->find() following symlinks. Anyway, currently find would do an
endless recursion on circular symlinks.
```
--- lib/modules/Filesystem.pmod/System.pike Sun Aug 27 20:37:21 2000
+++ /opt/pike/pike/7.1.6/lib/modules/Filesystem.pmod/System.pike Thu
Aug 31 00:59:47 2000
@@ -74,7 +74,7 @@
Stdio.Stat a;
string full = combine_path(wd, file);
- if((a = file_stat("/"+root+full)))
+ if((a = file_stat("/"+root+full,1)))
{
Filesystem.Stat s = Filesystem.Stat();
s->fullpath = full;
```Pike 7.0Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/164No Servlet support found in Pike2009-04-16T14:11:39ZPeter BortasNo Servlet support found in PikeImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=164
Reported by Saeed Balevasi, Roxen AB <saeed@roxen.com>
No Servlet support found in Pike.
This massege I get when I start the server.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=164
Reported by Saeed Balevasi, Roxen AB <saeed@roxen.com>
No Servlet support found in Pike.
This massege I get when I start the server.Pike 7.0https://git.lysator.liu.se/pikelang/pike/-/issues/150Gmp linking problems2009-04-16T14:11:39ZPeter BortasGmp linking problemsImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=150
Reported by @grubba
```
From: Niels Möller <nisse@lysator.liu.se>
To: pike@roxen.com
Date: 29 Aug 2000 23:07:56 +0200
Subject: Linking problem with Pike-7.0.36
```
I...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=150
Reported by @grubba
```
From: Niels Möller <nisse@lysator.liu.se>
To: pike@roxen.com
Date: 29 Aug 2000 23:07:56 +0200
Subject: Linking problem with Pike-7.0.36
```
I there's a bug in the Gmp module in Pike-7.0.36. Its definition of
__MPN in src/modules/Gmp/my_gmp.h is bogus, and doesn't work at all
with gmp-3.x.
I just had a look at Pike-7.0.78 as well. The definition is still
present, although the code tries to avoid using mpn_popcount (and uses
mpz_popcount instead if that function exists).
Anyway, the symptoms are a runtime link arror as soon as popcount is
used; the Gmp.so has a reference to __mpn_popcount, while the symbol
name in the library is __gmpn_popcount (from gmp-3.x).
Furthermore, the behavior of popcount(x) when x < 0 seems not to be
well defined. In pike-7.0.36 it is an error, while in pike-7.0.78 it
returns -1 or UMAX_LONG casted to a pike integer.
/NielsPike 7.0Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/149GC fatal: An unreferenced thing got missed by gc_is_referenced().2009-04-16T14:11:39ZPeter BortasGC fatal: An unreferenced thing got missed by gc_is_referenced().Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=149
Reported by @grubba
From AutoBuild:
```
/private/tmp/autobuild/pike7.1-20000829204108.tar/src/gc.c:875: GC fatal:
**An unreferenced thing got missed by gc_is_referen...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=149
Reported by @grubba
From AutoBuild:
```
/private/tmp/autobuild/pike7.1-20000829204108.tar/src/gc.c:875: GC fatal:
**An unreferenced thing got missed by gc_is_referenced().
**Location: 0xd3c744 Type: object Refs: 1
**Got gc marker at 0x1151dc4: flags=0x00091e, refs=1, weak=-1, xrefs=0,
saved=1, frame=0xc2448c [back=0xc24474, prev=0xffffffff, next=0xffffffff,
data=0xd3c744, cycle=0, flags=0x39]
**Parent identifier: 8
**Program id: 207736
**The object is destructed.
**Attempting to describe program object was instantiated from:
**Got gc marker at 0x1143044: flags=0x000101, refs=2, weak=0, xrefs=0,
saved=3, frame=0x0
**Program id: 207736
**Location:
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1173
**There is no parent (any longer?)
*******************
/private/tmp/autobuild/pike7.1-20000829204108.tar/src/gc.c:2139: Fatal
error:
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1266:
mark 68, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1266:
call lfun & pop(7) 68, 2
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1169:
mark 69, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1162:
global(2) 69, 2
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1169:
call(0) 70, 2
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1169:
& local(0) 70, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1169:
push 0 72, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1169:
branch(0) 73, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1169:
foreach(0) 73, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1169:
return 0 73, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1267:
mark 68, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1162:
global(2) 68, 2
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1267:
call(0) 69, 2
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1267:
& local(0) 69, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1267:
push 0 71, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1267:
branch(0) 72, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1267:
foreach(0) 72, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1267:
pop_n_elems(4) 72, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1270:
mark 68, 1
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1270:
global(6) 68, 2
/private/tmp/autobuild/pike7.1-20000829204108.tar/build/mac-os-10.0-power-macintosh/test-install/pike/7.1.6/lib/modules/Yabu.pmod/module.pmod:1270:
call(1) 69, 2
Pike was in GC stage 510 when this fatal occured:
Fatal in garbage collector.
Attempting to dump backlog (may fail)...
Backtrace at time of fatal:
testsuite: Test 9375 (shift 0):1: a()
/private/tmp/autobuild/pike7.1-20000829204108.tar/bin/test_pike.pike:568:
main(3,({"/private/tmp/autobuild/pike7.1-20000829204108.tar/bin/test_pike.pike","modules/_Charset/module_testsuite","modules/_Crypto/module_testsuite","modules/_Image_JPEG/module_testsuite",,,34}))
make[2]: *** [verify] Abort trap
```Pike 7.2https://git.lysator.liu.se/pikelang/pike/-/issues/138Compilation error in Pike 7.02009-04-16T14:11:39ZPeter BortasCompilation error in Pike 7.0Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=138
Reported by Hans-Göran Wendelhof, Roxen Internet Software <hg@roxen.com>
Följande kompileringsfel uppkommer i en färsk cvs-version av Pike 7.0:
```
...
Creating lib...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=138
Reported by Hans-Göran Wendelhof, Roxen Internet Software <hg@roxen.com>
Följande kompileringsfel uppkommer i en färsk cvs-version av Pike 7.0:
```
...
Creating lib/.
Creating lib/modules/.
Creating lib/include/.
Creating symlink ./share
Done.
t="$PIKE_PATH_TRANSLATE"; if test "x$t" = "x"; then t=s,x,x,; else :; fi; \
sed -e 's!¤lib_prefix¤!'"`echo
\"/home/distmaker/Distmaker/2.1/build/pike.srcbuild/lib\" | sed -e
\"$t\"`"'!' </home/distmaker/Distmaker/2.1/build/pike/lib/master.pike.in \
| sed -e 's!¤share_prefix¤!'"`echo
\"/home/distmaker/Distmaker/2.1/build/pike/lib\" | sed -e \"$t\"`"'!'
>master.pike
make[3]: Entering directory
`/home/distmaker/Distmaker/2.1/build/pike.srcbuild/modules'
Making CommonLog
make[4]: Entering directory
`/home/distmaker/Distmaker/2.1/build/pike.srcbuild/modules/CommonLog'
Compiling
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/clf.c
In file included from
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../global.h:288,
from
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/clf.c:2:
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../port.h:52:
parse error before `TIME'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../port.h:52:
parse error before `*'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../port.h:52:
warning: type defaults to `int' in declaration of `TIME'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../port.h:52:
warning: data definition has no type or storage class
In file included from
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../global.h:288,
from
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/clf.c:2:
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../port.h:
In function `EXTRACT_CHAR':
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../port.h:180:
warning: comparison is always false due to limited range of data type
In file included from
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../pike_memory.h:14,
from
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../pike_macros.h:19,
from
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../fdlib.h:8,
from
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/clf.c:4:
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:
At top level:
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:153:
parse error before `0'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:153:
parse error before `('
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:153:
warning: type defaults to `int' in declaration of `to'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:153:
parse error before `const'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:153:
parse error before `0'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:153:
parse error before `1'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:153:
warning: type defaults to `int' in declaration of `PIKE_CONCAT'
/home/distmaker/Distmaker/2.1/build/pike.srcbuild/../pike/src/modules/CommonLog/../../stralloc.h:153:
warning: type defaults to `int' in declaration of `from'
...
```Pike 7.0