dbTalk Databases Forums  

SE on Red Hat 9

comp.databases.informix comp.databases.informix


Discuss SE on Red Hat 9 in the comp.databases.informix forum.



Reply
 
Thread Tools Search this Thread Display Modes
  #1  
Old   
Scott Burns
 
Posts: n/a

Default SE on Red Hat 9 - 06-24-2003 , 07:30 PM






Hi,
this is a slightly modified repost of something I sent to the IIUG
Linux forum.

As the subject suggests I'm trying to deploy SE 7.2 on Red Hat 9
using Client SDK 2.80 or 2.81. The significant problem is that glibc
changed the way __ctype_b, __ctype_tolower and __ctype_toupper actually
funciton ( I assume for internationalisation ).

The way they used to work was that they were global variables you
could reference into. They way these work now is that you have the
functions __ctype_b_loc, __ctype_tolower_loc and __ctype_toupper_loc
which return int arrays to represent these variables. Glibc 2.3.2
contains versioning symbols for the old style functions but I have no
idea how to reference these. They appear through nm as <symbol>@GLIBC_2.0

All this is a little low level - it appears Informix reach into the
guts of glibc rather than going through the front end, and now I'm
paying the penalty...

I have got programs to work compiled under RH9 by providing a
compatibility .c to provide the old style globals and compiling
statically. I cannot compile dynamically as the program errors with a
-23101 which in our case means that when it ran my locale through the
glib ctype toupper or tolower functions. The result was garbage and
thus I can't connect. I assume that when I compile statically my
versions of the globals take precedence. When I compile dynamically the
inbuilt ones, which don't function quite right in this case, take
precedence.

If I compile under RH7.2 I usually get the programs hanging in the
function futex, but sometimes core dumping elsewhere.

So, my questions are:
1) Has anyone else managed to compile SE applications on RH9?
2) Am I likely to have more luck with a later version of SE?
3) Does anyone know how to compile applications to use the old glibc
2.0 compatibility functions?
4) Is SE being actively developed by IBM, or is it just being left
alone until enough emails come in and they then assign someone to fix
the latest problem?

Thanks for your time.

Scott Burns
Mirrabooka Systems


Reply With Quote
  #2  
Old   
Michael Krzepkowski
 
Posts: n/a

Default Re: SE on Red Hat 9 - 06-24-2003 , 08:37 PM






Scott Burns wrote:

Quote:
Hi,
this is a slightly modified repost of something I sent to the IIUG
Linux forum.

As the subject suggests I'm trying to deploy SE 7.2 on Red Hat 9
using Client SDK 2.80 or 2.81. The significant problem is that glibc
changed the way __ctype_b, __ctype_tolower and __ctype_toupper
actually funciton ( I assume for internationalisation ).
The way they used to work was that they were global variables you
could reference into. They way these work now is that you have the
functions __ctype_b_loc, __ctype_tolower_loc and __ctype_toupper_loc
which return int arrays to represent these variables. Glibc 2.3.2
contains versioning symbols for the old style functions but I have no
idea how to reference these. They appear through nm as
symbol>@GLIBC_2.0
All this is a little low level - it appears Informix reach into the
guts of glibc rather than going through the front end, and now I'm
paying the penalty...

I have got programs to work compiled under RH9 by providing a
compatibility .c to provide the old style globals and compiling
statically. I cannot compile dynamically as the program errors with a
-23101 which in our case means that when it ran my locale through the
glib ctype toupper or tolower functions. The result was garbage and
thus I can't connect. I assume that when I compile statically my
versions of the globals take precedence. When I compile dynamically
the inbuilt ones, which don't function quite right in this case, take
precedence.

If I compile under RH7.2 I usually get the programs hanging in the
function futex, but sometimes core dumping elsewhere.

So, my questions are:
1) Has anyone else managed to compile SE applications on RH9?
2) Am I likely to have more luck with a later version of SE?
3) Does anyone know how to compile applications to use the old glibc
2.0 compatibility functions?
4) Is SE being actively developed by IBM, or is it just being left
alone until enough emails come in and they then assign someone to fix
the latest problem?

Thanks for your time.

Scott Burns
Mirrabooka Systems

FYI: I have SE 7.25 running under RH 8.0 fully patched. My application
only used isql not esql/c so I cannot comment
on libraries. However, I had lots of problems with 7.3x isql and
ended-up using ancient 7.20.
The engine itself is OK.

Michael



Reply With Quote
Reply




Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off



Powered by vBulletin Version 3.5.3
Copyright ©2000 - 2009, Jelsoft Enterprises Ltd.