Bug #60130 for DBD-DB2: get SQL_INVALID_HANDLE when $dbh->disconect()

Fri Aug 06 02:33:33 2010 pinkee999 [...] gmail.com - Ticket created

Subject:

get SQL_INVALID_HANDLE when $dbh->disconect()

Hello, I faced a problem in disconnect. I can't free the handle. My environment is as follows: ----------------- OS: AIX level: 6100-04 Perl: v5.8.8 built for aix-thread-multi DBI: DBI-1.609 DBD: DBD-DB2-1.79 DB: DB2 version: server:AIX64 9.7.2, client:AIX64 9.7.1 ----------------- sample script: ----------------- use DBI; use strict; # Open a connection my $dbh = DBI->connect("dbi:DB2:testdb", "test","test",{RaiseError => 1}); # use VALUES to retrieve value from special register my $stmt = "Values CURRENT DATE"; my $sth = $dbh->prepare($stmt); $sth->execute(); # associate variables with output columns... my $col1; $sth->bind_col(1,\$col1); while ($sth->fetch) { print "Today is: $col1\n"; } $sth->finish(); $dbh->disconnect() || die $DBI::errstr; ----------------- result: ----------------- # export DBI_TRACE=9=dbitrace.log # perl -w -f db2Connect.pl Today is: 2010-08-06 Invalid Handle Passed to retrieve Diagnostic Information at db2Connect.pl line 14. # ----------------- How can I solve this problem? Regards, Hiroyuki

Sat Aug 07 12:43:59 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Hello, On Solaris Environment, I faced the same problem, but it is okay on Linux Environment. I attached files of cli-trace and dbi-trace on Solaris Env. ----------------- OS: Solaris uname -a: SunOS xxx 5.10 Generic_142901-07 i86pc i386 i86pc Perl: v5.8.4 built for i86pc-solaris-64int DBI: DBI-1.611 DBD: DBD-DB2-1.79 DB: DB2 version: server: LINUXX8664 9.7.1, client:Solaris x86 64 9.7.1 ------------ OS: RHEL uname -a: Linux xxx 2.6.18-128.el5 #1 SMP Wed Dec 17 11:41:38 EST 2008 x86_64 x86_64 x86_64 GNU/Linux Perl: v5.8.8 built for x86_64-linux-thread-multi DBI: DBI-1.611 DBD: DBD-DB2-1.79 DB: DB2 version: server: LINUXX8664 9.7.1, client:Solaris x86 64 9.7.1 ----------------- Regards, Hiroyuki On Fri Aug 06 02:33:33 2010, pinkee wrote: Show quoted text

> Hello, > > I faced a problem in disconnect. > I can't free the handle. > > My environment is as follows: > ----------------- > OS: AIX > level: 6100-04 > Perl: v5.8.8 built for aix-thread-multi > DBI: DBI-1.609 > DBD: DBD-DB2-1.79 > DB: DB2 > version: server:AIX64 9.7.2, client:AIX64 9.7.1 > ----------------- > > sample script: > ----------------- > use DBI; > use strict; > # Open a connection > my $dbh = DBI->connect("dbi:DB2:testdb", "test","test",{RaiseError => 1}); > # use VALUES to retrieve value from special register > my $stmt = "Values CURRENT DATE"; > my $sth = $dbh->prepare($stmt); > $sth->execute(); > # associate variables with output columns... > my $col1; > $sth->bind_col(1,\$col1); > while ($sth->fetch) { print "Today is: $col1\n"; } > $sth->finish(); > $dbh->disconnect() || die $DBI::errstr; > ----------------- > > result: > ----------------- > # export DBI_TRACE=9=dbitrace.log > # perl -w -f db2Connect.pl > Today is: 2010-08-06 > Invalid Handle Passed to retrieve Diagnostic Information at > db2Connect.pl line 14. > # > ----------------- > > How can I solve this problem? > > > Regards, > Hiroyuki

Subject:

dbitrace.log

Download dbitrace.log
application/octet-stream 11.5k

Message body not shown because it is not plain text.

Subject:

p2732t1.cli.txt

Message body is not shown because it is too large.

Mon Aug 09 04:33:47 2010 opendev [...] us.ibm.com - Correspondence added

Hi Hiroyuki, Given that CLI trace shows a exit without any errors, we will need to check from where is the error being emitted. Hence, could you apply the patch attached and let me know the output. This patch would tell us if the call to SQLGetDiagrec was made from disconnect method or not. With the info we could try diagnosing further. Also I see that perl version on Solaris is 5.8.4 and Linux is 5.8.8. Though it should not matter could you upgrade your Perl on Solaris box and check what is the behaviour? -- Thanks Praveen IBM OpenSource Application Development Team India Software Labs, Bangalore (India)

Subject:

debug_patch.diff

--- perl/DBD-DB2/dbdimp.c version 1.79 Wed Jul 14 10:04:46 2010 +++ perl/DBD-DB2-1.79/dbdimp.c patched version 1.79 Mon Aug 09 13:36:10 2010 @@ -51,10 +51,16 @@ if(DBIc_TRACE_LEVEL(imp_xxh) >=3){ case SQL_SUCCESS_WITH_INFO: case SQL_NO_DATA_FOUND: + PerlIO_printf( DBILOGFP, + "For trace level > 3 - rc = %d\n", + rc ); setErrorFromDiagRecInfo(perlHandle, handleType, handle, ""); break; } case SQL_ERROR: + PerlIO_printf( DBILOGFP, + "For condition SQL_ERROR - rc =%d\n", + rc ); setErrorFromDiagRecInfo(perlHandle, handleType, handle, Nullch); break; case SQL_INVALID_HANDLE: @@ -644,6 +650,9 @@ SQLRETURN ret; ret = SQLDisconnect(imp_dbh->hdbc); + PerlIO_printf( DBILOGFP, + "return code after SQLDisconnect %d\n", + ret ); CHECK_ERROR(dbh, SQL_HANDLE_DBC, imp_dbh->hdbc, ret, "Disconnect Failed"); EOI(ret); @@ -654,6 +663,9 @@ DBIc_ACTIVE_off(imp_dbh); ret = SQLFreeHandle( SQL_HANDLE_DBC, imp_dbh->hdbc ); + PerlIO_printf( DBILOGFP, + "return code from Freeing Connect Handle %d\n", + ret ); CHECK_ERROR(dbh, SQL_HANDLE_DBC, imp_dbh->hdbc, ret, "Free Connect Failed"); EOI(ret); @@ -666,11 +678,16 @@ imp_drh->svNUM_OF_FIELDS = NULL; } ret = SQLFreeHandle( SQL_HANDLE_ENV, imp_drh->henv ); + PerlIO_printf( DBILOGFP, + "return code from Freeing ENV Handle %d\n", + ret ); CHECK_ERROR(dbh, SQL_HANDLE_ENV, imp_drh->henv, ret, "Free HENV Failed"); EOI(ret); imp_drh->henv = SQL_NULL_HENV; } - + PerlIO_printf( DBILOGFP, + "Returning from disconnect %d\n", + ret ); /* We don't free imp_dbh since a reference still exists */ /* The DESTROY method is the only one to 'free' memory. */ /* Note that statement objects may still exist for this dbh! */

Mon Aug 09 04:33:47 2010 The RT System itself - Status changed from 'new' to 'open'

Mon Aug 09 07:32:37 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Dear Praveen-san, Thank you for your debug patch. I patched and got trace logs. Please see those. First of all, I faced this problem on AIX 6.1 with perl 5.8.8. Regards, Hiroyuki On Mon Aug 09 04:33:47 2010, IBMTORDB2 wrote: Show quoted text

> Hi Hiroyuki, > > Given that CLI trace shows a exit without any errors, we will need to > check from where is the error being emitted. Hence, could you apply the > patch attached and let me know the output. This patch would tell us if > the call to SQLGetDiagrec was made from disconnect method or not. With > the info we could try diagnosing further. > > Also I see that perl version on Solaris is 5.8.4 and Linux is 5.8.8. > Though it should not matter could you upgrade your Perl on Solaris box > and check what is the behaviour?

Subject:

dbitrace.log

Download dbitrace.log
application/octet-stream 11.7k

Message body not shown because it is not plain text.

Subject:

p2714t1.cli

Download p2714t1.cli
application/octet-stream 13.2k

Message body not shown because it is not plain text.

Tue Aug 10 11:54:13 2010 opendev [...] us.ibm.com - Correspondence added

Hi Hiroyuki, Thanks for the log. Although the CLI trace shows a SQL_SUCCESS after freeing ENV handle a -1 (SQL_ERROR) [log in dbi trace: "return code from Freeing ENV Handle -1"] is been received in the driver, hence the error during disconnect is seen. To narrow down the problem, I have added one more log statement and have initialised the "ret" variable to 0, to check if this is DBD-DB2 problem or a problem in CLI. Could you apply the new patch attached and let me know the dbi trace. If this turns out to be a CLI problem then I will take this up with the CLI team and check what could have gone wrong. -- Thanks Praveen IBM OpenSource Application Development Team India Software Labs, Bangalore (India)

Subject:

debug_patch2.diff

--- perl/DBD-DB2/dbdimp.c version 1.79 Wed Jul 14 10:04:46 2010 +++ perl/DBD-DB2-1.79/dbdimp.c patched version 1.79 Tue Aug 10 21:05:41 2010 @@ -51,10 +51,16 @@ if(DBIc_TRACE_LEVEL(imp_xxh) >=3){ case SQL_SUCCESS_WITH_INFO: case SQL_NO_DATA_FOUND: + PerlIO_printf( DBILOGFP, + "For trace level > 3 - rc = %d\n", + rc ); setErrorFromDiagRecInfo(perlHandle, handleType, handle, ""); break; } case SQL_ERROR: + PerlIO_printf( DBILOGFP, + "For condition SQL_ERROR - rc =%d\n", + rc ); setErrorFromDiagRecInfo(perlHandle, handleType, handle, Nullch); break; case SQL_INVALID_HANDLE: @@ -644,6 +650,9 @@ SQLRETURN ret; ret = SQLDisconnect(imp_dbh->hdbc); + PerlIO_printf( DBILOGFP, + "return code after SQLDisconnect %d\n", + ret ); CHECK_ERROR(dbh, SQL_HANDLE_DBC, imp_dbh->hdbc, ret, "Disconnect Failed"); EOI(ret); @@ -654,6 +663,9 @@ DBIc_ACTIVE_off(imp_dbh); ret = SQLFreeHandle( SQL_HANDLE_DBC, imp_dbh->hdbc ); + PerlIO_printf( DBILOGFP, + "return code from Freeing Connect Handle %d\n", + ret ); CHECK_ERROR(dbh, SQL_HANDLE_DBC, imp_dbh->hdbc, ret, "Free Connect Failed"); EOI(ret); @@ -665,12 +677,23 @@ SvREFCNT_dec( imp_drh->svNUM_OF_FIELDS ); imp_drh->svNUM_OF_FIELDS = NULL; } + PerlIO_printf( DBILOGFP, + "return code Before Freeing ENV Handle %d\n", + ret ); + + ret = 0; ret = SQLFreeHandle( SQL_HANDLE_ENV, imp_drh->henv ); + + PerlIO_printf( DBILOGFP, + "return code from Freeing ENV Handle %d\n", + ret ); CHECK_ERROR(dbh, SQL_HANDLE_ENV, imp_drh->henv, ret, "Free HENV Failed"); EOI(ret); imp_drh->henv = SQL_NULL_HENV; } - + PerlIO_printf( DBILOGFP, + "Returning from disconnect %d\n", + ret ); /* We don't free imp_dbh since a reference still exists */ /* The DESTROY method is the only one to 'free' memory. */ /* Note that statement objects may still exist for this dbh! */

Tue Aug 10 17:44:15 2010 pinkee999 [...] gmail.com - Correspondence added

Subject:	Re: [rt.cpan.org #60130] get SQL_INVALID_HANDLE when $dbh->disconect()
Date:	Wed, 11 Aug 2010 05:35:02 +0900
To:	bug-DBD-DB2 [...] rt.cpan.org
From:	Hiroyuki Ujiie <pinkee999 [...] gmail.com>

Dear Praveen-san, Thank you for your cooperation. I attach logs. I can't see cpan.org this morning, I will contribute it later. I extract a few lines of dbitrace.log as below: ------ Show quoted text

>> disconnect DISPATCH (DBI::db=HASH(0x81f3e1c) rc1/1 @1 g0

ima10c01 pid#2643) at - line 14 -> disconnect for DBD::DB2::db (DBI::db=HASH(0x81f3e1c)~0x81f3d98) return code after SQLDisconnect 0 return code from Freeing Connect Handle 0 return code Before Freeing ENV Handle 0 return code from Freeing ENV Handle -1 For condition SQL_ERROR - rc =-1 !! info: '' 'Invalid Handle Passed to retrieve Diagnostic Information' (err#0) <- disconnect= '' at - line 14 ------ Sincerely, Hiroyuki (2010/08/11 0:54), IBM OpenDev via RT wrote: Show quoted text

> <URL: https://rt.cpan.org/Ticket/Display.html?id=60130> > > Hi Hiroyuki, > > Thanks for the log. Although the CLI trace shows a SQL_SUCCESS after > freeing ENV handle a -1 (SQL_ERROR) [log in dbi trace: "return code from > Freeing ENV Handle -1"] is been received in the driver, hence the error > during disconnect is seen. > > To narrow down the problem, I have added one more log statement and have > initialised the "ret" variable to 0, to check if this is DBD-DB2 problem > or a problem in CLI. Could you apply the new patch attached and let me > know the dbi trace. > > If this turns out to be a CLI problem then I will take this up with the > CLI team and check what could have gone wrong. >

Download 20100810.zip
application/octet-stream 5.6k

Message body not shown because it is not plain text.

Wed Aug 11 00:43:08 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Dear Praveen-san, I added log file of perl5.8.8, AIX. It seems to the same. Regards, Hiroyuki On Tue Aug 10 17:44:15 2010, pinkee wrote: Show quoted text

> Dear Praveen-san, > > Thank you for your cooperation. > I attach logs. > I can't see cpan.org this morning, I will contribute it later. > > I extract a few lines of dbitrace.log as below: > ------

> >> disconnect DISPATCH (DBI::db=HASH(0x81f3e1c) rc1/1 @1 g0

> ima10c01 pid#2643) at - line 14 > -> disconnect for DBD::DB2::db (DBI::db=HASH(0x81f3e1c)~0x81f3d98) > return code after SQLDisconnect 0 > return code from Freeing Connect Handle 0 > return code Before Freeing ENV Handle 0 > return code from Freeing ENV Handle -1 > For condition SQL_ERROR - rc =-1 > !! info: '' 'Invalid Handle Passed to retrieve Diagnostic > Information' (err#0) > <- disconnect= '' at - line 14 > ------ > > Sincerely, > Hiroyuki > > (2010/08/11 0:54), IBM OpenDev via RT wrote:

> > <URL: https://rt.cpan.org/Ticket/Display.html?id=60130> > > > > Hi Hiroyuki, > > > > Thanks for the log. Although the CLI trace shows a SQL_SUCCESS after > > freeing ENV handle a -1 (SQL_ERROR) [log in dbi trace: "return code from > > Freeing ENV Handle -1"] is been received in the driver, hence the error > > during disconnect is seen. > > > > To narrow down the problem, I have added one more log statement and have > > initialised the "ret" variable to 0, to check if this is DBD-DB2 problem > > or a problem in CLI. Could you apply the new patch attached and let me > > know the dbi trace. > > > > If this turns out to be a CLI problem then I will take this up with the > > CLI team and check what could have gone wrong. > >

>

Subject:

dbitrace_AIX.log

Download dbitrace_AIX.log
application/octet-stream 12.9k

Message body not shown because it is not plain text.

Mon Aug 16 00:45:46 2010 opendev [...] us.ibm.com - Correspondence added

Hi Hiroyuki, Looks like I will need the CLI teams help, as the return code is not the same as shown in CLI trace. Will contact the CLI team and get back to you on how it goes. Let me up ahead tell you that this could take some time. Will try to get back to you with a solution as soon as possible. -- Thanks Praveen IBM OpenSource Application Development Team India Software Labs, Bangalore (India)

Mon Aug 16 02:42:16 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Hello Praveen, Thank you for your investigation. I patiently wait for resolving this problem. We are starting to change our program. Some program can't run properly with defective　free "ENV　Handle." Sincerely, Hiroyuki On Mon Aug 16 00:45:46 2010, IBMTORDB2 wrote: Show quoted text

> Hi Hiroyuki, > > Looks like I will need the CLI teams help, as the return code is not the > same as shown in CLI trace. Will contact the CLI team and get back to > you on how it goes. > > Let me up ahead tell you that this could take some time. Will try to get > back to you with a solution as soon as possible.

Mon Aug 16 03:56:02 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Hello Praveen, I succeed free "Env handle" on AIX environment. I compiled the perl 5.8.9 in 64-bit and used it. Last time, I tested with perl OS bundled. Perhaps it run in 32bit and Solaris's also run in 32bit. I attached dbitrace.log and outputs of each 'perl -V.' There is minor version difference, so I will try 5.8.8 version tomorrow. Regards, Hiroyuki On Mon Aug 16 02:42:16 2010, pinkee wrote: Show quoted text

> Hello Praveen, > > Thank you for your investigation. > I patiently wait for resolving this problem. > > We are starting to change our program. > Some program can't run properly with defective　free "ENV　Handle." > > Sincerely, > Hiroyuki > > On Mon Aug 16 00:45:46 2010, IBMTORDB2 wrote:

> > Hi Hiroyuki, > > > > Looks like I will need the CLI teams help, as the return code is not the > > same as shown in CLI trace. Will contact the CLI team and get back to > > you on how it goes. > > > > Let me up ahead tell you that this could take some time. Will try to get > > back to you with a solution as soon as possible.

Subject:

dbitrace_AIX64.log

Download dbitrace_AIX64.log
application/octet-stream 12.3k

Message body not shown because it is not plain text.

Subject:

perl_V_OS.out

Download perl_V_OS.out
application/octet-stream 2.2k

Message body not shown because it is not plain text.

Subject:

perl_V_64bit.out

Download perl_V_64bit.out
application/octet-stream 2.4k

Message body not shown because it is not plain text.

Tue Aug 17 00:29:24 2010 opendev [...] us.ibm.com - Correspondence added

Hi Hiroyuki, Looks like this was a problem due to architecture (32, 64) issues. We could easily verify if this was a architecture issue by checking for what architecture is the binaries built, using the following command. file <filename> Here you can check for the perl executable [<perl_install_dir>/bin/perl]and the DBD-DB2 [<perl_install_dir>/lib/site_perl/5.10.0/x86_64-linux-thread-multi/auto/DBD/DB2/DB2.so] binary file, if they have been built for the same architecture type. -- Thanks Praveen IBM OpenSource Application Development Team India Software Labs, Bangalore (India)

Tue Aug 17 01:36:45 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Hello Praveen, Thank you for your cool advice. The consistence between perl and DBD-DB2 module was broken. DB.so was depended 64bit library. Please see attached files. I'll try to fix it. Regards, Hiroyuki

Subject:

bad_file+ldd.txt

# file /usr/bin/perl /usr/bin/perl: executable (RISC System/6000) or object module not stripped # file /usr/opt/perl5/lib/site_perl/5.8.8/aix-thread-multi/auto/DBD/DB2/DB2.so /usr/opt/perl5/lib/site_perl/5.8.8/aix-thread-multi/auto/DBD/DB2/DB2.so: executable (RISC System/6000) or object module not stripped # ldd /usr/bin/perl /usr/bin/perl needs: /usr/opt/perl5/lib/5.8.8/aix-thread-multi/CORE/libperl.a(libperl.o) /usr/lib/libnsl.a(shr.o) /usr/lib/libdl.a(shr.o) /usr/lib/libpthreads.a(shr_xpg5.o) /usr/lib/threads/libc.a(shr.o) /usr/lib/libbsd.a(shr.o) /usr/lib/librtl.a(shr.o) /usr/lib/libbind.a(shr.o) /usr/lib/libcrypt.a(shr.o) /usr/lib/libthread.a(shr.o) /usr/lib/libpthreads_compat.a(shr.o) /usr/lib/libtli.a(shr.o) /usr/lib/libpthreads.a(shr_comm.o) /unix /usr/lib/libpthreads.a(shr.o) /usr/lib/threads/libc.a(pse.o) # ldd /usr/opt/perl5/lib/site_perl/5.8.8/aix-thread-multi/auto/DBD/DB2/DB2.so /usr/opt/perl5/lib/site_perl/5.8.8/aix-thread-multi/auto/DBD/DB2/DB2.so needs: /usr/lib/libpthreads.a(shr_xpg5.o) /usr/lib/libc.a(shr.o) /usr/opt/db2_09_07/lib64/libdb2.a(shr.o) /usr/lib/libpthreads.a(shr_comm.o) /unix /usr/lib/libcrypt.a(shr.o) /usr/opt/db2_09_07/lib64/libdb2osse.a(shr.o) /usr/opt/db2_09_07/lib64/libdb2locale.a(shr.o) /usr/opt/db2_09_07/lib64/libdb2g11n.a(shr.o) /usr/lib/libC.a(shr3.o) /usr/lib/libC.a(ansi_32.o) /usr/opt/db2_09_07/lib64/libdb2genreg.a(shr.o) /usr/lib/libC.a(shr2.o) /usr/lib/libC.a(shr.o) /usr/opt/db2_09_07/lib64/libdb2osse_db2.a(shr.o) /usr/opt/db2_09_07/lib64/libdb2install.a(shr.o) /usr/opt/db2_09_07/lib64/libdb2trcapi.a(shr.o) /usr/opt/db2_09_07/lib64/libdb2dascmn.a(shr.o) /usr/lib/libpthread.a(shr_xpg5.o) /usr/lib/libC.a(shrcore.o) /usr/lib/libC.a(ansicore_32.o) #

Subject:

good_file+ldd.txt

# file perl5/bin/perl perl5/bin/perl: 64-bit XCOFF executable or object module not stripped # file ./perl5/lib/site_perl/5.8.9/aix-thread-multi-64all/auto/DBD/DB2/DB2.so ./perl5/lib/site_perl/5.8.9/aix-thread-multi-64all/auto/DBD/DB2/DB2.so: 64-bit XCOFF executable or object module not stripped # ldd perl5/bin/perl perl5/bin/perl needs: /work/tmp/ujiie/local/perl5/lib/5.8.9/aix-thread-multi-64all/CORE/libperl.a(libperl.o) /usr/lib/libnsl.a(shr_64.o) /usr/lib/libpthreads.a(shr_xpg5_64.o) /usr/lib/libc.a(shr_64.o) /usr/lib/libbind.a(shr_64.o) /usr/lib/libdl.a(shr_64.o) /usr/lib/libcrypt.a(shr_64.o) /usr/lib/libbsd.a(shr_64.o) /usr/lib/libthread.a(shr_64.o) /usr/lib/libtli.a(shr_64.o) /usr/lib/libC.a(shr3_64.o) /usr/lib/libC.a(shr2_64.o) /unix /usr/lib/libc.a(pse_64.o) /usr/lib/libC.a(shrcore_64.o) /usr/lib/libC.a(ansicore_64.o) # ldd ./perl5/lib/site_perl/5.8.9/aix-thread-multi-64all/auto/DBD/DB2/DB2.so ./perl5/lib/site_perl/5.8.9/aix-thread-multi-64all/auto/DBD/DB2/DB2.so needs: /usr/lib/libpthreads.a(shr_xpg5_64.o) /usr/lib/libc.a(shr_64.o) /home/pdm01p/sqllib/lib64/libdb2.a(shr_64.o) /unix /usr/lib/libcrypt.a(shr_64.o) /home/pdm01p/sqllib/lib64/libdb2osse.a(shr_64.o) /usr/lib/libC.a(shr2_64.o) /usr/lib/libC.a(shr3_64.o) /home/pdm01p/sqllib/lib64/libdb2g11n.a(shr_64.o) /home/pdm01p/sqllib/lib64/libdb2osse_db2.a(shr_64.o) /usr/lib/libC.a(ansi_64.o) /home/pdm01p/sqllib/lib64/libdb2genreg.a(shr_64.o) /home/pdm01p/sqllib/lib64/libdb2locale.a(shr_64.o) /home/pdm01p/sqllib/lib64/libdb2install.a(shr_64.o) /usr/lib/libC.a(shr_64.o) /home/pdm01p/sqllib/lib64/libdb2trcapi.a(shr_64.o) /home/pdm01p/sqllib/lib64/libdb2dascmn.a(shr_64.o) /usr/lib/libpthread.a(shr_xpg5_64.o) /usr/lib/libC.a(shrcore_64.o) /usr/lib/libC.a(ansicore_64.o) #

Tue Aug 17 06:15:15 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Hello Praveen, Sorry, my last story was wrong. The architecture type of perl and DB2.so is the same. AIX, 32bit perl5.8.8(included with OS) is "executable (RISC System/6000) or object module not stripped." ------------- # file /usr/bin/perl /usr/bin/perl: executable (RISC System/6000) or object module not stripped # file /usr/opt/perl5/lib/site_perl/5.8.8/aix-thread-multi/auto/DBD/DB2/DB2.so /usr/opt/perl5/lib/site_perl/5.8.8/aix-thread-multi/auto/DBD/DB2/DB2.so: executable (RISC System/6000) or object module not stripped ------------- AIX, 64bit perl5.8.9(I compiled) is "64-bit XCOFF executable or object module not stripped." ------------- # file perl5/bin/perl perl5/bin/perl: 64-bit XCOFF executable or object module not stripped # file ./perl5/lib/site_perl/5.8.9/aix-thread-multi-64all/auto/DBD/DB2/DB2.so ./perl5/lib/site_perl/5.8.9/aix-thread-multi-64all/auto/DBD/DB2/DB2.so: 64-bit XCOFF executable or object module not stripped ------------- Those depend libraries are located lib64, but they have symbolic link from lib32. Ex. /usr/opt/db2_09_07/lib32/libdb2.a is symbolic link as "../lib64/libdb2.a" 64bit perl and DBD::DB2 on AIX and Solaris run properly. 32bit modules don't run properly. I think that I have to seek a another breakthrough. Sincerely, Hiroyuki

Tue Aug 24 00:11:44 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Hello Praveen, We are going to test 64-bit perl I compiled on AIX environment. It takes about 2 weeks to complete the test. I'll inform you the result of our test. Regards, Hiroyuki

Wed Aug 25 02:02:53 2010 opendev [...] us.ibm.com - Correspondence added

Sure Hiroyuki. Let me know if you are facing any problems. -- Thanks Praveen IBM OpenSource Application Development Team India Software Labs, Bangalore (India)

Thu Sep 16 01:53:17 2010 pinkee999 [...] gmail.com - Correspondence added

From:

pinkee999 [...] gmail.com

Dear Praveen, We have just finished our test in successfully. Thank you for your cooperation and smart advice. Regards, Hiroyuki