Archived
This forum has been archived. Please start a new discussion on GitHub.
something about BerkeleyDB
emoly
in Help Center
We will develop a set of middleware on AMD64+suse sles8.0 in one month. We choose ice to support this project, but now we meet a problem about db.
We know maybe this is not your work but we need your help really.
We have tried several db versions to resolve it,
1. db-4.2.52
We build it but at last
/usr/lib/../lib64/crti.o: In function `_init':
/tmp/ccZWpbY5.s:26: multiple definition of `_init'
/usr/lib/../lib64/crti.o:/tmp/ccZWpbY5.s:26: first defined here
/usr/lib/../lib64/crti.o: In function `_fini':
/tmp/ccZWpbY5.s:26: multiple definition of `_fini'
/usr/lib/../lib64/crti.o:/tmp/ccZWpbY5.s:26: first defined here
/opt/gcc33/lib64/gcc-lib/x86_64-suse-linux/3.3/crtbeginS.o(.data.rel+0x0): multiple definition of `__dso_handle'
/opt/gcc33/lib64/gcc-lib/x86_64-suse-linux/3.3/crtbeginS.o(.data.rel+0x0): first defined here
collect2: ld returned 1 exit status
I ever built it on linux-32bit, it works very well.
2. db-4.3.27
we build it successfully and build ice-2.1.0 with it successfully too. But when I run "make test", it says"./test/Freeze/dbmap failed with exit status 256". In all the tests of Freeze.
Then when I run ./demo/IcePack/simple ,icepacknode always report"segmentation defualt" here and there randomly.
I ever tried to copy db-4.2 from linux-32bit to present host and built ice-2.1 with 32 option, but it didn't work either.
Our design is based on ice and the project is on the way .This problem really makes us in trouble. We can not make sure "db-4.3+ice2.1" will work dependably.
Thank you!
We know maybe this is not your work but we need your help really.
We have tried several db versions to resolve it,
1. db-4.2.52
We build it but at last
/usr/lib/../lib64/crti.o: In function `_init':
/tmp/ccZWpbY5.s:26: multiple definition of `_init'
/usr/lib/../lib64/crti.o:/tmp/ccZWpbY5.s:26: first defined here
/usr/lib/../lib64/crti.o: In function `_fini':
/tmp/ccZWpbY5.s:26: multiple definition of `_fini'
/usr/lib/../lib64/crti.o:/tmp/ccZWpbY5.s:26: first defined here
/opt/gcc33/lib64/gcc-lib/x86_64-suse-linux/3.3/crtbeginS.o(.data.rel+0x0): multiple definition of `__dso_handle'
/opt/gcc33/lib64/gcc-lib/x86_64-suse-linux/3.3/crtbeginS.o(.data.rel+0x0): first defined here
collect2: ld returned 1 exit status
I ever built it on linux-32bit, it works very well.
2. db-4.3.27
we build it successfully and build ice-2.1.0 with it successfully too. But when I run "make test", it says"./test/Freeze/dbmap failed with exit status 256". In all the tests of Freeze.
Then when I run ./demo/IcePack/simple ,icepacknode always report"segmentation defualt" here and there randomly.
I ever tried to copy db-4.2 from linux-32bit to present host and built ice-2.1 with 32 option, but it didn't work either.
Our design is based on ice and the project is on the way .This problem really makes us in trouble. We can not make sure "db-4.3+ice2.1" will work dependably.
Thank you!
0
Comments
-
2.4.19 is indeed quite old. The oldest Linux officially supported with Ice 2.1 (http://www.zeroc.com/platforms.html) is RedHat 9 -- kernel 2.4.20 I believe.
Cheers,
Bernard0 -
From similar reports found with Google (search for "multiple __dso_handle amd64" on Google), it sounds like a problem with the libtool script (it's generated with the BerkeleyDB configure script.)
From your report -- it sounds like this might have been fixed in the latest BerkeleyDB configure script.
It's difficult to say however why you're getting runtime errors such as segmentation faults with the IcePack processes when you use BerkelyDB 4.3.27, this might be for other reasons (old libc perhaps?)... if you send us the stack traces of the crash we might be able to figure out where it's coming from (make sure to compile Ice with debug information, i.e.: no optimization!).
Of course, I would recommend to also consider upgrading your machine to a more recent OS version, it's quite possible that the 64 bits support is more stable these days
Benoit.0
