On May 6, 2008, at 12:49 PM, Unnikrishnan Udinoor wrote:
Hi,
We still don’t have a solution yet for this problem.
Can someone give some idea/ tips for fixing this issue.
Regards
Unni
I spent some time yesterday and today building a VM for OpenSolaris
2008.05. I was able to build and test log4cxx using Sun Studio 11 in
my VirtualBox VM on Mac and was able to fix some of the reported
compiler warnings on Solaris before I had to do a hard restart of my
host machine and lost that VM. I thought it best to fix all the
easily addressed previously reported Solaris issues just in case one
of them was responsible. I haven't finished them all, but I don't
think that is likely at this point. I'm trying to rebuild a VM
(trying VMWare this time) and will try a gcc build, but will be likely
18 hours or so from now.
Could you do the following:
If you built your own apr and apr-util, provide the results on "make
check". I saw some failures when I ran that on OpenSolaris 2008.05,
but they didn't seem to be in parts of APR that we use. However, it
you get a different set of failures that I do, then that might be a
problem.
Run "make check" on an unmodified log4cxx (either 0.10.0 or SVN HEAD)
and report the results.
Run the trivial sample program (src/examples/cpp/trivial) and report
the results. That program should be pretty similar to your sample
program and if trivial runs and your sample fails then should look at
the differences between the apps.
If you have access to a comparable Solaris on x86 or x86_64, do you
have the same failures or is it Sparc-specific.
Probably should file a bug report and add the reports as attachments.
