Re: [Wrapper-user] I have added support for HP/AIX

[Leif M. <le...@ta...>]

  Ashish & William,

Thanks for the patches.  It took me a little while to work through all 
the changes
because I had already implemented the ability to daemonize the wrapper 
as well
as to set the pid file on all platforms.  Take a look at the new 
sh.script.in and
bash.script.in files for examples of how they are used.

I think I got all of your patches implemented correctly but could you 
please download
the following source file and try doing the following builds on the AIX 
and HPUX
platforms to make sure that they work correctly.
./build.sh
./build.sh release  ( should create the src and binary releases for the 
platform )
./build.sh clean
./build.sh total-clean

You can get your hands on the latest source by doing a CVS checkout.  Or 
if you
prefer, I placed a copy of the source at the following URL:
http://www.tanukisoftware.com/wrapper_2.3.0d_src_with_doc_src.tar.gz

Let me know if you find any problems,
Cheers,
Leif

Ashish Gawarikar wrote:

> Leif Mortenson wrote:
>
>> Ashish Gawarikar wrote:
>>
>>> But this is in the package 2.2.8 and not in 2.2.9.
>>>
>>> Me and my colleague have also added few bug fixes:
>>> 1. Added run as daemon option for this wrapper to fix the problem 
>>> that the wrapper dies when the shell exits.
>>> 2. clean up the add_srv before binding.
>>>
>>> This has been tested on AIX and HP. I can send the package to the 
>>> Admin/maintainer.
>>>
>>> I did rather send the full tar.gz. Let me know the email id. 
>>
>>
>>
>> Great, please send the source to me directly, not to the list.   I 
>> will look through the changes
>> you made and get them merged in with the trunk.   Once I have done 
>> that, I will ask you to
>> download the patched source again and verify that I did everything 
>> correctly before a release.
>>
>> As to your additional patches:
>> 1) I think that I already got this implemented thanks to a patch by 
>> Rajiv a few days ago.  I would
>> like to see how you implemented it and make and modifications necessary.
>> 2) Could you explain to me what the "add_srv" problem was that you 
>> had been encountering.
>> It may be obvious from the source. 
>
>
> Before doing the bind (in the wrapper.c), the code was not doing a
> memset(&addr_srv, 0, sizeof(addr_srv));
> to clean the addr_srv, as a result I had some problems in the bind. It 
> would always
>
>
> Here I am attaching the source (not a diff). Please let me know if 
> this looks good.
> Thanks,
>
> Ashish Gawarikar
>
> PS: My colleagues name is William Lee.