SASL ,HP-UX and dynamic loaders

2006-10-25 Thread Biswatosh
Hi, In dlopen.c, we have dlopen() as wrapper on shl_load(), dlclose() on shl_unload() etc. Any reason for that? Today,in HP-UX,we can directly use dlopen() and all. The reason I am asking is, one of my clients is complaining that symbol relocation is happening because of this. dl* functions

Re: SASL ,HP-UX and dynamic loaders

2006-10-25 Thread Biswatosh
Ok. But what is the reason for still keeping shl_load and unloads? Is it to support some old HP-UX platforms when they didnt have dl* functions? Otherwise,we could simply do away with these sh* functions,is not it? --Biswatosh --- Howard Chu [EMAIL PROTECTED] wrote: Biswatosh wrote: Hi,

RE: Can't get SASL Authentication to work

2006-10-25 Thread Doug Campbell
After much additional testing I think I have solved the SASL related problems. The first issue I discovered was that the reason I wasn't seeing anything interesting in my slapd logs was because slapd wasn't being contacted. I was using unix sockets and it turns out that I had missed a step

Re: SASL ,HP-UX and dynamic loaders

2006-10-25 Thread Howard Chu
Biswatosh wrote: Ok. But what is the reason for still keeping shl_load and unloads? Is it to support some old HP-UX platforms when they didnt have dl* functions? Otherwise,we could simply do away with these sh* functions,is not it? --Biswatosh Right, HPUX systems older than 11.00 only have