Re: [BugDB] Segfault on graceful restart (PR#434)

[Cliff Woolley]

>>> [EMAIL PROTECTED] 07/26/00 05:37PM >>> >I'd still suggest trying to >remove those three lines from ssl_engine_init.c first - just to see if it >makes a difference. The patch was originally applied to fix a memory leak that was happening upon graceful restart. (Does this sound familiar? <grin>)   I found the same patch in the CVS repository a while ago, and the whole conversation came back to memory.  My suspicion (yet to be verified) is that the correct solution is neither backing out this change nor keeping it... we do want to avoid the leak, but we probably need to add something ELSE to re-initialize/re-allocate something that is lost in the process.  Or it could be that part of the patch was correct, but it went too far (ie, part of it might need to be removed).  Just a hunch.   Backing out the patch implies that there was never any leak in the first place. (Perhaps just the illusion of one?)  I suppose that that might be the case as well... will have to look in to this more.   --Cliff   Cliff Woolley Central Systems Software Administrator Washington and Lee University http://www.wlu.edu/~jwoolley/   Work: (540) 463-8089 Pager: (540) 462-2303