Re: New libc malloc patch

On Dec 4, 2005, at 4:51 AM, Claus Guttesen wrote:
> I was able to do a buildworld on current with this patch, but I had
> problems getting X to run and kldxref took all my space on the
> root-partition doing a installkernel.

I've fixed the offending bug in kldxref in the latest patch:

http://www.canonware.com/~jasone/jemalloc/jemalloc_20051211b.diff

I spent several hours poking at X, but was never able to determine  
why it goes into an infinite loop.  The infinite loop happens rather  
early, during the load of the libbitmap module.  My best guess is  
that it is stuck trying to acquire the Xlib lock, but cannot be sure,  
since I don't know how to get debug symbols for the loaded X module.   
In any case, malloc is nowhere in the backtrace.  I do not have the  
time to acquire the X expertise that is likely needed to track down  
this problem.  Hopefully someone else will be willing to do so.

No new problems in the malloc code have been found for some time  
now.  It has been tested on i386, sparc64, arm, and amd64.  In my  
opinion, the malloc patch is ready to be committed.  I am now working  
on the assumption that new problems are more likely application bugs  
than malloc bugs.  This seems like a good time to start sharing the  
debugging load with the community. =)

So, how about it?  Can this patch go in now?

Thanks,
Jason