Re: dumping bug explained?

[Chip Coldwell]

On Mon, 21 May 2007, Ulrich Drepper wrote:

> Chong Yidong wrote:
> > Do you know if Emacs will still need to be patched after these changes
> > to glibc are done?
> 
> Not all the symptoms described in the bug can be explained by the
> changes.  It might very well be that changes are needed.

I don't know if this is the behavior you are thinking about, but IIUC
this message

*** glibc detected *** emacs: malloc(): memory corruption: 0x0000000002904d10 
***

was generated by an assert failing in malloc that Jakub Jelinek
described in his libc-hacker posting containing his patch to solve
this problem thus "removes a bogus assert mp_.n_mmaps <=
mp_.n_mmaps_max."  This would be the assert that would succeed if you
dumped emacs with MALLOC_MMAP_MAX_=0, right?

Chip

-- 
Charles M. "Chip" Coldwell
Senior Software Engineer
Red Hat, Inc
978-392-2426