weird error with nmh

[Poul-Henning Kamp]

In message <E1I2luR-0001kJ-25 at clue.co.za>, Ian FREISLICH writes:
>Mark Murray wrote:
>> Ian FREISLICH writes:
>> > So I'm really not sure why the linker is choosing libc's version
>> > over this one.
>> 
>> If you turn on debug building (-g), then nmh works again.
>
>I'll give that a try.  Last time I tried (Jun 11) -g didn't help
>either with preventing the core dump or making sense of the stack
>trace.
>
>> Even weirder. GCC bug?
>
>I suspected the rude function aliasing deprecation, but I couldn't
>find a warning to that effect.

It worked for me to remove -O2

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.