Re: cvsup broken on amd64?

On Fri, Sep 09, 2011 at 11:30:46AM +0200, Oliver Lehmann wrote:
> 
> Chris Rees <crees_at_freebsd.org> wrote:
> 
> >On 9 September 2011 06:33, Oliver Lehmann <lehmann_at_ans-netz.de> wrote:
> >>I got used to it in the past 12 years? But this is not realy the question.
> >>If it is "BROKEN" it should be marked as BROKEN or there should be a
> >>statement that it will not work with FreeBSD 9 on at least amd64 or we
> >>will have other users complaining about the same at least when
> >>9.0 RELEASE is out - right?
> >
> >The cvsup port is normally used now only for cvsupd, for which there
> >is no csupd analog. As far as I know, and perhaps mux (CC'd) could
> >confirm every feature present in cvsup is present in csup-- and it's a
> >fair amount faster too.
> 
> Ok, but this again is not the point of my email ;)
> This is not about "just me". I know how to help me in that case. I want
> to prevent users facing the same problem and writing mails like my initial
> mail.
> I'm quiet sure that there are numerous users out there still using cvsup
> as client so they will start like me with cvsup on ther new instaled system.
> It would be better if they just would not be able to install cvsup if it
> will not run and we don't want it to run.
> I was also curious if it is only me where it fails on amd64 or if it is
> because it was compiled on an Atom 330 with some amd64 flags determined by
> the system which does not fit the Atom 330 (gcc 4.2 is older than the CPU
> design).
> With other words: If the support for cvsup on amd64 is dropped, it has
> to be marked as BROKEN/IGNORE/whatever. Otherwise it need to get fixed for
> 9.0?

For start, you should provide the information what exactly is the
instruction that caused the fault. Show the disassembly from gdb
for the function that caused the fault.