error making world: stage 1

[MikeM]

Hi,

This is the first time I've tried to install current,
so if the following issue has already been discussed,
please let me know where to find the logs.

I am trying to upgrade from 3.2 Release to -current
but the build fails in the bootstrap stage.

my supfile has the following
release=cvs tag=.

I synced my source tree at about 10 a.m. Pacific this
morning (05/19).

I ran make world.

As it tried to compile new.cc in gperf I got the
following error:

/pub/src/gnu/usr.bin/gperf/../../../contrib/gperf/src/new.cc:
In function `void operator delete(void *)':
/pub/src/gnu/usr.bin/gperf/../../../contrib/gperf/src/new.cc:82:
declaration of `operator delete(void *)' throws
different exceptions...
:82: ...from previous declaration here
*** Error code 1

A look at /usr/src/contrib/gperf/src/new.cc yielded
this:

/* We need this deletion operator in order to make the
linker happy.
   Because `operator new' and `operator delete' always
come together.  */
void
operator delete (void *ptr)
#ifdef HAVE_THROW_DECL
  throw()
#endif
{
  T (Trace t ("operator delete");) 
  // We cannot call free here, as it doesn't match the
mallocs.
  // free ((char *) ptr);
  (void) ptr;
}

So, apparently somewhere else delete is declared
without throw() or with something inbetween the ().
My understanding is that throw() is an optional
compile time thing, and acts simply as an "interface
contract" between the calling and the called function
specifying what exceptions will be thrown by the
function, so I took the liberty of adding:
#undef HAVE_THROW_DECL
just before the function declaration. That seems to
have fixed my problem (it's in "stage 3: cross tools"
right now). 

However, that only fixed the symptom, not the cause.
Presumably, the problem will still be there next time
I cvsup.  I would rather not have to muck with the
code every time I sync my source tree, and would
appreciate it if someone could give me a heads up as
to anything I might be doing wrong on my end, or if it
is a problem with the code that the maintainer be made
aware.

I'd appreciate any help/response.

Thanks,

Mike_TM.
--"The beatings will continue until morale improves!"

__
Do You Yahoo!?
Send instant messages & get email alerts with Yahoo! Messenger.
http://im.yahoo.com/


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message

Re: spam

[mikem]

my 2 cents worth,

I don't think the effort is worth the reward. If the frequency increases, then
the issue should be revisited. 

If we do decide this problem merits action...

A simple solution would be to restrict posts to subscribers of the lists
(-questions could be an exception). Another simple solution (if we are using
mail filters) is to designate a group of people as spam watchers. When they
spot a spam they forward it to a special addres ( i.e. [EMAIL PROTECTED]),
which then includes it in some sort of filter recipe.

The disadvantage of such a scheme is that only subsequent spams from the 
poster or containing the same (body|header) are filtered. That is, the spam
gets through to th list the first time. The advantage is that this whole process 
is only triggered when a spam is spotted. The other solutions have the 
disadvantage of being triggered on every post or every post by a non-subscriber.
I would venture to guess that most non-subscriber posts are legitimate. Besides,
most people are too busy to take on moderator roles.

Although this solution obviously provides less than 100% spam free lists, I 
think it will prove itself pretty effective because I've noticed that
most spams get sent to the list more than once-- the recent "Linus Torvalds..."
spam comes to mind.

Ok, my 2 cents worth has gone on too long :-)

mikem.



To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message