Re: binutils bug (Debian bug 336022 )
Ok, after all my badmouthing I finally get it.
The new binutils fully supports 16 bit operations so it has become pickier
when it comes to syntax.
saying movl when you mean movw no longer cuts it because it really will try to
move a long.
I guess this makes sense.
I wish it would have been announced with a little more fanfare though, instead
of just quietly breaking EVERYTHING.
you guys are sure lucky you have crazy flamers like me to ramble to themselves
and explain these things.
Where was Bryan during all this? I need his equal and oppositely insane flames
to balance my actions.
On Wednesday 21 December 2005 5:28 pm, NZG wrote:
> Ah, there they are, the mythical seg patches.
> I just needed to follow the thread into the bug report.
> Sorry if I've seemed OT here, but anyone who builds the kernel is probably
> going to run into this eventually when they upgrade binutils.
> I assure you it will seem more interesting then.
> "The time has come," the Walrus said,
> "To talk of many things:
> Of shoes--and ships--and sealing-wax--
> Of cabbages--and kings--
> And why the sea is boiling hot--
> And whether pigs have wings."
> (the walrus)
> On Wednesday 21 December 2005 5:21 pm, NZG wrote:
> > Hmm, it appears the the debian testing binutils "works fine"
> > but all previous kernel versions have invalid assembly code, which
> > binutils just now decided to start noticing.
> > http://lists.gnu.org/archive/html/bug-binutils/2005-04/msg00152.html
> > It would be funny if I wasn't breaking all my stuff.
> > This finger pointing has apparently been going on for several months. Has
> > anyone else run into this? What did you do?
> > Are there kernel patches out there to "fix" all the old kernels
> > somewhere? I haven't found them yet.
> > Where would you put a fix to a kernel that only applies after binutils
> > gets changed. How does that kernel numbering work?
> > thx,
> > NZG.