It seems that I've finally found the mysterious bug that caused some SMP
machines to lock up at bootup if they had no keyboard enabled. It turns
out that the keyboard was a complete red herring, and that it just changed
timings of bottom half handling in particular. The real culprit was some
misguided locking attempts by the console driver at a really bad time.

Anyway, that means that the last of my personal show-stopper bugs in 2.1.x
seems to be finally history. I still expect to sync up with Alan Cox's
patches in particular, but I'm mentally getting ready for a real 2.2.

I still haven't decided on whether I'll make the same kind of "pre-2.2"
that I did before the 2.0 release, but there are strong psychological
reasons to do so to get people to more actively test it out with a "this
really should be stable" mindset.

In the meantime, there's now 2.1.125. Most of 2.1.125 is driver updates
for various things, most notably perhaps joystick and the new 5.10 version
of the Adaptec aic7xxx driver by Doug Ledford (but there are various other
driver updates). The fix for the mysterious lock-up is a few embarrassing
lines removed, but makes me feel a lot better ;)

Go forth, multiply, and fill the earth

.. is out there now, and includes:

 - subtle fix for lazy FP save and restore on x86. The bug has been there
   for a long time, but was apparently triggered by the re-write of the
   low-level scheduling function. It could result in corrupted i387 state
   under certain (admittedly fairly unlikely) circumstances.
 - various networking updates. Some of the bugs fixed could result in
   kernel Oopses. None of them were common, though.
 - fixes for both filesystem accounting and quota handling.
 - the much-ado-about-little video driver merge.
 - PPC and Sparc updates
 - i386/SMP interrupt handling falls back on the safe mode.. Please tell
   me whether there are still machines with problems.
 - some new network drivers and updates
 - final (we hope) IP masquerade update

I still have a problem with certain machines that apparently don't want to
boot with the keyboard not plugged in even though they should. Kill me
now. If you have problems with i386/SMP on a machine without a keyboard,
plug one in and send me a report..

This may or may not fix the APM problems, and the INITRD ones.  The
INITRD one in particular was a case of a fairly inexplicable test that
shouldn't have been there in the first place breaking when something
completely unrelated was cleaned up..

The APM breakage was simply due to it being in the wrong place.  The
patch looks bigger than it really is - it really only moves the file to
the proper directory, and makes sure that it should compile with the
standard assembler..

                Linus

I made a 117 to fix the silly things left in 116 in my excitement over it
passing all my crashtests. This should fix the things with the kernel
thinking it was out of memory much sooner than it actually was etc.

Alan still reports some funnies with unix domain sockets, but he's
reportedly fixed the behaviour of NFS over TCP. He didn't make it sound as
if you really want to use it yet, though ;)

                Linus

I just released Linux-2.1.116. I've tested it fairly extensively on my SMP
box, both with little memory and much, and I cannot make it lock up any
more.

Special thanks to Dean Gaudet who helped me set up a apache configuration
that finally made me able to repeat the lockup, and made me able to debug
the thing.

Most of the 2.1.116 patches are "just" alpha and m68k updates, and can be
ignored by most people. The bugfixes are, roughly:

 - fixed serious low-memory situation problem, where a critical resource
   allocation problem could result in nasy behaviour. Notably, doing TCP
   under low memory could result in TCP trying to allocate memory in a
   tight loop and locking out kswapd completely so that the situation
   would never be rectified. In short, the machine hung.
   This problem has been there forever, the only reason it doesn't show up
   under 2.0.x seems to be because under 2.0.x the TCP allocation was
   always for a single page, for which this situation never arises. Under
   2.1.x the slab code forced multi-page allocations.
   If you've seen lockups with 2.1.x, this may be the cause. This was what
   held up 2.1.116 for so long.
 - various minor driver updates. Networking, radio, bttv.
 - NFS over TCP still doesn't work, but at least it fails due to new
   reasons.

Alan, try your squid thing under 2.1.116. I suspect it will hold up now,

                Linus

Ok, we've been in a tentative code freeze for a long time, and now it's
final. I've made a 2.1.115 that I hope is good enough, and I won't be
accepting anything but bug-fixes until 2.2..
There are two long-standing patches that I'm still considering:

 - devfs
 - dynamic fd's

and I kind of expect that they'll go in (devfs is configurable, so if you
don't want it you don't need to care, and the dynamic fd's save some
memory and speed certain things up a bit). The reason they're not in now
is mainly that I've been trying to get everything else off my plate, and I
want to ruminate on them in peace for a while.
Bug-fixes are still (and will always be) accepted,

                Linus