- USB: ohci controller update, round-robin device numbering
    - ksymoops moved: document
    - sparc updates
    - sg.c: get rid of more #ifdef MODULE code

    - USB: documentation.
    - Yeah. MD/LVM should really be fixed this time.
    - SH architecture update
    - i810 RNG driver update
    - IDE-PCI: make sure we initialize the chipsets correctly
    - VIA IDE driver fixes
    - VM balancing, part 53761 of 798321
    - SCHED_YIELD cleanups

    - initialize to zero -> put it in the .bss instead
    - no extended dumb serial driver options, if no dumb serial driver
    - access() on a special file on a read-only filesystem is special.
    - DRM update
    - fix SCHED_YIELD problems.
    - quintela: fix the synchronous wait on kmem_cache_shrink().
      This should fix the mmap02 lockup.
    - syncppp got lost in the Makefile reshuffle. Unlose it.
    - firewire update
    - flock blocking list fix
    - correct watchdog initialization order
    - USB-storage: reset fixes. Race condition fixes.
    - USB: fix freeing already free'd device.
    - minix truncate fixes
    - USB: pack only the relevant subset, not the whole descriptor (so
      as to not create extra unaligned fields).
    - nfsfh: DCACHE_NFSD_DISCONNECTED checking typo
    - dquota silly bugfix
    - sound updates (get rid of check_region, check request_region() instead)
    - scsihosts boottime parameter passing
    - avoid double init of MD
    - eicon ISDN driver update
    - fix Cyrix MTRR thinko
    - toshiba driver 2.4.x update
    - Makefile subdirectory traversal cleanup and documentation
    - cciss typos from bad merge fixed
    - cdrom driver oops fix for CONFIG_SYSCTL=y CONFIG_PROC_FS=n
    - coda initialization - we already did the module_init, no need for
      the extra double init.

    - USB: remember to release the kernel lock and other updates..
    - recognize the k6 model 13: it's a K6-2+ mobile processor.
    - file locking deadlock detection bugfix..
    - NFSv3 is not really really experimental any more.
    - don't raise privileges when re-trying a failed NFS RPM request
    - alpha cross-compile fixes..
    - sound init cleanups
    - shm statistics bugfix.
    - nfsd: mark us as a O_LARGEFILE case, so that the VFS allows
      the full 64-bit access..
    - fix up ac97 codec initialization
    - Ingo: clean up VM handling, improve balancing.
    - add SGI PCI ID's.
    - export the new lock copy/init functions
    - cs4281 sound driver
    - official Compaq CISS driver.

    - TUN/TAP driver: use proper device number (misc device, minor=200).
    - teach st.c about some SCSI tapes that really aren't SCSI tapes (OnStream)
    - samba 2.2 needs leases for efficient file sharing.  They are kind
      of like file locks with async IO notification.
    - broadcast I/O APIC interrupt MP-tables are legal..
    - alpha RTC year magic again..
    - careful memory ordering by Andrea..
    - make the scsi-generic module work properly again.
    - file locking fixes
    - update atp ISA net driver
    - VIA IDE driver bugfixes
    - more linux-2.2 driver sync-ups
    - new PCI ids
    - emu10k stereo sound fix.
    - makefile documentation update
    - USB uhci updates
    - networking updates
    - codafs fixups
    - VM UP deadlock fix
    - Add Camino chipset ID to eepro100 driver.

    - Make SCSI initialization order be same as before.
    - fix cardbus bridge resources..
    - don't disallow Onstream ide-scsi devices
    - byteorder: use statement expressions instead of macros, to avoid argument re-use.
    - codafs update
    - more USB updates
    - _fput/__fput are no longer used.
    - ixj telephony driver fixes
    - pmac SCSI driver init update
    - Andries: net device name allocation as in 2.2.x
    - sis900 driver update
    - more drivers synced to Alan's 2.2.x changes

    - continued SCSI cleanup
    - more USB updates

    - USB updates
    - NFS over TCP - handle TCP socket writability right..
    - NFS cache coherency across file locking fix
    - floppy: we'd better hold the io_request_lock when playing with "CURRENT".
    - acenic driver update
    - ARM update (including ARM drivers)
    - adfs correct dentry operations
    - netfilter update
    - networking updates (iipv6 works non-modular etc)
    - Sync up with Alans 2.2.x driver changes
    - SCSI initialization - move over to the modular case. No more
      double initialization.
    - block_prepare_write and block_truncate_page: if the page is
      up-to-date, then so are the buffer heads inside it once they
      are mapped..
    - uninitialized == zero. Remove extra initializers.

    - scsi fixes
    - network updates
    - PCI bridge scanning fix: assign numbers properly
    - sparc updates
    - Riel VM update
    - disallow re-mounting same filesystem in same place multiple times.
      Too confusing. And /etc/mtab gets strange.
    - PPC updates (including PPC-related drivers etc)
    - more initcall updates
    - various net drvr updates and fixes
    - "extern inline" -> "static inline".  It doesn't matter right now,
      but it's proactive for future gcc versions.

    - MM balancing (Rik Riel)
    - pcmcia debounce logic. Ugh.
    - 3c59x driver update
    - SysKonnect gigabit driver update
    - ACPI interpreter updates (and file renames - making this part big)
    - don't do notifier locking at low level: higher levels do (or
      should do) this already.
    - truncate unmapped/uptodate case handled correctly
    - IDE driver picks up master/slave relationships on its own.
    - misc small fixes to random drivers by Tigran
    - mktime(). Do it just once - not 16 times.
    - NFS sillyrename fixups
    - acenic update
    - epic100 credits, 8139too driver update, sr.c initcalls
    - name clash in hamradio/pi2.c and hamradio/pt.c
    - sparc64: register window race. Non-deadlock rwlocks.
    - USB: storage update
    - USB: OHCI controller unlink and bandwidth reclamation fixes

Ok, as the truncate problems really seem to be fixed, it's time to do an
official test8, the first development kernel in about a year and a half
that should have a working truncate() again. Thanks to everybody who
tested, and especially to Al Viro who did a lot of the heavy lifting.

There are a number of other various changes there too - the truncate fix
itself was fairly small, it was just more complex than most problems that
can be solved in 50 lines of code.

The only one of any note that I'd like to point out directly is the
clarification in the COPYING file, making it clear that it's only _that_
particular version of the GPL that is valid for the kernel. This should
not come as any surprise, as that's the same license that has been there
since 0.12 or so, but I thought I'd make that explicit.

Why? There's been some discussions of a GPL v3 which would limit licensing
to certain "well-behaved" parties, and I'm not sure I'd agree with such
restrictions - and the GPL itself allows for "any version" so I wanted to
make this part unambigious as far as my personal code is concerned.

The reason I wanted to mention that particular issue here explicitly
(rather than as just a one-liner in the changelog) is that code written by
others is obviously under _their_ discretion, and not limited by my
personal foibles, fears and misgivings.

If anybody wants to explicitly state that their code will be valid under
any version of the GPL (current or future - whatever they may look like),
please send patches to say so for the code in question. If you've used the
FSF boiler-place copyright notice, you already have this in place (it says
"v2 or later" - the FSF itself doesn't recommend v1 any more).

(Me, I'm taking the careful "wait and see" approach. I don't know if a GPL
v3 is imminent, and I don't know if the issues discussed will even
_become_ real issues, so you might as well consider me a paranoid, if
careful, bastard).

		Linus

 - final:
    - sparc updates
    - routing cleanups
    - fix up warnings
    - USB: dsbr driver update and warning cleanup
    - USB: update the status file timestamps when changes occur
    - USB: clean up module init/exit. No more explicit inits.
    - USB: bluetooth and microtek driver updates
    - serial.c: remove duplicated code. Add SGI serial port signature.
    - /proc/kcore and ncpfs - more care with kernel virtual addresses
    - discontiguous memory initialization fixup
    - sd oom fixes
    - pcmcia state machine fixes.
    - microcode driver update - make sure we get a sane revision level
    - copyright license version clarification (make v2 explicit)
    - epic100, 8139too and tulip driver updates
    - clean up truncate_inode_pages (no, this has nothing to do with the
      rest of the truncate issues)
    - jffs update

    - truncate - the never-ending story.  Makes me feel like a long
      Kurosawa movie.  But in this one the hero _will_ survive, or my
      name isn't Maxwell.
    - SCSI tape driver potential memory leak.
    - XMM FP handler bug fix: we really must not change the FP error
      mask on exceptions.  People care.

    - truncate. Guess what? We threw away the key to the clue-box.
    - simplify signal notification. And remember the spinlock.
    - VIA ide driver update (well, rewrite - the old one was buggy and broken)
    - network driver fixes (not checking for oom etc)
    - USB serial driver SMP locking fixes
    - fix memory leak on failed USB configuration queries
    - USB initialization using proper "init()" calls.
    - dvd capacity bug fix and other cdrom driver cleanups
    - sis5513 IDE chipset update
    - do_fork() - add "stack-top" for ia64 (and potentially other
      architectures that may care)
    - devfs support for LVM
    - quota transfer miscount fix
    - x86 checksum/copy prefetch
    - NFS sillydelete fix
    - mark_buffer_dirty() doesn't actually use the second argument. Delete it.
    - SCSI communications device - no need to complain about it
    - SCSI WP test fix (all pages, not just the first one)
    - epic driver update
    - fix pcmcia "driver asserts interrupts when booting up" issues

    - truncate really fixed this time. Everybody agrees.

    - ext2: final truncate piece - fix the innd problem.
    - use "sfence" for x86 memory barrier when available.
    - remove the thread-group signal code for now: no feedback.
      Leave the cleanups in place so that we can add it back in
      cleanly later, but remove the new features.
    - ARM update.
    - released for Al Viro to check the truncate thing

    - The TCP socket leak patch _really_ went in this time.
    - get rid of more suser() checks in networking.. It's "capable(CAP_NET_ADMIN)".
    - sparc updates
    - alpha updates. Fast alpha xor for raid. AP1000 updates.
    - Wonders never cease. digiboard driver updates. Christoph Lameter is BAAACK!
    - SiS frame buffer driver updates. Can be used without a BIOS.
    - nfsd interface cleanup.
    - fix potential buffer overruns in get_partition_list.  Remove
      limitation of one page.
    - floppy driver capability cleanups. Use "request_region()".
    - handle dcache flushing when there are shared user mappings that
      may be dirty.
    - get rid of the "xxxx_ret()" user access macros. They are more complex than
      just doing the return directly and they hide what's going on.
    - fix up broken BIOSes that don't give unique ID's to different APICs
    - make more of the drm drivers compile on other platforms and know
      about the signal blocking issues.
    - net/atm/mpoa_proc.c: user-space access thinko
    - pcmcia: David Hinds: merge updates from 3.1.20
    - pcmcia: non-ISA machines really shouldn't use ISA interrupts ;)
    - ext2: truncate races and error code return fixes
    - true shared signals for pthreads..

    - Oops. Moved back stallion.c to drivers/char. It's not a TV driver.
      Never has been, and I don't see it ever really becoming one ;)
    - mca.c: outp_b() should be outb_p().  Obviously nobody actually
      _uses_ the MCA bus any more ;)
    - umsdos should be ok again after the page_address() type-changes.
    - re-enable asynchronous read-ahead code.
    - Sun ESP driver update
    - netfilter debug fixes
    - IPv6 needs to register before proto_init()
    - socket() error code fix (EAFNOSUPPORT instead of EINVAL)
    - potential TCP socket leak fix
    - don't self-deadlock on the kbd_controller_lock when probing for the mouse
    - CONFIG_SMB_NLS_REMOTE didn't work. Silly typo.
    - scheduler wakeup race condition could cause delayed scheduling on SMP..
    - net/packet/af_packet.c: use the standard macros for marking page resevredness
    - ncpfs buffer-overflow fix
    - thread groups, take 1.
    - USB storage driver update

Ok, getting close to the real test7.

The patch looks bigger than it is due to the drivers re-organization
(discussed to death on this very list ;).

		Linus

test7:
 - pre7
    - block_all_signals()/unblock_all_signals() interface to allow drivers to
      react nicely to signals that happen to a process that owns driver data.
      Read: direct rendering lock handling.
    - ThunderLAN update (timer fixes, full-duplex, activity-led)
    - Fix NFS oops on removing negative dentry. Honour rsize for directory read.
    - usb updates
    - scheduler wakeup race fix.
    - move radio/tv cards to drivers/media, cleaning up drivers/char
    - move "input" layer to drivers/input, cleaning up drivers/usb
    - Cirrus SoundFusion CS4280/461x sound driver.
    - proper camera locking in usb/dc2xx.c
    - USB printer driver update (Printer Protocol 3 and timeout handling)

    - Pentium/MMX has a good IO-APIC
    - APIC detection (aka "try to boot the SMP kernel even on buggy UP
      boards that claim SMP.")
    - alpha time.c epoch calculations. Again.
    - report where the memory region information came from.
    - sparc updates
    - file locking: fix deadlock detection and EOVERFLOW
    - network bugfixes
    - md readahead was getting confused about bytes vs pages
    - DRM and AGP fixes
    - xircom_tulip.c: don't corrupt the timer list on suspend
    - PCI /proc interface buffer overrun fix
    - ISDN updates
    - DAC960 update
    - fork: get CLONE_PTRACE right (SIGSTOP the child)

    - i2c-old compile fix.
    - sparc updates (pte_pagenr cleanups)
    - make ipv6 loadable as a module together with CONFIG_INET_ECN
    - get rid of bdf_prm.b_un.age_super - it just confused the buffer aging
    - smbfs knows about nls and ascii mount options
    - small joystick driver updates
    - make sure to mark inode dirty after having changed the size!
    - hgafb documentation update

    - "USE_STANDARD_AS_RULE" - generic Rules.make as rule
    - arm update (arch/arm, asm-arm, drivers/acorn, Documentation/arm etc)
    - eicon ISDN driver update (big).
    - serial.c warnings removal.
    - compilation fixes under different configurations..
    - bounds checking for hpfs code page index.
    - sparc64 bugfix for atomic_dec_and_lock. Oops. And use flock64.
    - FAT missed the d_type thing from readdir.
    - fix tsk->files race fixes from -pre3 ("struct files_struct", not
      "struct file" and make sure to register the socket fs before we
      use a pointer to it)
    - ns558.c: don't leave the driver registered after a failed module
      load.  Either return success, or unregister the PCI driver. And
      don't leak IO port allocations.
    - USB OHCI controller fixes for oopses due to races..
    - usb updates
    - 3c59x driver update
    - VIA KX-133/KT-133 chipset detection and AGP bridge support
    - raid/raw-io cleanup: use generic_make_request instead of ll_rw_block.
    - Emu10k1 sound driver update

    - nfs_commit_rpcsetup() signed comparison bugfix and cleanup
    - sparc updates and TLB invalidation fix
    - networking updates (less verbose on the new reordering messages)
    - network driver Makefile cleanup
    - Fix segment copy on fork.
    - tsk->files race fixes: close-on-exec etc.
    - sound #define cleanups
    - fs/proc/array.c task_lock cleanup

    - fix modversions.h generation ("make -j dep" works now)
    - finish 64-bit VFS: getdents64 and fcntl64 (getdents64 also adds
      the "file type" to the readdir data - VFS layer change.  fcntl64
      allows 64-bit file locking)
    - Intel i810 watchdog driver and NS DP83810 network driver
    - dup2() cannot screw up the file table with threads any more.

    - fix PCI resource bug that crept in in test6 due to the new
      requirements to handle multiple bus regions transparently
    - ll_rw_block documentation
    - sound driver module counting bugfix and cleanup (move to named
      initializers)
    - directory rename bug fix for busy directories (oops)
    - allow "init_new_context()" to fail - it can do so on some
      architectures when out of memory.
    - networking updates - TCP retransmission and ordering logic
    - fix strsep(). Not that anybody cared.

- speling fixces.
 - fix drm/agp initialization issue
 - saner modules installation
        (*) NOTE! This may/will break some module setups.  Files go in
        different places. Better places.
 - per-CPU irq count area. Better for caches, simpler code.
 - "mem_map + MAP_NR(x)" => virt_to_page(x)
        (*) Purely syntactic change at this point. NUMA memory handling
        will take advantage of this during 2.5.x
 - page_address() returns (void *) to make it clearer that it is a
   virtual address (it's the reverse of "virt_to_page()", see above).
 - zimage builds should work again.
 - Make current gcc's able to compile the kernel.
 - fix irq probing in IDE driver: this caused strange irq problems for
   other drivers later on (notably PCMCIA, which is one of the few
   drivers to still probe for ISA interrupts on modern machines).
 - Intel microcode update update.
 - mips/mips64/sh/sparc/sparc64/acorn updates
 - DAC960 driver update
 - floppy shouldn't scream on open/close
 - console driver does correct palette setting.  No more black screens
   with XF86-4.x
 - ISDN updates
 - PCI layer can assign resources from multiple IO and memory windows
 - yenta_socket driver no longer oopsable on unload.
 - flush_dcache_page() for more virtual dcache coherency issues
 - ext2_get_block() races fixed
 - jffs bugfixes galore.
 - user resource tracking infrastructure re-organization.
 - umsdos works again.
 - loopback shouldn't deadlock

The diff is pretty huge, to a
large degree due to a bttv driver syntactic split-up and due to the NLS
forward-port from 2.2.x.

Other notable bugfixes:
 - the buggy Toshiba (and possibly others) BIOS memory reporting thing
   is fixed. Just ignore RAM that the BIOS reports in the 640k-1M range.
   The BIOS is confused.
 - Manfred Spraul found and fixed a SMP TLB invalidation problem with threads.
 - various architecture updates (arm, ia64, sparc, sh..)
 - MD driver cleanups
 - Toshiba floppy controller problem workaround
 - updated DRI code (works with XF86-4.0.1)
 - various driver updates (ToPIC CardBus should work, ide updates, etc)
 - "kfree_s()" is gone. It hasn't existed for several years, but people
   still used it. No more.
 - USB driver updates and fbcon cleanups
 - various othe rupdates I've repressed.