This patch has already made it into 2.4.21pre3-ac4. Please apply.

Actually exporting the symbol introduced in 2.5.57 makes module users
happy.

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

The CPUFREQ initialisation now registers an interface with the device
model, and thus needs to initialise after postcore.  We use the
arch level for this.  This does, however, impose the restriction
that cpufreq may not be available for other architecture
initialisated code.

into flint.arm.linux.org.uk:/usr/src/linux-bk-2.5/linux-2.5-rmk

Remove ide_release_lock and ide_get_lock definitions from asm-arm/ide.h;
they're defined in include/linux/ide.h.

Add entries for sendfile64, futex, async io, etc system calls to
both unistd.h and the system call handler table.

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

In addition to the diff from the 1.7 i8kutils release to the 1.13
release, I made the new globals static as per Rusty's namespace
pollution patch yesterday, and removed the reference to an include no
longer in 2.5.

	An old AMD 486DX4-SE(as reported by bios) here crashes on boot in
amd_init, doing rdmsr in k6 case. Elan fixes added case 4 there but
without break. This break allows it to boot.

Northwood P4's have one more bit in the CPUID processor info set: bit
31.  Intel calls the feature PBE (Pending Break Enable).

The attached patch adds the necessary entry.

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

SGI Modid: 2.5.x-xfs:slinx:136827a

This doesn't mean dmapi is scheduled for inclusion, just adding
the reserved minor number to miscdevice.h for documentation purposes.

SGI Modid: 2.5.x-xfs:slinx:136795a

apparently the root file system are not being synced 
correctly during system shutdown

SGI Modid: 2.5.x-xfs:slinx:136269a

SGI Modid: 2.5.x-xfs:slinx:136126a

SGI Modid: 2.5.x-xfs:slinx:135869a

SGI Modid: 2.5.x-xfs:slinx:135920a

SGI Modid: 2.5.x-xfs:slinx:135919a3

SGI Modid: 2.5.x-xfs:slinx:135168a

These shouldn't be there, probably the result of corruption.
However, if we find one, handle it specially so that we don't
deadlock during unlinked list processing in recovery.  Without
xfs_iput_new, we'd be waiting on an inode lock we already hold.

SGI Modid: 2.5.x-xfs:slinx:136535a

SGI Modid: 2.5.x-xfs:slinx:135917a

SGI Modid: 2.5.x-xfs:slinx:135881a

SGI Modid: 2.5.x-xfs:slinx:135847a

SGI Modid: 2.5.x-xfs:slinx:135827a

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

into tetrachloride.(none):/mnt/stuff/kernel/2.5/watchdog

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

patch from Bill Irwin

has_stopped_jobs() is completely unused. This patch removes
has_stopped_jobs() and renames __has_stopped_jobs() to has_stopped_jobs().

proc_fill_super() simply wants a count of processes, not threads.
This creates a per-cpu counter for it to use to determine that.