There's nopoint in walking through a lot of tmpfs or ramdisk pages when
we're trying to clean memory.  So if a memory-backed inode is
discovered during writeback, skip the entire superblock.

Patch from Mark Fasheh <mark.fasheh@oracle.com> (plus a few cleanups
and a speedup from yours truly)

Adds the semtimedop() function - semop with a timeout.  Solaris has
this.  It's apparently worth a couple of percent to Oracle throughput
and given the simplicity, that is sufficient benefit for inclusion IMO.

This patch hooks up semtimedop() only for ia64 and ia32.

The pte_chain_unlock() needs to be outside the ifdef.

The read_zero() implementation for !CONFIG_MMU was very inefficient.
This sped-up version has been tested and acked by Greg Ungerer.

Back out the sys_syslog()-based printk-from-userspace and replace
it with Ben's /proc/kmsg version.

Requires a `mknod /dev/kmsg c 1 11'.

Patch from Robert Love <rml@tech9.net>

We can never get rid of it if we do not deprecate it - so do so and
print a stern warning to those who still run bdflush daemons.

The PF_MEMALLOC handling got broken somewhere, and it is now possible
for a PF_MEMALLOC process to reenter page reclaim.

Change it to fail the allocation if we're PF_MEMALLOC and there are
zero pages free.

This removes calls to function which disappeared during last Iphase
driver update.  Since this update, Iphase driver has been using plain
modern pci style init.

Problem wasn't noticed until Adrian Bunk tried to build non-modular kernel
(I only tested the modularized driver). Everybody else seemed happy :o)

scrup is using memcpy even when the memory areas src, dest overlap.  The
key is to use memmove which handles overlapping memory gracefully.

into home.transmeta.com:/home/torvalds/v2.5/linux

2.5 currently tries to register disk sda twice. Not nice and now we use
sysfs to do name to dev_t mapping, I couldnt mount my root filesystem.

This driver does not need to use atomic operations on local variables.

Trivial but annoying: two printk() calls in drivers/scsi/hosts.c are
missing '\n'

Also, for some reason I have not yet investigated, shost_tp->name is
NULL here.  This should not be, eh?  If it can be, we should do
something to tidy up the printing of it.

This error message is uber annoying and needs to go.  Non-root can flood
the console with this junk on invalid SCSI CD-ROM ioctl(), and that is
exactly what gnome-cd does.

An illegal ioctl() returns an error to the program.  That is sufficient
- we do not need KERN_ERROR warnings all over the place.  Especially
when any user can cause them at any rate.

In some configurations, parport and bttv request a module inside their
module_init function.  Drop the lock around mod->init(), change
module->live to module->state so we can detect modules which are in
init.

module-init-tools 0.9 and newer supply a replacement depmod, so
it's safe to run again.

Also, some external programs like PCMCIA and mkinitrd really want the
directory hierarchy in /lib/modules back again: it makes no difference
to the tools (since 0.9), so revert it.

into home.transmeta.com:/home/torvalds/v2.5/linux

Relatively straightforward fixes for intermezzo problems in 2.5.50.  I
think all of them related to:
 - two missing headers
 - use of timespec instead of time_t.

into home.transmeta.com:/home/torvalds/v2.5/linux

into home.transmeta.com:/home/torvalds/v2.5/linux

I made the changes to the set/get_registers code.

Cset exclude: rth@dorothy.sfbay.redhat.com|ChangeSet|20021207231352|30637

 - remove NATIVE_CHAR typedef
 - remove ACPI_{GET,VALID}_ADDRESS macros
 - fix memory corruption in deletion of a static AML buffer
 - fix fault caused by 0-length AML
 - fix user-buffer overwrite/corruption of buffer is too small
 - fix buffer-to-string conversion

This allowed a lock to be removed.
Also removed the MOD_* functions, and some remove logic was cleaned
up by Oliver Neukum.

into groveronline.com:/root/bk/linux-acpi

here is an id-patch to get the Sony Cybershot F717 6meg pixel digital
camera working with the standard usb-storage device driver.

Here are minimal usb_find_interface() patches for the core, usblp and
scanner.

Basic design is:
- device major (USB_MAJOR for now) and minor are stored in probe()
   function to struct usb_interface as kdev_t
- open() can use new core function usb_find_interface() to find matching
  device in drivers device list
- disconnect() will set kdev_t struct usb_interface to NODEV, so open
  wont open it anymore without new probe()

I tested these patches and they work for me. I will work on small patches
of other work in these drivers (like removal of lock_kernel/unlock_kernel
in usblp, fixing the disconnect problems in both drivers etc.). Those
patches would be very small too, but there will be quite many.

Thanks to Pat Mochel for pointing this out to me.

This patch disables the USB Bluetooth TTY driver (bluetty.o) from
the drivers/usb/class directory if the Linux Bluetooth subsystem
is selected.

The logic surrounding COMMIT_Holdlock was well-intentioned to reduce
latency when deleting files, but it can also result in a hang that I don't
have good fix for.  I don't think removing this will hurt our overall
performance very much.

jfs_destroy_inode was assuming that active_ag should never be set coming into
this routine.  Since it's possible for file to be extended after the file
descriptor has been closed, we need to allow the possibility.  (Dirty pages
of memory mapped files can be written after the file has been closed.)

This means that all individual usb-serial ports show up as their
own devices in the driver model tree.

sorry to bother you again - now that 2.4.20 is out, is there any
chance to include this in 2.4.21?  I've been trying since 2.4.19,
a few other UNUSUAL_DEV entries were added, but not this one...

The device works fine with the patch (and doesn't work at all without
it) for me and a few other people (devices with different "marketing"
names, the same vendor:device id), no one has reported any problems.
The patch has been in the 2.4-ac tree for a while, too.