Thanks to John Moses <jmoses@lanl.gov> for the information.

This patch clears out the device queue when a unit is removed.

This code implements the setting of devices offline during the removal
phase.

This patch makes us hold the host reference count a little bit longer in
the /proc interface code.  We were releasing it too early before.

  speedtouch: reject outgoing packets earlier when the firmware is not loaded.

  speedtouch: allocate send urbs in udsl_usb_probe rather than in udsl_usb_data_init.
  Since this diminishes udsl_usb_data_init down to almost nothing, roll it into the one
  place it was used.  Get rid of the semaphore Oliver put it - it is no longer needed.


 speedtouch.c |   86 ++++++++++++++++++++++++++---------------------------------
 1 files changed, 38 insertions(+), 48 deletions(-)

  speedtouch: change data_started to firmware_loaded, which is what it actually
  means, plus some minor related changes.

  speedtouch: a pile of cosmetic changes to make me feel happier (no code changes).

  speedtouch: wait for receive urb completion handlers to finish after calling
  usb_unlink_urb.

  speedtouch: more robust handling of receive urb failure: retry failed urbs whenever
  a new connection is opened.  This should work well with pppd's persist option.

Rediffed version of the original patch - no sk_buff on the stack this time.

  speedtouch: recycle the receive urb's buffer.  Currently, every time a receive urb
    completes, its old buffer is thrown away and replaced with a new one.  This patch
    performs the minor changes needed to reuse the old buffer.

  speedtouch: move all processing of receive urbs to udsl_atm_processqueue.  This has
  several advantages, as will be seen in the next few patches.  The most important is
  that it makes it easy to reuse of the urb's buffer (right now a new buffer is
  allocated every time the urb completes).  By the way, this patch is much smaller than
  it looks: most of the bulk is due to indentation changes.

This patch prints the vendor + product ids of the scanner after it has
been successfully detected.

Also the annoying error message about "Scanner device is already open"
was downgraded to a dbg. Scanning for devices while one scanner device
was open produced several 100 error messages in syslog.

This patch changes the maintainer from Brian Beattie to Henning
Meier-Geinitz and adds a link to the documentation and website.

This is a minor cleanup to convert 8 spaces into tabs.  There is no
functional change here.

This patch removes the US_FL_DEV_ATTACHED flag, which is now rendered
obsolete by the new hotplug system.

It also adds a comment or two about areas of code that need to be
re-examined.

This patch does the following:
(o) Add comments showing what needs to be done to complete the hot-unplug
    system.
(o) Add a BUG_ON() for (what is now) a critical failure case.
(o) Make certain that a debug print happens even if a usb_get_intfdata()
    crashes.
(o) Add an un-necessary up() to balance a down, for the auto-code-checkers.

It should fix the OOPS on attach.

This fixes a silly error where I fail to initialize a pointer early enough
for the scanning code.  If this isn't a perfect example of why
scsi_register() and scsi_add_host() aren't two separate functions, I don't
know what is.  :)

Oh, and I added a couple of comments, too.

  -  Fix an OOPS by moving the setting of the hostdata[] pointer to _before_
     the device scan starts.

This patch goes on top of the last one.  It fixes a typo in the test for
scsi_register() failure.

  -- reversed the logic of failure test for scsi_register()

The attached patch is my first implementation of SCSI hotplugging.

It's only been tested that it compiles, as I can't get the current
linux-2.5 tree from linuxusb to boot.  It dies _very_ early.  Greg, I'm not
sure if you'll want to apply this.  Linus seemed to want this very much,
and it is 2.5.x... I say go for it, but I can understand if you have
reservations.

I would definately like to see this tested by anyone who can get a kernel
to boot.

This patch is quite large.  Lots of things had to be changed.  Among them:

(o) The proc interface now uses the host number to look up the SCSI host
    structure, and then finds the usb-storage structure from that.
(o) The SCSI interface has been changed.  The code flow is now much
    clearer, as more work is done from the USB probe/detach functions than
    from auxillary functions.
(o) Names have been changed for newer conventions
(o) GUIDs have been removed
(o) The linked-list of devices has been removed, and it's associated
    semaphore
(o) All code dealing with re-attaching a device to it's old association has
    been removed
(o) Some spaces changed to tabs
(o) usb-storage now takes one directory under /proc/scsi instead of
    one per virtual-HBA
(o) All control threads now have the same name.  This could be changed back
    to the old behavior, if enough people want it.

Known problems:
(o) Testing, testing, testing
(o) More dead code needs to be cut
(o) It's a unclear how a LLD is supposed to cut off the flow of
    commands, so that the unregister() call always succeeds.  SCSI folks
    need to work on this.
(o) Probing needs to be broken down into smaller functions, probably.

Thanks to Matt Dharm and David Brownell for tracking this bug down.

Truncates can take a very long time.  Especially if there is a lot of
writeout happening, because truncate must wait on in-progress I/O.

And sys_unlink() is performing that truncate while holding the parent
directory's i_sem.  This basically shuts down new accesses to the entire
directory until the synchronous I/O completes.

In the testing I've been doing, that directory is /tmp, and this hurts.

So change sys_unlink() to perform the actual truncate outside i_sem.

When there is a continuous streaming write to the same disk, this patch
reduces the time for `make -j4 bzImage' from 370 seconds to 220.

When a throttled writer is performing writeback, and it encounters an inode
which is already under writeback it is forced to wait on the inode.  So that
process sleeps until whoever is writing it out finishes the writeout.

Which is OK - we want to throttle that process, and another process is
currently pumping data at the disk anyway.

But in one situations the delays are excessive.  If one process is performing
a huge linear write, other processes end up waiting for a very long time
indeed.  It appears that this is because the writing process just keeps on
hogging the CPU, returning to userspace, generating more dirty data, writing
it out, sleeping in get_request_wait, etc.  All other throttled dirtiers get
starved.

So just remove the wait altogether if it is just a memory-cleansing writeout.
 The calling process will then throttle in balance_dirty_pages()'s call to
blk_congestion_wait().

We do not need to pass the `wait' argument down to __sync_single_inode().
That information is now present at wbc->sync_mode.

- Fix dead comment in load_elf_interp() (Dave Airlie)

- Add some (hard-won) commentary around the early SET_PERSONALITY() in
  load_elf_binary().

- Remove dead hugetlb prototype.

- Fix some silliness in hugetlbpage.c

Patch from Pavel Machek <pavel@ucw.cz>

scsi_eh_* needs to run even during suspend because suspend does not prevent a
hard disk from reporting an error.

Patch from miles@lsi.nec.co.jp (Miles Bader)

Otherwise it won't compile.  I guesss this used to work because <asm/page.h>
was included somewhere to get the BUG macros, but now with the advent of
<asm/bug.h> that's changed.

Patch from john stultz <johnstul@us.ibm.com>

Adds some lost-tick compensation code, which handles the case where time
accounting goes wrong due to interrupts being disabled for longer than two
ticks.

This patch solves the problem by checking when an interrupt occurs if
timer->get_offset() is a value greater then 2 ticks.  If so, it increments
jiffies appropriately.

I was concerned that we'd be better off finding and fixing the misbehaving
drivers, but it turns out that the main culprits are system management cards
over which the kernel has no control.

However John has added some debug code which will drop a backtrace on the
first five occurrences which will allow us to find-and-fix bad drivers if
overruns _are_ due to Linux software.  (I disabled this - it was irritating
me.  Dave Hansen has a patch which allows it to be turned on via a kernel
boot parameter, like the x86_64 equiv).

Patch from "Randy.Dunlap" <rddunlap@osdl.org>

This fixes a few #warning's that gcc 2.96 complains about having
unmatched single-quote marks.  (warnings on #warnings)

Patch from Russell King <rmk@arm.linux.org.uk>

Have a couple of extra warnings:

fs/binfmt_elf.c: In function `create_elf_tables':
fs/binfmt_elf.c:239: warning: initialization makes integer from pointer without a cast
fs/binfmt_elf.c:249: warning: initialization makes integer from pointer without a cast

#ifndef elf_addr_t
#define elf_addr_t unsigned long
#endif

        elf_addr_t *argv, *envp;

        __put_user(NULL, argv);
        __put_user(NULL, envp);

It would therefore appear that x86 __put_user is not properly type-checking
the arguments to __put_user().

Here's a patch which fixes the warning (but doesn't fix x86's type-check
challenged __put_user implementation).

Patch from Matthew Dobson <colpatch@us.ibm.com>

When I originally wrote the patches implementing the in-kernel topology
macros, they were meant to be called as a second layer of functions,
sans underbars.  This additional layer was deemed unnecessary and
summarily dropped.  As such, carrying around (and typing!) all these
extra underbars is quite pointless.  Here's a patch to nip this in the
(sorta) bud.  The macros only appear in 16 files so far, most of them
being the definitions themselves.

Patch from Dave Olien <dmo@osdl.org>

The following patch implements retry on media errors for the DAC960 driver.
On such media errors, the DAC960 apparently doesn't report how much of the
transfer may have been successful before the error was encountered.

This type of error should be rare on healthy hardware, especially if the
disks are stripped or mirrored.  But, when large transfers are submitted to
the controller, it's especially bad to have to fail the entire transfer
because one disk sector may have been bad.

Patch from William Lee Irwin III <wli@holomorphy.com>

CONFIG_X86_NUMA no longer exists.  This changes the MAX_IO_APICS definition
to 32, where it is required to be so large on NUMA-Q in order to boot.

Patch from William Lee Irwin III <wli@holomorphy.com>

will_become_orphaned_pgrp()'s sole use is is_orphaned_pgrp().  Fold its body
into is_orphaned_pgrp(), rename __will_become_orphaned_pgrp(), and adjust
callers.  Code shrinkage plus some relief from underscore-itis.