Because we do not reserve space for the pci-x and pci-e state in struct
pci dev we need to dynamically allocate it.  However because we need
to support restore being called multiple times after a single save
it is never safe to free the buffers we have allocated to hold the
state.

So this patch modifies the save routines to first check to see
if we have already allocated a state buffer before allocating
a new one.  Then the restore routines are modified to not free
the state after restoring it.  Simple and it fixes some subtle
error path handling bugs, that are hard to test for.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Acked-by: Auke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

There are two ways pci_save_state and pci_restore_state are used.  As
helper functions during suspend/resume, and as helper functions around
a hardware reset event.  When used as helper functions around a hardware
reset event there is no reason to believe the calls will be paired, nor
is there a good reason to believe that if we restore the msi state from
before the reset that it will match the current msi state.  Since arch
code may change the msi message without going through the driver, drivers
currently do not have enough information to even know when to call
pci_save_state to ensure they will have msi state in sync with the other
kernel irq reception data structures.

It turns out the solution is straight forward, cache the state in the
existing msi data structures (not the magic pci saved things) and
have the msi code update the cached state each time we write to the hardware.
This means we never need to read the hardware to figure out what the hardware
state should be.

By modifying the caching in this manner we get to remove our save_state
routines and only need to provide restore_state routines.

The only fields that were at all tricky to regenerate were the msi and msi-x
control registers and the way we regenerate them currently is a bit dependent
upon assumptions on how we use the allow msi registers to be configured and used
making the code a little bit brittle.  If we ever change what cases we allow
or how we configure the msi bits we can address the fragility then.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Acked-by: Auke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* master.kernel.org:/home/rmk/linux-2.6-arm:
  [ARM] 4263/1: fix IXP4XX_NPE[ABC]_BASE_VIRT address
  [ARM] 4256/1: i.MX/MX1 SDHC fix/workaround of SD card recognition problems
  [ARM] 4255/1: i.MX/MX1 Correct MPU PLL reference clock value.
  [ARM] 4254/1: i.MX/MX1 CPU Frequency scaling honor boot loader set BCLK_DIV.
  [ARM] 4251/1: Fix sharpsl_pm dependency
  [ARM] 4250/1: Fix locomo backlight conversion error/compile failure
  [ARM] 4249/1: Fix tosa compile failure
  [ARM] 4248/1: lh7a40x: fix missing definitions for get_irqnr_preamble
  [ARM] 4247/1: Fix long name for cc9p9360dev
  ARM: OMAP: Fix OMAP2 dss2 so clk_set_parent works
  ARM: OMAP: Fix missing workqueue include in board-h2.c
  ARM: OMAP: Include missing header

This fixs address defines for IXP4XX_NPE[ABC]_BASE_VIRT.
They are defined as (IXP4XX_PERIPHERAL_BASE_PHYS + 0x[678]000) now,
but they should be defined as (IXP4XX_PERIPHERAL_BASE_VIRT + 0x[678]000). Note PHYS vs VIRT in IXP4XX_PERIPHERAL_BASE...
Signed-off-by: Milan Svoboda <msvoboda@ra.rockwell.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

The SDHC controllers cannot process shorter transfers.
They has to be handled as longer ones, but it such case CRC
error is evaluated. There was a case in the code still,
where this error is not ignored as it should to be process
these transfers.
Signed-off-by: Pavel Pisa <pisa@cmp.felk.cvut.cz>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

Only System PLL clock source is selectable by CSCR_SYSTEM_SEL
bit. MPU PLL is driven by 512*CLK32 for each case.
Signed-off-by: Pavel Pisa <pisa@cmp.felk.cvut.cz>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

The minimal bus clock prescaler should be kept at value
selected by the board / boot loader designer.
Switching frequency above startup limit could
lead to the external memory/devices misbehave.
Signed-off-by: Pavel Pisa <pisa@cmp.felk.cvut.cz>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

This reverts commit d2487cb4.

Russell King points out that it's obviously bogus, and I have to agree.
Not only does "irq" not even exist in that scope, but we obviously need
to free the irq that we actually requested, and that's IRQ_USB.
Reported-by: Russell King <rmk@arm.linux.org.uk>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Greg KH <greg@kroah.com>
Cc: David Brownell <david-b@pacbell.net>,
Cc: Milan Svoboda <msvoboda@ra.rockwell.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6:
  [IPV6] fix ipv6_getsockopt_sticky copy_to_user leak
  [IPV6]: Fix for ipv6_setsockopt NULL dereference
  [DCCP]: Initialise write_xmit_timer also on passive sockets
  [IPV4]: Fix rtm_to_ifaddr() error handling.

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
  [IA64] refresh config files
  [IA64] put kdump_find_rsvd_region in __init
  [IA64] Remove sparse warning from unwind code
  [IA64] add missing syscall trace clear
  [IA64] Cleanup in crash.c
  [IA64] kexec: declare ia64_mca_pal_base in mca.h rather than kexec.h
  [IA64] pci_get_legacy_ide_irq should return irq (not GSI)
  [IA64] whitespace fixes for include/asm-ia64/sal.h
  [IA64] Cache error recovery
  [IA64] Proper handling of TLB errors from duplicate itr.d dropins

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/driver-2.6:
  kobject: new_device->kref wasn't putted after error in kobject_move()
  driver core: export device_rename
  Remove devfs from MAINTAINERS
  Driver core: add device symlink back to sysfs
  Revert "driver core: refcounting fix"

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6:
  pci: fix section mismatch warning
  PCI: aer: fix section mismatch warning
  pcie: fix section mismatch warning
  PCI: allow multiple calls to pcim_pin_device()

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/usb-2.6: (23 commits)
  USB Elan FTDI: check for workqueue creation
  USB: fix spinlock recursion in cdc-acm.c
  USB: fix Unaligned access in EHCI driver
  USB: Product ID for FT232RL in ftdi_sio
  USBNET: DM9501: Add Corega FEther USB-TXC support.
  USB: ipaq.c: Additional devices
  USB: further fix for usb-serial
  USB: fix usb-serial device naming bug
  USB: RTS/DTR signal patch for airprime driver
  USB: ftdi_sio: use port_probe / port_remove thereby fixing access to the latency_timer
  usb-serial: fix shutdown / device_unregister order
  USB: add Additional PIDs in ftdi_sio
  USB: add QL355P power supply ids to fdti_sio
  USB: New device IDs for cp2101 driver
  USB: kill dead code from hub.c
  USB: ratelimit debounce error messages
  USB: pxa2xx_udc: fix hardcoded irq number
  UHCI: fix port resume problem
  USB: set the correct interval for interrupt URBs
  USB: goku_udc: Remove crude cache coherency code
  ...

Delay the read of the EC status register until
after the event that caused it occurs -- otherwise
it is possible to read and act on stale status that was
associated with the previous event.

Do this with a perpetually incrementing "event_count" to detect
when a new event occurs and it is safe to read status.

There is no workaround for polling mode -- it is inherently
exposed to reading and acting on stale status, since it
doesn't have an interrupt to tell it the event completed.

http://bugzilla.kernel.org/show_bug.cgi?id=8110Signed-off-by: Alexey Starikovskiy <alexey.y.starikovskiy@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

Avoid NULL pointer usage if workqueue creation failed.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cc: Pete Zaitcev <zaitcev@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

this fixes the spinlock recursion issue. The older fix was incomplete.
Signed-off-by: Oliver Neukum <oneukum@suse.de>
Acked-by: Pete Zaitcev <zaitcev@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

I get following warnings on spar64:
Kernel unaligned access at TPC[1000c9e4] ehci_hub_control+0x54c/0x68c [ehci_hcd]

Despite of the comment in the patched code, the type cast used there
does make unaligned access. The fix was made as it's done in
ohci-hub.c.
Signed-off-by: Max Dmitrichenko <dmitrmax@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Here is a patch adding the PID for the FT232RL to ftdi_sio. The patch
generates a warning during compilation because get_ftdi_divisor doesn't
explicitly handle the FT232RL with this patch, so I guess you don't want
to use it in its current state. It is all I could come up with with the
knowledge I have of the drivers at the moment, though, and I hope you
can have some use for it at least. It works fine with my DLP-TILT with
an FT232RL.

From: Gard Spreemann <spreeman@stud.ntnu.no>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Additional devices
Signed-off-by: Andre Spahlinger <uenz@gmx.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

this is a fix for the outstanding usb-serial issues.
Signed-off-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Am Montag, 26. Februar 2007 15:16 schrieb Craig Schlenter:
> Hi Greg
> 
> 34ef50e5 is definitely
> the source of the problem. Reverting that makes the
> ftdi port show up as ttyUSB0 again for me and it
> can actually be opened.

This patch should fix the issue.
Signed-off-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Craig Schlenter <craig@codefountain.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

I encountered some problems with the airprime driver in use with a Novatel
Merlin XU870:

Closing an open Connection to e.g. /dev/ttyUSB0 doesn't reset the
RTS/DTR lines of the Modem.  Consequently, when I use minicom to
establish a connection by "ATD*99#" the modem doesn't hang up even if i
exit minicom and so I cannot reuse the modem unless I remove it and plug
it in again.

With the attached patch, the RTS/DTR lines are resetted on a close.  The
code was mainly taken from the option.c driver.

Convert all the port specific code in attach / shutdown to use the new
port_probe / port_register callbacks from device_register /
device_unregister allowing adding the sysfs attributes to be added at
the correct time and to the serial port device itself, instead of to
the unadorned usb device, avoiding a NULL dereference.
Signed-off-by: Jim Radford <radford@blackbean.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Ensure that the ->port_remove() callbacks get called before the
->shutdown() callback which makeing the order symmetric with
->attach() being called before ->port_probe().
Signed-off-by: Jim Radford <radford@blackbean.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

I've developed some devices with FTDI chips (FT232xx). FTDI was so kind
to give some own PID's which I can use together with their VID. Some of
the devices are already very popular here and I have customers from
universities, institutes .....

I use the FTDI VID 0x0403. My PID's are:

0xff38 - IBS US485 (USB<-->RS422/485 interface)
0xff39 - IBS PIC-Programmer
0xff3a - IBS Card reader for PCMCIA SRAM-cards
0xff3b - IBS PK1 - Particel counter
0xff3c - IBS RS232 - Monitor
0xff3d - APP 70 (dust monitoring system)
0xff3e - IBS PEDO-Modem (RF modem 868.35 MHz)
0xff3f - future device

The company is "IBS Ing.-Buero Schleusener".


From: Thomas Schleusener <thomas@be-schl.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Here are two new device IDs for the cp2101 driver.
The diff is with linus's tree as of this evending.

From: Jon K Hellan <hellan@acm.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

this kills residual obsoletet code from hub.c
Signed-off-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

flaky hardware can cause a lot of debounce failed messages. To limit
the performance impact, a ratelimit should be used.
Signed-off-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>