[PATCH 1/2] pci/irq: restore mask_bits in msi shutdown -v3

Yinghai found that kexec'ing a RHEL 5.1 kernel with 2.6.25-rc3+ kernels
prevents his NIC from working.  He bisected to

| commit 89d694b9
| Author: Thomas Gleixner <tglx@linutronix.de>
| Date:   Mon Feb 18 18:25:17 2008 +0100
|
|   genirq: do not leave interupts enabled on free_irq
|
|   The default_disable() function was changed in commit:
|
|    76d21601
|    genirq: do not mask interrupts by default
|

For MSI, default_shutdown will call mask_bit for msi device.  All mask bits
will left disabled after free_irq.  Then in the kexec case, the next kernel
can only use msi_enable bit, so all device's MSI can not be used.

So lets to restore the mask bit to its pci reset defined value (enabled) when
we disable the kernels use of msi to be a little friendlier to kexec'd kernels.

Extend msi_set_mask_bit to msi_set_mask_bits to take mask, so we can fully
restore that to 0x00 instead of 0xfe.
Signed-off-by: NYinghai Lu <yhlu.kernel@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@hobbes.lan>

Replace "dev" with "pdev" for consistency in DMA-mapping.txt.
Signed-off-by: NMatti Linnanvuori <mattilinnanvuori@yahoo.com>
Signed-off-by: NJesse Barnes <jbarnes@hobbes.lan>

We just need to forward declare it for struct pci_dev, not expose it outside of
__KERNEL__.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Fix an incorrect suggestion to pass NULL to pci_alloc_consistent
for PCI like buses where devices don't have struct pci_dev (like ISA, EISA).
Signed-off-by: NMatti Linnanvuori <mattilinnanvuori@yahoo.com>
Acked-by: NMatthew Wilcox <willy@linux.intel.com>
Signed-off-by: NJesse Barnes <jbarnes@hobbes.lan>

Update DMA mapping documentation to use 'pdev' rather than 'dev' in
example code that calls routines expecting 'struct pci_device *', since 'dev'
might make readers think they're passing 'struct device *' parameters.
Bug 10397.
Signed-off-by: NMatti Linnanvuori <mattilinnanvuori@yahoo.com>
Acked-by: NMatthew Wilcox <willy@linux.intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Current implementation of pciehp_write_cmd() always enables command
completed interrupt. But pciehp_write_cmd() is also used for clearing
command completed interrupt enable bit. In this case, we must not set
the command completed interrupt enable bit. To fix this bug, this
patch add the check to see if caller wants to change command complete
interrupt enable bit.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Current shpchp uses the combination of bus number and slot number as a
slot name. But it is not a good idea because bus number is not a
physical identifier but a logical identifier. This is against the shpc
specification. So remove the bus number from the physical identifier.

However, there are some platforms with the problem that it provides
the same slot number. For those platforms, this patch also introduces
new module option 'shpchp_slot_with_bus'. If it is specified, shpchp
uses the combination of bus number and slot number as a slot name.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

this_patch_makes_the_needlessly_global_pciehp_acpi_get_hp_hw_control_from_firmware_static

;)
Signed-off-by: NAdrian Bunk <bunk@kernel.org>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Clean up pciehp_ini(). This patch is trying to

  - Remove redundant capablity checks that were already done in PCIe
    port bus driver.
  - Separate the code only for debugging and make debug information
    easier to read.
  - Make the entire code easier to read and understand what it is doing.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

We must disable hotplug interrupt at controller relase time, otherwise
spurious interrupts might happen if any slot events occured (e.g. MRL
change) after unloading pciehp driver.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Hotplug interrupt is enabled at initialization and nobody clears it.
So we need to setup it in each command. This patch removes redundant
codes about this.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Current pciehp saves only 8bits of Slot Capability registers in
ctrl->ctrlcap. But it refers more than 8bit for checking EMI capability.
It is clearly a bug and EMI would never work. To fix this problem,
this patch saves full Slot Capability contens in ctrl->slot_cap. It also
reduce the redundant reads of Slot Capability register. And this pach
also cleans up the macros to check the slot capabilitys (e.g. MRL_SENS(),
and so on).
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Current pciehp implementaion clears hotplug events without waiting for
command completion. Because of this, events might not be cleared properly.
To prevent this problem, we must use pciehp_write_cmd() to write to
Slot Control register.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Fix the possible race condition between pcie_isr() and pciehp_write_cmd()
because of the lack of memory barrier.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Current pciehp implementation disables and re-enables hotplug interrupts
in its interrupt handler. This operation might be intend to guarantee
that interrupts for the events newly occured during previous events are
being handled will be successfully generated. But current implementaion
has the following prolems.

  - Current interrupt service routin clears status changes without
    waiting command completion. Because of this, events might not be
    cleared properly.
  - Current interrupt service routine clears status changes caused by
    disabling or enabling hotplug interrupts itself. This will lose new
    events that occurs during previous interrupts are being handled.
  - Current implementation doesn't have any serialization mechanism
    between the code to wait for command completion and the interrupt
    handler that clears the command completion events caused by itself.
    There is clearly race conditions between them, and it may cause
    the problem that waiting for command completion doesn't work for
    example.

To fix those problems, this patch stops disabling/re-enabling hotplug
interrupts in interrupt service routine. Instead of this, this patch
re-inspects Slot Status register after clearing what is presumed to
be the last bending interrupt in order to guarantee that all interrupt
events are serviced.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Current pciehp uses the combination of bus number and slot number as a
slot name. But it is not a good idea because bus number is not a
physical identifier but a logical identifier. This is against the PCIE
specification. So remove the bus number from the physical identifier.

However, there are some platforms with the problem that it provides
the same slot number. For those platforms, this patch also introduces
new module option 'pciehp_slot_with_bus'. If it is specified, pciehp
uses the combination of bus number and slot number as a slot name.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

The PCI tree is now in git at
kernel.org:/pub/scm/linux/kernel/git/jbarnes/pci-2.6.git; add that info to
MAINTAINERS.

This patch adds Intel SCH chipsets (US15W, US15L, UL11L) PCI IDs, these
IDs will be used by following SCH driver patches.
Signed-off-by: NAlek Du <alek.du@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@jbarnes-t61.(none)>

Running 'make oldconfig' I just noticed that PCIEASPM defaults to
'y' in Kconfig even though the feature is both experimental and the
help text recommends that if you are unsure you say 'n'.
It seems to me that this really should default to 'n', not 'y' at the
moment.
The following patch makes that change. Please consider applying.
Signed-off-by: NJesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Signed-off-by: NChristoph Lameter <clameter@sgi.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
  [patch 7/7] vfs: mountinfo: show dominating group id
  [patch 6/7] vfs: mountinfo: add /proc/<pid>/mountinfo
  [patch 5/7] vfs: mountinfo: allow using process root
  [patch 4/7] vfs: mountinfo: add mount peer group ID
  [patch 3/7] vfs: mountinfo: add mount ID
  [patch 2/7] vfs: mountinfo: add seq_file_root()
  [patch 1/7] vfs: mountinfo: add dentry_path()
  [PATCH] remove unused label in xattr.c (noise from ro-bind)

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
  iwlwifi: Fix built-in compilation of iwlcore
  net: Unexport move_addr_to_{kernel,user}
  rt2x00: Select LEDS_CLASS.
  iwlwifi: Select LEDS_CLASS.
  leds: Do not guard NEW_LEDS with HAS_IOMEM
  [IPSEC]: Fix catch-22 with algorithm IDs above 31
  time: Export set_normalized_timespec.
  tcp: Make use of before macro in tcp_input.c
  hamradio: Remove unneeded and deprecated cli()/sti() calls in dmascc.c
  [NETNS]: Remove empty ->init callback.
  [DCCP]: Convert do_gettimeofday() to getnstimeofday().
  [NETNS]: Don't initialize err variable twice.
  [NETNS]: The ip6_fib_timer can work with garbage on net namespace stop.
  [IPV4]: Convert do_gettimeofday() to getnstimeofday().
  [IPV4]: Make icmp_sk_init() static.
  [IPV6]: Make struct ip6_prohibit_entry_template static.
  tcp: Trivial fix to correct function name in a comment in net/ipv4/tcp.c
  [NET]: Expose netdevice dev_id through sysfs
  skbuff: fix missing kernel-doc notation
  [ROSE]: Fix soft lockup wrt. rose_node_list_lock

PNP_MAX_IRQ is 2
If a device invokes pnpacpi_parse_allocated_irqresource() 0, 1, or 2 times, we are happy.
The 3rd time, we will fail and print "pnpacpi: exceeded the max number of IRQ resources: 2"
The 4th and subsequent calls (if this ever happened) would silently scribble on
irq_resource[2], which doesn't actualy exist.
Found-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This patch fixes problem in Makefile that prevented
built-in compilation of iwlcore

Commit that caused this problem: eadd3c4b ("iwlwifi: make Makefile
more concise")
Signed-off-by: NTomas Winkler <tomas.winkler@intel.com>
Signed-off-by: NYi Zhu <yi.zhu@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After the removal of the Solaris binary emulation the exports of 
move_addr_to_{kernel,user} are no longer used.
Signed-off-by: NAdrian Bunk <bunk@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The LEDS infrastructure itself does not require anything
that a platform dependant upon HAS_IOMEM.

The individual drivers do, but they are properly guarded
with the necessary platform dependencies.

One can even imagine a hypervisor based LED driver that
a platform without HAS_IOMEM might have.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Show peer group ID of nearest dominating group that has intersection
with the mount's namespace.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

[mszeredi@suse.cz] rewrite and split big patch into managable chunks

/proc/mounts in its current form lacks important information:

 - propagation state
 - root of mount for bind mounts
 - the st_dev value used within the filesystem
 - identifier for each mount and it's parent

It also suffers from the following problems:

 - not easily extendable
 - ambiguity of mountpoints within a chrooted environment
 - doesn't distinguish between filesystem dependent and independent options
 - doesn't distinguish between per mount and per super block options

This patch introduces /proc/<pid>/mountinfo which attempts to address
all these deficiencies.

Code shared between /proc/<pid>/mounts and /proc/<pid>/mountinfo is
extracted into separate functions.

Thanks to Al Viro for the help in getting the design right.
Signed-off-by: NRam Pai <linuxram@us.ibm.com>
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Allow /proc/<pid>/mountinfo to use the root of <pid> to calculate
mountpoints.

 - move definition of 'struct proc_mounts' to <linux/mnt_namespace.h>
 - add the process's namespace and root to this structure
 - pass a pointer to 'struct proc_mounts' into seq_operations

In addition the following cleanups are made:

 - use a common open function for /proc/<pid>/{mounts,mountstat}
 - surround namespace.c part of these proc files with #ifdef CONFIG_PROC_FS
 - make the seq_operations structures const
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Add a unique ID to each peer group using the IDR infrastructure.  The
identifiers are reused after the peer group dissolves.

The IDR structures are protected by holding namepspace_sem for write
while allocating or deallocating IDs.

IDs are allocated when a previously unshared vfsmount becomes the
first member of a peer group.  When a new member is added to an
existing group, the ID is copied from one of the old members.

IDs are freed when the last member of a peer group is unshared.

Setting the MNT_SHARED flag on members of a subtree is done as a
separate step, after all the IDs have been allocated.  This way an
allocation failure can be cleaned up easilty, without affecting the
propagation state.

Based on design sketch by Al Viro.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Add a unique ID to each vfsmount using the IDR infrastructure.  The
identifiers are reused after the vfsmount is freed.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Add a new function:

  seq_file_root()

This is similar to seq_path(), but calculates the path relative to the
given root, instead of current->fs->root.  If the path was unreachable
from root, then modify the root parameter to reflect this.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

[mszeredi@suse.cz] split big patch into managable chunks

Add the following functions:

  dentry_path()
  seq_dentry()

These are similar to d_path() and seq_path().  But instead of
calculating the path within a mount namespace, they calculate the path
from the root of the filesystem to a given dentry, ignoring mounts
completely.
Signed-off-by: NRam Pai <linuxram@us.ibm.com>
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
  [PATCH] get rid of __exit_files(), __exit_fs() and __put_fs_struct()
  [PATCH] proc_readfd_common() race fix
  [PATCH] double-free of inode on alloc_file() failure exit in create_write_pipe()
  [PATCH] teach seq_file to discard entries
  [PATCH] umount_tree() will unhash everything itself
  [PATCH] get rid of more nameidata passing in namespace.c
  [PATCH] switch a bunch of LSM hooks from nameidata to path
  [PATCH] lock exclusively in collect_mounts() and drop_collected_mounts()
  [PATCH] move a bunch of declarations to fs/internal.h

Commit 88452565 ("PCI: clean up resource
alignment management") didn't set the alignment information for the
cardbus window resources, causing their subsequent allocations to fail
miserably with a message like

  yenta_cardbus 0000:15:00.0: device not available because of BAR 7 [100:1ff] collisions
  yenta_cardbus: probe of 0000:15:00.0 failed with error -16

or similar.

This fixes it and clarifies the code a bit too (we used to have to use
the insane PCI bridge alignment logic that put the alignment in the
"start" field, this makes it use the slightly easier-to-understand
size-based alignment, and allows us to set the resource start to zero
until it gets allocated).
Reported-and-tested-by: NJeff Chua <jeff.chua.linux@gmail.com>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Prevent bus_remove_device() from crashing if dev->knode_bus has not been
initialized before it's called.

This can happen if the device_add() ended up breaking out early due to
an error, for example.
Signed-off-by: NRafael J. Wysocki <rjw@sisk.pl>
Acked-by: NGreg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Do not refuse to actually register children of suspended devices,
but still warn about attempts to do that.
Signed-off-by: NRafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>