PCI/ASPM: Fix deadlock when enabling ASPM

A last minute revert in 6.7-final introduced a potential deadlock when
enabling ASPM during probe of Qualcomm PCIe controllers as reported by
lockdep:

============================================
WARNING: possible recursive locking detected
6.7.0 #40 Not tainted
--------------------------------------------
kworker/u16:5/90 is trying to acquire lock:
ffffacfa78ced000 (pci_bus_sem){++++}-{3:3}, at: pcie_aspm_pm_state_change+0x58/0xdc

but task is already holding lock:
ffffacfa78ced000 (pci_bus_sem){++++}-{3:3}, at: pci_walk_bus+0x34/0xbc

other info that might help us debug this:
Possible unsafe locking scenario:

CPU0
----
lock(pci_bus_sem);
lock(pci_bus_sem);

*** DEADLOCK ***

Call trace:
print_deadlock_bug+0x25c/0x348
__lock_acquire+0x10a4/0x2064
lock_acquire+0x1e8/0x318
down_read+0x60/0x184
pcie_aspm_pm_state_change+0x58/0xdc
pci_set_full_power_state+0xa8/0x114
pci_set_power_state+0xc4/0x120
qcom_pcie_enable_aspm+0x1c/0x3c [pcie_qcom]
pci_walk_bus+0x64/0xbc
qcom_pcie_host_post_init_2_7_0+0x28/0x34 [pcie_qcom]

The deadlock can easily be reproduced on machines like the Lenovo ThinkPad
X13s by adding a delay to increase the race window during asynchronous
probe where another thread can take a write lock.

Add a new pci_set_power_state_locked() and associated helper functions that
can be called with the PCI bus semaphore held to avoid taking the read lock
twice.

Link: https://lore.kernel.org/r/ZZu0qx2cmn7IwTyQ@hovoldconsulting.com
Link: https://lore.kernel.org/r/20240130100243.11011-1-johan+linaro@kernel.org
Fixes: f93e71aea6c6 ("Revert "PCI/ASPM: Remove pcie_aspm_pm_state_change()"")
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: <stable@vger.kernel.org> # 6.7

PCI: Create device tree node for bridge

The PCI endpoint device such as Xilinx Alveo PCI card maps the register
spaces from multiple hardware peripherals to its PCI BAR. Normally,
the PCI core discovers devices and BARs using the PCI enumeration process.
There is no infrastructure to discover the hardware peripherals that are
present in a PCI device, and which can be accessed through the PCI BARs.

Apparently, the device tree framework requires a device tree node for the
PCI device. Thus, it can generate the device tree nodes for hardware
peripherals underneath. Because PCI is self discoverable bus, there might
not be a device tree node created for PCI devices. Furthermore, if the PCI
device is hot pluggable, when it is plugged in, the device tree nodes for
its parent bridges are required. Add support to generate device tree node
for PCI bridges.

Add an of_pci_make_dev_node() interface that can be used to create device
tree node for PCI devices.

Add a PCI_DYNAMIC_OF_NODES config option. When the option is turned on,
the kernel will generate device tree nodes for PCI bridges unconditionally.

Initially, add the basic properties for the dynamically generated device
tree nodes which include #address-cells, #size-cells, device_type,
compatible, ranges, reg.

Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Lizhi Hou <lizhi.hou@amd.com>
Link: https://lore.kernel.org/r/1692120000-46900-3-git-send-email-lizhi.hou@amd.com
Signed-off-by: Rob Herring <robh@kernel.org>

PCI: move OF status = "disabled" detection to dev->match_driver

The blamed commit has broken probing on
arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi when &enetc_port0
(PCI function 0) has status = "disabled".

Background: pci_scan_slot() has logic to say that if the function 0 of a
device is absent, the entire device is absent and we can skip the other
functions entirely. Traditionally, this has meant that
pci_bus_read_dev_vendor_id() returns an error code for that function.

However, since the blamed commit, there is an extra confounding
condition: function 0 of the device exists and has a valid vendor id,
but it is disabled in the device tree. In that case, pci_scan_slot()
would incorrectly skip the entire device instead of just that function.

In the case of NXP LS1028A, status = "disabled" does not mean that the
PCI function's config space is not available for reading. It is, but the
Ethernet port is just not functionally useful with a particular SerDes
protocol configuration (0x9999) due to pinmuxing constraints of the Soc.
So, pci_scan_slot() skips all other functions on the ENETC ECAM
(enetc_port1, enetc_port2, enetc_mdio_pf3 etc) when just enetc_port0 had
to not be probed.

There is an additional regression introduced by the change, caused by
its fundamental premise. The enetc driver needs to run code for all PCI
functions, regardless of whether they're enabled or not in the device
tree. That is no longer possible if the driver's probe function is no
longer called. But Rob recommends that we move the of_device_is_available()
detection to dev->match_driver, and this makes the PCI fixups still run
on all functions, while just probing drivers for those functions that
are enabled. So, a separate change in the enetc driver will have to move
the workarounds to a PCI fixup.

Fixes: 6fffbc7ae137 ("PCI: Honor firmware's device disabled status")
Link: https://lore.kernel.org/netdev/CAL_JsqLsVYiPLx2kcHkDQ4t=hQVCR7NHziDwi9cCFUFhx48Qow@mail.gmail.com/
Suggested-by: Rob Herring <robh@kernel.org>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

PCI: Make pci_bus_for_each_resource() index optional

Refactor pci_bus_for_each_resource() in the same way as
pci_dev_for_each_resource(). This allows the index to be hidden inside the
implementation so the caller can omit it when it's not used otherwise.

No functional changes intended.

Link: https://lore.kernel.org/r/20230330162434.35055-6-andriy.shevchenko@linux.intel.com
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Krzysztof Wilczyński <kw@linux.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>

PCI: s390: Fix use-after-free of PCI resources with per-function hotplug

On s390 PCI functions may be hotplugged individually even when they
belong to a multi-function device. In particular on an SR-IOV device VFs
may be removed and later re-added.

In commit a50297cf8235 ("s390/pci: separate zbus creation from
scanning") it was missed however that struct pci_bus and struct
zpci_bus's resource list retained a reference to the PCI functions MMIO
resources even though those resources are released and freed on
hot-unplug. These stale resources may subsequently be claimed when the
PCI function re-appears resulting in use-after-free.

One idea of fixing this use-after-free in s390 specific code that was
investigated was to simply keep resources around from the moment a PCI
function first appeared until the whole virtual PCI bus created for
a multi-function device disappears. The problem with this however is
that due to the requirement of artificial MMIO addreesses (address
cookies) extra logic is then needed to keep the address cookies
compatible on re-plug. At the same time the MMIO resources semantically
belong to the PCI function so tying their lifecycle to the function
seems more logical.

Instead a simpler approach is to remove the resources of an individually
hot-unplugged PCI function from the PCI bus's resource list while
keeping the resources of other PCI functions on the PCI bus untouched.

This is done by introducing pci_bus_remove_resource() to remove an
individual resource. Similarly the resource also needs to be removed
from the struct zpci_bus's resource list. It turns out however, that
there is really no need to add the MMIO resources to the struct
zpci_bus's resource list at all and instead we can simply use the
zpci_bar_struct's resource pointer directly.

Fixes: a50297cf8235 ("s390/pci: separate zbus creation from scanning")
Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com>
Reviewed-by: Matthew Rosato <mjrosato@linux.ibm.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://lore.kernel.org/r/20230306151014.60913-2-schnelle@linux.ibm.com
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>

PCI: Skip allocate_resource() if too little space available

pci_bus_alloc_from_region() allocates MMIO space by iterating through all
the resources available on the bus. The available resource might be
reduced if the caller requires 32-bit space or we're avoiding BIOS or E820
areas.

Don't bother calling allocate_resource() if we need more space than is
available in this resource. This prevents some pointless and annoying
messages about avoided areas.

Link: https://lore.kernel.org/r/20221208190341.1560157-3-helgaas@kernel.org
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>

PCI: Add device even if driver attach failed

device_attach() returning failure indicates a driver error while trying to
probe the device. In such a scenario, the PCI device should still be added
in the system and be visible to the user.

When device_attach() fails, merely warn about it and keep the PCI device in
the system.

This partially reverts ab1a187bba5c ("PCI: Check device_attach() return
value always").

Link: https://lore.kernel.org/r/20200706233240.3245512-1-rajatja@google.com
Signed-off-by: Rajat Jain <rajatja@google.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: stable@vger.kernel.org # v4.6+

PCI: Unexport pci_bus_get() and pci_bus_put()

pci_bus_get() and pci_bus_put() are not used by a loadable kernel module
and do not need to be exported.

These were exported by fe830ef62ac6 ("PCI: Introduce pci_bus_{get|put}() to
manage PCI bus reference count"), but there are no loadable modules in the
tree that use them.

Link: https://lore.kernel.org/r/20190717182353.45557-1-skunberg.kelsey@gmail.com
Signed-off-by: Kelsey Skunberg <skunberg.kelsey@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

PCI: Replace dev_printk(KERN_DEBUG) with dev_info(), etc

Replace dev_printk(KERN_DEBUG) with dev_info(), etc to be more consistent
with other logging and avoid checkpatch warnings.

The KERN_DEBUG messages could be converted to dev_dbg(), but that depends
on CONFIG_DYNAMIC_DEBUG and DEBUG, and we want most of these messages to
*always* be in the dmesg log.

Link: https://lore.kernel.org/lkml/1555733240-19875-1-git-send-email-mohankumar718@gmail.com
Signed-off-by: Mohan Kumar <mohankumar718@gmail.com>
[bhelgaas: commit log]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Replace printk(KERN_INFO) with pr_info(), etc

Replace printk() with pr_*() to be more consistent with other logging and
avoid checkpatch warnings.

Link: https://lore.kernel.org/lkml/1555733026-19609-1-git-send-email-mohankumar718@gmail.com
Link: https://lore.kernel.org/lkml/1555733130-19804-1-git-send-email-mohankumar718@gmail.com
Signed-off-by: Mohan Kumar <mohankumar718@gmail.com>
[bhelgaas: squash in similar changes from second patch in series]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Fix is_added/is_busmaster race condition

When a PCI device is detected, pdev->is_added is set to 1 and proc and
sysfs entries are created.

When the device is removed, pdev->is_added is checked for one and then
device is detached with clearing of proc and sys entries and at end,
pdev->is_added is set to 0.

is_added and is_busmaster are bit fields in pci_dev structure sharing same
memory location.

A strange issue was observed with multiple removal and rescan of a PCIe
NVMe device using sysfs commands where is_added flag was observed as zero
instead of one while removing device and proc,sys entries are not cleared.
This causes issue in later device addition with warning message
"proc_dir_entry" already registered.

Debugging revealed a race condition between the PCI core setting the
is_added bit in pci_bus_add_device() and the NVMe driver reset work-queue
setting the is_busmaster bit in pci_set_master(). As these fields are not
handled atomically, that clears the is_added bit.

Move the is_added bit to a separate private flag variable and use atomic
functions to set and retrieve the device addition state. This avoids the
race because is_added no longer shares a memory location with is_busmaster.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=200283
Signed-off-by: Hari Vyas <hari.vyas@broadcom.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Lukas Wunner <lukas@wunner.de>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>

PCI: remove CONFIG_PCI_BUS_ADDR_T_64BIT

This symbol is now always identical to CONFIG_ARCH_DMA_ADDR_T_64BIT, so
remove it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Tidy comments

Remove pointless comments that tell us the file name, remove blank line
comments, follow multi-line comment conventions. No functional change
intended.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Add SPDX GPL-2.0 when no license was specified

b24413180f56 ("License cleanup: add SPDX GPL-2.0 license identifier to
files with no license") added SPDX GPL-2.0 to several PCI files that
previously contained no license information.

Add SPDX GPL-2.0 to all other PCI files that did not contain any license
information and hence were under the default GPL version 2 license of the
kernel.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

PCI: Add wrappers for dev_printk()

Add PCI-specific dev_printk() wrappers and use them to simplify the code
slightly. No functional change intended.

Signed-off-by: Frederick Lawler <fred@fredlawl.com>
[bhelgaas: squash into one patch]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Autosense device removal in pci_bridge_d3_update()

The algorithm to update the flag indicating whether a bridge may go to D3
makes a few optimizations based on whether the update was caused by the
removal of a device on the one hand, versus the addition of a device or the
change of its D3cold flags on the other hand.

The information whether the update pertains to a removal is currently
passed in by the caller, but the function may as well determine that itself
by examining the device in question, thereby allowing for a considerable
simplification and reduction of the code.

Out of several options to determine removal, I've chosen the function
device_is_registered() because it's cheap: It merely returns the
dev->kobj.state_in_sysfs flag. That flag is set through device_add() when
the root bus is scanned and cleared through device_remove(). The call to
pci_bridge_d3_update() happens after each of these calls, respectively, so
the ordering is correct.

No functional change intended.

Tested-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

PCI: Put PCIe ports into D3 during suspend

Currently the Linux PCI core does not touch power state of PCI bridges and
PCIe ports when system suspend is entered. Leaving them in D0 consumes
power unnecessarily and may prevent the CPU from entering deeper C-states.

With recent PCIe hardware we can power down the ports to save power given
that we take into account few restrictions:

- The PCIe port hardware is recent enough, starting from 2015.

- Devices connected to PCIe ports are effectively in D3cold once the port
is transitioned to D3 (the config space is not accessible anymore and
the link may be powered down).

- Devices behind the PCIe port need to be allowed to transition to D3cold
and back. There is a way both drivers and userspace can forbid this.

- If the device behind the PCIe port is capable of waking the system it
needs to be able to do so from D3cold.

This patch adds a new flag to struct pci_device called 'bridge_d3'. This
flag is set and cleared by the PCI core whenever there is a change in power
management state of any of the devices behind the PCIe port. When system
later on is suspended we only need to check this flag and if it is true
transition the port to D3 otherwise we leave it in D0.

Also provide override mechanism via command line parameter
"pcie_port_pm=[off|force]" that can be used to disable or enable the
feature regardless of the BIOS manufacturing date.

Tested-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

PCI: Add devm_request_pci_bus_resources()

Several host bridge drivers iterate through the list of bridge windows to
request resources. Several others don't request the window resources at
all.

Add a devm_request_pci_bus_resources() interface to make it easier for
drivers to request all the window resources. Export to GPL modules (from
Arnd Bergmann <arnd@arndb.de>).

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Do not treat EPROBE_DEFER as device attach failure

Linux 4.5 introduced a behavioral change in device probing during the
suspend process with commit 013c074f8642 ("PM / sleep: prohibit devices
probing during suspend/hibernation"): It defers device probing during the
entire suspend process, starting from the prepare phase and ending with the
complete phase. A rule existed before that "we rely on subsystems not to
do any probing once a device is suspended" but it is enforced only now
(Alan Stern, https://lkml.org/lkml/2015/9/15/908).

This resulted in a WARN splat if a PCI device (e.g., Thunderbolt) is
plugged in while the system is asleep: Upon waking up, pciehp_resume()
discovers new devices in the resume phase and immediately tries to bind
them to a driver. Since probing is now deferred, device_attach() returns
-EPROBE_DEFER, which provoked a WARN in pci_bus_add_device().

Linux 4.6-rc1 aggravates the situation with commit ab1a187bba5c ("PCI:
Check device_attach() return value always"): If device_attach() returns a
negative value, pci_bus_add_device() now removes the sysfs and procfs
entries for the device and pci_bus_add_devices() subsequently locks up with
a BUG. Even with the BUG fixed we're still in trouble because the device
remains on the deferred probing list even though its sysfs and procfs
entries are gone and its children won't be added.

Fix by not interpreting -EPROBE_DEFER as failure. The device will be
probed eventually (through device_unblock_probing() in dpm_complete()) and
there is proper locking in place to avoid races (e.g., if devices are
unplugged again und thus deleted from the system before deferred probing
happens, I have tested this). Also, those functions which dereference
dev->driver (e.g. pci_pm_*()) do contain proper NULL pointer checks. So it
seems safe to ignore -EPROBE_DEFER.

Fixes: ab1a187bba5c ("PCI: Check device_attach() return value always")
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: Grygorii Strashko <grygorii.strashko@ti.com>
Cc: Alan Stern <stern@rowland.harvard.edu>

PCI: Fix BUG on device attach failure

Previously when pci_bus_add_device() called device_attach() and it returned
a negative value, we emitted a WARN but carried on.

Commit ab1a187bba5c ("PCI: Check device_attach() return value always"),
introduced in Linux 4.6-rc1, changed this to unwind all steps preceding
device_attach() and to not set dev->is_added = 1.

The latter leads to a BUG if pci_bus_add_device() was called from
pci_bus_add_devices(). Fix by not recursing to a child bus if
device_attach() failed for the bridge leading to it.

This can be triggered by plugging in a PCI device (e.g. Thunderbolt) while
the system is asleep. The system locks up when woken because
device_attach() returns -EPROBE_DEFER.

Fixes: ab1a187bba5c ("PCI: Check device_attach() return value always")
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

PCI: Add pcibios_bus_add_device() weak function

This adds weak function pcibios_bus_add_device() for arch dependent
code could do proper setup. For example, powerpc could setup EEH
related resources for SRIOV VFs.

Signed-off-by: Wei Yang <weiyang@linux.vnet.ibm.com>
Reviewed-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

PCI: Check device_attach() return value always

Previously we checked the device_attach() return value only when
CONFIG_BUG=y. That caused this warning in builds where CONFIG_BUG is not
set:

drivers/pci/bus.c:237:6: warning: variable 'retval' set but not used [-Wunused-but-set-variable]

Check the return value of device_attach() always and clean up after
failure.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

PCI: Fix minimum allocation address overwrite

Commit 36e097a8a297 ("PCI: Split out bridge window override of minimum
allocation address") claimed to do no functional changes but unfortunately
did: The "min" variable is altered. At least the AVM A1 PCMCIA adapter was
no longer detected, breaking ISDN operation.

Use a local copy of "min" to restore the previous behaviour.

[bhelgaas: avoid gcc "?:" extension for portability and readability]
Fixes: 36e097a8a297 ("PCI: Split out bridge window override of minimum allocation address")
Signed-off-by: Christoph Biedl <linux-kernel.bfrz@manchmal.in-ulm.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: stable@vger.kernel.org # v3.14+

PCI: Clear IORESOURCE_UNSET when clipping a bridge window

c770cb4cb505 ("PCI: Mark invalid BARs as unassigned") sets IORESOURCE_UNSET
if we fail to claim a resource. If we tried to claim a bridge window,
failed, clipped the window, and tried to claim the clipped window, we
failed again because of IORESOURCE_UNSET:

pci_bus 0000:00: root bus resource [mem 0xc0000000-0xffffffff window]
pci 0000:00:01.0: can't claim BAR 15 [mem 0xbdf00000-0xddefffff 64bit pref]: no compatible bridge window
pci 0000:00:01.0: [mem size 0x20000000 64bit pref] clipped to [mem size 0x1df00000 64bit pref]
pci 0000:00:01.0: bridge window [mem size 0x1df00000 64bit pref]
pci 0000:00:01.0: can't claim BAR 15 [mem size 0x1df00000 64bit pref]: no address assigned

The 00:01.0 window started as [mem 0xbdf00000-0xddefffff 64bit pref]. That
starts before the host bridge window [mem 0xc0000000-0xffffffff window], so
we clipped the 00:01.0 window to [mem 0xc0000000-0xddefffff 64bit pref].
But we left it marked IORESOURCE_UNSET, so the second claim failed when it
should have succeeded.

This means downstream devices will also fail for lack of resources, e.g.,
in the bugzilla below,

radeon 0000:01:00.0: Fatal error during GPU init

Clear IORESOURCE_UNSET when we clip a bridge window. Also clear
IORESOURCE_UNSET in our copy of the unclipped window so we can see exactly
what the original window was and how it now fits inside the upstream
window.

Fixes: c770cb4cb505 ("PCI: Mark invalid BARs as unassigned")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=85491#c47
Based-on-patch-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Based-on-patch-by: Yinghai Lu <yinghai@kernel.org>
Tested-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
CC: stable@vger.kernel.org # v4.1+

PCI: Add pci_bus_addr_t

David Ahern reported that d63e2e1f3df9 ("sparc/PCI: Clip bridge windows
to fit in upstream windows") fails to boot on sparc/T5-8:

pci 0000:06:00.0: reg 0x184: can't handle BAR above 4GB (bus address 0x110204000)

The problem is that sparc64 assumed that dma_addr_t only needed to hold DMA
addresses, i.e., bus addresses returned via the DMA API (dma_map_single(),
etc.), while the PCI core assumed dma_addr_t could hold *any* bus address,
including raw BAR values. On sparc64, all DMA addresses fit in 32 bits, so
dma_addr_t is a 32-bit type. However, BAR values can be 64 bits wide, so
they don't fit in a dma_addr_t. d63e2e1f3df9 added new checking that
tripped over this mismatch.

Add pci_bus_addr_t, which is wide enough to hold any PCI bus address,
including both raw BAR values and DMA addresses. This will be 64 bits
on 64-bit platforms and on platforms with a 64-bit dma_addr_t. Then
dma_addr_t only needs to be wide enough to hold addresses from the DMA API.

[bhelgaas: changelog, bugzilla, Kconfig to ensure pci_bus_addr_t is at
least as wide as dma_addr_t, documentation]
Fixes: d63e2e1f3df9 ("sparc/PCI: Clip bridge windows to fit in upstream windows")
Fixes: 23b13bc76f35 ("PCI: Fail safely if we can't handle BARs larger than 4GB")
Link: http://lkml.kernel.org/r/CAE9FiQU1gJY1LYrxs+ma5LCTEEe4xmtjRG0aXJ9K_Tsu+m9Wuw@mail.gmail.com
Link: http://lkml.kernel.org/r/1427857069-6789-1-git-send-email-yinghai@kernel.org
Link: https://bugzilla.kernel.org/show_bug.cgi?id=96231
Reported-by: David Ahern <david.ahern@oracle.com>
Tested-by: David Ahern <david.ahern@oracle.com>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: David S. Miller <davem@davemloft.net>
CC: stable@vger.kernel.org # v3.19+

PCI: Use common resource list management code instead of private implementation

Use common resource list management data structure and interfaces
instead of private implementation.

Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

PCI: Add pci_bus_clip_resource() to clip to fit upstream window

Add pci_bus_clip_resource(). If a PCI-PCI bridge window overlaps an
upstream bridge window but is not completely contained by it, this clips
the downstream window so it fits inside the upstream one.

No functional change (this adds the function but no callers).

[bhelgaas: changelog, split into separate patch]
Link: https://bugzilla.kernel.org/show_bug.cgi?id=85491
Reported-by: Marek Kordik <kordikmarek@gmail.com>
Fixes: 5b28541552ef ("PCI: Restrict 64-bit prefetchable bridge windows to 64-bit resources")
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: stable@vger.kernel.org # v3.16+

PCI: Move EXPORT_SYMBOL so it immediately follows function/variable

Move EXPORT_SYMBOL so it immediately follows the function or variable.

No functional change.

[bhelgaas: squash similar changes, fix hotplug, probe, rom, search, too]
Signed-off-by: Ryan Desfosses <ryan@desfo.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Make pci_bus_add_device() void

pci_bus_add_device() always returns 0, so there's no point in returning
anything at all. Make it a void function and remove the tests of the
return value from the callers.

[bhelgaas: changelog, remove unused "err" from i82875p_setup_overfl_dev()]
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Remove unnecessary includes of <linux/init.h>

None of these files are actually using any __init type directives and hence
don't need to include <linux/init.h>. Most are just a left over from
__devinit and __cpuinit removal, or simply due to code getting copied from
one driver to the next.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Change pci_bus_alloc_resource() type_mask to unsigned long

The pci_bus_alloc_resource() "type_mask" parameter is used to compare with
the "flags" member of a struct resource, so it should be the same type,
namely "unsigned long".

No functional change because all current IORESOURCE_* flags fit in 32 bits.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Check all IORESOURCE_TYPE_BITS in pci_bus_alloc_from_region()

When allocating space from a bus resource, i.e., from apertures leading to
this bus, make sure the entire resource type matches. The previous code
assumed the IORESOURCE_TYPE_BITS field was a bitmask with only a single bit
set, but this is not true. IORESOURCE_TYPE_BITS is really an enumeration,
and we have to check all the bits.

See 72dcb1197228 ("resources: Add register address resource type").

No functional change. If we used this path for allocating IRQs, DMA
channels, or bus numbers, this would fix a bug because those types are
indistinguishable when masked by IORESOURCE_IO | IORESOURCE_MEM. But we
don't, so this shouldn't make any difference.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Don't check resource_size() in pci_bus_alloc_resource()

Paul reported that after f75b99d5a77d ("PCI: Enforce bus address limits in
resource allocation") on a 32-bit kernel (CONFIG_PHYS_ADDR_T_64BIT not
set), intel-gtt complained "can't ioremap flush page - no chipset
flushing". In addition, other PCI resource allocations, e.g., for bridge
windows, failed.

This happens because we incorrectly skip bus resources of
[mem 0x00000000-0xffffffff] because we think they are of size zero.
When resource_size_t is 32 bits wide, resource_size() on
[mem 0x00000000-0xffffffff] returns 0 because (r->end - r->start + 1)
overflows.

Therefore, we can't use "resource_size() == 0" to decide that allocation
from this resource will fail. allocate_resource() should fail anyway if it
can't satisfy the address constraints, so we should just depend on that.

A [mem 0x00000000-0xffffffff] bus resource is obviously not really valid,
but we do fall back to it as a default when we don't have information about
host bridge apertures.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=71611
Fixes: f75b99d5a77d PCI: Enforce bus address limits in resource allocation
Reported-and-tested-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Don't check resource_size() in pci_bus_alloc_resource()

Paul reported that after f75b99d5a77d ("PCI: Enforce bus address limits in
resource allocation") on a 32-bit kernel (CONFIG_PHYS_ADDR_T_64BIT not
set), intel-gtt complained "can't ioremap flush page - no chipset
flushing". In addition, other PCI resource allocations, e.g., for bridge
windows, failed.

This happens because we incorrectly skip bus resources of
[mem 0x00000000-0xffffffff] because we think they are of size zero.
When resource_size_t is 32 bits wide, resource_size() on
[mem 0x00000000-0xffffffff] returns 0 because (r->end - r->start + 1)
overflows.

Therefore, we can't use "resource_size() == 0" to decide that allocation
from this resource will fail. allocate_resource() should fail anyway if it
can't satisfy the address constraints, so we should just depend on that.

A [mem 0x00000000-0xffffffff] bus resource is obviously not really valid,
but we do fall back to it as a default when we don't have information about
host bridge apertures.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=71611
Fixes: f75b99d5a77d PCI: Enforce bus address limits in resource allocation
Reported-and-tested-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Allocate 64-bit BARs above 4G when possible

Try to allocate space for 64-bit BARs above 4G first, to preserve the space
below 4G for 32-bit BARs. If there's no space above 4G available, fall
back to allocating anywhere.

[bhelgaas: reworked starting from http://lkml.kernel.org/r/1387485843-17403-2-git-send-email-yinghai@kernel.org]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Enforce bus address limits in resource allocation

When allocating space for 32-bit BARs, we previously limited RESOURCE
addresses so they would fit in 32 bits. However, the BUS address need not
be the same as the resource address, and it's the bus address that must fit
in the 32-bit BAR.

This patch adds:

- pci_clip_resource_to_region(), which clips a resource so it contains
only the range that maps to the specified bus address region, e.g., to
clip a resource to 32-bit bus addresses, and

- pci_bus_alloc_from_region(), which allocates space for a resource from
the specified bus address region,

and changes pci_bus_alloc_resource() to allocate space for 64-bit BARs from
the entire bus address region, and space for 32-bit BARs from only the bus
address region below 4GB.

If we had this window:

pci_root HWP0002:0a: host bridge window [mem 0xf0180000000-0xf01fedfffff] (bus address [0x80000000-0xfedfffff])

we previously could not put a 32-bit BAR there, because the CPU addresses
don't fit in 32 bits. This patch fixes this, so we can use this space for
32-bit BARs.

It's also possible (though unlikely) to have resources with 32-bit CPU
addresses but bus addresses above 4GB. In this case the previous code
would allocate space that a 32-bit BAR could not map.

Remove PCIBIOS_MAX_MEM_32, which is no longer used.

[bhelgaas: reworked starting from http://lkml.kernel.org/r/1386658484-15774-3-git-send-email-yinghai@kernel.org]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Split out bridge window override of minimum allocation address

pci_bus_alloc_resource() avoids allocating space below the "min" supplied
by the caller (usually PCIBIOS_MIN_IO or PCIBIOS_MIN_MEM). This is to
protect badly documented motherboard resources. But if we're allocating
space inside an already-configured PCI-PCI bridge window, we ignore "min".

See 688d191821de ("pci: make bus resource start address override minimum IO
address").

This patch moves the check to make it more visible and simplify future
patches. No functional change.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Move pci_proc_attach_device() to pci_bus_add_device()

4f535093cf8f ("PCI: Put pci_dev in device tree as early as possible")
moved pci_proc_attach_device() from pci_bus_add_device() to
pci_device_add().

This moves it back to pci_bus_add_device(), essentially reverting that
part of 4f535093cf8f. This makes it symmetric with pci_stop_dev(),
where we call pci_proc_detach_device() and pci_remove_sysfs_dev_files()
and set dev->is_added = 0.

[bhelgaas: changelog, create sysfs then attach proc for symmetry]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Delay enabling bridges until they're needed

We currently enable PCI bridges after scanning a bus and assigning
resources. This is often done in arch code.

This patch changes this so we don't enable a bridge until necessary, i.e.,
until we enable a PCI device behind the bridge. We do this in the generic
pci_enable_device() path, so this also removes the arch-specific code to
enable bridges.

[bhelgaas: changelog]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Introduce pci_bus_{get|put}() to manage PCI bus reference count

Introduce helper functions pci_bus_{get|put}() to manage PCI bus
reference count.

Signed-off-by: Jiang Liu <jiang.liu@huawei.com>
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Delay final fixups until resources are assigned

Commit 4f535093cf "PCI: Put pci_dev in device tree as early as possible"
moved final fixups from pci_bus_add_device() to pci_device_add(). But
pci_device_add() happens before resource assignment, so BARs may not be
valid yet.

Typical flow for hot-add:

pciehp_configure_device
pci_scan_slot
pci_scan_single_device
pci_device_add
pci_fixup_device(pci_fixup_final, dev) # previous location
# resource assignment happens here
pci_bus_add_devices
pci_bus_add_device
pci_fixup_device(pci_fixup_final, dev) # new location

[bhelgaas: changelog, move fixups to pci_bus_add_device()]
Reference: https://lkml.kernel.org/r/20130415182614.GB9224@xanatos
Reported-by: David Bulkow <David.Bulkow@stratus.com>
Tested-by: David Bulkow <David.Bulkow@stratus.com>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: stable@vger.kernel.org # v3.9+

PCI: Warn about failures instead of "must_check" functions

These places capture return values to avoid "must_check" warnings,
but we didn't *do* anything with the return values, which causes
"set but not used" warnings. We might as well do something instead
of just trying to evade the "must_check" warnings.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Clean up usages of pci_bus->is_added

Now pci_bus->is_added is only used to guard invoking of
pcibios_fixup_bus() in pci_scan_child_bus(), so just set
it directly after the fixups and remove the other test
and set in pci_bus_add_devices().

[bhelgaas: changelog]
Signed-off-by: Jiang Liu <jiang.liu@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Yinghai Lu <yinghai@kernel.org>

PCI: Put pci_dev in device tree as early as possible

We want to put pci_dev structs in the device tree as soon as possible so
for_each_pci_dev() iteration will not miss them, but driver attachment
needs to be delayed until after pci_assign_unassigned_resources() to make
sure all devices have resources assigned first.

This patch moves device registering from pci_bus_add_devices() to
pci_device_add(), which happens earlier, leaving driver attachment in
pci_bus_add_devices().

It also removes unattached child bus handling in pci_bus_add_devices().
That's not needed because child bus via pci_add_new_bus() is already
in parent bus children list.

[bhelgaas: changelog]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

PCI: Skip attaching driver in device_add()

We want to add PCI devices to the device tree as early as possible but
delay attaching drivers.

device_add() adds a device to the device hierarchy and (via
device_attach()) attaches a matching driver and calls its .probe() method.
We want to separate adding the device to the hierarchy from attaching the
driver.

This patch does that by adding "match_driver" in struct pci_dev. When
false, we return failure from pci_bus_match(), which makes device_attach()
believe there's no matching driver.

Later, we set "match_driver = true" and call device_attach() again, which
now attaches the driver and calls its .probe() method.

[bhelgaas: changelog, explicitly init dev->match_driver,
fold device_attach() call into pci_bus_add_device()]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

PCI/ACPI: Reserve firmware-allocated resources for hot-added root buses

Firmware may have assigned PCI BARs for hot-added devices, so reserve
those resources before trying to allocate more.

[bhelgaas: move empty weak definition here]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: Add pcibios_add_device

Platforms may want to provide architecture-specific functionality during
PCI enumeration. Add a pcibios_add_device() call that architectures can
override to do so.

Signed-off-by: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Seth Forshee <seth.forshee@canonical.com>

PCI/PM: Fix deadlock when unbinding device if parent in D3cold

If a PCI device and its parents are put into D3cold, unbinding the
device will trigger deadlock as follow:

- driver_unbind
- device_release_driver
- device_lock(dev) <--- previous lock here
- __device_release_driver
- pm_runtime_get_sync
...
- rpm_resume(dev)
- rpm_resume(dev->parent)
...
- pci_pm_runtime_resume
...
- pci_set_power_state
- __pci_start_power_transition
- pci_wakeup_bus(dev->parent->subordinate)
- pci_walk_bus
- device_lock(dev) <--- deadlock here


If we do not do device_lock in pci_walk_bus, we can avoid deadlock.
Device_lock in pci_walk_bus is introduced in commit:
d71374dafbba7ec3f67371d3b7e9f6310a588808, corresponding email thread
is: https://lkml.org/lkml/2006/5/26/38. The patch author Zhang Yanmin
said device_lock is added to pci_walk_bus because:

Some error handling functions call pci_walk_bus. For example, PCIe
aer. Here we lock the device, so the driver wouldn't detach from the
device, as the cb might call driver's callback function.

So I fixed the deadlock as follows:

- remove device_lock from pci_walk_bus
- add device_lock into callback if callback will call driver's callback

I checked pci_walk_bus users one by one, and found only PCIe aer needs
device lock.

Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
CC: stable@vger.kernel.org # v3.6+
CC: Zhang Yanmin <yanmin.zhang@intel.com>

PCI: Use correct type when freeing bus resource list

Should use struct pci_bus_resource instead of struct pci_host_bridge_window

Commit 45ca9e9730 ("PCI: add helpers for building PCI bus resource lists")
added pci_free_resource_list() and used it in pci_bus_remove_resources().
Later it was also used for host bridge aperture lists, which was fine until
commit 0efd5aab41 ("PCI: add struct pci_host_bridge_window with CPU/bus
address offset"). That commit added offset information, so we needed a
struct pci_host_bridge_window that was separate from struct
pci_bus_resource.

Commit 0efd5aab41 should have split the host bridge aperture users of
pci_free_resource_list() from the pci_bus_resource user
(pci_bus_remove_resources()), but it did not.

[bhelgaas: changelog -- 0efd5aab41 was mine, so this is all my fault]
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: fix undefined reference to 'pci_fixup_final_inited'

My "PCI: Integrate 'pci_fixup_final' quirks into hot-plug paths" patch
introduced an undefined reference to 'pci_fixup_final_inited' when
CONFIG_PCI_QUIRKS is not enabled (on x86_64):
drivers/built-in.o: In function `pci_bus_add_device':
(.text+0x4f62): undefined reference to `pci_fixup_final_inited'

This patch removes the external reference ending up with a result closer
to what we ultimately want when the boot path issues described in the
original patch are resolved.

References:
https://lkml.org/lkml/2012/7/9/542 Original, offending, patch
https://lkml.org/lkml/2012/7/12/338 Randy's catch

Reported-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Myron Stowe <myron.stowe@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Randy Dunlap <rdunlap@xenotime.net>

PCI: call final fixups hot-added devices

Final fixups are currently applied only at boot-time by
pci_apply_final_quirks(), which is an fs_initcall(). Hot-added devices
don't get these fixups, so they may not be completely initialized.

This patch makes us run final fixups for hot-added devices in
pci_bus_add_device() just before the new device becomes eligible for driver
binding.

This patch keeps the fs_initcall() for devices present at boot because we
do resource assignment between pci_bus_add_device and the fs_initcall(),
and we don't want to break any fixups that depend on that assignment. This
is a design issue that may be addressed in the future -- any resource
assignment should be done *before* device_add().

[bhelgaas: changelog]
Signed-off-by: Myron Stowe <myron.stowe@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: add struct pci_host_bridge_window with CPU/bus address offset

Some PCI host bridges apply an address offset, so bus addresses on PCI are
different from CPU addresses. This patch adds a way for architectures to
tell the PCI core about this offset. For example:

LIST_HEAD(resources);
pci_add_resource_offset(&resources, host->io_space, host->io_offset);
pci_add_resource_offset(&resources, host->mem_space, host->mem_offset);
pci_scan_root_bus(parent, bus, ops, sysdata, &resources);

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

PCI: add helpers for building PCI bus resource lists

We'd like to supply a list of resources when we create a new PCI bus,
e.g., the root bus under a PCI host bridge. These are helpers for
constructing that list.

These are exported because the plan is to replace this exported interface:
pci_scan_bus_parented()
with this one:
pci_add_resource(resources, ...)
pci_scan_root_bus(..., resources)

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI/sysfs: move bus cpuaffinity to class dev_attrs

Requested by Greg KH to fix a race condition in the creating of PCI bus
cpuaffinity files.

Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

Revert "PCI: allocate bus resources from the top down"

This reverts commit b126b4703afa4010b161784a43650337676dd03b.

We're going back to the old behavior of allocating from bus resources
in _CRS order.

Acked-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

Revert "PCI: fix pci_bus_alloc_resource() hang, prefer positive decode"

This reverts commit 82e3e767c21fef2b1b38868e20eb4e470a1e38e3.

We're going back to considering bus resources in the order we found
them (in _CRS order, when we're using _CRS), so we don't need to
define any ordering.

Acked-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: fix pci_bus_alloc_resource() hang, prefer positive decode

When a PCI bus has two resources with the same start/end, e.g.,

pci_bus 0000:04: resource 2 [mem 0xd0000000-0xd7ffffff pref]
pci_bus 0000:04: resource 7 [mem 0xd0000000-0xd7ffffff]

the previous pci_bus_find_resource_prev() implementation would alternate
between them forever:

pci_bus_find_resource_prev(... [mem 0xd0000000-0xd7ffffff pref])
returns [mem 0xd0000000-0xd7ffffff]
pci_bus_find_resource_prev(... [mem 0xd0000000-0xd7ffffff])
returns [mem 0xd0000000-0xd7ffffff pref]
pci_bus_find_resource_prev(... [mem 0xd0000000-0xd7ffffff pref])
returns [mem 0xd0000000-0xd7ffffff]
...

This happened because there was no ordering between two resources with the
same start and end. A resource that had the same start and end as the
cursor, but was not itself the cursor, was considered to be before the
cursor.

This patch fixes the hang by making a fixed ordering between any two
resources.

In addition, it tries to allocate from positively decoded regions before
using any subtractively decoded resources. This means we will use a
positive decode region before a subtractive decode one, even if it means
using a smaller address.

Reference: https://bugzilla.kernel.org/show_bug.cgi?id=22062
Reported-by: Borislav Petkov <bp@amd64.org>
Tested-by: Borislav Petkov <bp@amd64.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: allocate bus resources from the top down

Allocate space from the highest-address PCI bus resource first, then work
downward.

Previously, we looked for space in PCI host bridge windows in the order
we discovered the windows. For example, given the following windows
(discovered via an ACPI _CRS method):

pci_root PNP0A03:00: host bridge window [mem 0x000a0000-0x000bffff]
pci_root PNP0A03:00: host bridge window [mem 0x000c0000-0x000effff]
pci_root PNP0A03:00: host bridge window [mem 0x000f0000-0x000fffff]
pci_root PNP0A03:00: host bridge window [mem 0xbff00000-0xf7ffffff]
pci_root PNP0A03:00: host bridge window [mem 0xff980000-0xff980fff]
pci_root PNP0A03:00: host bridge window [mem 0xff97c000-0xff97ffff]
pci_root PNP0A03:00: host bridge window [mem 0xfed20000-0xfed9ffff]

we attempted to allocate from [mem 0x000a0000-0x000bffff] first, then
[mem 0x000c0000-0x000effff], and so on.

With this patch, we allocate from [mem 0xff980000-0xff980fff] first, then
[mem 0xff97c000-0xff97ffff], [mem 0xfed20000-0xfed9ffff], etc.

Allocating top-down follows Windows practice, so we're less likely to
trip over BIOS defects in the _CRS description.

On the machine above (a Dell T3500), the [mem 0xbff00000-0xbfffffff] region
doesn't actually work and is likely a BIOS defect. The symptom is that we
move the AHCI controller to 0xbff00000, which leads to "Boot has failed,
sleeping forever," a BUG in ahci_stop_engine(), or some other boot failure.

Reference: https://bugzilla.kernel.org/show_bug.cgi?id=16228#c43
Reference: https://bugzilla.redhat.com/show_bug.cgi?id=620313
Reference: https://bugzilla.redhat.com/show_bug.cgi?id=629933
Reported-by: Brian Bloniarz <phunge0@hotmail.com>
Reported-and-tested-by: Stefan Becker <chemobejk@gmail.com>
Reported-by: Denys Vlasenko <dvlasenk@redhat.com>
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

x86/PCI: Export pci_walk_bus function.

In preperation of modularizing Xen-pcifront the pci_walk_bus
needs to be exported so that the xen-pcifront module can walk
call the pci subsystem to walk the PCI devices and claim them.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org> [http://marc.info/?l=linux-pci&m=126149958010298&w=2]

PCI: check return value of pci_enable_device() when enabling bridges

pci_enable_device can fail. In that case, a printed warning would be
more appropriate.

Signed-off-by: Justin P. Mattock <justinmattock@gmail.com>
Signed-off-by: Junchang Wang <junchangwang@gmail.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: sparse warning (trivial)

Assigning zero where NULL should be used.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h

percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.

percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.

http://userweb.kernel.org/~tj/misc/slabh-sweep.py

The script does the followings.

* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.

* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.

* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.

The conversion was done in the following steps.

1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.

2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.

3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.

4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.

5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.

6. percpu.h was updated not to include slab.h.

7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).

* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig

8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.

Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.

Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>

Driver core: create lock/unlock functions for struct device

In the future, we are going to be changing the lock type for struct
device (once we get the lockdep infrastructure properly worked out) To
make that changeover easier, and to possibly burry the lock in a
different part of struct device, let's create some functions to lock and
unlock a device so that no out-of-core code needs to be changed in the
future.

This patch creates the device_lock/unlock/trylock() functions, and
converts all in-tree users to them.

Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Jean Delvare <khali@linux-fr.org>
Cc: Dave Young <hidave.darkstar@gmail.com>
Cc: Ming Lei <tom.leiming@gmail.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Phil Carmody <ext-phil.2.carmody@nokia.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Cornelia Huck <cornelia.huck@de.ibm.com>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Len Brown <len.brown@intel.com>
Cc: Magnus Damm <damm@igel.co.jp>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Stefan Richter <stefanr@s5r6.in-berlin.de>
Cc: David Brownell <dbrownell@users.sourceforge.net>
Cc: Vegard Nossum <vegard.nossum@gmail.com>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: Alex Chiang <achiang@hp.com>
Cc: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andrew Patterson <andrew.patterson@hp.com>
Cc: Yu Zhao <yu.zhao@intel.com>
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Cc: Samuel Ortiz <sameo@linux.intel.com>
Cc: Wolfram Sang <w.sang@pengutronix.de>
Cc: CHENG Renquan <rqcheng@smu.edu.sg>
Cc: Oliver Neukum <oliver@neukum.org>
Cc: Frans Pop <elendil@planet.nl>
Cc: David Vrabel <david.vrabel@csr.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

PCI: augment bus resource table with a list

Previously we used a table of size PCI_BUS_NUM_RESOURCES (16) for resources
forwarded to a bus by its upstream bridge. We've increased this size
several times when the table overflowed.

But there's no good limit on the number of resources because host bridges
and subtractive decode bridges can forward any number of ranges to their
secondary buses.

This patch reduces the table to only PCI_BRIDGE_RESOURCE_NUM (4) entries,
which corresponds to the number of windows a PCI-to-PCI (3) or CardBus (4)
bridge can positively decode. Any additional resources, e.g., PCI host
bridge windows or subtractively-decoded regions, are kept in a list.

I'd prefer a single list rather than this split table/list approach, but
that requires simultaneous changes to every architecture. This approach
only requires immediate changes where we set up (a) host bridges with more
than four windows and (b) subtractive-decode P2P bridges, and we can
incrementally change other architectures to use the list.

Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: add pci_bus_for_each_resource(), remove direct bus->resource[] refs

No functional change; this converts loops that iterate from 0 to
PCI_BUS_NUM_RESOURCES through pci_bus resource[] table to use the
pci_bus_for_each_resource() iterator instead.

This doesn't change the way resources are stored; it merely removes
dependencies on the fact that they're in a table.

Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

resource/PCI: mark struct resource as const

Now that we return the new resource start position, there is no
need to update "struct resource" inside the align function.
Therefore, mark the struct resource as const.

Cc: Bjorn Helgaas <bjorn.helgaas@hp.com>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

resource/PCI: align functions now return start of resource

As suggested by Linus, align functions should return the start
of a resource, not void. An update of "res->start" is no longer
necessary.

Cc: Bjorn Helgaas <bjorn.helgaas@hp.com>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI AER: support Multiple Error Received and no error source id

Based on PCI Express AER specs, a root port might receive multiple
TLP errors while it could only save a correctable error source id
and an uncorrectable error source id at the same time. In addition,
some root port hardware might be unable to provide a correct source
id, i.e., the source id, or the bus id part of the source id provided
by root port might be equal to 0.

The patchset implements the support in kernel by searching the device
tree under the root port.

Patch 1 changes parameter cb of function pci_walk_bus to return a value.
When cb return non-zero, pci_walk_bus stops more searching on the
device tree.

Reviewed-by: Andrew Patterson <andrew.patterson@hp.com>
Signed-off-by: Zhang Yanmin <yanmin_zhang@linux.intel.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI/x86: don't assume prefetchable ranges are 64bit

We should not assign 64bit ranges to PCI devices that only take 32bit
prefetchable addresses.

Try to set IORESOURCE_MEM_64 in 64bit resource of pci_device/pci_bridge
and make the bus resource only have that bit set when all devices under
it support 64bit prefetchable memory. Use that flag to allocate
resources from that range.

Reported-by: Yannick <yannick.roehlly@free.fr>
Reviewed-by: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: Setup disabled bridges even if buses are added

This patch sets up disabled bridges even if buses have already been
added.

pci_assign_unassigned_resources is called after buses are added.
pci_assign_unassigned_resources calls pci_bus_assign_resources.
pci_bus_assign_resources calls pci_setup_bridge to configure BARs of
bridges.

Currently pci_setup_bridge returns immediately if the bus have already
been added. So pci_assign_unassigned_resources can't configure BARs of
bridges that were added in a disabled state; this patch fixes the issue.

On logical hot-add, we need to prevent the kernel from re-initializing
bridges that have already been initialized. To achieve this,
pci_setup_bridge returns immediately if the bridge have already been
enabled.

We don't need to check whether the specified bus is a root bus or not.
pci_setup_bridge is not called on a root bus, because a root bus does
not have a bridge.

The patch adds a new helper function, pci_is_enabled. I made the
function name similar to pci_is_managed. The codes which use
enable_cnt directly are changed to use pci_is_enabled.

Acked-by: Alex Chiang <achiang@hp.com>
Signed-off-by: Yuji Shimada <shimada-yxb@necst.nec.co.jp>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: do not enable bridges more than once

In preparation for PCI core hotplug, we need to ensure that we do
not attempt to re-enable bridges that have already been enabled.

Reported-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: Alex Chiang <achiang@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: constify pci_bus_add_devices()

drivers/pci/hotplug/fakephp.c:283: warning: passing argument 1 of 'pci_bus_add_devices' discards qualifiers from pointer target type

Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: factor pci_bus_add_child() from pci_bus_add_devices()

This patch splits a new function, pci_bus_add_child(), from
pci_bus_add_devices(). The new function can be used to register PCI
buses to the device core.

Signed-off-by: Yu Zhao <yu.zhao@intel.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: cleanup pci_bus_add_devices()

Cleanup pci_bus_add_devices() by negating the conditional and
continuing, rather than having a single conditional take up the whole
body.

Signed-off-by: Yu Zhao <yu.zhao@intel.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: Add legacy_io/mem to all busses

Currently, only PHBs get the legacy_* files, which makes it tricky for
userland to get access to the legacy space. This commit exposes them in
every bus, since even child buses may forward legacy cycles if
configured properly.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: make CPU list affinity visible

Stephen Hemminger wrote:
> Looks like Mike created cpulistaffinty in sysfs but never completed
> the job.

This patch hooks things up correctly, taking care to remove the new file
when the bus is destroyed.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

PCI: remove global list of PCI devices

This patch finally removes the global list of PCI devices. We are
relying entirely on the list held in the driver core now, and do not
need a separate "shadow" list as no one uses it.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

PCI: add is_added flag to struct pci_dev

This lets us check if the device is really added to the driver core or
not, which is what we need when walking some of the bus lists. The flag
is there in anticipation of getting rid of the other PCI device list,
which is what we used to check in this situation.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

PCI: fix issue with busses registering multiple times in sysfs

PCI busses can be registered multiple times, so we need to detect if we
have registered our bus structure in sysfs already. If so, don't do it
again.

Thanks to Guennadi Liakhovetski <g.liakhovetski@gmx.de> for reporting
the problem, and to Linus for poking me to get me to believe that it was
a real problem.

Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

PCI: fix up error messages for pci_bus registering

Due to the class_device cleanup of pci_bus, the error messages when
things go wrong are incorrect. So fix this up to properly report what
is really happening, if things go wrong.

Thanks to Kay for pointing out the issue.

Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

PCI: make pci_bus a struct device

This moves the pci_bus class device to be a real struct device and at
the same time, place it in the device tree in the correct location.

Note, the old "bridge" symlink is now gone, but this was a non-standard
link and no userspace program used it. If you need to determine the
device that the bus is on, follow the standard device symlink, or walk
up the device tree.


Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

PCI: drivers/pci/: remove unused exports

This patch removes the following unused exports:
- remove the following unused EXPORT_SYMBOL's:
- pci-acpi.c: pci_osc_support_set
- proc.c: pci_proc_detach_bus
- remove the following unused EXPORT_SYMBOL_GPL's:
- bus.c: pci_walk_bus
- probe.c: pci_create_bus
- setup-res.c: pci_claim_resource

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

pci: do not mark exported functions as __devinit

Functions marked __devinit will be removed after kernel init. But being
exported they are potentially called by a module much later.

So the safer choice seems to be to keep the function even in the non
CONFIG_HOTPLUG case.

This silence the follwoing section mismatch warnings:
WARNING: drivers/built-in.o - Section mismatch: reference to .init.text:pci_bus_add_device from __ksymtab_gpl between '__ksymtab_pci_bus_add_device' (at offset 0x20) and '__ksymtab_pci_walk_bus'
WARNING: drivers/built-in.o - Section mismatch: reference to .init.text:pci_create_bus from __ksymtab_gpl between '__ksymtab_pci_create_bus' (at offset 0x40) and '__ksymtab_pci_stop_bus_device'
WARNING: drivers/built-in.o - Section mismatch: reference to .init.text:pci_bus_max_busnr from __ksymtab_gpl between '__ksymtab_pci_bus_max_busnr' (at offset 0xc0) and '__ksymtab_pci_assign_resource_fixed'
WARNING: drivers/built-in.o - Section mismatch: reference to .init.text:pci_claim_resource from __ksymtab_gpl between '__ksymtab_pci_claim_resource' (at offset 0xe0) and '__ksymtab_pcie_port_bus_type'
WARNING: drivers/built-in.o - Section mismatch: reference to .init.text:pci_bus_add_devices from __ksymtab between '__ksymtab_pci_bus_add_devices' (at offset 0x70) and '__ksymtab_pci_bus_alloc_resource'
WARNING: drivers/built-in.o - Section mismatch: reference to .init.text:pci_scan_bus_parented from __ksymtab between '__ksymtab_pci_scan_bus_parented' (at offset 0x90) and '__ksymtab_pci_root_buses'
WARNING: drivers/built-in.o - Section mismatch: reference to .init.text:pci_bus_assign_resources from __ksymtab between '__ksymtab_pci_bus_assign_resources' (at offset 0x4d0) and '__ksymtab_pci_bus_size_bridges'
WARNING: drivers/built-in.o - Section mismatch: reference to .init.text:pci_bus_size_bridges from __ksymtab between '__ksymtab_pci_bus_size_bridges' (at offset 0x4e0) and '__ksymtab_pci_setup_cardbus'

Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

PCI: fix __must_check warnings

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

[PATCH] 64bit resource: change pci core and arch code to use resource_size_t

Based on a patch series originally from Vivek Goyal <vgoyal@in.ibm.com>

Cc: Vivek Goyal <vgoyal@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

[PATCH] PCI: fix race with pci_walk_bus and pci_destroy_dev

pci_walk_bus has a race with pci_destroy_dev. When cb is called
in pci_walk_bus, pci_destroy_dev might unlink the dev pointed by next.
Later on in the next loop, pointer next becomes NULL and cause
kernel panic.

Below patch against 2.6.17-rc4 fixes it by changing pci_bus_lock (spin_lock)
to pci_bus_sem (rw_semaphore).

Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

[PATCH] PCI: Add pci_walk_bus function to PCI core (nonrecursive)

The PCI error recovery infrastructure needs to be able to contact all
the drivers affected by a PCI error event, which may mean traversing
all the devices under a given PCI-PCI bridge. This patch adds a
function to the PCI core that traverses all the PCI devices on a PCI
bus and under any PCI-PCI bridges on that bus (and so on), calling a
given function for each device. This provides a way for the error
recovery code to iterate through all devices that are affected by an
error event.

This version is not implemented as a recursive function. Instead,
when we reach a PCI-PCI bridge, we set the pointers to start doing the
devices on the bus under the bridge, and when we reach the end of a
bus's devices, we use the bus->self pointer to go back up to the next
higher bus and continue doing its devices.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

[PATCH] PCI: start paying attention to a lot of pci function return values

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

pci: make bus resource start address override minimum IO address

The reason we have PCIBIOS_MIN_IO and PCIBIOS_MIN_CARDBUS_IO is because
we want to protect badly documented motherboard PCI resources and thus
don't want to allocate new resources in low IO/MEM space.

However, if we have already discovered a PCI bridge with a specified
resource base, that should override that decision.

This change will allow us to move the "careful" region upwards without
resulting in problems allocating resources in low mappings. This was
brought on by us having allocated a bus resource at 0x1000, conflicting
with a undocumented VAIO Sony PI resources.

[PATCH] acpi bridge hotadd: Link newly created pci child bus to its parent on creation

When a pci child bus is created, add it to the parent's children list
immediately rather than waiting till pci_bus_add_devices(). For hot-plug
bridges/devices, pci_bus_add_devices() may be called much later, after they
have been properly configured. In the meantime, this allows us to use the
normal pci bus search functions for the hot-plug bridges/buses.

Signed-off-by: Rajesh Shah <rajesh.shah@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Linux-2.6.12-rc2

Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!