2008-10-08 22:21:29

by Jesse Barnes

[permalink] [raw]
Subject: What's in pci-2.6.git

Quite a few patches have queued up in my linux-next branch for 2.6.28, and a
couple of more are still pending (they've seen lots of review so should be
just about ready), specifically the I/O virtualization code and more PCI slot
driver fixes from Alex and Kenji-san.

Full shortlog below, there's stuff all over the map...

Jesse

Alex Chiang (1):
PCI: connect struct pci_dev to struct pci_slot

Andi Kleen (1):
PCI: Document that most pci options are shared between i386 and x86-64

Arjan van de Ven (1):
PCI: introduce an pci_ioremap(pdev, barnr) function

Bjorn Helgaas (3):
x86/PCI: follow lspci device/vendor style
PCI: follow lspci device/vendor style
PCI: tidy PME support messages

Chuck Ebbert (1):
PCI: document the pcie_aspm kernel parameter

Greg Kroah-Hartman (1):
PCI: fix compiler warnings in pci_get_subsys()

Jean Delvare (1):
PCI: Check dynids driver_data value for validity

Jeff Garzik (1):
PCI: documentation fix for pci_dma_mapping_error

Jesse Barnes (5):
PCI: fix -Wakpm warnings in pci_pm_init debug output
PCI: remove #ifdef DEBUG around dev_dbg call
PCI: pcie whitespace cleanup
PCI: fix PCI sysfs rom file size for shadowed ROMs
PCI: use pci_find_ext_capability everywhere

Jike Song (1):
PCI: utilize calculated results when detecting MSI features

Kenji Kaneshige (2):
PCI: pciehp: fix irq initialization
PCI: fix get_##name merge error

Mike Travis (1):
PCI: make CPU list affinity visible

Milton Miller (1):
PCI: remove dynids.use_driver_data

Rafael J. Wysocki (1):
PCI PM: Introduce function pci_wake_from_d3

Roland Dreier (1):
PCI: fix MSI-HOWTO.txt info about MSI-X MMIO space

Seth Heasley (1):
x86/PCI: irq and pci_ids patch for Intel Ibex Peak DeviceIDs

Stephen Hemminger (1):
PCI: fix sparse warning in pci_remove_behind_bridge

Stephen Rothwell (1):
PCI: hotplug/rpaphp: remove unused error path code

Taku Izumi (2):
PCI: pciehp: change name tag of "hpdriver_portdrv" variable
PCI: pciehp: replace printk with dev_printk

Thomas Petazzoni (1):
PCI: allow quirks to be compiled out

Yinghai Lu (2):
PCI: change MSI-x vector to 32bit
PCI: use %pF instead of print_fn_descriptor_symbol() in quirks.c

Zhang, Yanmin (1):
PCI: merge define of PCI_CFG_SPACE_SIZE

Zhao, Yu (1):
PCI: fix hotplug get_##name return value problem

[email protected] (2):
PCI: ibmphp: list_for_each to list_for_each_entry
PCI: ibmphp: list_for_each to list_for_each_entry-checkpatch cleanups


2008-10-21 05:18:50

by Anirban Chakraborty

[permalink] [raw]
Subject: Re: What's in pci-2.6.git


On Oct 8, 2008, at 3:21 PM, Jesse Barnes wrote:

> Quite a few patches have queued up in my linux-next branch for
> 2.6.28, and a
> couple of more are still pending (they've seen lots of review so
> should be
> just about ready), specifically the I/O virtualization code and more
> PCI slot
> driver fixes from Alex and Kenji-san.
>
> Full shortlog below, there's stuff all over the map...
>
> Jesse
>
> Alex Chiang (1):
> PCI: connect struct pci_dev to struct pci_slot
>
> Andi Kleen (1):
> PCI: Document that most pci options are shared between i386 and
> x86-64
>
> Arjan van de Ven (1):
> PCI: introduce an pci_ioremap(pdev, barnr) function

Does it make sense to rename iounmap to pci_iounmap, just to be
consistent with pci_ioremap_bar?

-Anirban

2008-10-21 05:51:36

by Matthew Wilcox

[permalink] [raw]
Subject: Re: What's in pci-2.6.git

On Mon, Oct 20, 2008 at 10:18:38PM -0700, Anirban Chakraborty wrote:
> >Arjan van de Ven (1):
> > PCI: introduce an pci_ioremap(pdev, barnr) function
>
> Does it make sense to rename iounmap to pci_iounmap, just to be
> consistent with pci_ioremap_bar?

iounmap can be used for ioremaps that aren't necessarily from PCI devices.
Since there's no length parameter to iounmap, there's nothing to get
wrong, so there's no natural counterpart to pci_ioremap().

--
Matthew Wilcox Intel Open Source Technology Centre
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours. We can't possibly take such
a retrograde step."

2008-10-21 05:55:52

by Anirban Chakraborty

[permalink] [raw]
Subject: Re: What's in pci-2.6.git


On Oct 20, 2008, at 10:51 PM, Matthew Wilcox wrote:

> On Mon, Oct 20, 2008 at 10:18:38PM -0700, Anirban Chakraborty wrote:
>>> Arjan van de Ven (1):
>>> PCI: introduce an pci_ioremap(pdev, barnr) function
>>
>> Does it make sense to rename iounmap to pci_iounmap, just to be
>> consistent with pci_ioremap_bar?
>
> iounmap can be used for ioremaps that aren't necessarily from PCI
> devices.
> Since there's no length parameter to iounmap, there's nothing to get
> wrong, so there's no natural counterpart to pci_ioremap().

Yeah, that makes sense.
Thanks,
-Anirban