PCI changes
The following changes since commitf423fc627b05f47bc9305f9661630fce30f208f9:
Revert "rtc: Expire alarms after the time is set." (2012-01-04 07:57:22 -0800)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci linux-next
There's a fix pending for a build problem when CONFIG_AMD_NB=n, but I didn't want to delay the pull request any longer; I'll include it in the next fixes pull.
This one doesn't include Yinghai's resource management stuff either, I'll put that in my -next tree for the next merge window when Yinghai has it ready. That might be a good time to revert 12c22d6ef299 (the revert of the transparent bridge sizing code we talked about), unless you want to try it now.
It *does* include some nice improvements from Bjorn on how the root bus and bus resource code is structured, some good boot fixes from Fedora, and lots of misc fixes elsewhere.
There was definitely more rebasing than I'd like fairly late on, but that was mainly to merge in fixes to Bjorn's series for various build errors that got reported, so I don't think it invalidated testing (I generally don't pull from you and rebase except after you've pulled from me).
Thanks,
Jesse
Alessandro Rubini (1):
PCI: pci_ids: add device ids for STA2X11 device (aka ConneXT)
Alex Williamson (2):
PCI: Fix PRI and PASID consistency
PCI: Fix PCI_EXP_TYPE_RC_EC value
Andreas Herrmann (1):
x86/PCI: amd: Kill misleading message about enablement of IO access to PCI ECS]
Bjorn Helgaas (36):
PCI: add helpers for building PCI bus resource lists
PCI: show host bridges and root bus resources
PCI: add pci_create_root_bus() that accepts resource list
PCI: add pci_scan_root_bus() that accepts resource list
PCI: convert pci_scan_bus() to use pci_create_root_bus()
PCI: convert pci_scan_bus_parented() to use pci_create_root_bus()
PCI: deprecate pci_scan_bus_parented()
alpha/PCI: convert to pci_scan_root_bus() for correct root bus resources arm/PCI: convert to pci_scan_root_bus() for correct root bus resources frv/PCI: convert to pci_scan_root_bus() for correct root bus resources ia64/PCI: use pci_create_bus() instead of pci_scan_bus_parented() ia64/PCI: convert to pci_create_root_bus() for correct root bus resources microblaze/PCI: fix pci_bus_for_each_resource() usage microblaze/PCI: make pcibios_setup_phb_resources() static microblaze/PCI: convert to pci_create_root_bus() for correct root bus resources microblaze/PCI: use pci_scan_root_bus()
mips/PCI: convert to pci_scan_root_bus() for correct root bus resources mn10300/PCI: convert to pci_scan_root_bus() for correct root bus resources parisc/PCI: dino: use pci_create_bus() instead of pci_scan_bus_parented() parisc/PCI: dino: convert to pci_create_root_bus() for correct root bus resources parisc/PCI: lba: deal with LMMIO/PAT overlaps before creating PCI root bus parisc/PCI: lba: use pci_create_bus() instead of pci_scan_bus_parented() parisc/PCI: lba: convert to pci_create_root_bus() for correct root bus resources powerpc/PCI: make pcibios_setup_phb_resources() static
powerpc/PCI: split PHB part out of pcibios_map_io_space() powerpc/PCI: convert to pci_create_root_bus()
sh/PCI: convert to pci_scan_root_bus() for correct root bus resources sparc/PCI: convert to pci_create_root_bus()
sparc32, leon/PCI: convert to pci_scan_root_bus() for correct root bus resources x86/PCI: read Broadcom CNB20LE host bridge info before PCI scan x86/PCI: use pci_scan_bus() instead of pci_scan_bus_parented() x86/PCI: convert to pci_create_root_bus() and pci_scan_root_bus() xtensa/PCI: convert to pci_scan_root_bus() for correct root bus resources
PCI: remove pci_create_bus()
x86/PCI: amd: factor out MMCONFIG discovery
PNP: work around Dell 1536/1546 BIOS MMCONFIG bug that breaks USB
Dave Jones (3):
PCI: add set_nouse_crs for use by a pci=nocrs blacklist
PCI: Add Dell Studio 1557 to pci=nocrs blacklist
PCI: Add Thinkpad SL510 to pci=nocrs blacklist
David Fries (1):
PCI: pci_has_legacy_pm_support add driver and device to WARN
Eric W. Biederman (1):
PCI: msi: Disable msi interrupts when we initialize a pci device
Gary Hade (1):
x86/PCI: Ignore CPU non-addressable _CRS reserved memory resources
Hao, Xudong (1):
PCI: Enable ATS at the device state restore
Jan Kiszka (3):
PCI: Rework config space blocking services
PCI: Introduce INTx check & mask API
uio: Convert uio_generic_pci to new intx masking API
Kenji Kaneshige (2):
PCI: pciehp: Fix wrong workqueue cleanup
PCI: pciehp: Handle push button event asynchronously
Konrad Rzeszutek Wilk (1):
x86/PCI: Expand the x86_msi_ops to have a restore MSIs.
Matthew Garrett (1):
PCI: Rework ASPM disable code
Myron Stowe (16):
PCI: add declaration for pcibios_set_master() to pci core
PCI: ARM: convert pcibios_set_master() to a non-inlined function
PCI: IA64: convert pcibios_set_master() to a non-inlined function
PCI: MicroBlaze: convert pcibios_set_master() to a non-inlined function
PCI: PowerPC: convert pcibios_set_master() to a non-inlined function
PCI: SPARC: convert pcibios_set_master() to a non-inlined function
PCI: TILE: convert pcibios_set_master() to a non-inlined function
PCI: UniCore: convert pcibios_set_master() to a non-inlined function
PCI: Xtensa: convert pcibios_set_master() to a non-inlined function
PCI: Pull PCI 'latency timer' setup up into the core
PCI: frv: use generic pcibios_set_master()
PCI: MIPS: use generic pcibios_set_master()
PCI: mn10300: use generic pcibios_set_master()
PCI: sh: use generic pcibios_set_master()
PCI: x86: use generic pcibios_set_master()
PCI: latency timer doesn't apply to PCIe
Neil Horman (2):
PCI/sysfs: add per pci device msi[x] irq listing (v5)
PCI: msi: fix imbalanced refcount of msi irq sysfs objects
P. Christeas (1):
PCI: kconfig: English typo in pci/pcie/Kconfig
Rafael J. Wysocki (1):
PCI/ACPI/PM: Avoid resuming devices that don't signal PME
Ram Pai (1):
PCI: delay configuration of SRIOV capability
Vincent Palatin (1):
PCI/PM/Runtime: make PCI traces quieter
Yinghai Lu (3):
PCI: Only call pci_stop_bus_device() one time for child devices at remove
PCI: DEVICE_COUNT_RESOURCE should be equal to PCI_NUM_RESOURCES
PCI: Increase resource array mask bit size in pcim_iomap_regions()
Zac Storer (1):
PCI: fix a brace coding style issue in probe.c
Documentation/ABI/testing/sysfs-bus-pci | 18 +++ Documentation/feature-removal-schedule.txt | 12 ++ arch/alpha/kernel/pci.c | 40 ++++---- arch/arm/common/it8152.c | 9 +- arch/arm/common/via82c505.c | 3 +- arch/arm/include/asm/mach/pci.h | 2 +- arch/arm/include/asm/pci.h | 12 -- arch/arm/kernel/bios32.c | 34 +++---- arch/arm/mach-cns3xxx/pcie.c | 8 +- arch/arm/mach-dove/pcie.c | 9 +- arch/arm/mach-footbridge/dc21285.c | 8 +- arch/arm/mach-integrator/pci_v3.c | 19 ++-- arch/arm/mach-iop13xx/pci.c | 17 ++-- arch/arm/mach-ixp2000/enp2611.c | 3 +- arch/arm/mach-ixp2000/pci.c | 8 +- arch/arm/mach-ixp23xx/pci.c | 8 +- arch/arm/mach-ixp4xx/common-pci.c | 8 +- arch/arm/mach-kirkwood/pcie.c | 8 +- arch/arm/mach-ks8695/pci.c | 8 +- arch/arm/mach-mv78xx0/pcie.c | 8 +- arch/arm/mach-orion5x/pci.c | 16 ++-- arch/arm/mach-sa1100/pci-nanoengine.c | 13 ++- arch/arm/mach-tegra/pcie.c | 9 +- arch/arm/mach-versatile/pci.c | 19 ++-- arch/arm/plat-iop/pci.c | 8 +- arch/blackfin/include/asm/pci.h | 4 - arch/frv/mb93090-mb00/pci-frv.c | 20 ---- arch/frv/mb93090-mb00/pci-frv.h | 2 - arch/frv/mb93090-mb00/pci-vdk.c | 11 +- arch/h8300/include/asm/pci.h | 5 - arch/ia64/include/asm/pci.h | 6 - arch/ia64/pci/pci.c | 43 ++++---- arch/microblaze/include/asm/pci-bridge.h | 1 - arch/microblaze/include/asm/pci.h | 5 - arch/microblaze/pci/pci-common.c | 43 ++++---- arch/mips/pci/pci.c | 38 ++----- arch/mn10300/unit-asb2305/pci-asb2305.c | 22 ---- arch/mn10300/unit-asb2305/pci-asb2305.h | 2 - arch/mn10300/unit-asb2305/pci.c | 12 +- arch/powerpc/include/asm/pci-bridge.h | 1 - arch/powerpc/include/asm/pci.h | 5 - arch/powerpc/kernel/pci-common.c | 30 ++++-- arch/powerpc/kernel/pci_64.c | 40 ++++---- arch/sh/drivers/pci/pci.c | 41 ++------ arch/sparc/include/asm/pci_32.h | 5 - arch/sparc/include/asm/pci_64.h | 5 - arch/sparc/kernel/leon_pci.c | 25 ++--- arch/sparc/kernel/pci.c | 22 ++-- arch/tile/include/asm/pci.h | 7 -- arch/tile/kernel/pci.c | 5 + arch/unicore32/include/asm/pci.h | 5 - arch/unicore32/kernel/pci.c | 5 + arch/x86/include/asm/amd_nb.h | 2 + arch/x86/include/asm/pci.h | 9 ++ arch/x86/include/asm/pci_x86.h | 2 - arch/x86/include/asm/topology.h | 2 +- arch/x86/include/asm/x86_init.h | 1 + arch/x86/kernel/amd_nb.c | 31 ++++++ arch/x86/kernel/x86_init.c | 1 + arch/x86/pci/acpi.c | 75 ++++++++++--- arch/x86/pci/amd_bus.c | 43 ++------ arch/x86/pci/broadcom_bus.c | 62 +++++++---- arch/x86/pci/bus_numa.c | 31 +++--- arch/x86/pci/common.c | 19 ++-- arch/x86/pci/i386.c | 20 ---- arch/x86/pci/legacy.c | 3 - arch/x86/pci/numaq_32.c | 2 - arch/xtensa/include/asm/pci.h | 5 - arch/xtensa/kernel/pci.c | 90 ++++++++-------- drivers/acpi/pci_irq.c | 10 +- drivers/acpi/pci_root.c | 7 ++ drivers/parisc/dino.c | 47 ++++---- drivers/parisc/lba_pci.c | 72 +++++++------ drivers/pci/access.c | 76 +++++++++----- drivers/pci/ats.c | 37 +++++-- drivers/pci/bus.c | 32 +++++- drivers/pci/hotplug/pciehp.h | 1 - drivers/pci/hotplug/pciehp_core.c | 11 +-- drivers/pci/hotplug/pciehp_ctrl.c | 4 +- drivers/pci/hotplug/pciehp_hpc.c | 1 - drivers/pci/iov.c | 16 ++-- drivers/pci/msi.c | 160 +++++++++++++++++++++++++++- drivers/pci/pci-acpi.c | 13 ++- drivers/pci/pci-driver.c | 3 +- drivers/pci/pci.c | 153 +++++++++++++++++++++++++- drivers/pci/pci.h | 10 ++ drivers/pci/pcie/Kconfig | 2 +- drivers/pci/pcie/aspm.c | 58 +++++++---- drivers/pci/probe.c | 68 +++++++++++-- drivers/pci/remove.c | 10 ++- drivers/pci/setup-res.c | 6 +- drivers/pnp/quirks.c | 42 +++++++ drivers/scsi/ipr.c | 67 +++++++++++-- drivers/scsi/ipr.h | 1 + drivers/uio/uio_pci_generic.c | 76 +------------ include/linux/msi.h | 3 + include/linux/pci-aspm.h | 4 +- include/linux/pci.h | 51 ++++++---- include/linux/pci_ids.h | 23 ++++ include/linux/pci_regs.h | 6 +- lib/devres.c | 6 +- 101 files changed, 1334 insertions(+), 857 deletions(-)
Quoted 3 lines · Jesse Barnes
Matthew Garrett (1):
PCI: Rework ASPM disable code
Yay.
However, looking this through some more, I'm still a bit confused by the pcie_clear_aspm() call in acpi_pci_root_add().
It seems to be explicitly against the rules elsewhere - we clear ASPM even though ACPI_FADT_NO_ASPM is set - which the changelog (and the other parts of that commit) says means "don't touch pre-existing ASPM state").
So why is acpi_pci_root_add() special? Cna you explain that part to me?
Linus
Quoted 12 lines · Jesse Barnes
The following changes since commit
f423fc627b05f47bc9305f9661630fce30f208f9:Revert "rtc: Expire alarms after the time is set." (2012-01-04 07:57:22 -0800)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci linux-nextThere's a fix pending for a build problem when
CONFIG_AMD_NB=n, but I didn't want to delay the pull request any longer; I'll include it in the next fixes pull.
Quoted 5 lines
This one doesn't include Yinghai's resource management stuff either, I'll put that in my -next tree for the next merge window when Yinghai has it ready. That might be a good time to revert
12c22d6ef299(the revert of the transparent bridge sizing code we talked about), unless you want to try it now.
Can you please push following patch to linus in this merging window?
https://lkml.org/lkml/2011/11/23/2
[PATCH -v4] PCI: Make sriov work with hotplug remove
Thanks
Yinghai
Quoted 10 lines · Linus Torvalds
Quoted 3 lines · Jesse Barnes
Matthew Garrett (1):
PCI: Rework ASPM disable codeYay.
However, looking this through some more, I'm still a bit confused by the
pcie_clear_aspm()call inacpi_pci_root_add().
Quoted 6 lines
It seems to be explicitly against the rules elsewhere - we clear ASPM even though
ACPI_FADT_NO_ASPMis set - which the changelog (and the other parts of that commit) says means "don't touch pre-existing ASPM state").So why is
acpi_pci_root_add()special? Cna you explain that part to me?
pcie_no_aspm() means "Do not permit ASPM to be enabled" - it doesn't alter the existing state. pcie_clear_aspm() does that.
Signature · 1 line
Matthew Garrett | [email protected]
Quoted 5 lines · Matthew Garrett
Quoted 2 lines
So why is
acpi_pci_root_add()special? Cna you explain that part to me?
pcie_no_aspm()means "Do not permit ASPM to be enabled" - it doesn't alter the existing state.pcie_clear_aspm()does that.
I *know*.
So look again.
Why is acpi_pci_root_add() special?
Because dammit, it does exactly that pcie_clear_aspm() for the ACPI_FADT_NO_ASPM case.
That was my question. Why does it do that, when the commit log (and the other users) say that ACPI_FADT_NO_ASPM means "don't touch".
It touches.
Linus
Quoted 7 lines · Linus Torvalds
Quoted 5 lines · Matthew Garrett
Quoted 2 lines
So why is
acpi_pci_root_add()special? Cna you explain that part to me?
pcie_no_aspm()means "Do not permit ASPM to be enabled" - it doesn't alter the existing state.pcie_clear_aspm()does that.
Quoted 8 lines
I *know*.
So look again.
Why is
acpi_pci_root_add()special?Because dammit, it does exactly that
pcie_clear_aspm()for theACPI_FADT_NO_ASPMcase.
I'm sorry, I thought you were referring to the status quo rather than the patch - I see what you mean now. The intent here is to clear ASPM state if the FADT bit has been set *and* we've got _OSC control, whereas previously we'd clear ASPM state any time the FADT bit was set. There are some machines which (a) set the FADT bit, (b) hand over _OSC control, (c) program the device and (d) crash under certain circumstances if you then try to use it. My assumption here is that the FADT bit *does* mean "clear ASPM state", but only if we've been granted _OSC control. That's the behaviour described in the changelog.
Signature · 1 line
Matthew Garrett | [email protected]