All of lore.kernel.org
 help / color / mirror / Atom feed
From: Bjorn Helgaas <bjorn.helgaas@hp.com>
To: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>,
	Tony Luck <tony.luck@intel.com>,
	linux-acpi@vger.kernel.org, linux-pci@vger.kernel.org,
	Peter Haight <peterh@sapros.com>, Gary Hade <garyhade@us.ibm.com>,
	linux-kernel@vger.kernel.org, linuxppc-dev@ozlabs.org,
	linux-am33-list@redhat.com, Yinghai Lu <yinghai@kernel.org>,
	Ingo Molnar <mingo@elte.hu>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Larry Finger <Larry.Finger@lwfinger.net>
Subject: [PATCH v5 0/5] PCI: try "pci=use_crs" again
Date: Tue, 23 Feb 2010 10:24:15 -0700	[thread overview]
Message-ID: <20100223172355.17242.55927.stgit@bob.kio> (raw)

Historically, Linux has assumed a single PCI host bridge, with that bridge
claiming all the address space left after RAM and legacy devices are taken out.

If the system contains multiple host bridges, we can no longer operate under
that assumption.  We have to know what parts of the address space are claimed
by each bridge so that when we assign resources to a PCI device, we take them
from a range claimed by the upstream host bridge.

On x86 and ia64, we use ACPI to enumerate all the PCI host bridges in the
system, and part of the host bridge description is the "_CRS" (current resource
settings" property, which lists the address space used by the bridge.  On x86,
we currently ignore most of the _CRS information.  This patch series changes
this, so we will use _CRS to learn about the host bridge windows.

Since most x86 machines with multiple host bridges are relatively new, this
series only turns this on for machines with BIOS dates of 2008 or newer.

Changes from v4 to v5:
    - Add EXPORT_SYMBOL_GPL(pci_bus_resource_n) for module builds.

Changes from v3 to v4:
    - Keep the bus->resource[] table to avoid having to change all arches
      at once, but reduce it to size 4 (minimum required for P2P & CardBus
      bridges).
    - Store additional host bridge or subtractive decode windows in a list.
    - Add pci_bus_for_each_resource() iterator to hide PCI_BUS_NUM_RESOURCES
      and the table/list split.  This reduces the size of the main patch.
    - Remove tidy-up patches completely; they feel like distractions to this
      series.

Changes from v2 to v3:
    - Fix alpha, powerpc, and mn10300 references to pci_bus resource table
    - Turn on "pci=use_crs" for 2008 and newer, not 2010 and newer
    - Remove Toshiba A355 quirk (BIOS date is 2009, so automatically included)
    - Remove IBM x3850 and x3950 quirks (BIOS dates in 2008, so automatically
      included)
    - Leave IBM x3800 quirk (pre-2008 BIOS)
    - Use "bool" for pci_use_crs (new to me, but I see akpm suggesting it)
    - Reorder so the important patches are first

Changes from v1 to v2:
    - Rebase to be6e9f7853e
    - Add patch to clean up "disabled window" printk
    - Add bugzilla reference comment in use_crs DMI quirk
---

Bjorn Helgaas (5):
      PCI: split up pci_read_bridge_bases()
      PCI: read bridge windows before filling in subtractive decode resources
      PCI: add pci_bus_for_each_resource(), remove direct bus->resource[] refs
      PCI: augment bus resource table with a list
      x86/PCI: use host bridge _CRS info by default on 2008 and newer machines


 Documentation/kernel-parameters.txt  |    8 ++-
 arch/ia64/include/asm/acpi.h         |    1 
 arch/ia64/pci/pci.c                  |   17 ++-----
 arch/mn10300/unit-asb2305/pci.c      |    6 +-
 arch/powerpc/kernel/pci-common.c     |   11 ++---
 arch/powerpc/platforms/fsl_uli1575.c |   12 ++---
 arch/x86/include/asm/pci_x86.h       |    1 
 arch/x86/pci/acpi.c                  |   82 ++++++++++++++++++++--------------
 arch/x86/pci/bus_numa.c              |    3 +
 arch/x86/pci/bus_numa.h              |    3 -
 arch/x86/pci/common.c                |    3 +
 drivers/acpi/pci_root.c              |    1 
 drivers/pci/bus.c                    |   49 +++++++++++++++++++-
 drivers/pci/hotplug/shpchp_sysfs.c   |    9 +---
 drivers/pci/pci.c                    |    5 +-
 drivers/pci/probe.c                  |   68 ++++++++++++++++++++++------
 drivers/pci/setup-bus.c              |   10 ++--
 drivers/pcmcia/rsrc_nonstatic.c      |    3 -
 drivers/pcmcia/yenta_socket.c        |    5 +-
 include/acpi/acpi_drivers.h          |    1 
 include/linux/pci.h                  |   36 +++++++++++++--
 21 files changed, 226 insertions(+), 108 deletions(-)

WARNING: multiple messages have this Message-ID (diff)
From: Bjorn Helgaas <bjorn.helgaas@hp.com>
To: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>,
	Tony Luck <tony.luck@intel.com>,
	linuxppc-dev@ozlabs.org, linux-pci@vger.kernel.org,
	Peter Haight <peterh@sapros.com>, Gary Hade <garyhade@us.ibm.com>,
	linux-kernel@vger.kernel.org, linux-acpi@vger.kernel.org,
	linux-am33-list@redhat.com,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Ingo Molnar <mingo@elte.hu>, Yinghai Lu <yinghai@kernel.org>,
	Larry Finger <Larry.Finger@lwfinger.net>
Subject: [PATCH v5 0/5] PCI: try "pci=use_crs" again
Date: Tue, 23 Feb 2010 10:24:15 -0700	[thread overview]
Message-ID: <20100223172355.17242.55927.stgit@bob.kio> (raw)

Historically, Linux has assumed a single PCI host bridge, with that bridge
claiming all the address space left after RAM and legacy devices are taken out.

If the system contains multiple host bridges, we can no longer operate under
that assumption.  We have to know what parts of the address space are claimed
by each bridge so that when we assign resources to a PCI device, we take them
from a range claimed by the upstream host bridge.

On x86 and ia64, we use ACPI to enumerate all the PCI host bridges in the
system, and part of the host bridge description is the "_CRS" (current resource
settings" property, which lists the address space used by the bridge.  On x86,
we currently ignore most of the _CRS information.  This patch series changes
this, so we will use _CRS to learn about the host bridge windows.

Since most x86 machines with multiple host bridges are relatively new, this
series only turns this on for machines with BIOS dates of 2008 or newer.

Changes from v4 to v5:
    - Add EXPORT_SYMBOL_GPL(pci_bus_resource_n) for module builds.

Changes from v3 to v4:
    - Keep the bus->resource[] table to avoid having to change all arches
      at once, but reduce it to size 4 (minimum required for P2P & CardBus
      bridges).
    - Store additional host bridge or subtractive decode windows in a list.
    - Add pci_bus_for_each_resource() iterator to hide PCI_BUS_NUM_RESOURCES
      and the table/list split.  This reduces the size of the main patch.
    - Remove tidy-up patches completely; they feel like distractions to this
      series.

Changes from v2 to v3:
    - Fix alpha, powerpc, and mn10300 references to pci_bus resource table
    - Turn on "pci=use_crs" for 2008 and newer, not 2010 and newer
    - Remove Toshiba A355 quirk (BIOS date is 2009, so automatically included)
    - Remove IBM x3850 and x3950 quirks (BIOS dates in 2008, so automatically
      included)
    - Leave IBM x3800 quirk (pre-2008 BIOS)
    - Use "bool" for pci_use_crs (new to me, but I see akpm suggesting it)
    - Reorder so the important patches are first

Changes from v1 to v2:
    - Rebase to be6e9f7853e
    - Add patch to clean up "disabled window" printk
    - Add bugzilla reference comment in use_crs DMI quirk
---

Bjorn Helgaas (5):
      PCI: split up pci_read_bridge_bases()
      PCI: read bridge windows before filling in subtractive decode resources
      PCI: add pci_bus_for_each_resource(), remove direct bus->resource[] refs
      PCI: augment bus resource table with a list
      x86/PCI: use host bridge _CRS info by default on 2008 and newer machines


 Documentation/kernel-parameters.txt  |    8 ++-
 arch/ia64/include/asm/acpi.h         |    1 
 arch/ia64/pci/pci.c                  |   17 ++-----
 arch/mn10300/unit-asb2305/pci.c      |    6 +-
 arch/powerpc/kernel/pci-common.c     |   11 ++---
 arch/powerpc/platforms/fsl_uli1575.c |   12 ++---
 arch/x86/include/asm/pci_x86.h       |    1 
 arch/x86/pci/acpi.c                  |   82 ++++++++++++++++++++--------------
 arch/x86/pci/bus_numa.c              |    3 +
 arch/x86/pci/bus_numa.h              |    3 -
 arch/x86/pci/common.c                |    3 +
 drivers/acpi/pci_root.c              |    1 
 drivers/pci/bus.c                    |   49 +++++++++++++++++++-
 drivers/pci/hotplug/shpchp_sysfs.c   |    9 +---
 drivers/pci/pci.c                    |    5 +-
 drivers/pci/probe.c                  |   68 ++++++++++++++++++++++------
 drivers/pci/setup-bus.c              |   10 ++--
 drivers/pcmcia/rsrc_nonstatic.c      |    3 -
 drivers/pcmcia/yenta_socket.c        |    5 +-
 include/acpi/acpi_drivers.h          |    1 
 include/linux/pci.h                  |   36 +++++++++++++--
 21 files changed, 226 insertions(+), 108 deletions(-)

             reply	other threads:[~2010-02-23 17:24 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-02-23 17:24 Bjorn Helgaas [this message]
2010-02-23 17:24 ` [PATCH v5 0/5] PCI: try "pci=use_crs" again Bjorn Helgaas
2010-02-23 17:24 ` [PATCH v5 1/5] PCI: split up pci_read_bridge_bases() Bjorn Helgaas
2010-02-23 17:24   ` Bjorn Helgaas
2010-02-23 17:24 ` [PATCH v5 2/5] PCI: read bridge windows before filling in subtractive decode resources Bjorn Helgaas
2010-02-23 17:24   ` Bjorn Helgaas
2010-02-23 17:24 ` [PATCH v5 3/5] PCI: add pci_bus_for_each_resource(), remove direct bus->resource[] refs Bjorn Helgaas
2010-02-23 17:24   ` Bjorn Helgaas
2010-02-23 17:24 ` [PATCH v5 4/5] PCI: augment bus resource table with a list Bjorn Helgaas
2010-02-23 17:24   ` Bjorn Helgaas
2010-02-23 17:24 ` [PATCH v5 5/5] x86/PCI: use host bridge _CRS info by default on 2008 and newer machines Bjorn Helgaas
2010-02-23 17:24   ` Bjorn Helgaas
2010-02-23 18:06 ` [PATCH v5 0/5] PCI: try "pci=use_crs" again Jesse Barnes
2010-02-23 18:06   ` Jesse Barnes

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20100223172355.17242.55927.stgit@bob.kio \
    --to=bjorn.helgaas@hp.com \
    --cc=Larry.Finger@lwfinger.net \
    --cc=garyhade@us.ibm.com \
    --cc=jbarnes@virtuousgeek.org \
    --cc=linux-acpi@vger.kernel.org \
    --cc=linux-am33-list@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=linuxppc-dev@ozlabs.org \
    --cc=mingo@elte.hu \
    --cc=mjg59@srcf.ucam.org \
    --cc=peterh@sapros.com \
    --cc=tony.luck@intel.com \
    --cc=torvalds@linux-foundation.org \
    --cc=yinghai@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.