From: Lukas Wunner <lukas@wunner.de>
To: Bjorn Helgaas <bhelgaas@google.com>
Cc: Mika Westerberg <mika.westerberg@linux.intel.com>,
"Rafael J. Wysocki" <rafael.j.wysocki@intel.com>,
Ashok Raj <ashok.raj@intel.com>,
Keith Busch <keith.busch@intel.com>,
Yinghai Lu <yinghai@kernel.org>, Sinan Kaya <okaya@kernel.org>,
linux-pci@vger.kernel.org, Greg Kroah-Hartman <greg@kroah.com>,
Thomas Gleixner <tglx@linutronix.de>,
Mayurkumar Patel <mayurkumar.patel@intel.com>,
Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>,
Stefan Roese <sr@denx.de>, Rajat Jain <rajatja@google.com>,
Alex Williamson <alex.williamson@redhat.com>,
Andreas Noever <andreas.noever@gmail.com>
Subject: [PATCH 00/32] Rework pciehp event handling & add runtime PM
Date: Sat, 16 Jun 2018 21:25:00 +0200 [thread overview]
Message-ID: <cover.1529173804.git.lukas@wunner.de> (raw)
Rework pciehp to use modern, threaded IRQ handling. The slot is powered
on and off synchronously in the IRQ thread, no indirection via a work
queue anymore.
When the slot is enabled/disabled by the user via sysfs or an Attention
Button press, a request is sent to the IRQ thread. The IRQ thread is
thus the sole entity enabling/disabling the slot.
The IRQ thread can cope with missed events, e.g. if a card is inserted
and immediately pulled out before the IRQ thread had a chance to react.
It also tolerates an initially unstable link as observed in the wild by
Stefan Roese.
Finally, runtime PM support is added. This was the original motivation
of the series because runtime suspending hotplug ports is needed to power
down Thunderbolt controllers on idle, which saves ~1.5W per controller.
Runtime resuming ports takes tenths of milliseconds during which events
may be missed, this in turn necessitated the event handling rework.
I've pushed the series to GitHub to ease reviewing/fetching:
https://github.com/l1k/linux/commits/pciehp_runpm_v2
Please review and test.
Thanks,
Lukas
Lukas Wunner (32):
PCI: hotplug: Don't leak pci_slot on registration failure
PCI: pciehp: Fix UAF on unplug
PCI: pciehp: Fix deadlock on unplug
PCI: pciehp: Fix unprotected list iteration in IRQ handler
PCI: pciehp: Drop unnecessary NULL pointer check
PCI: pciehp: Declare pciehp_unconfigure_device() void
PCI: pciehp: Document struct slot and struct controller
genirq: Synchronize only with single thread on free_irq()
PCI: pciehp: Convert to threaded IRQ
PCI: pciehp: Convert to threaded polling
PCI: pciehp: Stop blinking on slot enable failure
PCI: pciehp: Handle events synchronously
PCI: pciehp: Drop slot workqueue
PCI: hotplug: Demidlayer registration with the core
PCI: pciehp: Publish to user space last on probe
PCI: pciehp: Track enable/disable status
PCI: pciehp: Enable/disable exclusively from IRQ thread
PCI: pciehp: Drop enable/disable lock
PCI: pciehp: Declare pciehp_enable/disable_slot() static
PCI: pciehp: Tolerate initially unstable link
PCI: pciehp: Become resilient to missed events
PCI: pciehp: Always enable occupied slot on probe
PCI: pciehp: Avoid slot access during reset
PCI: portdrv: Deduplicate PM callback iterator
PCI: pciehp: Clear spurious events earlier on resume
PCI: pciehp: Obey compulsory command delay after resume
PCI: pciehp: Support interrupts sent from D3hot
PCI: pciehp: Resume to D0 on enable/disable
PCI: pciehp: Resume parent to D0 on config space access
PCI: sysfs: Resume to D0 on function reset
PCI: Whitelist native hotplug ports for runtime D3
PCI: Whitelist Thunderbolt ports for runtime D3
drivers/pci/hotplug/acpiphp_core.c | 22 +-
drivers/pci/hotplug/cpci_hotplug_core.c | 14 +-
drivers/pci/hotplug/cpqphp_core.c | 16 +-
drivers/pci/hotplug/ibmphp_core.c | 15 +-
drivers/pci/hotplug/ibmphp_ebda.c | 20 --
drivers/pci/hotplug/pci_hotplug_core.c | 134 +++++++--
drivers/pci/hotplug/pciehp.h | 117 ++++++--
drivers/pci/hotplug/pciehp_core.c | 104 ++++---
drivers/pci/hotplug/pciehp_ctrl.c | 352 ++++++++++--------------
drivers/pci/hotplug/pciehp_hpc.c | 261 ++++++++++--------
drivers/pci/hotplug/pciehp_pci.c | 42 ++-
drivers/pci/hotplug/pcihp_skeleton.c | 16 +-
drivers/pci/hotplug/pnv_php.c | 5 +-
drivers/pci/hotplug/rpaphp_core.c | 2 +-
drivers/pci/hotplug/rpaphp_slot.c | 13 +-
drivers/pci/hotplug/s390_pci_hpc.c | 13 +-
drivers/pci/hotplug/sgi_hotplug.c | 9 +-
drivers/pci/hotplug/shpchp_core.c | 20 +-
drivers/pci/pci-sysfs.c | 2 +
drivers/pci/pci.c | 21 +-
drivers/pci/pcie/portdrv.h | 2 +
drivers/pci/pcie/portdrv_core.c | 30 +-
drivers/pci/pcie/portdrv_pci.c | 2 +
drivers/platform/x86/asus-wmi.c | 12 +-
drivers/platform/x86/eeepc-laptop.c | 12 +-
include/linux/pci_hotplug.h | 15 +-
kernel/irq/manage.c | 21 +-
27 files changed, 687 insertions(+), 605 deletions(-)
--
2.17.1
next reply other threads:[~2018-06-16 19:25 UTC|newest]
Thread overview: 66+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-06-16 19:25 Lukas Wunner [this message]
2018-06-16 19:25 ` [PATCH 26/32] PCI: pciehp: Obey compulsory command delay after resume Lukas Wunner
2018-06-16 19:25 ` [PATCH 30/32] PCI: sysfs: Resume to D0 on function reset Lukas Wunner
2018-06-16 19:25 ` [PATCH 17/32] PCI: pciehp: Enable/disable exclusively from IRQ thread Lukas Wunner
2018-06-21 11:58 ` Mika Westerberg
2018-06-16 19:25 ` [PATCH 10/32] PCI: pciehp: Convert to threaded polling Lukas Wunner
2018-06-16 19:25 ` [PATCH 08/32] genirq: Synchronize only with single thread on free_irq() Lukas Wunner
2018-07-12 22:21 ` Bjorn Helgaas
2018-07-13 7:21 ` Lukas Wunner
2018-07-13 11:44 ` Bjorn Helgaas
2018-07-16 12:37 ` Bjorn Helgaas
2018-07-16 13:37 ` Lukas Wunner
2018-06-16 19:25 ` [PATCH 28/32] PCI: pciehp: Resume to D0 on enable/disable Lukas Wunner
2018-06-16 19:25 ` [PATCH 09/32] PCI: pciehp: Convert to threaded IRQ Lukas Wunner
2018-06-19 23:16 ` Keith Busch
2018-06-20 11:01 ` Lukas Wunner
2018-06-16 19:25 ` [PATCH 22/32] PCI: pciehp: Always enable occupied slot on probe Lukas Wunner
2018-06-16 19:25 ` [PATCH 25/32] PCI: pciehp: Clear spurious events earlier on resume Lukas Wunner
2018-06-16 19:25 ` [PATCH 19/32] PCI: pciehp: Declare pciehp_enable/disable_slot() static Lukas Wunner
2018-06-16 19:25 ` [PATCH 06/32] PCI: pciehp: Declare pciehp_unconfigure_device() void Lukas Wunner
2018-06-16 19:25 ` [PATCH 29/32] PCI: pciehp: Resume parent to D0 on config space access Lukas Wunner
2018-06-16 19:25 ` [PATCH 31/32] PCI: Whitelist native hotplug ports for runtime D3 Lukas Wunner
2018-06-16 19:25 ` [PATCH 20/32] PCI: pciehp: Tolerate initially unstable link Lukas Wunner
2018-06-16 19:25 ` [PATCH 01/32] PCI: hotplug: Don't leak pci_slot on registration failure Lukas Wunner
2018-06-16 19:25 ` [PATCH 05/32] PCI: pciehp: Drop unnecessary NULL pointer check Lukas Wunner
2018-06-16 19:25 ` [PATCH 14/32] PCI: hotplug: Demidlayer registration with the core Lukas Wunner
2018-06-17 16:44 ` Andy Shevchenko
2018-07-16 12:46 ` Bjorn Helgaas
2018-07-16 14:14 ` Andy Shevchenko
2018-06-16 19:25 ` [PATCH 13/32] PCI: pciehp: Drop slot workqueue Lukas Wunner
2018-06-16 19:25 ` [PATCH 15/32] PCI: pciehp: Publish to user space last on probe Lukas Wunner
2018-06-16 19:25 ` [PATCH 12/32] PCI: pciehp: Handle events synchronously Lukas Wunner
2018-06-16 19:25 ` [PATCH 32/32] PCI: Whitelist Thunderbolt ports for runtime D3 Lukas Wunner
2018-06-21 11:13 ` Mika Westerberg
2018-07-18 19:30 ` Lukas Wunner
2018-07-20 15:23 ` Mika Westerberg
2018-07-20 16:00 ` Mika Westerberg
2018-07-20 20:33 ` Bjorn Helgaas
2018-06-16 19:25 ` [PATCH 16/32] PCI: pciehp: Track enable/disable status Lukas Wunner
2018-06-16 19:25 ` [PATCH 03/32] PCI: pciehp: Fix deadlock on unplug Lukas Wunner
2018-09-06 16:01 ` Mika Westerberg
2018-09-06 16:26 ` Lukas Wunner
2018-09-06 18:08 ` Mika Westerberg
2018-06-16 19:25 ` [PATCH 04/32] PCI: pciehp: Fix unprotected list iteration in IRQ handler Lukas Wunner
2018-06-16 19:25 ` [PATCH 02/32] PCI: pciehp: Fix UAF on unplug Lukas Wunner
2018-06-16 19:25 ` [PATCH 24/32] PCI: portdrv: Deduplicate PM callback iterator Lukas Wunner
2018-06-16 19:25 ` [PATCH 18/32] PCI: pciehp: Drop enable/disable lock Lukas Wunner
2018-06-16 19:25 ` [PATCH 23/32] PCI: pciehp: Avoid slot access during reset Lukas Wunner
2018-06-21 12:06 ` Mika Westerberg
2018-06-22 9:23 ` Lukas Wunner
2018-06-25 13:10 ` Mika Westerberg
2018-06-16 19:25 ` [PATCH 27/32] PCI: pciehp: Support interrupts sent from D3hot Lukas Wunner
2018-07-12 23:03 ` Bjorn Helgaas
2018-06-16 19:25 ` [PATCH 07/32] PCI: pciehp: Document struct slot and struct controller Lukas Wunner
2018-06-16 19:25 ` [PATCH 11/32] PCI: pciehp: Stop blinking on slot enable failure Lukas Wunner
2018-06-16 19:25 ` [PATCH 21/32] PCI: pciehp: Become resilient to missed events Lukas Wunner
2018-06-21 12:19 ` [PATCH 00/32] Rework pciehp event handling & add runtime PM Mika Westerberg
2018-06-27 13:35 ` Patel, Mayurkumar
2018-07-12 22:28 ` Bjorn Helgaas
2018-07-13 7:54 ` Lukas Wunner
2018-07-13 11:43 ` Bjorn Helgaas
2018-07-16 14:20 ` Bjorn Helgaas
2018-07-19 9:43 ` Lukas Wunner
2018-07-19 19:05 ` Bjorn Helgaas
2018-07-19 22:50 ` Bjorn Helgaas
2018-07-28 5:44 ` Lukas Wunner
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=cover.1529173804.git.lukas@wunner.de \
--to=lukas@wunner.de \
--cc=alex.williamson@redhat.com \
--cc=andreas.noever@gmail.com \
--cc=ashok.raj@intel.com \
--cc=bhelgaas@google.com \
--cc=greg@kroah.com \
--cc=kaneshige.kenji@jp.fujitsu.com \
--cc=keith.busch@intel.com \
--cc=linux-pci@vger.kernel.org \
--cc=mayurkumar.patel@intel.com \
--cc=mika.westerberg@linux.intel.com \
--cc=okaya@kernel.org \
--cc=rafael.j.wysocki@intel.com \
--cc=rajatja@google.com \
--cc=sr@denx.de \
--cc=tglx@linutronix.de \
--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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).