Skip to contentSkip to conversation

ACPI & driver patches for 2.6.31-rc1

GIT PULL Superseded Len Brown 4 participants 8 messages Jun 26, 2009 – Jun 29, 2009

Subsystems (2)
lore
Share & tools
lore

Len Brown

author reviewer Tested-by Darren Salt
#0 lore raw

Hi Linus,

please pull from:

git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6.git release

This includes a workaround for a just-merged oops regression -- reducing the bug's severity while we get a proper fix in place.

It also includes some driver-specific patches that were staged for the merge window, but missed-the bus because of upstream conflicts with patches from another tree.

This will update the files shown below.

thanks!

Len Brown
Intel Open Source Technology Center

ps. individual patches are available on [email protected] and a consolidated plain patch is available here:
http://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/patches/2.6.31/acpi-release-20090521-2.6.31-rc1.diff.gz

 drivers/acpi/pci_root.c             |    2 +-
 drivers/platform/x86/Kconfig        |    2 +
 drivers/platform/x86/eeepc-laptop.c |  346 ++++++++++++++++++++++++-----------
 3 files changed, 239 insertions(+), 111 deletions(-)

through these commits:

Corentin Chary (6):
eeepc-laptop: Register as a pci-hotplug device
eeepc-laptop: rfkill refactoring
eeepc-laptop: right parent device
eeepc-laptop: makes get_acpi() returns -ENODEV
eeepc-laptop: get the right value for CMSG
eeepc-laptop: add rfkill support for the 3G modem in Eee PC 901 Go

Joe Perches (1):
eeepc-laptop.c: use pr_fmt and pr_<level>

Troy Moure (1):
ACPI: video: prevent NULL deref in acpi_get_pci_dev()

with this log:

commit 412af97838828bc6d035a1902c8974f944663da6

Author: Troy Moure <[email protected]>
Date:   Thu Jun 25 17:05:35 2009 -0600

ACPI: video: prevent NULL deref in acpi_get_pci_dev()

ref: http://thread.gmane.org/gmane.linux.kernel/857228/focus=857468

When the ACPI video driver initializes, it does a namespace walk looking for for supported devices. When we find an appropriate handle, we walk up the ACPI tree looking for a PCI root bus, and then walk back down the PCI bus, assuming that every device inbetween is a P2P bridge.

This assumption is not correct, and is reported broken on at least:

Dell Latitude E6400
ThinkPad X61
Dell XPS M1330

Add a NULL deref check to prevent boot panics.

commit 3cd530b5aaffd27b231f9717730f2f6684c00bda

Author: Corentin Chary <[email protected]>
Date:   Thu Jun 25 13:25:42 2009 +0200

eeepc-laptop: add rfkill support for the 3G modem in Eee PC 901 Go

commit dbfa3ba90dfe353a56e107cff5bce9fb7976f06f

Author: Corentin Chary <[email protected]>
Date:   Thu Jun 25 13:25:41 2009 +0200

eeepc-laptop: get the right value for CMSG

CMSG is an ACPI method used to find features available on an Eee PC. But some features are never repported, even if present.

If the getter of a feature is present, this patch will set the corresponding bit in cmsg.

commit f36509e7248631671d02f48d1a88f56cdeb54ed8

Author: Corentin Chary <[email protected]>
Date:   Thu Jun 25 13:25:40 2009 +0200

eeepc-laptop: makes get_acpi() returns -ENODEV

If there is there is no getter defined, get_acpi() will return -ENODEV.

commit 1ddec2f9435e77b4d3f50eced68c4c942f2bcd4b

Author: Corentin Chary <[email protected]>
Date:   Thu Jun 25 13:25:39 2009 +0200

eeepc-laptop: right parent device

commit 7de39389d8f61aa517ce2a8b4d925acc62696ae5

Author: Corentin Chary <[email protected]>
Date:   Thu Jun 25 13:25:38 2009 +0200

eeepc-laptop: rfkill refactoring

Refactor rfkill code, because we'll add another rfkill for wwan3g later.

commit 19b532892834b7f1c04b2940ac73177dc566fed5

Author: Joe Perches <[email protected]>
Date:   Thu Jun 25 13:25:37 2009 +0200

eeepc-laptop.c: use pr_fmt and pr_<level>

Convert the unusual printk(EEEPC_<level> uses to the more standard pr_fmt and pr_<level>(.

commit 2b121bc262fa03c94e653b2d44356c2f86c1bcdc

Author: Corentin Chary <[email protected]>
Date:   Thu Jun 25 13:25:36 2009 +0200

eeepc-laptop: Register as a pci-hotplug device

The eee contains a logically (but not physically) hotpluggable PCIe slot. Currently this is handled by adding or removing the PCI device in response to rfkill events, but if a user has forced pciehp to bind to it (with the force=1 argument) then both drivers will try to handle the event and hilarity (in the form of oopses) will ensue. This can be avoided by having eee-laptop register the slot as a hotplug slot. Only one of pciehp and eee-laptop will successfully register this, avoiding the problem.

#1 lore raw

PATCH 1/1 eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: [git pull request] ACPI & driver patches for 2.6.31-rc1)

FYI, there's a post-rc1 build regression with certain configs:

drivers/built-in.o: In function `pci_hp_deregister':
(.text+0xb166): undefined reference to pci_hp_remove_module_link' drivers/built-in.o: In function pci_hp_deregister':
(.text+0xb19f): undefined reference to pci_destroy_slot' drivers/built-in.o: In function __pci_hp_register':
(.text+0xb583): undefined reference to pci_create_slot' drivers/built-in.o: In function __pci_hp_register':
(.text+0xb5b1): undefined reference to `pci_hp_create_module_link'
make: *** [.tmp_vmlinux1] Error 1

Caused by:

| 2b121bc262fa03c94e653b2d44356c2f86c1bcdc is first bad commit
| commit 2b121bc262fa03c94e653b2d44356c2f86c1bcdc
| Date:   Thu Jun 25 13:25:36 2009 +0200
|
|     eeepc-laptop: Register as a pci-hotplug device

The problem is caused by this portion of the commit:

drivers/platform/x86/Kconfig +2 -0
@@ -357,6 +357,8 @@ config EEEPC_LAPTOP
        depends on RFKILL || RFKILL = n
        select BACKLIGHT_CLASS_DEVICE
        select HWMON
+       select HOTPLUG
+       select HOTPLUG_PCI if PCI
        ---help---
          This driver supports the Fn-Fx keys on Eee PC laptops.
 

We never select CONFIG_HOTPLUG_PCI, as it is a non-trivial, interactive, many-child-dependencies configuration option.

The select caused partial dependencies to be instantiated by the Kconfig language, and allowed the PCI hotplug core code to be built without SYSFS - so drivers/pci/slot.c wasnt built but was being relied on.

The complexity of select statements is already a warning sign that something unusual is done here. They should be avoided generally, except for clear-cut 'leaf' modules.

The patch below fixes the build failure by turning the select lines into a single 'depends on' line. There's no significant reduction in utility of the driver: it already has a list of dependencies, and basically all distros enable hotplug-pci anyway.

Ingo

drivers/platform/x86/Kconfig +1 -2
index fee6a40..46dad12 100644
@@ -355,10 +355,9 @@ config EEEPC_LAPTOP
 	depends on INPUT
 	depends on EXPERIMENTAL
 	depends on RFKILL || RFKILL = n
+	depends on HOTPLUG_PCI
 	select BACKLIGHT_CLASS_DEVICE
 	select HWMON
-	select HOTPLUG
-	select HOTPLUG_PCI if PCI
 	---help---
 	  This driver supports the Fn-Fx keys on Eee PC laptops.
#2 lore raw

PATCH v2 1/1 eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: [git pull request] ACPI & driver patches for 2.6.31-rc1)

Quoted 2 lines · Ingo Molnar

The patch below fixes the build failure by turning the select lines into a single 'depends on' line. [...]

i did the fix against the bisection point and the patch has a conflict against later mainline - below is a merged up patch.

Ingo

------------------>

Quoted 1 line

From 85ee49b3ff6608e66122d17d370d0a7e5d008198 Mon Sep 17 00:00:00 2001

From: Ingo Molnar <[email protected]>
Date: Mon, 29 Jun 2009 10:07:54 +0200
Subject: [PATCH] eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: ACPI & driver patches for 2.6.31-rc1)

FYI, there's a post-rc1 build regression with certain configs:

drivers/built-in.o: In function `pci_hp_deregister':
(.text+0xb166): undefined reference to pci_hp_remove_module_link' drivers/built-in.o: In function pci_hp_deregister':
(.text+0xb19f): undefined reference to pci_destroy_slot' drivers/built-in.o: In function __pci_hp_register':
(.text+0xb583): undefined reference to pci_create_slot' drivers/built-in.o: In function __pci_hp_register':
(.text+0xb5b1): undefined reference to `pci_hp_create_module_link'
make: *** [.tmp_vmlinux1] Error 1

Caused by:

| 2b121bc262fa03c94e653b2d44356c2f86c1bcdc is first bad commit
| commit 2b121bc262fa03c94e653b2d44356c2f86c1bcdc
| Date:   Thu Jun 25 13:25:36 2009 +0200
|
|     eeepc-laptop: Register as a pci-hotplug device

The problem is caused by this portion of the commit:

---
 drivers/platform/x86/Kconfig |    5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)
drivers/platform/x86/Kconfig +3 -2
index 4bc72d0..e120be8 100644
@@ -356,10 +356,11 @@ config EEEPC_LAPTOP
 	depends on INPUT
 	depends on EXPERIMENTAL
 	depends on RFKILL || RFKILL = n
+	depends on HOTPLUG_PCI
 	select BACKLIGHT_CLASS_DEVICE
 	select HWMON
-	select HOTPLUG
-	select HOTPLUG_PCI if PCI
+       select HOTPLUG
+       select HOTPLUG_PCI if PCI
 	---help---
 	  This driver supports the Fn-Fx keys on Eee PC laptops.
 
#3 lore raw

PATCH v3 1/1 eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: [git pull request] ACPI & driver patches for 2.6.31-rc1)

Review

Quoted 10 lines · Ingo Molnar
Quoted 2 lines

The patch below fixes the build failure by turning the select lines into a single 'depends on' line. [...]

i did the fix against the bisection point and the patch has a conflict against later mainline - below is a merged up patch.

(quoted diff) +2 -2
-	select HOTPLUG
-	select HOTPLUG_PCI if PCI
+       select HOTPLUG
+       select HOTPLUG_PCI if PCI

uhm, not a very nice conflict resolution this is.

The one below is actually tested ;-)

Ingo

-------------------->

Quoted 1 line

From de3e1cacc93ca48a288e27af5e82a365ba90020f Mon Sep 17 00:00:00 2001

From: Ingo Molnar <[email protected]>
Date: Mon, 29 Jun 2009 10:07:54 +0200
Subject: [PATCH] eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: ACPI & driver patches for 2.6.31-rc1)

FYI, there's a post-rc1 build regression with certain configs:

drivers/built-in.o: In function `pci_hp_deregister':
(.text+0xb166): undefined reference to pci_hp_remove_module_link' drivers/built-in.o: In function pci_hp_deregister':
(.text+0xb19f): undefined reference to pci_destroy_slot' drivers/built-in.o: In function __pci_hp_register':
(.text+0xb583): undefined reference to pci_create_slot' drivers/built-in.o: In function __pci_hp_register':
(.text+0xb5b1): undefined reference to `pci_hp_create_module_link'
make: *** [.tmp_vmlinux1] Error 1

Caused by:

| 2b121bc262fa03c94e653b2d44356c2f86c1bcdc is first bad commit
| commit 2b121bc262fa03c94e653b2d44356c2f86c1bcdc
| Date:   Thu Jun 25 13:25:36 2009 +0200
|
|     eeepc-laptop: Register as a pci-hotplug device

The problem is caused by this portion of the commit:

---
 drivers/platform/x86/Kconfig |    3 +--
 1 files changed, 1 insertions(+), 2 deletions(-)
drivers/platform/x86/Kconfig +1 -2
index 4bc72d0..c0be866 100644
@@ -356,10 +356,9 @@ config EEEPC_LAPTOP
 	depends on INPUT
 	depends on EXPERIMENTAL
 	depends on RFKILL || RFKILL = n
+	depends on HOTPLUG_PCI
 	select BACKLIGHT_CLASS_DEVICE
 	select HWMON
-	select HOTPLUG
-	select HOTPLUG_PCI if PCI
 	---help---
 	  This driver supports the Fn-Fx keys on Eee PC laptops.
 
#4 lore raw

Re: [PATCH, v3] eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS

Quoted 44 lines · Ingo Molnar

drivers/platform/x86/Kconfig config EEEPC_LAPTOP

Quoted 9 lines · Ingo Molnar
Quoted 2 lines

The patch below fixes the build failure by turning the select lines into a single 'depends on' line. [...]

i did the fix against the bisection point and the patch has a conflict against later mainline - below is a merged up patch.

drivers/platform/x86/Kconfig +2 -2
-	select HOTPLUG
-	select HOTPLUG_PCI if PCI
+       select HOTPLUG
+       select HOTPLUG_PCI if PCI

uhm, not a very nice conflict resolution this is.

The one below is actually tested ;-)

Ingo

-------------------->
From de3e1cacc93ca48a288e27af5e82a365ba90020f Mon Sep 17 00:00:00 2001
From: Ingo Molnar <[email protected]>
Date: Mon, 29 Jun 2009 10:07:54 +0200
Subject: [PATCH] eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: ACPI & driver patches for 2.6.31-rc1)

FYI, there's a post-rc1 build regression with certain configs:

drivers/built-in.o: In function `pci_hp_deregister':
(.text+0xb166): undefined reference to pci_hp_remove_module_link' drivers/built-in.o: In function pci_hp_deregister':
(.text+0xb19f): undefined reference to pci_destroy_slot' drivers/built-in.o: In function __pci_hp_register':
(.text+0xb583): undefined reference to pci_create_slot' drivers/built-in.o: In function __pci_hp_register':
(.text+0xb5b1): undefined reference to `pci_hp_create_module_link'
make: *** [.tmp_vmlinux1] Error 1

Caused by:

| 2b121bc262fa03c94e653b2d44356c2f86c1bcdc is first bad commit
| commit 2b121bc262fa03c94e653b2d44356c2f86c1bcdc
| Date:   Thu Jun 25 13:25:36 2009 +0200
|
|     eeepc-laptop: Register as a pci-hotplug device
Quoted 3 lines

The problem is caused by this portion of the commit:

Ack. Thanks for fixing this.

Quoted 20 lines
---
 drivers/platform/x86/Kconfig |    3 +--
 1 files changed, 1 insertions(+), 2 deletions(-)
drivers/platform/x86/Kconfig +1 -2
index 4bc72d0..c0be866 100644
@@ -356,10 +356,9 @@ config EEEPC_LAPTOP
 	depends on INPUT
 	depends on EXPERIMENTAL
 	depends on RFKILL || RFKILL = n
+	depends on HOTPLUG_PCI
 	select BACKLIGHT_CLASS_DEVICE
 	select HWMON
-	select HOTPLUG
-	select HOTPLUG_PCI if PCI
 	---help---
 	  This driver supports the Fn-Fx keys on Eee PC laptops.
 
Signature · 3 lines
~Randy
LPC 2009, Sept. 23-25, Portland, Oregon
http://linuxplumbersconf.org/2009/
author reviewer Acked-by on v3 1/1
#5 lore raw

Re: [PATCH, v3] eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: [git pull request] ACPI & driver patches for 2.6.31-rc1)

In reply to Ingo Molnar

Thanks for fixing this config, Ingo.

I agree, our build system can handle "depends", but "select" does not deliver on what it advertises, making it a trap for programmers.

thanks,
Len Brown, Intel Open Source Technology Center

#6 lore raw

Re: [PATCH, v3] eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: [git pull request] ACPI & driver patches for 2.6.31-rc1)

Quoted 5 lines · Len Brown

Thanks for fixing this config, Ingo.

I agree, our build system can handle "depends", but "select" does not deliver on what it advertises, making it a trap for programmers.

Hm, this portion of Kconfig is a trap in a way indeed - on the other hand it is also a useful tool and the simplistic approach to disallow select altogether is a step backwards.

IIRC either Vegard or Steve has come up with a patch series that improved Kconfig to warn about problem selects while still allowing the simple ones. That would deflect most of the bad uses while still allowing the good uses. Not sure what happened to that series.

Ingo

#7 lore raw

Re: [PATCH, v3] eeepc-laptop: Fix build failure with HOTPLUG_PCI && !SYSFS (was: [git pull request] ACPI & driver patches for 2.6.31-rc1)

Quoted 9 lines · Ingo Molnar
Quoted 5 lines · Len Brown

Thanks for fixing this config, Ingo.

I agree, our build system can handle "depends", but "select" does not deliver on what it advertises, making it a trap for programmers.

Quoted 8 lines

Hm, this portion of Kconfig is a trap in a way indeed - on the other hand it is also a useful tool and the simplistic approach to disallow select altogether is a step backwards.

IIRC either Vegard or Steve has come up with a patch series that improved Kconfig to warn about problem selects while still allowing the simple ones. That would deflect most of the bad uses while still allowing the good uses. Not sure what happened to that series.

Steve came up with a patch that warned about problematic select only in a defconfig. It did not warn properly for all the problematic selects.
I ran (as usual) out of time last time I looked :-(
I would likely need a full day or two to come up with somthing, but these days finding an hour is not easy.

Sam