ACPI patches for 2.6.22 - part 2

Hi Linus,

please pull from:

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

Please test this on your Evo before pushing it upstream.

Due to a crash on some ACER laptops, this reverts the Notify-on-stack fix from 2.6.21 and goes back to an updated notify work-queue fix that we've discussed (and struggled with) before.

I'm not totally satisfied that we've reverse engineered all the bizarre Notify cases that BIOS send us, but this should work on the machines we've seen to date...

The bulk of the text changes are to restore the invariant that ACPICA is Lindent clean.

thanks!

-Len

ps. individual patches are available on [email protected]
and a consolidated plain patch is available here:
ftp://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/patches/release/2.6.21/acpi-release-20070126-2.6.21.diff.gz

 drivers/acpi/dispatcher/dsmethod.c |   12 
 drivers/acpi/dispatcher/dsopcode.c |    3 
 drivers/acpi/dispatcher/dsutils.c  |    7 
 drivers/acpi/dispatcher/dswstate.c |    9 
 drivers/acpi/ec.c                  |   39 -
 drivers/acpi/events/evgpe.c        |    5 
 drivers/acpi/events/evgpeblk.c     |    3 
 drivers/acpi/events/evmisc.c       |   20 
 drivers/acpi/events/evregion.c     |   15 
 drivers/acpi/events/evrgnini.c     |    3 
 drivers/acpi/events/evxface.c      |    7 
 drivers/acpi/events/evxfevnt.c     |    2 
 drivers/acpi/executer/exconvrt.c   |    5 
 drivers/acpi/executer/excreate.c   |    6 
 drivers/acpi/executer/exdump.c     |   17 
 drivers/acpi/executer/exmutex.c    |   37 -
 drivers/acpi/executer/exnames.c    |    3 
 drivers/acpi/executer/exprep.c     |    2 
 drivers/acpi/executer/exresop.c    |    3 
 drivers/acpi/executer/exsystem.c   |   30 -
 drivers/acpi/executer/exutils.c    |  104 +++-
 drivers/acpi/hardware/hwsleep.c    |    1 
 drivers/acpi/namespace/nseval.c    |   13 
 drivers/acpi/namespace/nsinit.c    |    7 
 drivers/acpi/namespace/nswalk.c    |    6 
 drivers/acpi/namespace/nsxfeval.c  |   17 
 drivers/acpi/osl.c                 |   45 +
 drivers/acpi/parser/psopcode.c     |  618 ++++++++++++-------------
 drivers/acpi/resources/rscalc.c    |    3 
 drivers/acpi/resources/rscreate.c  |   13 
 drivers/acpi/resources/rsdump.c    |    8 
 drivers/acpi/resources/rsinfo.c    |    2 
 drivers/acpi/resources/rslist.c    |    7 
 drivers/acpi/resources/rsmisc.c    |    4 
 drivers/acpi/resources/rsutils.c   |    6 
 drivers/acpi/resources/rsxface.c   |    3 
 drivers/acpi/sleep/main.c          |    3 
 drivers/acpi/sleep/proc.c          |   11 
 drivers/acpi/tables/tbfadt.c       |    6 
 drivers/acpi/tables/tbxface.c      |   16 
 drivers/acpi/thermal.c             |  104 ----
 drivers/acpi/utilities/utalloc.c   |    1 
 drivers/acpi/utilities/utcache.c   |    3 
 drivers/acpi/utilities/utcopy.c    |    4 
 drivers/acpi/utilities/utdebug.c   |    4 
 drivers/acpi/utilities/utdelete.c  |    1 
 drivers/acpi/utilities/utglobal.c  |    6 
 drivers/acpi/utilities/utmisc.c    |    6 
 drivers/acpi/utilities/utmutex.c   |    8 
 drivers/acpi/utilities/utresrc.c   |    1 
 drivers/acpi/utilities/utxface.c   |    2 
 drivers/misc/Kconfig               |    2 
 drivers/misc/asus-laptop.c         |   66 ++
 drivers/misc/msi-laptop.c          |   12 
 drivers/misc/sony-laptop.c         |    8 
 include/acpi/acdispat.h            |    9 
 include/acpi/acglobal.h            |    2 
 include/acpi/acinterp.h            |    9 
 include/acpi/aclocal.h             |    2 
 include/acpi/acnamesp.h            |    6 
 include/acpi/acobject.h            |    4 
 include/acpi/acpi_bus.h            |    9 
 include/acpi/acpi_drivers.h        |    6 
 include/acpi/acpi_numa.h           |    2 
 include/acpi/acpiosxf.h            |    3 
 include/acpi/actypes.h             |    2 
 include/acpi/acutils.h             |    3 
 include/acpi/platform/aclinux.h    |   21 
 include/acpi/processor.h           |   18 
 include/linux/acpi.h               |    3 
 70 files changed, 721 insertions(+), 727 deletions(-)

through these commits:

Alexey Starikovskiy (1):
ACPI: created a dedicated workqueue for notify() execution

Corentin Chary (4):
asus-laptop: notify ALL events
asus-laptop: add GPS support
asus-laptop: fix light sens init
asus-laptop: version bump and lindent

Len Brown (6):
ACPI: thermal trip points are read-only
ACPI: delete un-reliable concept of cooling mode Revert "ACPICA: revert "acpi_serialize" changes" Revert "Execute AML Notify() requests on stack." Revert "ACPICA: fix AML mutex re-entrancy"
ACPICA: Lindent

Lennart Poettering (1):
acpi,msi-laptop: Fall back to EC polling mode for MSI laptop specific EC commands

Mattia Dongili (1):
sony-laptop: rename SONY_LAPTOP_OLD to a more meaningful SONYPI_COMPAT

with this log:

commit f685648e7d6520653fa2641840cccc58ba8ef8e0
Merge: f142051... 00eb43a...
Author: Len Brown <[email protected]>
Date: Thu May 10 04:06:12 2007 -0400

Pull misc-for-upstream into release branch

commit 00eb43a1897a8845d0edb198cec69ac5f1f299dd

Author: Lennart Poettering <[email protected]>
Date:   Fri May 4 14:16:19 2007 +0200

acpi,msi-laptop: Fall back to EC polling mode for MSI laptop specific EC commands

The ACPI EC that is used in MSI laptops knows some non-standard commands for changing the screen brighntess and a few other things, which are used by the msi-laptop.c driver. Unfortunately for these commands no GPE events for IBF and OBF are triggered. Since nowadays the EC code uses the ec_intr=1 mode by default, this causes these operations to timeout, although they don't fail. In result, all operations that you can do with the msi-laptop.c driver take more or less 1s to complete, which is awfully slow.

In one of the more recent kernels (2.6.20?) the EC subsystem has been revamped. With that change the EC timeout has been increased. before that increase the MSI EC accesses were slow -- but not *that* slow, hence I took notice of this limitation of the MSI EC hardware only very recently.

The standard EC operations on the MSI EC as defined in the ACPI spec support GPE events properly.

The following patch adds a new argument "force_poll" to the ec_transaction() function (and friends). If set to 1, the function will poll for IBF/OBF even if ec_intr=1 is enabled. If set to 0 the current behaviour is used. The msi-laptop driver is modified to make use of this new flag, so that OBF/IBF is polled for the special MSI EC transactions -- but only for them.

commit f142051505154da4ba689e129deb06524c5ece8a
Merge: 19ed0d6... a64e62a...
Author: Len Brown <[email protected]>
Date: Thu May 10 02:50:09 2007 -0400

Pull misc-for-upstream into release branch

commit 19ed0d6fa894aec0050ab75185e69785a656da91
Merge: 71b43ca... f398532...
Author: Len Brown <[email protected]>
Date: Thu May 10 02:50:01 2007 -0400

Pull asus into release branch

commit 71b43ca46fc5f0588ef64a14c6133c19a9b3f1ea
Merge: 3dd6786... fd35094...
Author: Len Brown <[email protected]>
Date: Thu May 10 02:49:34 2007 -0400

Pull acpica into release branch

commit 3dd6786f55219f87fd33e153669600d715ca9270
Merge: f697b67... 88db5e1...
Author: Len Brown <[email protected]>
Date: Thu May 10 02:49:21 2007 -0400

Pull bugzilla-8385 into release branch

commit f697b677620d04d8c77841745727de85f7e948b1
Merge: de56037... eaca2d3...
Author: Len Brown <[email protected]>
Date: Thu May 10 02:49:01 2007 -0400

Pull thermal into release branch

commit a64e62a07097f67108f0b68bc15216c3a4a5299b

Author: Mattia Dongili <[email protected]>
Date:   Tue May 1 11:19:53 2007 +0900

sony-laptop: rename SONY_LAPTOP_OLD to a more meaningful SONYPI_COMPAT

commit f398532722b44aa797915d6b784e8694b277c2c7

Author: Corentin Chary <[email protected]>
Date:   Sun May 6 14:48:22 2007 +0200

asus-laptop: version bump and lindent

Version bump, lindent, etc ..

commit 832d995029466e12aef66187bc9103487bf7be9f

Author: Corentin Chary <[email protected]>
Date:   Sun May 6 14:47:29 2007 +0200

asus-laptop: fix light sens init

Fix a stupid light sens detection bug.

commit e539c2f69ca17f71dd41ba89cde67612a1326c57

Author: Corentin Chary <[email protected]>
Date:   Sun May 6 14:47:06 2007 +0200

asus-laptop: add GPS support

Just adds GPS support found in R2H thanks to Sam Lin. It will make a "gps" file in /sys/devices/platform/asus-laptop/.

commit a9dbcb37a45f45813fc64d672a1f7ae6135650b1

Author: Corentin Chary <[email protected]>
Date:   Sun May 6 14:46:41 2007 +0200

asus-laptop: notify ALL events

We need to handle all events, because some dsdt use events >= 0x80

commit fd3509436fde38d4c854bf5a6b83d2c779904f8e

Author: Len Brown <[email protected]>
Date:   Wed May 9 23:34:35 2007 -0400

ACPICA: Lindent

commit 88db5e1489f23876a226f5393fd978ddc09dc5f9

Author: Alexey Starikovskiy <[email protected]>
Date:   Wed May 9 23:31:03 2007 -0400

ACPI: created a dedicated workqueue for notify() execution

HP nx6125/nx6325/... machines have a _GPE handler with an infinite loop sending Notify() events to different ACPI subsystems.

Notify handler in ACPI driver is a C-routine, which may call ACPI interpreter again to get access to some ACPI variables (acpi_evaluate_xxx).
On these HP machines such an evaluation changes state of some variable and lets the loop above break.

In the current ACPI implementation Notify requests are being deferred to the same kacpid workqueue on which the above GPE handler with infinite loop is executing. Thus we have a deadlock -- loop will continue to spin, sending notify events, and at the same time preventing these notify events from being run on a workqueue. All notify events are deferred, thus we see increase in memory consumption noticed by author of the thread. Also as GPE handling is bloked, machines overheat. Eventually by external poll of the same acpi_evaluate, kacpid is released and all the queued notify events are free to run, thus 100% cpu utilization by kacpid for several seconds or more.

To prevent all these horrors it's needed to not put notify events to kacpid workqueue by either executing them immediately or putting them on some other thread. It's dangerous to execute notify events in place, as it will put several ACPI interpreter stacks on top of each other (at least 4 in case of nx6125), thus causing kernel stack overflow.

First attempt to create a new thread was done by Peter Wainwright He created a bunch of threads, which were stealing work from a kacpid workqueue.
This patch appeared in 2.6.15 kernel shipped with Ubuntu 6.06 LTS.

Second attempt was done by me, I created a new thread for each Notify event. This worked OK on HP nx machines, but broke Linus' Compaq n620c, by producing threads with a speed what they stopped the machine completely. Thus this patch was reverted from 18-rc2 as I remember. I re-made the patch to create second workqueue just for notify events, thus hopping it will not break Linus' machine. Patch was tested on the same HP nx machines in #5534 and #7122, but I did not received reply from Linus on a test patch sent to him.
Patch went to 19-rc and was rejected with much fanfare again. There was 4th patch, which inserted schedule_timeout(1) into deferred execution of kacpid, if we had any notify requests pending, but Linus decided that it was too complex (involved either changes to workqueue to see if it's empty or atomic inc/dec).
Now you see last variant which adds yield() to every GPE execution.

http://bugzilla.kernel.org/show_bug.cgi?id=5534
http://bugzilla.kernel.org/show_bug.cgi?id=8385

commit 262a7a28de060f3a63cae20035876d6f22fd7670

Author: Len Brown <[email protected]>
Date:   Wed May 9 23:01:59 2007 -0400

Revert "ACPICA: fix AML mutex re-entrancy"

This reverts commit c0d127b56937c3e72c2b1819161d2f6718eee877.

These changes to AML locking were made to allow Notify handlers to be called on the stack
and not deadlock. However, that scheme turns out to be flawed and was reverted by the previous commit, so this commit restores the locking to it previous design.

commit 40d07080e585396dc58bc64befa1de0695318b3b

Author: Len Brown <[email protected]>
Date:   Wed May 9 22:59:38 2007 -0400

Revert "Execute AML Notify() requests on stack."

This reverts commit 5f7748cf91558a5026ded5be93c5bf6c1ac34edf.

While that change fixed the HP
http://bugzilla.kernel.org/show_bug.cgi?id=5534

it broke the ACER:
http://bugzilla.kernel.org/show_bug.cgi?id=8385
which as AML that caused Linux go recursive and stack fault.

So this commit by itself will restore the ACER and again break the HP, which we'll fix another way.

commit 4d2acd9ea539e0f59178b126f6750ccc41eefcdd

Author: Len Brown <[email protected]>
Date:   Wed May 9 22:56:38 2007 -0400

Revert "ACPICA: revert "acpi_serialize" changes"

This reverts commit a8f4af6dc6600980885c594f52eecd60edd62013. Thus restoring ACPICA's new acpi_serialize code.

This commit by itself may cause a regression, but it is reverted in this order so that subsequent reverts reverts under this one can be made without conflict.

commit eaca2d3f6c4de9d4274a4e2be54c9693e76b0303

Author: Len Brown <[email protected]>
Date:   Mon Apr 30 23:27:43 2007 -0400

ACPI: delete un-reliable concept of cooling mode

The scheme where the thermal driver displayed the cooling mode /proc/acpi/thermal_zone/*/cooling_mode was flawed in two ways.

First, the success of _SCP doesn't actually mean that the BIOS moved any trip points.
On many BIOS, _SCP is present, but does nothing. So displaying what _SCP executed actually was wrong more times than it was right.

Second, examining the relative position of the trip points when the thermal_zone is added is insufficient -- as the BIOS reserves the right to change the trip points at run-time.

The only reliable way for the user to determine if the thermal zone is in active, passive, or critical mode is to examine the relative position of the trip points. The user can do this without the kernel doing it for them by looking in /proc/acpi/thermal_zone/*/trip_points

New contents for /proc/acpi/thermal_zone/*/cooling_mode:

If _SCP available:
"0 - Active; 1 - Passive\n"

If _SCP unavailable:
"<setting not supported>\n"

commit 11ccc0f249cb01a129f54760b8ff087f242935d4

Author: Len Brown <[email protected]>
Date:   Mon Apr 30 22:36:01 2007 -0400

ACPI: thermal trip points are read-only

/proc/acpi/thermal_zone/*/trip_points displays what the kernel reads from the BIOS via ACPI.

If you echo a string of ':' deliminted numbers to this file then it will change what it displays.

But it shouldn't, since the kernel has no way to communicate these changes to ACPI thermal zones. ACPI thermal zone trip points are read-only.

The kernel does have the opportunity to ask the BIOS to change the trip points with _SCP - Set Cooling Policy.

Request Active Cooling Mode:
# echo 0 > /proc/acpi/thermal_zone/*/cooling_policy

Request Passive Cooling Mode:
# echo 1 > /proc/acpi/thermal_zone/*/cooling_policy

However, in practice it is quite rare for the BIOS to support the optional _SCP, and it is even more rare for the BIOS to export an _SCP that actually changes the trip points.

Seems to work for me. My evo correctly started the fan, and stopped it when the temperature went down again.

I forget if there were other issues on that machine that some other ACPI updates have triggered, but at least there isn't anything *obviously* broken on it.

Linus

Looking at things in "top", I do end up occasionally seeing spikes where kacpid takes 17% of CPU time, and kacpi_notify takes a few percent too. But the machine works ok, and it doesn't seem to be horrible:

   64 ?        S<     0:15 [kacpid]
   65 ?        S<     0:08 [kacpi_notify]

so they've gotten 23 seconds of CPU time over the 37 minutes that laptop has been up now. That's arguably too much, but on the other hand, I did end up trying to stress it out by doing some 3D stuff while compiling the kernel and doing "git grep" over the kernel tree etc.

Linus

Thanks, I noticed the same thing on an nx6325.
The goal at the moment is to revert to the simplest functional & stable solution -- as what is shipping today crashes on some boxes.

We've got a couple of tweaks in mind where we think Linux can get smarter -- but this is an area where several platform vendors are taking advantage of Windows' implementation in (different) twisted ways. For us to reach our goal of Linux handling any system out there in as optimal a way as possible, we need to study each one in detail.

That said, can you send me or point me to the acpidump output for your EVO. Yes, I'm sure you've sent it before a long time ago, but that was about probably 2,000,000 e-mail messages and a couple of disk crashes ago:-)

thanks,
-Len

Sure. If you send me a pointer to "acpidump" again, because I've long since updated that machine, and no longer have it.

Linus

[...]

[...]

I've just tried booting up 2.6.22-rc1-+a-day-or-two and my nx6125 has lost fan control again. It doesn't sound like that's an expected result of this patch series, but it seems unlikely that it's due to anything but the ACPI merge.

Shall I bisect, or do you have enough to go on, given the long history of grief that this laptop has caused you?