Battery Status Detected Incorrectly (MSI-1719 / GX700)

Bug #323823 reported by bkloppenborg
28
This bug affects 2 people
Affects Status Importance Assigned to Milestone
acpi (Ubuntu)
Fix Released
Medium
Unassigned

Bug Description

Although it was closed and a patch released, bug 147560(https://bugs.launchpad.net/ubuntu/+source/linux/+bug/147560) pertaining to the battery status being incorrectly detected still persists in Ubuntu 8.10 x86_64, kernel 2.6.27-11-generic and in Ubuntu 9.04 alpha3 except now the acpi=noirq kernel parameter appears to do nothing to rectify the situation.

While running without the acpi=noirq kernel parameter, the battery erroneously disappears.

The battery is detected during boot:
$ dmesg | grep battery
[ 12.305116] ACPI: Battery Slot [BAT1] (battery present)

and appears to work fine for a while:

$ cat /proc/acpi/battery/BAT1/info
present: yes
design capacity: 4400 mAh
last full capacity: 3255 mAh
battery technology: rechargeable
design voltage: 10800 mV
design capacity warning: 0 mAh
design capacity low: 0 mAh
capacity granularity 1: 1 mAh
capacity granularity 2: 1 mAh
model number: MS-1719
serial number:
battery type: LION
OEM info: MSI Corp.

$ cat /proc/acpi/battery/BAT1/state
present: yes
capacity state: ok
charging state: charged
present rate: 0 mA
remaining capacity: 3255 mAh
present voltage: 12506 mV

$ acpi -bs

     Battery 0: Full, 100%

But after ~10 minutes, it disappears

$ cat /proc/acpi/battery/BAT1/info
present: no

$ cat /proc/acpi/battery/BAT1/state
present: no

$ acpi -bs
[nothing]

I receive similar results WITH the acpi=noirq kernel parameters, except it will take ~1-2 hours before the battery disappears. Problem persists whether or not the AC power is plugged in.

When I boot into Ubuntu 8.10, kernel 2.6.27-9-generic, the acpi=noirq kernel parameter works and I never lose the battery. Please let me know what other information I can provide or if you need someone to test out patches, I would be willing to do so.

description: updated
Revision history for this message
bkloppenborg (bkloppenborg) wrote :

Because this got moved to the acpi section, I've included the standard information.

$ uname -a
Linux Lappy 2.6.27-11-generic #1 SMP Thu Jan 29 19:28:32 UTC 2009 x86_64 GNU/Linux

Attached is lspci output

Revision history for this message
bkloppenborg (bkloppenborg) wrote :

From $ sudo dmidecode > dmidecode.txt

Revision history for this message
bkloppenborg (bkloppenborg) wrote :

Tarball of the /proc/acpi directory with the acpi=noirq kernel parameter specified while the battery is detected (I'll post another copy of this when the battery is not detected).

cp -r /proc/acpi /tmp
sudo tar -cvjf ~/acpi.tar.bz /tmp/acpi

Revision history for this message
Will Law (williumbillium) wrote :

This issue also affects the MSI PR200, also sold by System 76 as the Darter Ultra 2.

Revision history for this message
bkloppenborg (bkloppenborg) wrote :
Revision history for this message
bkloppenborg (bkloppenborg) wrote :
Revision history for this message
bkloppenborg (bkloppenborg) wrote :

The battery disappeared after the laptop was running for ~10 hours on AC power WITH the acpi=noirq kernel parameter specified.

Revision history for this message
reh4c (gene-hoffler) wrote :

I can confirm that this annoying bug exists with 9.04 alpha 3.

Changed in acpi:
importance: Undecided → Medium
status: New → Triaged
Revision history for this message
Timo Tretter (timo-tretter) wrote :

these laptops want everything a little bit slower ;-)
the patch from http://bugzilla.kernel.org/show_bug.cgi?id=12011 works great.
just increase the ACPI_EC_UDELAY to 500 and added delays at various places in acpi_ec_transaction_unlocked()

Revision history for this message
bkloppenborg (bkloppenborg) wrote :

Given that Zygfryd (http://bugzilla.kernel.org/show_bug.cgi?id=12011#c110) isn't having luck with his GX700, this doesn't seem like an option at the moment. I'll continue checking in on that bug report from time to time.

Revision history for this message
Timo Tretter (timo-tretter) wrote :

so, we think the patch is ready for upstream. The problem still exists but occurs way less often. I don't think we figure out whats the "real" problem is.

Revision history for this message
bkloppenborg (bkloppenborg) wrote :

Timo, thank you for the update. Do you need another tester or do you anticipate that the patch will be released in the mainstream kernel anytime soon? Lastly, does this patch fix both issues I discussed above, namely can I run my laptop without the acpi=noirq kernel parameter or does this just fix the battery disappearing issue? Thanks for your hard work.

Revision history for this message
Timo Tretter (timo-tretter) wrote :

the patch is now shipped to the acpi testers. hopfully it will shipped in the next kernel verion.
with the patch you don't need the acpi=noirq parameter anymore, brightness keys works normally.

Revision history for this message
T-Lo (ausserirdischer) wrote :

I am also using the patched kernel from Timo and it's working great for me. Battery status never broke an I don't have to use acpi=noirq, so I can use my powerbuttion again!

Great work Timo, I hope the patch will make it into the new kernel.

Revision history for this message
bkloppenborg (bkloppenborg) wrote :

I just finished compiling the kernel with Timo's patch. One hour and 30 minutes of uptime (without any kernel parameters) and my battery is still detected. I'll post back if it disappears. Way to go!

Revision history for this message
Timo Tretter (timo-tretter) wrote :

the patch is shipped in 2.6.29-rc5-git7 (but this is maybe a mistake and means rc7-gid5)
is there a way to apply this patch to the older ubuntu kernel or at least to the jaunty kernel, whitch is still 2.6.28?

Revision history for this message
bkloppenborg (bkloppenborg) wrote :

I've been running the kernel with Timo's patch for four days now without any kernel parameters. The battery status has always been detected correctly.

The patch in question is found here:

http://bugzilla.kernel.org/show_bug.cgi?id=12011#c105

And is compiled by normal means (see discussion and links in that thread). If anyone else would like to try the kernel, I have built a .deb for the kernel and headers (2.6.27-11 amd64) and can place it in a public place. Send me an email if you are interested.

Provided that the patch stays in the kernel and is not overwritten (unlike the previous patch), I think this problem is solved.

Changed in acpi:
status: Triaged → Fix Released
Revision history for this message
Timo Tretter (timo-tretter) wrote :

i'm running a ppa with a patched kernel. it could be found at https://launchpad.net/~timo-tretter/+archive/ppa see the instructions on the page to add it to your apt sources.list

Revision history for this message
martenjanderuiter (martenjanderuiter) wrote :

Hi Timo,

I am attempting to install your kernel for Intrepid to see whether it fixes the battery status. I am able to find the kernel image for the installer kernel-image-2.6.27-11-generic-di, but not linux-image-2.6.27-11-generic.
Your page anounces [FAILEDTOBUILD] lpia, and I suspect that to be the cause for the partial availability of the packages mentioned on your page.
Is my suspicion correct, or am I overlooking something? (I am not very familiar with the debian system; I just know how to use synaptic to install new kernel images during update, keeping the current as a backup and deinstall the older ones.)

Is there a CLI command to install
https://launchpad.net/~timo-tretter/+archive/ppa/+files/linux-image-2.6.27-11-generic_2.6.27-11.27~ppa0_i386.deb
or are there more strings attached?
I guess there is something like "apt-get install https://URL", but I am afraid I'll overlook something that is obvious for the experienced user.

Revision history for this message
Timo Tretter (timo-tretter) wrote :

take a look at the first comment from here: http://ubuntuforums.org/showthread.php?t=1055188&page=3

if you dont want do add it to your sources list you can download the file via "wget https://launchpad.net/~timo-tretter/+archive/ppa/+files/linux-image-2.6.27-11-generic_2.6.27-11.27~ppa0_i386.deb " and install it with sudo dpkg -i linux-image-*.deb

Revision history for this message
martenjanderuiter (martenjanderuiter) wrote :

Thanks.
The 'sudo dpkg' command was what I needed. I inferred that the package can be installed independently.

It seems the kernel works like a charm. The upgrade manager complains that an updated kernel by the Ubuntu Kernel Team is available, which I happily ignore. Furthermore, I made the kernel the default in grub.

Based on the behaviour of the laptop, the patch is not merged into the 2.6.27-1[2-4], but maybe later (Jaunty? 2.6.29 that was released today?)

Revision history for this message
Bill Gjestvang (bgjestvang) wrote :

It's not looking like 2.6.29 will be in jaunty. Can we get this patch merged into the jaunty kernel?

Revision history for this message
bkloppenborg (bkloppenborg) wrote :

I just upgraded to 9.04 over the weekend. The persons who solved this problem in 8.10 (kernel 2.6.27-11) are again working towards a solution for 9.04 (kernel 2.6.28-11). I suggest that all further discussion of this bug be directed to Bug 12011 instead (http://bugzilla.kernel.org/show_bug.cgi?id=12011).

Revision history for this message
bkloppenborg (bkloppenborg) wrote :

Please note that the current work to fix the bug is hosted on bugzilla.kernel.org not on launchpad. Follow the explicit http link above (http://bugzilla.kernel.org/show_bug.cgi?id=12011), not the "Bug 12011" link.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.