Bug 40222 - Laptop running hot and toshiba_acpi doesn't insert TOSHIBA L670-1JN
Laptop running hot and toshiba_acpi doesn't insert TOSHIBA L670-1JN
Status: RESOLVED PATCH_ALREADY_AVAILABLE
Product: Drivers
Classification: Unclassified
Component: Platform_x86
All Linux
: P1 normal
Assigned To: drivers_platform_x86@kernel-bugs.osdl.org
:
Depends on:
Blocks: 56331
  Show dependency treegraph
 
Reported: 2011-07-28 07:49 UTC by Mathieu Malaterre
Modified: 2013-04-09 06:23 UTC (History)
4 users (show)

See Also:
Kernel Version: 3.0.0
Tree: Mainline
Regression: No


Attachments
output of acpidump (329.43 KB, application/octet-stream)
2011-07-28 07:49 UTC, Mathieu Malaterre
Details
output of dmidecode (8.07 KB, application/octet-stream)
2011-07-28 07:51 UTC, Mathieu Malaterre
Details
debug.patch (1.17 KB, patch)
2011-08-07 13:55 UTC, Lan Tianyu
Details | Diff
proposed patch (25.27 KB, patch)
2011-08-30 11:35 UTC, Mathieu Malaterre
Details | Diff

Description Mathieu Malaterre 2011-07-28 07:49:19 UTC
Created attachment 66952 [details]
output of acpidump

I cannot insert the toshiba_acpi on my laptop:

$ sudo modprobe toshiba_acpi
FATAL: Error inserting toshiba_acpi (/lib/modules/3.0.0-1-amd64/kernel/drivers/platform/x86/toshiba_acpi.ko): No such device

as such computer is runnning with fan all time long. Sometimes, it suddenly reboots for no reason when doing compilation (make -j4).
Comment 1 Mathieu Malaterre 2011-07-28 07:51:23 UTC
Created attachment 66962 [details]
output of dmidecode
Comment 2 Lan Tianyu 2011-07-29 01:44:50 UTC
Please attach output of the dmesg after inserting toshiba_acpi.ko.
Comment 3 Mathieu Malaterre 2011-07-29 16:07:58 UTC
There is absolutely nothing added in the output of dmesg.


$ sudo insmod /lib/modules/3.0.0-1-amd64/kernel/drivers/platform/x86/toshiba_acpi.ko
insmod: error inserting '/lib/modules/3.0.0-1-amd64/kernel/drivers/platform/x86/toshiba_acpi.ko': -1 No such device
Comment 4 Lan Tianyu 2011-08-07 13:55:26 UTC
Created attachment 67862 [details]
debug.patch

Please apply the patch. load and attach the output of dmesg.
Comment 5 Mathieu Malaterre 2011-08-18 08:04:16 UTC
Here is the new output of dmesg:

[53933.532862] toshiba_acpi: no simple device 


$ sudo modinfo toshiba_acpi.ko
filename:       toshiba_acpi.ko
license:        GPL
description:    Toshiba Laptop ACPI Extras Driver
author:         John Belmonte
alias:          acpi*:TOS1900:*
alias:          acpi*:TOS6208:*
alias:          acpi*:TOS6200:*
depends:        rfkill,sparse-keymap
vermagic:       3.0.0-1-amd64 SMP mod_unload modversions
Comment 6 Lan Tianyu 2011-08-22 01:38:42 UTC
In the DSDT, there is not "GHCI" method under device "VALZ". So the toshiba_acpi driver can't be loaded.

The driver first check whether there is a valid path for hci method in the DSDT. 

static int __init toshiba_acpi_init(void)
{
        ...

	/* simple device detection: look for HCI method */
	if (is_valid_acpi_path(TOSH_INTERFACE_1 GHCI_METHOD)) {
		method_hci = TOSH_INTERFACE_1 GHCI_METHOD;
		if (toshiba_acpi_setup_keyboard(TOSH_INTERFACE_1))
			pr_info("Unable to activate hotkeys\n");
	} else if (is_valid_acpi_path(TOSH_INTERFACE_2 GHCI_METHOD)) {
		method_hci = TOSH_INTERFACE_2 GHCI_METHOD;
		if (toshiba_acpi_setup_keyboard(TOSH_INTERFACE_2))
			pr_info("Unable to activate hotkeys\n");
	} else
		return -ENODEV;

        ...
}

if not, return back.
Comment 7 Mathieu Malaterre 2011-08-22 06:58:51 UTC
I did follow instructions from:

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/735756

which pointed me to:

http://people.canonical.com/~sforshee/lp735756/

When loading this new toshiba_acpi module one gets in dmesg:


[150040.763112] toshiba_acpi: Toshiba Laptop ACPI Extras version 0.20
[150040.763141] toshiba_acpi: HCI type 2 (SPFC)
[150040.763972] input: Toshiba input device as /devices/virtual/input/input15
[150040.776320] toshiba_acpi: Illumination not supported


Let me know if you need anything else.
Comment 8 Mathieu Malaterre 2011-08-30 11:35:10 UTC
Created attachment 70882 [details]
proposed patch
Comment 9 Lan Tianyu 2011-09-01 07:24:21 UTC
http://marc.info/?l=linux-kernel&m=131465192113854&w=2

Seth has sent related patchs to lkml. Just waitting.
Comment 10 Zhang Rui 2012-01-18 05:29:53 UTC
Tianyu, please update the status of this bug.
Comment 11 Lan Tianyu 2012-01-18 05:45:29 UTC
Seth's patch has been merged into upstream kernel.

commit 6e02cc7eb61aeafadb91f7b591a768cdb7a57740
Author: Seth Forshee <seth.forshee@canonical.com>
Date:   Tue Sep 20 16:55:51 2011 -0500

    toshiba_acpi: Use handle for HCI calls
    
    Now that we're using an acpi driver we already have a handle to
    the namespace of the HCI call, so there's no need to test various
    paths to the HCI call or even be aware of the path at all.

So the bug should be resolved in the upstream kernel. Please check the bug in the lastest upstream kernel.
Comment 12 Mathieu Malaterre 2012-01-30 15:01:30 UTC
Looks good to me:


$ sudo modinfo toshiba_acpi
filename:       /lib/modules/3.2.0-1-amd64/kernel/drivers/platform/x86/toshiba_acpi.ko
license:        GPL
description:    Toshiba Laptop ACPI Extras Driver
author:         John Belmonte
alias:          acpi*:TOS1900:*
alias:          acpi*:TOS6208:*
alias:          acpi*:TOS6200:*
depends:        rfkill,sparse-keymap
intree:         Y
vermagic:       3.2.0-1-amd64 SMP mod_unload modversions 

Thanks for working on this
Comment 13 Len Brown 2012-06-05 04:08:06 UTC
6e02cc7eb61aeafadb91f7b591a768cdb7a57740
toshiba_acpi: Use handle for HCI calls

shipped in linux 3.2

closed

Note You need to log in before you can comment on or make changes to this bug.