Revision 43e58856585f8c61e6a4a0f1fd6996d78799a973 authored by Emmanuel Grumbach on 10 November 2011, 00:50:50 UTC, committed by John W. Linville on 11 November 2011, 16:03:24 UTC
When HW RF kill switch is set to kill the radio, our NIC issues an
interrupt after we stop the APM module. When we unload the module,
the driver disables and cleans the interrupts before stopping the
APM. So we have a real interrupt (inta not zero) pending.
When this interrupts pops up the tasklet has already been killed
and we crash.

Here is a logical description of the flow:

disable and clean interrupts
synchronize interrupts
kill the tasklet

stop the APM <<== creates an RF kill interrupt

free_irq <<== somehow our ISR is called here and we crash

Here is the panic message:

[  201.313636] BUG: unable to handle kernel paging request at ffff8800911b7150
[  201.314541] IP: [<ffffffff8106d652>] tasklet_action+0x62/0x130
[  201.315149] PGD 1c06063 PUD db37f067 PMD db408067 PTE 80000000911b7160
[  201.316456] Oops: 0000 [#1] SMP DEBUG_PAGEALLOC
[  201.317324] CPU 1
[  201.317495] Modules linked in: arc4 iwlwifi(-) mac80211 cfg80211 netconsole configfs binfmt_misc i915 drm_kms_helper drm uvcvideo i2c_algo_bit videodev dell_laptop dcdbas intel_agp dell_wmi intel_ips psmouse intel_gtt v4l2_compat_ioctl32 asix usbnet mii serio_raw video sparse_keymap firewire_ohci sdhci_pci sdhci firewire_core e1000e crc_itu_t [last unloaded: configfs]
[  201.323839]
[  201.324015] Pid: 2061, comm: modprobe Not tainted 3.1.0-rc9-wl #4 Dell Inc. Latitude E6410/0667CC
[  201.324736] RIP: 0010:[<ffffffff8106d652>]  [<ffffffff8106d652>] tasklet_action+0x62/0x130
[  201.325128] RSP: 0018:ffff88011bc43ea0  EFLAGS: 00010286
[  201.325338] RAX: ffff88008ae70000 RBX: ffff8800911b7150 RCX: ffff88008ae70028
[  201.325555] RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff88008ae70000
[  201.325775] RBP: ffff88011bc43ec0 R08: 0000000000000000 R09: 0000000000000000
[  201.325994] R10: 0000000000000002 R11: 0000000000000001 R12: 0000000000000001
[  201.326212] R13: 0000000000000006 R14: 0000000000000100 R15: ffff88008e259fd8
[  201.326431] FS:  00007f4b90ea9700(0000) GS:ffff88011bc40000(0000) knlGS:0000000000000000
[  201.326657] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[  201.326864] CR2: ffff8800911b7150 CR3: 000000008fd6d000 CR4: 00000000000006e0
[  201.327083] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  201.327302] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[  201.327521] Process modprobe (pid: 2061, threadinfo ffff88008e258000, task ffff88008ae70000)
[  201.327747] Stack:
[  201.330494]  0000000000000046 0000000000000030 0000000000000001 0000000000000006
[  201.333870]  ffff88011bc43f30 ffffffff8106cd8a ffffffff811e1016 ffff88011bc43f08
[  201.337186]  0000000100000046 ffff88008e259fd8 0000000a10be2160 0000000000000006
[  201.340458] Call Trace:
[  201.342994]  <IRQ>
[  201.345656]  [<ffffffff8106cd8a>] __do_softirq+0xca/0x250
[  201.348185]  [<ffffffff811e1016>] ? pde_put+0x76/0x90
[  201.350730]  [<ffffffff8131aeae>] ? do_raw_spin_unlock+0x5e/0xb0
[  201.353261]  [<ffffffff811e1016>] ? pde_put+0x76/0x90
[  201.355776]  [<ffffffff8163ccfc>] call_softirq+0x1c/0x30
[  201.358287]  [<ffffffff8101531d>] do_softirq+0x9d/0xd0
[  201.360823]  [<ffffffff8106cb05>] irq_exit+0xd5/0xf0
[  201.363330]  [<ffffffff8163d5d6>] do_IRQ+0x66/0xe0
[  201.365819]  [<ffffffff81632673>] common_interrupt+0x73/0x73
[  201.368257]  <EOI>

Cc: <stable@kernel.org> 3.1+
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
1 parent 0ecfe80
Raw File
efirtc.txt
EFI Real Time Clock driver
-------------------------------
S. Eranian <eranian@hpl.hp.com>
March 2000

I/ Introduction

This document describes the efirtc.c driver has provided for
the IA-64 platform. 

The purpose of this driver is to supply an API for kernel and user applications
to get access to the Time Service offered by EFI version 0.92.

EFI provides 4 calls one can make once the OS is booted: GetTime(),
SetTime(), GetWakeupTime(), SetWakeupTime() which are all supported by this
driver. We describe those calls as well the design of the driver in the
following sections.

II/ Design Decisions

The original ideas was to provide a very simple driver to get access to, 
at first, the time of day service. This is required in order to access, in a 
portable way, the CMOS clock. A program like /sbin/hwclock uses such a clock 
to initialize the system view of the time during boot.

Because we wanted to minimize the impact on existing user-level apps using
the CMOS clock, we decided to expose an API that was very similar to the one
used today with the legacy RTC driver (driver/char/rtc.c). However, because 
EFI provides a simpler services, not all ioctl() are available. Also
new ioctl()s have been introduced for things that EFI provides but not the 
legacy.

EFI uses a slightly different way of representing the time, noticeably
the reference date is different. Year is the using the full 4-digit format.
The Epoch is January 1st 1998. For backward compatibility reasons we don't
expose this new way of representing time. Instead we use something very 
similar to the struct tm, i.e. struct rtc_time, as used by hwclock.
One of the reasons for doing it this way is to allow for EFI to still evolve
without necessarily impacting any of the user applications. The decoupling
enables flexibility and permits writing wrapper code is ncase things change.

The driver exposes two interfaces, one via the device file and a set of
ioctl()s. The other is read-only via the /proc filesystem. 

As of today we don't offer a /proc/sys interface.

To allow for a uniform interface between the legacy RTC and EFI time service,
we have created the include/linux/rtc.h header file to contain only the 
"public" API of the two drivers.  The specifics of the legacy RTC are still 
in include/linux/mc146818rtc.h.

 
III/ Time of day service

The part of the driver gives access to the time of day service of EFI.
Two ioctl()s, compatible with the legacy RTC calls:

	Read the CMOS clock: ioctl(d, RTC_RD_TIME, &rtc);

	Write the CMOS clock: ioctl(d, RTC_SET_TIME, &rtc);

The rtc is a pointer to a data structure defined in rtc.h which is close
to a struct tm:

struct rtc_time {
        int tm_sec;
        int tm_min;
        int tm_hour;
        int tm_mday;
        int tm_mon;
        int tm_year;
        int tm_wday;
        int tm_yday;
        int tm_isdst;
};

The driver takes care of converting back an forth between the EFI time and
this format.

Those two ioctl()s can be exercised with the hwclock command:

For reading:
# /sbin/hwclock --show
Mon Mar  6 15:32:32 2000  -0.910248 seconds

For setting:
# /sbin/hwclock --systohc

Root privileges are required to be able to set the time of day.

IV/ Wakeup Alarm service

EFI provides an API by which one can program when a machine should wakeup,
i.e. reboot. This is very different from the alarm provided by the legacy
RTC which is some kind of interval timer alarm. For this reason we don't use
the same ioctl()s to get access to the service. Instead we have
introduced 2 news ioctl()s to the interface of an RTC. 

We have added 2 new ioctl()s that are specific to the EFI driver:

	Read the current state of the alarm
	ioctl(d, RTC_WKLAM_RD, &wkt)

	Set the alarm or change its status
	ioctl(d, RTC_WKALM_SET, &wkt)

The wkt structure encapsulates a struct rtc_time + 2 extra fields to get 
status information:
	
struct rtc_wkalrm {

        unsigned char enabled; /* =1 if alarm is enabled */
        unsigned char pending; /* =1 if alarm is pending  */

        struct rtc_time time;
} 

As of today, none of the existing user-level apps supports this feature.
However writing such a program should be hard by simply using those two 
ioctl(). 

Root privileges are required to be able to set the alarm.

V/ References.

Checkout the following Web site for more information on EFI:

http://developer.intel.com/technology/efi/
back to top