Skip to contentSkip to conversation

BUG: sleeping function called from invalid context at kernel/sched.c:5558

DISCUSSION Thomas Meyer 3 participants 3 messages Nov 3, 2008 – Nov 4, 2008 ACPI

lore
Share & tools
lore
#0 lore raw

False positive, while suspend/resume-to/from-disk?

[19332.352863] BUG: sleeping function called from invalid context at kernel/sched.c:5558
[19332.352866] in_atomic(): 0, irqs_disabled(): 1, pid: 3287, name: pm-hibernate
[19332.352869] Pid: 3287, comm: pm-hibernate Tainted: G        W  2.6.28-rc2 #97
[19332.352871] Call Trace:
[19332.352878]  [<c0436408>] __might_sleep+0xdc/0xe3
[19332.352882]  [<c043bbde>] __cond_resched+0x12/0x36
[19332.352886]  [<c09d29a8>] _cond_resched+0x27/0x32
[19332.352889]  [<c0697d9c>] acpi_ps_complete_op+0x1e0/0x1ef
[19332.352892]  [<c069884e>] ? acpi_ps_alloc_op+0x82/0x84
[19332.352896]  [<c0689814>] ? acpi_ds_exec_end_op+0x30c/0x370
[19332.352899]  [<c069839f>] acpi_ps_parse_loop+0x5f4/0x721
[19332.352902]  [<c069779c>] acpi_ps_parse_aml+0x6d/0x267
[19332.352905]  [<c06989e1>] acpi_ps_execute_method+0x11c/0x1d7
[19332.352909]  [<c069563c>] acpi_ns_evaluate+0x11c/0x1e4
[19332.352913]  [<c069d223>] ? acpi_ut_create_internal_object_dbg+0x15/0x6d
[19332.352916]  [<c0699c0b>] acpi_rs_set_srs_method_data+0xb7/0xd2
[19332.352919]  [<c0699768>] acpi_set_current_resources+0x2e/0x39
[19332.352923]  [<c06a5cfb>] acpi_pci_link_set+0xff/0x1a4
[19332.352926]  [<c06a5de9>] irqrouter_resume+0x49/0x60
[19332.352930]  [<c06fb1ac>] __sysdev_resume+0x14/0x38
[19332.352933]  [<c06fb1f1>] sysdev_resume+0x21/0x54
[19332.352936]  [<c06ff712>] device_power_up+0xb/0x15
[19332.352940]  [<c0463819>] hibernation_snapshot+0x149/0x1ba
[19332.352943]  [<c0463936>] hibernate+0xac/0x17d
[19332.352945]  [<c04628c9>] ? state_store+0x0/0xa8
[19332.352948]  [<c046291f>] state_store+0x56/0xa8
[19332.352951]  [<c04628c9>] ? state_store+0x0/0xa8
[19332.352954]  [<c0645565>] kobj_attr_store+0x1a/0x22
[19332.352958]  [<c04e08cb>] sysfs_write_file+0xb4/0xdf
[19332.352961]  [<c04e0817>] ? sysfs_write_file+0x0/0xdf
[19332.352964]  [<c04a557d>] vfs_write+0x84/0xdf
[19332.352967]  [<c04a5671>] sys_write+0x3b/0x60
[19332.352970]  [<c0417b15>] sysenter_do_call+0x12/0x21
[19332.352986] Enabling non-boot CPUs ...
#1 lore raw
Quoted 14 lines · Thomas Meyer

False positive, while suspend/resume-to/from-disk?

[19332.352863] BUG: sleeping function called from invalid context at kernel/sched.c:5558
[19332.352866] in_atomic(): 0, irqs_disabled(): 1, pid: 3287, name: pm-hibernate
[19332.352869] Pid: 3287, comm: pm-hibernate Tainted: G        W  2.6.28-rc2 #97
[19332.352871] Call Trace:
[19332.352878]  [<c0436408>] __might_sleep+0xdc/0xe3
[19332.352882]  [<c043bbde>] __cond_resched+0x12/0x36
[19332.352886]  [<c09d29a8>] _cond_resched+0x27/0x32
[19332.352889]  [<c0697d9c>] acpi_ps_complete_op+0x1e0/0x1ef
[19332.352892]  [<c069884e>] ? acpi_ps_alloc_op+0x82/0x84
[19332.352896]  [<c0689814>] ? acpi_ds_exec_end_op+0x30c/0x370
[19332.352899]  [<c069839f>] acpi_ps_parse_loop+0x5f4/0x721
[19332.352902]  [<c069779c>] acpi_ps_parse_aml+0x6d/0x267

Such messages are never false positives, this is a real bug.

Ingo

#2 lore raw
Quoted 18 lines · Ingo Molnar
Quoted 14 lines · Thomas Meyer

False positive, while suspend/resume-to/from-disk?

[19332.352863] BUG: sleeping function called from invalid context at kernel/sched.c:5558
[19332.352866] in_atomic(): 0, irqs_disabled(): 1, pid: 3287, name: pm-hibernate
[19332.352869] Pid: 3287, comm: pm-hibernate Tainted: G        W  2.6.28-rc2 #97
[19332.352871] Call Trace:
[19332.352878]  [<c0436408>] __might_sleep+0xdc/0xe3
[19332.352882]  [<c043bbde>] __cond_resched+0x12/0x36
[19332.352886]  [<c09d29a8>] _cond_resched+0x27/0x32
[19332.352889]  [<c0697d9c>] acpi_ps_complete_op+0x1e0/0x1ef
[19332.352892]  [<c069884e>] ? acpi_ps_alloc_op+0x82/0x84
[19332.352896]  [<c0689814>] ? acpi_ds_exec_end_op+0x30c/0x370
[19332.352899]  [<c069839f>] acpi_ps_parse_loop+0x5f4/0x721
[19332.352902]  [<c069779c>] acpi_ps_parse_aml+0x6d/0x267
Quoted 1 line

Such messages are never false positives, this is a real bug.

Well, not sure.

This is the ACPI AML interpreter executing something with interrupts disabled due to resume in progress (effectively atomic context).

Thanks,
Rafael