Distribution: Slackware 9.1 Hardware Environment: Centrino 1.30GHz - 512MB DDR Software Environment: KDE 3.2.1 Problem Description:After boot things are fine for a while, then without any visible diagnostic output, the machine gets very slow. Running top shows that events/0 is consuming the CPU. (94%) and no way to stop process. Steps to reproduce: leave laptop to idle for 30 mins or more.
Please run a kernel profile when this is happening. - boot with "profile=1" on the kernel command line - Do: readprofile -r readprofile -M10 sleep 10 readprofile -n -v -m /boot/System.map | sort -n +2 | tail -40
ok and can you explain where exactly i put "profile=1" - i've never done profile. Thanks!
the "profile=1" goes into your kernel boot commandline parameters If you use grub, in /boot/grub/grub.conf: title local kernel (2.6.5-rc2-mm5) root (hd0,0) kernel /bzImage-2.6.5-rc2-mm5 ro root=/dev/sda2 profile=1 vga=extended initrd /initrd-2.4.22-1.2115.nptlsmp.img If you use LILO, in /etc/lilo.conf: image=/boot/bzImage-2.6.5-rc2-mm3 label=linux2.6.5 append="profile=1 gdbbaud=115200 gdbttyS=0 nmi_watchdog=2 vga=773" vga=extended read-only root=/dev/hda2 (remember to rerun lilo after editing lilo.conf)
Thanks! will try that --- radius
Well here it is .... root@averatec:~# readprofile -n -v -m /boot/System.map-2.6.5-rc3 | sort -n +2 | tail -40 c02658bb acpi_ns_search_and_enter 79 0.2937 c025e0a9 acpi_ev_asynch_execute_gpe_method 81 0.4241 c026e16e acpi_ut_create_generic_state 82 4.1000 c0268b78 acpi_ps_get_arg 83 1.2206 c0268e9f acpi_ps_delete_parse_tree 86 0.5478 c0268977 acpi_ps_parse_aml 103 0.2008 c026dc10 acpi_ut_valid_acpi_name 103 2.0600 c02675b5 acpi_ns_walk_namespace 116 0.4479 c025f3b8 acpi_ex_name_segment 119 0.8207 c02652a4 acpi_ns_lookup 119 0.1526 c0269120 acpi_ps_alloc_op 128 0.9624 c013f560 kfree 133 1.4149 c025c1f3 acpi_ds_create_walk_state 136 1.2035 c0264845 acpi_hw_low_level_read 155 0.9873 c025df91 acpi_ev_gpe_detect 165 0.5893 c012d768 __queue_work 169 1.3203 c0268d64 acpi_ps_get_next_walk_op 185 0.6293 c026580c acpi_ns_search_node 186 2.4474 c012d8e8 worker_thread 220 0.3537 c0268f3c acpi_ps_get_opcode_info 235 3.8525 c02680f3 acpi_ps_parse_loop 268 0.1229 c025988c acpi_ds_method_data_init 326 2.9636 c0266cec acpi_get_handle 369 1.7324 c0258f2e acpi_os_signal_semaphore 404 8.7826 c0121e70 do_softirq 648 4.1274 c0258f98 acpi_os_get_thread_id 1030 21.0204 c0258e44 acpi_os_wait_semaphore 1255 5.3632 c011b7b1 __might_sleep 1271 6.2611 c026e072 acpi_ut_release_mutex 1359 11.7155 c026dfff acpi_ut_acquire_mutex 1437 12.4957 c0108c28 handle_IRQ_event 2260 23.0612 c026cf38 acpi_ut_release_to_cache 2406 16.1477 c02589b6 acpi_os_write_port 2496 46.2222 c026cfcd acpi_ut_acquire_from_cache 2637 14.9830 c0266ca7 acpi_ns_get_parent_node 3818 90.9048 c0266cd1 acpi_ns_get_next_valid_node 4473 165.6667 c025896d acpi_os_read_port 5134 70.3288 c0266152 acpi_ns_delete_namespace_by_owner 7247 42.6294 c0267564 acpi_ns_get_next_node 8846 109.2099 00000000 total 53201 0.0128
Len, can you take a look at this one? Looks like ACPI has gone nuts doing schedule_work() continuously.
I have the same problem on my laptop (BLISS, Intel Pentium 2400MHz). It seems to me that problem occurs when it overheats. Putting a real cooler (one of those which are used to cool large rooms in the summer) at the laptop solved the problem. But: if the problem has occured already, you have to cool your laptop, then turn it off an on (reboot doesn't help).
Is this issue still present in recent 2.6 kernels?
Please reopen this bug if it's still present in kernel 2.6.16.