Bug 21252
Summary: | unable to handle paging request while using two KDE 4 sessions | ||
---|---|---|---|
Product: | Memory Management | Reporter: | Martin Steigerwald (Martin) |
Component: | Page Allocator | Assignee: | Andrew Morton (akpm) |
Status: | RESOLVED UNREPRODUCIBLE | ||
Severity: | normal | ||
Priority: | P1 | ||
Hardware: | All | ||
OS: | Linux | ||
Kernel Version: | 2.6.36 | Subsystem: | |
Regression: | No | Bisected commit-id: | |
Attachments: |
syslog
ps -aux output prior to rebooting the machine some other stuff from proc that might be useful |
Description
Martin Steigerwald
2010-10-27 10:24:48 UTC
Created attachment 35212 [details]
ps -aux output prior to rebooting the machine
BTW I have no idea on the reproducability on the bug since it happened out of a sudden and only once so far.
The memory situation appeared to be fine short after the bug happened: martin@shambhala:~/Linux/Kernel/Mainline/Bugs/unable-to-handle-paging-request> cat free-m.txt total used free shared buffers cached Mem: 2014 1047 966 0 121 443 -/+ buffers/cache: 482 1531 Swap: 3906 41 3864 I am using Radeon DRM KMS driver with OpenGL compositing on both desktops. In the time between the last reboot before the bug happened and the happening of the bug I once had some graphics corruption on my first KDE session. I also already had this quite some times with 2.6.35 but only when I had two KDE sessions, never when only one was running. I just restarted the desktop to get rid of the corruption. I also wondered whether the system was short on lowmem pages, but it doesn't appear so. martin@shambhala:~/Linux/Kernel/Mainline/Bugs/unable-to-handle-paging-request> cat proc/zoneinfo Node 0, zone DMA pages free 953 min 16 low 20 high 24 scanned 0 spanned 4095 present 3966 nr_free_pages 953 nr_inactive_anon 35 nr_active_anon 0 nr_inactive_file 1891 nr_active_file 866 nr_unevictable 0 nr_mlock 0 nr_anon_pages 31 nr_mapped 42 nr_file_pages 2762 nr_dirty 0 nr_writeback 0 nr_slab_reclaimable 171 nr_slab_unreclaimable 13 nr_page_table_pages 0 nr_kernel_stack 0 nr_unstable 0 nr_bounce 0 nr_vmscan_write 20 nr_writeback_temp 0 nr_isolated_anon 0 nr_isolated_file 0 nr_shmem 0 protection: (0, 865, 2015, 2015) pagesets cpu: 0 count: 0 high: 0 batch: 1 all_unreclaimable: 0 start_pfn: 1 inactive_ratio: 1 Node 0, zone Normal pages free 80137 min 932 low 1165 high 1398 scanned 0 spanned 223230 present 221486 nr_free_pages 80137 nr_inactive_anon 82 nr_active_anon 1059 nr_inactive_file 22327 nr_active_file 20378 nr_unevictable 0 nr_mlock 0 nr_anon_pages 1079 nr_mapped 113 nr_file_pages 42770 nr_dirty 13 nr_writeback 0 nr_slab_reclaimable 26184 nr_slab_unreclaimable 5173 nr_page_table_pages 740 nr_kernel_stack 177 nr_unstable 0 nr_bounce 0 nr_vmscan_write 59358 nr_writeback_temp 0 nr_isolated_anon 0 nr_isolated_file 0 nr_shmem 3 protection: (0, 0, 9202, 9202) pagesets cpu: 0 count: 109 high: 186 batch: 31 all_unreclaimable: 0 start_pfn: 4096 inactive_ratio: 1 Node 0, zone HighMem pages free 166527 min 128 low 437 high 747 scanned 0 spanned 296786 present 294467 nr_free_pages 166527 nr_inactive_anon 12693 nr_active_anon 17015 nr_inactive_file 55286 nr_active_file 43250 nr_unevictable 0 nr_mlock 0 nr_anon_pages 26796 nr_mapped 7685 nr_file_pages 102673 nr_dirty 3 nr_writeback 0 nr_slab_reclaimable 0 nr_slab_unreclaimable 0 nr_page_table_pages 0 nr_kernel_stack 0 nr_unstable 0 nr_bounce 0 nr_vmscan_write 298518 nr_writeback_temp 0 nr_isolated_anon 0 nr_isolated_file 32 nr_shmem 439 protection: (0, 0, 0, 0) pagesets cpu: 0 count: 34 high: 186 batch: 31 all_unreclaimable: 0 start_pfn: 227326 inactive_ratio: 3 Created attachment 35222 [details]
some other stuff from proc that might be useful
martin@shambhala:~/Linux/Kernel/Mainline/Bugs/unable-to-handle-paging-request> ls -lh proc
insgesamt 10M
-r--r--r-- 1 martin martin 300 27. Okt 10:01 buddyinfo
-r--r--r-- 1 martin martin 2,0M 27. Okt 10:02 kallsyms
-r-------- 1 martin martin 4,0M 27. Okt 10:00 kpagecount
-r-------- 1 martin martin 4,0M 27. Okt 10:00 kpageflags
-r--r--r-- 1 martin martin 1,1K 27. Okt 10:00 meminfo
-r--r--r-- 1 martin martin 2,3K 27. Okt 10:00 pagetypeinfo
-r--r--r-- 1 martin martin 3,5K 27. Okt 10:02 sched_debug
-r--r--r-- 1 martin martin 12K 27. Okt 10:02 slabinfo
-r--r--r-- 1 martin martin 104 27. Okt 10:02 swaps
-r--r--r-- 1 martin martin 148 27. Okt 10:06 version
-r-------- 1 martin martin 12K 27. Okt 10:00 vmallocinfo
-r--r--r-- 1 martin martin 1,5K 27. Okt 10:00 vmstat
-r--r--r-- 1 martin martin 2,7K 27. Okt 10:03 zoneinfo
gee, hard. Could be a transient hardware failure, could be a sporadic memory scribble from some errant code (DRI would be a good suspect). All I can suggest is that you enable as many debug options as possible (especially memory-related ones) then see if it happens again, and if some pattern can be determined. Thanks. I didn't see this again so far, so closing as unreproducible. If it happens again I consider adding debug options and wait till it happens a third time. Which ones would you recommend? |