Skip to contentSkip to conversation

2.5.70-mm3

DISCUSSION Ed Tomlinson 1 participant 1 message Jun 2, 2003 2 messages not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

#2 lore raw

Andrew,

This from -mm2 on UP (k6-III 400, 512M):

May 31 09:31:24 oscar kernel: Bad page state at free_hot_cold_page
May 31 09:31:24 oscar kernel: flags:0x01000208 mapping:00000000 mapped:0 count:0
May 31 09:31:24 oscar kernel: Backtrace:
May 31 09:31:24 oscar kernel: Call Trace:
May 31 09:31:24 oscar kernel:  [bad_page+70/108] bad_page+0x46/0x6c
May 31 09:31:24 oscar kernel:  [free_hot_cold_page+87/212] free_hot_cold_page+0x57/0xd4
May 31 09:31:24 oscar kernel:  [__pagevec_free+31/40] __pagevec_free+0x1f/0x28
May 31 09:31:24 oscar kernel:  [__pagevec_release_nonlru+117/132] __pagevec_release_nonlru+0x75/0x84
May 31 09:31:24 oscar kernel:  [shrink_list+1016/1212] shrink_list+0x3f8/0x4bc
May 31 09:31:24 oscar kernel:  [shrink_cache+383/684] shrink_cache+0x17f/0x2ac
May 31 09:31:24 oscar kernel:  [shrink_zone+98/108] shrink_zone+0x62/0x6c
May 31 09:31:24 oscar kernel:  [balance_pgdat+214/352] balance_pgdat+0xd6/0x160
May 31 09:31:24 oscar kernel:  [kswapd+261/268] kswapd+0x105/0x10c
May 31 09:31:24 oscar kernel:  [kswapd+0/268] kswapd+0x0/0x10c
May 31 09:31:24 oscar kernel:  [autoremove_wake_function+0/56] autoremove_wake_function+0x0/0x38
May 31 09:31:24 oscar kernel:  [autoremove_wake_function+0/56] autoremove_wake_function+0x0/0x38
May 31 09:31:24 oscar kernel:  [kernel_thread_helper+5/12] kernel_thread_helper+0x5/0xc
May 31 09:31:24 oscar kernel:
May 31 09:31:24 oscar kernel: Trying to fix it up, but a reboot is needed

So far I've not seen this with mm3, but it looks like Alistair has :-( In my case,
this triggered when doing "apt-get -u upgrade". It was a large upgrade as kde was updated.

Ed

PS. I have more examples if they will help.

Quoted 26 lines · Andrew Morton
Quoted 15 lines · Alistair J Strachan

Bad page state at free_hot_cold_page
flags:0x01010000 mapping:00000000 mapped:1 count:0

 Backtrace:
 Call Trace:
  [bad_page+93/144] bad_page+0x5d/0x90
  [free_hot_cold_page+112/256] free_hot_cold_page+0x70/0x100
  [zap_pte_range+385/448] zap_pte_range+0x181/0x1c0
  [do_wp_page+437/848] do_wp_page+0x1b5/0x350
  [zap_pmd_range+75/112] zap_pmd_range+0x4b/0x70
  [unmap_page_range+75/128] unmap_page_range+0x4b/0x80
  [unmap_vmas+254/544] unmap_vmas+0xfe/0x220
  [exit_mmap+109/384] exit_mmap+0x6d/0x180
  [mmput+65/176] mmput+0x41/0xb0
  [do_exit+243/832] do_exit+0xf3/0x340

eww, that's a PageDirect page. Never seen that before - it's bad.

Is the box SMP?
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/