oops and memory leak with uml-patch-2.5.67-1

From: Vladimir B. Savkin (master@sectorb.msk.ru)
Date: Sun May 11 2003 - 10:52:27 EST


Hi!

I tried to play with development kernels (v 2.5.67 and 2.5.69) with UML
but found that they are unusable to me.

Here's what I get with 2.5.67 + uml-patch-2.5.67-1 just after boot

uml1:~# uname -a
Linux uml1.sectorb.msk.ru 2.5.67-1um #1 SMP Sat May 10 15:31:43 MSD 2003
i686 un
known
uml1:~# cat /proc/meminfo
MemTotal: 28360 kB
MemFree: 18576 kB
Buffers: 708 kB
Cached: 3712 kB
SwapCached: 0 kB
Active: 3864 kB
Inactive: 1508 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 28360 kB
LowFree: 18576 kB
SwapTotal: 0 kB
SwapFree: 0 kB
Dirty: 204 kB
Writeback: 0 kB
Mapped: 2184 kB
Slab: 1432 kB
Committed_AS: 1120 kB
PageTables: 124 kB
ReverseMaps: 1450
uml1:~# cat /proc/slabinfo | grep task
task_struct 144 144 1724 36 36 2 : 54 27

Now starting test load:

uml1:~# while /bin/true ; do /bin/true ; done

After a minute I get OOM and oops:

Out of Memory: Killed process 129 (atd).
buffer layer error at fs/buffer.c:127
Pass this trace through ksymoops for reporting
Call Trace: [<a0078bbe>] [<a0078ce3>] [<a000bfa7>] [<a002fd2c>]
[<a002fd2c>]
   [<a007aacb>] [<a00fc16c>] [<a007b3e1>] [<a00bfe1c>] [<a005a987>]
[<a00c01d5>]
 
   [<a00bfe1c>] [<a005d775>] [<a0010e40>] [<a0010e34>] [<a0010e40>]
[<a020a27f>]
 
   [<a020a27f>] [<a0016b3f>] [<a0016af1>] [<a005dfd7>] [<a0016ad6>]
[<a000c452>]
 
   [<a000c427>] [<a000bfa7>] [<a002b260>] [<a0019324>] [<a001a18a>]
[<a0016b28>]
 
   [<a005e1af>] [<a005e08c>] [<a0077440>] [<a0077647>] [<a00160f3>]
[<a0016032>]
 
   [<a0077784>] [<a0010dc9>] [<a00777f3>] [<a0018fc5>] [<a0010e40>]
[<a0010e34>]
 
   [<a0010e40>] [<a0010e34>] [<a001504a>] [<a001907a>] [<a0019071>]
[<a00169e9>]
 
   [<a0010e8c>] [<a001a0f9>] [<a001a0e9>] [<a0018379>] [<a0018358>]
[<a0016a0b>]
 
   [<a020a188>]
Out of Memory: Killed process 137 (bash).
bash: page allocation failure. order:2, mode:0xd0

Decoded oops and kernel config attached.
After that it can be found that all memory is gone:

uml1:~# cat /proc/meminfo
MemTotal: 28360 kB
MemFree: 640 kB
Buffers: 52 kB
Cached: 468 kB
SwapCached: 0 kB
Active: 1112 kB
Inactive: 372 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 28360 kB
LowFree: 640 kB
SwapTotal: 0 kB
SwapFree: 0 kB
Dirty: 0 kB
Writeback: 0 kB
Mapped: 1344 kB
Slab: 3672 kB
Committed_AS: 1032 kB
PageTables: 112 kB
ReverseMaps: 359
uml1:~# cat /proc/slabinfo | grep task
task_struct 1368 1368 1724 342 342 2 : 54 27

I have tried to merge uml-patch with kernel 2.5.69, it runs and gives no
oops, but memory leak is still there.

I'm willing to run any tests and to help debugging someway but I don't
know how to start.

:wq
                                        With best regards,
                                           Vladimir Savkin.


$ ksymoops -v ./linux -K -L -O -M < ../OOPS1
ksymoops 2.4.5 on i686 2.4.21-pre2. Options used
     -v ./linux (specified)
     -K (specified)
     -L (specified)
     -O (specified)
     -M (specified)

Call Trace: [<a0078bbe>] [<a0078ce3>] [<a0079945>] [<a000bfa7>] [<a002fd2c>]
   [<a002fd2c>] [<a007aacb>] [<a00fc16c>] [<a007b3e1>] [<a00bfe1c>] [<a005a987>]
   [<a00c01d5>] [<a00bfe1c>] [<a005d775>] [<a013a977>] [<a013aa88>] [<a020a27f>]
   [<a0016b3f>] [<a002b1e9>] [<a0016af1>] [<a005dfd7>] [<a0016ad6>] [<a000c427>]
   [<a002b260>] [<a0019324>] [<a000bfa7>] [<a001a18a>] [<a0016b28>] [<a005e1af>]
   [<a005e08c>] [<a0077440>] [<a0077647>] [<a00160f3>] [<a0016032>] [<a0077784>]
   [<a0010dc9>] [<a00777f3>] [<a0018fc5>] [<a0010e40>] [<a0010e34>] [<a001504a>]
   [<a001907a>] [<a0019071>] [<a00169e9>] [<a0010e8c>] [<a001a0f9>] [<a001a0e9>]
   [<a0018379>] [<a0018358>] [<a0016a0b>] [<a020a188>]
Warning (Oops_read): Code line not seen, dumping what data is available

Trace; a0078bbe <__buffer_error+3e/40>
Trace; a0078ce3 <__wait_on_buffer+6f/b0>
Trace; a0079945 <create_buffers+21/9c>
Trace; a000bfa7 <region_va+13/24>
Trace; a002fd2c <autoremove_wake_function+0/40>
Trace; a002fd2c <autoremove_wake_function+0/40>
Trace; a007aacb <__block_prepare_write+31b/440>
Trace; a00fc16c <radix_tree_node_alloc+1c/6c>
Trace; a007b3e1 <block_prepare_write+21/38>
Trace; a00bfe1c <ext2_get_block+0/350>
Trace; a005a987 <add_to_page_cache+4f/104>
Trace; a00c01d5 <ext2_prepare_write+19/1c>
Trace; a00bfe1c <ext2_get_block+0/350>
Trace; a005d775 <generic_file_aio_write_nolock+9dd/11cc>
Trace; a013a977 <kfree_skbmem+6f/a0>
Trace; a013aa88 <__kfree_skb+e0/e8>
Trace; a020a27f <sigemptyset+2f/3c>
Trace; a0016b3f <__do_copy+17/1c>
Trace; a002b1e9 <os_map_memory+39/50>
Trace; a0016af1 <__do_user_copy+41/78>
Trace; a005dfd7 <generic_file_write_nolock+73/90>
Trace; a0016ad6 <__do_user_copy+26/78>
Trace; a000c427 <map_memory+1b/50>
Trace; a002b260 <os_unmap_memory+14/2f>
Trace; a0019324 <fix_range+164/1f8>
Trace; a000bfa7 <region_va+13/24>
Trace; a001a18a <__do_copy_from_user+2a/44>
Trace; a0016b28 <__do_copy+0/1c>
Trace; a005e1af <generic_file_writev+33/48>
Trace; a005e08c <generic_file_write+0/5c>
Trace; a0077440 <do_readv_writev+bc/36c>
Trace; a0077647 <do_readv_writev+2c3/36c>
Trace; a00160f3 <flush_tlb_range+23/34>
Trace; a0016032 <flush_tlb_page+22/24>
Trace; a0077784 <vfs_writev+4c/50>
Trace; a0010dc9 <change_sig+45/64>
Trace; a00777f3 <sys_writev+2b/40>
Trace; a0018fc5 <execute_syscall_tt+121/188>
Trace; a0010e40 <change_signals+58/80>
Trace; a0010e34 <change_signals+4c/80>
Trace; a001504a <execute_syscall+26/28>
Trace; a001907a <syscall_handler_tt+4a/98>
Trace; a0019071 <syscall_handler_tt+41/98>
Trace; a00169e9 <usr2_handler+1d/20>
Trace; a0010e8c <unblock_signals+10/14>
Trace; a001a0f9 <Letext+9d/f6>
Trace; a001a0e9 <Letext+8d/f6>
Trace; a0018379 <finish_fork_handler+12d/130>
Trace; a0018358 <finish_fork_handler+10c/130>
Trace; a0016a0b <sig_handler+1f/34>
Trace; a020a188 <__restore+0/8>

1 warning issued. Results may not be reliable.



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Thu May 15 2003 - 22:00:37 EST