xfs
[Top] [All Lists]

A problem in xfs mount

To: linux-xfs@xxxxxxxxxxx
Subject: A problem in xfs mount
From: tom <twang@xxxxxxxxxxxxxxxxxxxx>
Date: Thu, 23 May 2002 13:54:43 +0800
Sender: owner-linux-xfs@xxxxxxxxxxx
Hi:
I wonder if someone can help me out of the problem.
when I tested mount function of xfs in lvm using a shell,which keeps on
mounting lvm partition, I found that mount process hang up, after a
while,
at that time cpu usages over 80%.  The test program as fllows,
(it mounts 255 lvm partitions one by one)

*******************************************************
#!/bin/sh

vg=vg2
lvm_base=test
mnt_base=mnt
share_size=8

lvremove -f /dev/$vg/${lvm_base}* >/dev/null
echo "Done"
for ((j=0; j < 255; j++));do
        mkdir -p /shares/${mnt_base}$j
done
for ((j=0; j< 255; j++));do
      lvcreate -L $share_size -n ${lvm_base}$j /dev/$vg >/dev/null
done
for ((j=0; j< 255; j++));do
      mkfs.xfs  -f /dev/$vg/${lvm_base}$j >/dev/null
      mount -t xfs /dev/$vg/${lvm_base}$j /shares/${mnt_base}$j >
/dev/null
done

**************************************************************

the trace  
********************************************************************
Stack traceback for pid 1448
    EBP       EIP         Function(args)
           0xc012df5a page_launder+0x24e (0x23, 0x1, 0x23, 0x0)
                               kernel .text 0xc0100000 0xc012dd0c
0xc012e614
           0xc012f63d __alloc_pages+0x13d
                               kernel .text 0xc0100000 0xc012f500
0xc012f730
           0xc012f748 __get_free_pages+0x18 (0x1, 0xc1117d10, 0x202,
0xc1117d10, 0xc012c262)
                               kernel .text 0xc0100000 0xc012f730
0xc012f754
           0xc012be89 kmem_cache_grow+0xd1
                               kernel .text 0xc0100000 0xc012bdb8
0xc012c000
           0xc012c2b8 kmalloc+0xc0 (0x1000, 0x8203, 0xc1a288f0, 0x10,
0x6)
                               kernel .text 0xc0100000 0xc012c1f8
0xc012c2e0
           0xc4ae7479 [xfs_support]kmem_alloc_Rsmp_dc01ff11+0x89
(0x1000, 0x8203, 0xc1a288f0, 0x0, 0xc26c9490)
                               xfs_support .text 0xc4ae7060 0xc4ae73f0
0xc4ae756c
           0xc4ae757c [xfs_support]kmem_zalloc_Rsmp_16428841+0x10
(0x8203, 0x10,0x80, 0xc18f8044, 0x8203)
                               xfs_support .text 0xc4ae7060 0xc4ae756c
0xc4ae75a4

           0xc4b454ab [xfs]xlog_alloc_log+0x20b (0xc18f8000, 0x3a17,
0x20, 0x0,
0x2580)
                               xfs .text 0xc4aed060 0xc4b452a0
0xc4b4560c
           0xc4b4491e [xfs]xfs_log_mount+0x86 (0xc18f8000, 0x3a17, 0x20,
0x0, 0x2580)
                               xfs .text 0xc4aed060 0xc4b44898
0xc4b44998
           0xc4b50d73 [xfs]xfs_mountfs+0xa8f (0xc1a289b0, 0x0,
0xc1a289b0, 0xc4b4fed3, 0xc1a289b0)
****************************************************************
Seems that the system has no memory, but I check  the system found that
             total       used       free     shared    buffers    
cached
Mem:         61672      57616       4056          0        972     
25400
-/+ buffers/cache:      31244      30428
Swap:       160608      34644     125964

The kernel I used is xfs-release-1.0.1 (kernel 2.4.5). I have tested it
in xfs-1.1,
it sees no problem. However, for some reson, I can not upgrade to the
xfs-1.1.
I wonder where I can find the patch to resolve the problem.
Any advice will be appreciated.


<Prev in Thread] Current Thread [Next in Thread>
  • A problem in xfs mount, tom <=