ssh connection closed when detecting hardware

I am not sure this is caused by hardware detection.
My connection is closed by the OOM killer (Out of memory killer)
I got this error today when running the installer, the following is a selection from syslog:

Jan 1 00:23:37 kernel: tar invoked oom-killer: gfp_mask=0x200d2, order=0, oomkilladj=0
Jan 1 00:23:37 kernel: Function entered at [<c02021b0>] from [<c0064360>]
Jan 1 00:23:37 kernel: Function entered at [<c0064308>] from [<c00649f0>]
Jan 1 00:23:37 kernel: Function entered at [<c0064850>] from [<c00674ec>]
Jan 1 00:23:37 kernel: Function entered at [<c00671d8>] from [<c0061ecc>]
Jan 1 00:23:37 kernel: Function entered at [<c0061e70>] from [<c009ff08>]
Jan 1 00:23:37 kernel: Function entered at [<c009fee0>] from [<c0061040>]
Jan 1 00:23:37 kernel: r5:c1989d80 r4:c1989d80
Jan 1 00:23:37 kernel: Function entered at [<c0060f84>] from [<c0062cec>]
Jan 1 00:23:37 kernel: Function entered at [<c0062c7c>] from [<c0063420>]
Jan 1 00:23:37 kernel: Function entered at [<c0062fcc>] from [<c00639f8>]
Jan 1 00:23:37 kernel: Function entered at [<c0063974>] from [<c0086500>]
Jan 1 00:23:37 kernel: Function entered at [<c0086444>] from [<c0086ec0>]
Jan 1 00:23:37 kernel: r8:be8ae878 r7:00000332 r6:c1989f70 r5:be8ae878 r4:c16d0b80
Jan 1 00:23:37 kernel: Function entered at [<c0086e0c>] from [<c0087014>]
Jan 1 00:23:37 kernel: r7:00000332 r6:c16d0b80 r5:00000000 r4:00000000
Jan 1 00:23:37 kernel: Function entered at [<c0086fd0>] from [<c0023e00>]
Jan 1 00:23:37 kernel: r8:c0023f84 r7:00000004 r6:00000009 r5:be8ae878 r4:00000332
Jan 1 00:23:37 kernel: Mem-info:
Jan 1 00:23:37 kernel: DMA per-cpu:
Jan 1 00:23:37 kernel: CPU 0: hi: 0, btch: 1 usd: 0
Jan 1 00:23:37 kernel: Active_anon:512 active_file:0 inactive_anon:489
Jan 1 00:23:37 kernel: inactive_file:0 unevictable:5405 dirty:0 writeback:0 unstable:0
Jan 1 00:23:37 kernel: free:262 slab:559 mapped:734 pagetables:79 bounce:0
Jan 1 00:23:37 kernel: DMA free:1048kB min:720kB low:900kB high:1080kB active_anon:2048kB inactive_anon:1956kB active_file:0kB inactive_file:0kB unevictable:21620kB present:32512kB pages_scanned:2160 all_unreclaimable? no
Jan 1 00:23:37 kernel: lowmem_reserve[]: 0 0 0
Jan 1 00:23:37 kernel: DMA: 22*4kB 4*8kB 4*16kB 1*32kB 3*64kB 1*128kB 0*256kB 1*512kB 0*1024kB 0*2048kB 0*4096kB = 1048kB
Jan 1 00:23:37 kernel: 5422 total pagecache pages
Jan 1 00:23:37 kernel: 0 pages in swap cache
Jan 1 00:23:37 kernel: Swap cache stats: add 0, delete 0, find 0/0
Jan 1 00:23:37 kernel: Free swap = 0kB
Jan 1 00:23:37 kernel: Total swap = 0kB
Jan 1 00:23:37 kernel: 8192 pages of RAM
Jan 1 00:23:37 kernel: 373 free pages
Jan 1 00:23:37 kernel: 682 reserved pages
Jan 1 00:23:37 kernel: 476 slab pages
Jan 1 00:23:37 kernel: 3111 pages shared
Jan 1 00:23:37 kernel: 0 pages swap cached
Jan 1 00:23:37 kernel: Out of memory: kill process 1993 (sshd) score 81 or a child
Jan 1 00:23:37 kernel: Killed process 2013 (sshd)

I was wrong - looks like it can be caused by hardware detection, but it does depend on what installer components are selected. I had included debconf preseed which had enough memory usage to invoke oom killer.

This has only been reported on the nslu2, which is not supported by more recent Ubuntu releases. If you can reproduce this on hardware supported by the current release, please reopen and provide detail.