OK, I see... The implementation is highly dependent on the device being a MTD device. It is no secret that my plan was to change the content of an android file system image. Could you elaborate a bit on "you might need to fiddle a bit to get the nandsim alignment figured out.", please. However in my case it would be sufficient to be able to extract the file system image to a local directory, change the content and then build the yaffs2 file system image again from the modified directory. Maybe you have some alternative suggestions instead of fiddling with getting a live yaffs2 file system up and running? I have found some tools on the net that seems to do what I need, but I find them unattractive because they are either binary only or lack of relation to this (reference) implementation. Some of these tools include: http://code.google.com/p/yaffey/ http://code.google.com/p/android-group-korea/downloads/detail?name=mkfs.yaffs2.x86 http://code.google.com/p/yafuse/ http://code.google.com/p/yaffs2utils/ What do you suggest in the approach to modify an android system image? Thanks in advance. Jarl 2013/1/27 Charles Manning : > On Saturday 19 January 2013 01:57:17 Jarl Friis wrote: >> OK. >> >> So I built the yaffs2 module for my ubuntu 12.10. >> I did a >> sudo insmod fs/yaffs2/yaffs.ko >> >> Now I would expect to be able to mount a yaffs2 file system from the >> Android SDK, >> so I copied >> cp ~/opt/android-sdks/system-images/android-15/x86/system.img /tmp/ >> and did >> sudo mkdir /mnt/android >> and tried to mount it >> sudo mount -t yaffs2 -o loop /tmp/system.img /mnt/android/ >> >> But nothing happens and dmesg shows: >> [21020.202539] yaffs: dev is 7340035 name is "loop3" rw >> [21020.202544] yaffs: passed flags "" >> [21020.202547] yaffs: yaffs: Attempting MTD mount of 7.3,"loop3" >> [21020.202577] BUG: unable to handle kernel NULL pointer dereference at >> 0000000000000001 >> [21020.202581] IP: [] >> yaffs_internal_read_super.isra.17+0x2cb/0x9b0 [yaffs] >> [21020.202593] PGD 5e2ec067 PUD 775b1067 PMD 0 >> [21020.202598] Oops: 0000 [#4] SMP >> [21020.202602] CPU 1 >> [21020.202603] Modules linked in: yaffs vmnet(O) vsock(O) vmci(O) vmmon(O) >> dm_crypt nvidia(PO) pata_pcmcia snd_hda_codec_analog arc4 snd_hda_intel >> iwlwifi snd_hda_codec snd_hwdep snd_pcm tpm_infineon mac80211 snd_seq_midi >> snd_rawmidi snd_seq_midi_event snd_seq pcmcia snd_timer r852 snd_seq_device >> sm_common nand yenta_socket nand_ids mtd pcmcia_rsrc cfg80211 nand_bch bch >> pcmcia_core bnep rfcomm snd coretemp nand_ecc r592 uvcvideo soundcore >> memstick snd_page_alloc bluetooth kvm_intel kvm joydev psmouse lpc_ich >> hp_wmi videobuf2_core hp_accel sparse_keymap videodev lis3lv02d cdc_acm >> videobuf2_vmalloc mxm_wmi videobuf2_memops microcode tpm_tis serio_raw wmi >> parport_pc mac_hid input_polldev ppdev nfsd nfs lockd fscache auth_rpcgss >> nfs_acl lp sunrpc parport binfmt_misc firewire_ohci sdhci_pci firewire_core >> crc_itu_t sdhci e1000e video >> [21020.202676] >> [21020.202678] Pid: 17928, comm: mount Tainted: P D O >> 3.5.0-22-generic #34-Ubuntu Hewlett-Packard HP EliteBook 8730w/30EC >> [21020.202683] RIP: 0010:[] [] >> yaffs_internal_read_super.isra.17+0x2cb/0x9b0 [yaffs] >> [21020.202690] RSP: 0018:ffff88007757fc28 EFLAGS: 00010246 >> [21020.202693] RAX: ffffffffffffffed RBX: ffffffffa10bb8e1 RCX: >> 000000000000c25f >> [21020.202695] RDX: 0000000000000000 RSI: 0000000000000003 RDI: >> 0000000000000001 >> [21020.202697] RBP: ffff88007757fcd8 R08: 00000000000005c4 R09: >> 33706f6f6c222c33 >> [21020.202699] R10: 332e3720666f2074 R11: 2233706f6f6c222c R12: >> ffff880002f67000 >> [21020.202701] R13: 0000000000000002 R14: ffffffffffffffed R15: >> 0000000000000000 >> [21020.202704] FS: 00007f09509b0800(0000) GS:ffff88023fc80000(0000) >> knlGS:0000000000000000 >> [21020.202706] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 >> [21020.202708] CR2: 0000000000000001 CR3: 0000000009907000 CR4: >> 00000000000407e0 >> [21020.202710] DR0: 0000000000000000 DR1: 0000000000000000 DR2: >> 0000000000000000 >> [21020.202712] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: >> 0000000000000400 >> [21020.202715] Process mount (pid: 17928, threadinfo ffff88007757e000, task >> ffff8801afb79700) >> [21020.202717] Stack: >> [21020.202719] 0000000000000000 ffff88022f0ad800 ffff880002f67000 >> 000000003017c9c0 >> [21020.202724] ffff88007757fca8 0000000000000000 0000000000000000 >> 0000000000000000 >> [21020.202728] 0000000000000000 0000000000000000 6cff880000000000 >> ffff880033706f6f >> [21020.202732] Call Trace: >> [21020.202741] [] >> yaffs2_internal_read_super_mtd+0x19/0x30 [yaffs] >> [21020.202748] [] mount_bdev+0x1cb/0x210 >> [21020.202755] [] ? ida_get_new_above+0x218/0x2a0 >> [21020.202761] [] ? >> yaffs_internal_read_super.isra.17+0x9b0/0x9b0 [yaffs] >> [21020.202767] [] yaffs2_mount+0x15/0x20 [yaffs] >> [21020.202770] [] mount_fs+0x43/0x1b0 >> [21020.202774] [] ? find_filesystem+0x63/0x80 >> [21020.202778] [] vfs_kern_mount+0x74/0x110 >> [21020.202782] [] do_kern_mount+0x54/0x110 >> [21020.202786] [] do_mount+0x26a/0x890 >> [21020.202791] [] ? strndup_user+0x5b/0x80 >> [21020.202794] [] sys_mount+0x8d/0xe0 >> [21020.202799] [] system_call_fastpath+0x16/0x1b >> [21020.202801] Code: 62 01 00 00 00 00 f0 74 0e 48 c7 c7 48 c9 0b a1 31 c0 >> e8 a5 c8 5c e0 8b 95 78 ff ff ff 41 bd 02 00 00 00 85 d2 0f 85 bb fe ff ff >> <41> 81 7e 14 00 02 00 00 0f 85 ad fe ff ff f7 05 b1 62 01 00 00 >> [21020.202845] RIP [] >> yaffs_internal_read_super.isra.17+0x2cb/0x9b0 [yaffs] >> [21020.202852] RSP >> [21020.202853] CR2: 0000000000000001 >> [21020.202861] ---[ end trace 66076b41e0516697 ]--- >> j >> >> >> So any hints of what I am missing? Should this not be possible? > > The problem here is that you are trying to mount yaffs on a block device (that > is what loop is). Yaffs works with mtd devices. > > For example, the PC-based Linux testing I do, I use the linux-tests/initramsim > script to create a nand sim, then use > mount -t yaffs2 /dev/mtdblock0 $MNTPT > > > You might get a bit further by using the mtd tools to load that image as a > simulated mtd device, but you might need to fiddle a bit to get the nandsim > alignment figured out. > > -- Charles > > > _______________________________________________ > yaffs mailing list > yaffs@lists.aleph1.co.uk > http://lists.aleph1.co.uk/cgi-bin/mailman/listinfo/yaffs -- Jarl Friis Softace ApS Rådhustorvet 7, 2. 3520 Farum LinkedIn: http://dk.linkedin.com/in/jarlfriis