Okay, this is kinda replying to a dead thread, but anyway. On Thu, 2004-08-19 at 01:09, Alan Stern wrote: > On Wed, 18 Aug 2004, Andrew Pilley wrote: > > > Alrighty. > > > > Ran overnight for approximately 6-7 hours, and was working fine this > > morning. > > > > Of course, since it's statistically impossible to prove a negative in > > this situation, it's at least potentially working. I'll let it run > > longer later today.
Looks like statistics have finally struck. Unfortunately, I don't have usb-storage's debug messages turned on for this, but I did manage to trigger a device hang while writing to my iriver ihp-140 over usb 2.0. It was copying data using rsync to the device, and at some point, it stopped processing half-way through a file. task stack traces as follows: usb-storage: kernel: usb-storage D 00000002 0 304 1 354 303 (L-TLB) kernel: cf991e40 00000046 cf991e24 00000002 00000012 cf9dbc18 cf991e38 c1325c00 kernel: cb6443e0 cf991e38 c1325c64 000001fa 8b54cd42 000005df cf9b1898 cffc0cec kernel: cf991000 cf991000 cf991e94 c02e89f8 00000000 cf9b16f0 c0113b40 00000000 kernel: Call Trace: kernel: [<c02e89f8>] wait_for_completion+0x78/0xd0 kernel: [<c027b68f>] usb_sg_wait+0xff/0x160 kernel: [<d08b51e2>] usb_stor_bulk_transfer_sglist+0x82/0xd0 [usb_storage] kernel: [<d08b5264>] usb_stor_bulk_transfer_sg+0x34/0x70 [usb_storage] kernel: [<d08b5974>] usb_stor_Bulk_transport+0x134/0x2b0 [usb_storage] kernel: [<d08b52c1>] usb_stor_invoke_transport+0x21/0x2d0 [usb_storage] kernel: [<d08b4936>] usb_stor_transparent_scsi_command+0x16/0x50 [usb_storage] kernel: [<d08b607f>] usb_stor_control_thread+0x19f/0x240 [usb_storage] kernel: [<c0103291>] kernel_thread_helper+0x5/0x14 scsi_eh_0: (just in case this is actually a scsi problem) kernel: scsi_eh_0 D C03CD960 0 303 1 304 161 (L-TLB) kernel: cf9bfef8 00000046 cf9b16f0 c03cd960 cf9bfee0 00000092 00000001 00000000 kernel: 00000092 cf991f60 cfed7100 00002eef 8841d5cf 000005e6 cf9b1308 cffc0d14 kernel: cf9bf000 cf9bf000 cf9bff4c c02e89f8 00000000 cf9b1160 c0113b40 00000000 kernel: Call Trace: kernel: [<c02e89f8>] wait_for_completion+0x78/0xd0 kernel: [<d08b41b2>] command_abort+0x62/0xb0 [usb_storage] kernel: [<d08c4ceb>] scsi_try_to_abort_cmd+0x5b/0x80 [scsi_mod] kernel: [<d08c4e11>] scsi_eh_abort_cmds+0x41/0x70 [scsi_mod] kernel: [<d08c594b>] scsi_unjam_host+0x9b/0xc0 [scsi_mod] kernel: [<d08c5a31>] scsi_error_handler+0xc1/0x100 [scsi_mod] kernel: [<c0103291>] kernel_thread_helper+0x5/0x14 the process doing the writing: (rsync) kernel: rsync D C0251DCA 0 1923 1 2248 948 (NOTLB) kernel: ce894bb0 00200082 ce894b90 c0251dca ce894000 cf492984 c133802c c133802c kernel: 8b309216 000005df cfed6050 000142b9 8b3acf69 000005df c2ce8418 ce5d135c kernel: ce894be4 c03dba38 ce894bb8 c02e8cfe ce894c0c c014ac5e ce5d135c 00000000 kernel: Call Trace: kernel: [<c02e8cfe>] io_schedule+0xe/0x20 kernel: [<c014ac5e>] __wait_on_buffer+0x8e/0x90 kernel: [<c014c05d>] __bread_slow+0x4d/0x70 kernel: [<c014c2bb>] __bread+0x1b/0x20 kernel: [<d09b706c>] __fat_access+0x6c/0x3d0 [fat] kernel: [<d09b742b>] fat_access+0x5b/0xb0 [fat] kernel: [<d09bc176>] fat_add_cluster+0xb6/0x1f0 [fat] kernel: [<d09b9e50>] fat_get_block+0xd0/0x170 [fat] kernel: [<c014c9d7>] __block_prepare_write+0x197/0x350 kernel: [<c014d03b>] cont_prepare_write+0xab/0x300 kernel: [<d09bb8c6>] fat_prepare_write+0x36/0x40 [fat] kernel: [<c012fe60>] generic_file_aio_write_nolock+0x3e0/0xb70 kernel: [<c0130646>] generic_file_write_nolock+0x56/0x70 kernel: [<c013071b>] generic_file_write+0x3b/0x50 kernel: [<d09b9f0d>] fat_file_write+0x1d/0x60 [fat] kernel: [<c0149a82>] vfs_write+0xa2/0xf0 kernel: [<c0149b81>] sys_write+0x41/0x70 kernel: [<c0104ecf>] syscall_call+0x7/0xb any other processes that might be relevent should be in http://www.icculus.org/~ashridah/messages.gz (i stupidly got 'ls' caught in the same state :( ) Hope that helps, but mail me back if it's not, and I'll setup something to keep writing in a continuous loop until I manage to reproduce it. Again, CC replies to me, since I'm not on the list at this time. thanks Andrew Pilley -- <Berawler> Is there any sanity or light left in this shrivelled husk of a world? <SingingDancingMoose> There was, but we had to trade it in for the internet. ------------------------------------------------------- This SF.Net email is sponsored by BEA Weblogic Workshop FREE Java Enterprise J2EE developer tools! Get your free copy of BEA WebLogic Workshop 8.1 today. http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
