Charles, I think you are right about the small partition not being special. I left a device running for two days copying and deleting a test file over and over with debugging flags turned on and indeed bad blocks were formed. I don't think the the bad blocks are valid, as this happens on new devices and not only on a specific (old) device. This is the relevant log part, I couldn't make much of it so I will be happy if any one could help me with this: May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 729 byte 4096 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk729 byte 4096 del 0 ser 0 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_write_chunk_tags chunk 15356 data d4811000 tags c5d69bc8 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 729 byte 4096 seq 40727 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk729 byte 4096 del 0 ser 1 seq 40727 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_read_chunk_tags chunk 16381 data d4811000 tags c5d69bc8 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 512 byte 4096 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 0 blkbad 0 chused 1 obj 275 chunk512 byte 4096 del 0 ser 0 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 512 byte 4096 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk512 byte 4096 del 0 ser 0 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_write_chunk_tags chunk 15357 data d4811000 tags c5d69bc8 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 512 byte 4096 seq 40727 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk512 byte 4096 del 0 ser 1 seq 40727 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_read_chunk_tags chunk 16382 data d4811000 tags c5d69bc8 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 513 byte 4096 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 0 blkbad 0 chused 1 obj 275 chunk513 byte 4096 del 0 ser 0 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 513 byte 4096 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk513 byte 4096 del 0 ser 0 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_write_chunk_tags chunk 15358 data d4811000 tags c5d69bc8 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 513 byte 4096 seq 40727 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk513 byte 4096 del 0 ser 1 seq 40727 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_read_chunk_tags chunk 16383 data d4811000 tags c5d69bc8 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 514 byte 4096 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 0 blkbad 0 chused 1 obj 275 chunk514 byte 4096 del 0 ser 0 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 514 byte 4096 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk514 byte 4096 del 0 ser 0 seq 40613 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_write_chunk_tags chunk 15359 data d4811000 tags c5d69bc8 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 514 byte 4096 seq 40727 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk514 byte 4096 del 0 ser 1 seq 40727 May 9 13:24:50 192.168.1.102 kernel: yaffs: yaffs_block_became_dirty block 128 state 8 needs retiring May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_mark_block_bad 127 May 9 13:24:50 192.168.1.102 kernel: Bad block table written to 0x00007ff80000, version 0x0D May 9 13:24:50 192.168.1.102 kernel: Bad block table written to 0x00007ff00000, version 0x0D May 9 13:24:50 192.168.1.102 kernel: yaffs: **>> Block 128 retired May 9 13:24:50 192.168.1.102 kernel: yaffs: Block 128 is in state 9 after gc, should be erased May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_write_chunk_tags chunk 15616 data c53a1000 tags c5d69c38 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 353 chunk 12675 byte 4096 seq 40728 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 0 blkbad 0 chused 1 obj 353 chunk12675 byte 4096 del 0 ser 1 seq 40728 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_read_chunk_tags chunk 1280 data d4811000 tags c5d69d28 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 1718 byte 4096 seq 40617 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 0 blkbad 0 chused 1 obj 275 chunk1718 byte 4096 del 0 ser 0 seq 40617 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 1718 byte 4096 seq 40617 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk1718 byte 4096 del 0 ser 0 seq 40617 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_write_chunk_tags chunk 15617 data d4811000 tags c5d69d28 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 1718 byte 4096 seq 40728 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 1 blkbad 0 chused 1 obj 275 chunk1718 byte 4096 del 0 ser 1 seq 40728 May 9 13:24:50 192.168.1.102 kernel: yaffs: nandmtd2_read_chunk_tags chunk 1281 data d4811000 tags c5d69d28 May 9 13:24:50 192.168.1.102 kernel: yaffs: packed tags obj 275 chunk 1719 byte 4096 seq 40617 May 9 13:24:50 192.168.1.102 kernel: yaffs: ext.tags eccres 0 blkbad 0 chused 1 obj 275 chunk1719 byte 4096 del 0 ser 0 seq 40617 I think the bad blocks are formed without a real reason, probably some issue with the mtd layer but I can't seem to find the problem here, I'm really pulling my hair here... Boaz. On 05/09/11 01:03, Charles Manning wrote: > On Thursday 05 May 2011 23:03:17 Boaz Ben-David wrote: >> Hi all, >> >> I am using Yaffs on MLC NAND with several partitions, two of them are >> 65MB and about 1.6GB in size. >> For some reason, the smaller one has a lot of bad blocks created on it >> under use when the same usage >> on the larger (1.6GB) partition does not generate such results. >> I haven't been able to determine the cause for this and I am using the >> default YAFFS2 configuration in the kernel, >> nothing too fancy. >> I've written to the mailing list before about this bad blocks creation >> issue but back then I didn't know this is not >> happening on bigger partitions, and since I wasn't able to find the >> cause for this I thought I should try again. >> BTW, I've already run all the built-in mtd tests on the device but no >> errors were found. >> >> Can anyone suggest a lead on this? >> > Do you see this on more than one device? If not, then perhaps there are just > more problems in the 65M area than the rest. > > Are the usage patterns the same? > > It is also worth considering that a fixed amount of write traffic will impact > the 65MB partition far more than the larger partition. > > > > > > _______________________________________________ > yaffs mailing list > yaffs@lists.aleph1.co.uk > http://lists.aleph1.co.uk/cgi-bin/mailman/listinfo/yaffs