diff options
Diffstat (limited to 'nuttx/fs/nxffs/nxffs_pack.c')
| -rw-r--r-- | nuttx/fs/nxffs/nxffs_pack.c | 37 |
1 files changed, 28 insertions, 9 deletions
diff --git a/nuttx/fs/nxffs/nxffs_pack.c b/nuttx/fs/nxffs/nxffs_pack.c index afbb55307..d6e31896e 100644 --- a/nuttx/fs/nxffs/nxffs_pack.c +++ b/nuttx/fs/nxffs/nxffs_pack.c @@ -1409,20 +1409,39 @@ start_pack: eblock < volume->geo.neraseblocks; eblock++) { - /* Read the erase block into the pack buffer. We need to do this even - * if we are overwriting the entire block so that we skip over - * previously marked bad blocks. - */ + /* Get the starting block number of the erase block */ pack.block0 = eblock * volume->blkper; - ret = MTD_BREAD(volume->mtd, pack.block0, volume->blkper, volume->pack); - if (ret < 0) + + /* Read the entire erase block into the pack buffer, one-block-at-a- + * time. We need to do this even if we are overwriting the entire + * block so that (1) we skip over previously marked bad blocks, and + * (2) we can handle individual block read failures. + * + * For most FLASH, a read failure indicates a fatal hardware failure. + * But for NAND FLASH, the read failure probably indicates a block + * with uncorrectable bit errors. + */ + + /* Pack each I/O block */ + + for (i = 0, block = pack.block0, pack.iobuffer = volume->pack; + i < volume->blkper; + i++, block++, pack.iobuffer += volume->geo.blocksize) { - fdbg("ERROR: Failed to read erase block %d: %d\n", eblock, -ret); - goto errout_with_pack; + /* Read the next block in the erase block */ + + ret = MTD_BREAD(volume->mtd, block, 1, pack.iobuffer); + if (ret < 0) + { + /* Force a the block to be an NXFFS bad block */ + + fdbg("ERROR: Failed to read block %d: %d\n", block, ret); + nxffs_blkinit(volume, pack.iobuffer, BLOCK_STATE_BAD); + } } - /* Pack each I/O block */ + /* Now ack each I/O block */ for (i = 0, block = pack.block0, pack.iobuffer = volume->pack; i < volume->blkper; |