Fixed issue with immediate exhaustion and small unaligned storage

This was a small hole in the logic that handles initializing the
lookahead buffer. To imitate exhaustion (so the block allocator
will trigger a scan), the lookahead buffer is rewound a full
lookahead and set up to look like it is exhausted. However,
unlike normal allocation, this rewind was not kept aligned to
a multiple of the scan size, which is limited by both the
lookahead buffer and the total storage size.

This bug went unnoticed for so long because it only causes
problems when the block device is both:
1. Not aligned to the lookahead buffer (not a power of 2)
2. Smaller than the lookahead buffer

While this seems like a strange corner case for a block device,
this turned out to be very common for internal flash, especially
when a handleful of blocks are reserved for code.

.travis.yml

@@ -12,10 +12,9 @@ script:
  - make test QUIET=1
 
  # run tests with a few different configurations
- - CFLAGS="-DLFS_READ_SIZE=1 -DLFS_PROG_SIZE=1" make test QUIET=1
- - CFLAGS="-DLFS_READ_SIZE=512 -DLFS_PROG_SIZE=512" make test QUIET=1
- - CFLAGS="-DLFS_BLOCK_COUNT=1023" make test QUIET=1
- - CFLAGS="-DLFS_LOOKAHEAD=2048" make test QUIET=1
+ - CFLAGS="-DLFS_READ_SIZE=1 -DLFS_PROG_SIZE=1" make test QUIET=1
+ - CFLAGS="-DLFS_READ_SIZE=512 -DLFS_PROG_SIZE=512" make test QUIET=1
+ - CFLAGS="-DLFS_BLOCK_COUNT=1023 -DLFS_LOOKAHEAD=2048" make test QUIET=1
 
  # self-host with littlefs-fuse for fuzz test
  - make -C littlefs-fuse

lfs.c

@@ -2040,8 +2040,8 @@ int lfs_mount(lfs_t *lfs, const struct lfs_config *cfg) {
  }
 
  // setup free lookahead
- lfs->free.begin = -lfs->cfg->lookahead;
- lfs->free.off = lfs->cfg->lookahead;
+ lfs->free.begin = -lfs_min(lfs->cfg->lookahead, lfs->cfg->block_count);
+ lfs->free.off = -lfs->free.begin;
  lfs->free.end = lfs->free.begin + lfs->free.off + lfs->cfg->block_count;
 
  // load superblock