Changed lfs_dir_alloc to maximize block cycles for new metadata pairs

Previously we only bumped the revision count if an eviction would occur
immediately (and possibly corrupt littlefs). This works, but does risk
an unoptimal superblock size if an almost-exhausted superblock was
allocated during lfs_format.

As pointed out by tim-nordell-nimbelink, we can align the revision count
to maximize the number of block cycles without breaking the existing
requirements of increasing revision counts.

As an added benefit, littlefs's wear-leveling should behave more
consistently after this change.

.travis.yml

@@ -208,22 +208,6 @@ jobs:
     script:
       - make test TFLAGS+="-k --valgrind"
 
-  # test compilation in thread-safe mode
-  - stage: test
-    env:
-      - NAME=littlefs-threadsafe
-      - CC="arm-linux-gnueabi-gcc --static -mthumb"
-      - CFLAGS="-Werror -DLFS_THREADSAFE"
-    if: branch !~ -prefix$
-    install:
-      - *install-common
-      - sudo apt-get install
-            gcc-arm-linux-gnueabi
-            libc6-dev-armel-cross
-      - arm-linux-gnueabi-gcc --version
-    # report-size will compile littlefs and report the size
-    script: [*report-size]
-
   # self-host with littlefs-fuse for fuzz test
   - stage: test
     env:

lfs.h

@@ -9,7 +9,6 @@
 
 #include <stdint.h>
 #include <stdbool.h>
-#include "lfs_util.h"
 
 #ifdef __cplusplus
 extern "C"
@@ -175,16 +174,6 @@ struct lfs_config {
     // are propogated to the user.
     int (*sync)(const struct lfs_config *c);
 
-#ifdef LFS_THREADSAFE
-    // Lock the underlying block device. Negative error codes
-    // are propogated to the user.
-    int (*lock)(const struct lfs_config *c);
-
-    // Unlock the underlying block device. Negative error codes
-    // are propogated to the user.
-    int (*unlock)(const struct lfs_config *c);
-#endif
-
     // Minimum size of a block read. All read operations will be a
     // multiple of this value.
     lfs_size_t read_size;