liblzma: Small fixes to comments in the API headers.
authorLasse Collin <lasse.collin@tukaani.org>
Sat, 2 Oct 2010 08:38:20 +0000 (11:38 +0300)
committerLasse Collin <lasse.collin@tukaani.org>
Sat, 2 Oct 2010 08:38:20 +0000 (11:38 +0300)
src/liblzma/api/lzma/lzma.h

index c17736d..9d31a5c 100644 (file)
@@ -49,7 +49,7 @@
  * The memory usage formulas are only rough estimates, which are closest to
  * reality when dict_size is a power of two. The formulas are  more complex
  * in reality, and can also change a little between liblzma versions. Use
- * lzma_memusage_encoder() to get more accurate estimate of memory usage.
+ * lzma_raw_encoder_memusage() to get more accurate estimate of memory usage.
  */
 typedef enum {
        LZMA_MF_HC3     = 0x03,
@@ -69,7 +69,9 @@ typedef enum {
                 *
                 * Minimum nice_len: 4
                 *
-                * Memory usage: dict_size * 7.5
+                * Memory usage:
+                *  - dict_size <= 32 MiB: dict_size * 7.5
+                *  - dict_size > 32 MiB: dict_size * 6.5
                 */
 
        LZMA_MF_BT2     = 0x12,
@@ -98,7 +100,9 @@ typedef enum {
                 *
                 * Minimum nice_len: 4
                 *
-                * Memory usage: dict_size * 11.5
+                * Memory usage:
+                *  - dict_size <= 32 MiB: dict_size * 11.5
+                *  - dict_size > 32 MiB: dict_size * 10.5
                 */
 } lzma_match_finder;