liblzma: Check that the first byte of range encoded data is 0x00.

It is just to be more pedantic and thus perhaps catch broken
files slightly earlier.

diff --git a/src/liblzma/lzma/lzma_decoder.c b/src/liblzma/lzma/lzma_decoder.c
index d61b7de..b6f1b6a 100644 (file)
--- a/src/liblzma/lzma/lzma_decoder.c
+++ b/src/liblzma/lzma/lzma_decoder.c
@@ -289,8 +289,12 @@ lzma_decode(lzma_coder *restrict coder, lzma_dict *restrict dictptr,
        // Initialization //
        ////////////////////
 
-       if (!rc_read_init(&coder->rc, in, in_pos, in_size))
-               return LZMA_OK;
+       {
+               const lzma_ret ret = rc_read_init(
+                               &coder->rc, in, in_pos, in_size);
+               if (ret != LZMA_STREAM_END)
+                       return ret;
+       }
 
        ///////////////
        // Variables //

diff --git a/src/liblzma/rangecoder/range_decoder.h b/src/liblzma/rangecoder/range_decoder.h
index fb96180..e0b051f 100644 (file)
--- a/src/liblzma/rangecoder/range_decoder.h
+++ b/src/liblzma/rangecoder/range_decoder.h
@@ -25,20 +25,26 @@ typedef struct {
 
 
 /// Reads the first five bytes to initialize the range decoder.
-static inline bool
+static inline lzma_ret
 rc_read_init(lzma_range_decoder *rc, const uint8_t *restrict in,
                size_t *restrict in_pos, size_t in_size)
 {
        while (rc->init_bytes_left > 0) {
                if (*in_pos == in_size)
-                       return false;
+                       return LZMA_OK;
+
+               // The first byte is always 0x00. It could have been omitted
+               // in LZMA2 but it wasn't, so one byte is wasted in every
+               // LZMA2 chunk.
+               if (rc->init_bytes_left == 5 && in[*in_pos] != 0x00)
+                       return LZMA_DATA_ERROR;
 
                rc->code = (rc->code << 8) | in[*in_pos];
                ++*in_pos;
                --rc->init_bytes_left;
        }
 
-       return true;
+       return LZMA_STREAM_END;
 }