[typo] Fix a double word in comments

Backport two parts of https://github.com/facebook/zstd/commit/b772f53952fa167e3c8d5630b26397e8d4fb4c5b
that were also reported on the LKML.

Reported-by: Slark Xiao <slark_xiao@163.com>
Reported-by: Xin Gao <gaoxin@cdjrlc.com>
Reported-by: Jilin Yuan <yuanjilin@cdjrlc.com>

lib/compress/zstd_compress.c

@@ -5764,7 +5764,7 @@ ZSTD_validateSequence(U32 offCode, U32 matchLength,
                       size_t posInSrc, U32 windowLog, size_t dictSize)
 {
     U32 const windowSize = 1 << windowLog;
-    /* posInSrc represents the amount of data the the decoder would decode up to this point.
+    /* posInSrc represents the amount of data the decoder would decode up to this point.
      * As long as the amount of data decoded is less than or equal to window size, offsets may be
      * larger than the total length of output decoded in order to reference the dict, even larger than
      * window size. After output surpasses windowSize, we're limited to windowSize offsets again.

lib/decompress/zstd_decompress.c

@@ -1108,7 +1108,7 @@ size_t ZSTD_decompress(void* dst, size_t dstCapacity, const void* src, size_t sr
 size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx* dctx) { return dctx->expected; }
 
 /**
- * Similar to ZSTD_nextSrcSizeToDecompress(), but when when a block input can be streamed,
+ * Similar to ZSTD_nextSrcSizeToDecompress(), but when a block input can be streamed,
  * we allow taking a partial block as the input. Currently only raw uncompressed blocks can
  * be streamed.
  *