Improved speed of ZSTD_decompressStream()

When ZSTD_decompressStream() detects
that there is enough space in dst
to complete decompression in a single pass,
delegates to ZSTD_decompress(),
for an extra ~5% speed boost

diff --git a/lib/decompress/zstd_decompress.c b/lib/decompress/zstd_decompress.c
index 2646c802883ecb1390e81688c756b81a0e8467d8..305a9a8762d82abf011545bf1c5921beb47a8639 100644 (file)
--- a/lib/decompress/zstd_decompress.c
+++ b/lib/decompress/zstd_decompress.c
@@ -2100,7 +2100,7 @@ size_t ZSTD_decompress_usingDDict(ZSTD_DCtx* dctx,
 {
     /* pass content and size in case legacy frames are encountered */
     return ZSTD_decompressMultiFrame(dctx, dst, dstCapacity, src, srcSize,
-                                     ddict->dictContent, ddict->dictSize,
+                                     NULL, 0,
                                      ddict);
 }
 
@@ -2301,6 +2301,20 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB
                     break;
             }   }
 
+            /* check for single-pass mode opportunity */
+            if (zds->fParams.frameContentSize
+                && (U64)(size_t)(oend-op) >= zds->fParams.frameContentSize) {
+                size_t const cSize = ZSTD_findFrameCompressedSize(istart, iend-istart);
+                if (cSize <= (size_t)(iend-istart)) {
+                    size_t const decompressedSize = ZSTD_decompress_usingDDict(zds->dctx, op, oend-op, istart, cSize, zds->ddict);
+                    if (ZSTD_isError(decompressedSize)) return decompressedSize;
+                    ip += cSize;
+                    op += decompressedSize;
+                    zds->stage = zdss_init;
+                    someMoreWork = 0;
+                    break;
+            }   }
+
             /* Consume header */
             ZSTD_refDDict(zds->dctx, zds->ddict);
             {   size_t const h1Size = ZSTD_nextSrcSizeToDecompress(zds->dctx);  /* == ZSTD_frameHeaderSize_prefix */