summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorYann Collet <cyan@fb.com>2018-05-06 01:24:11 (GMT)
committerYann Collet <cyan@fb.com>2018-05-06 01:24:11 (GMT)
commitaf127334670a5e7b710bbd6adb71aa7c3ef0cd72 (patch)
treebab2c44f789bcd9b92a80802ca4de791e60b6e17
parentd358e33faa87e0293f27f272a8579de73f4ba938 (diff)
downloadlz4-af127334670a5e7b710bbd6adb71aa7c3ef0cd72.zip
lz4-af127334670a5e7b710bbd6adb71aa7c3ef0cd72.tar.gz
lz4-af127334670a5e7b710bbd6adb71aa7c3ef0cd72.tar.bz2
fixed frametest error
The error can be reproduced using following command : ./frametest -v -i100000000 -s1659 -t31096808 It's actually a bug in the stream LZ4 API, when starting a new stream and providing a first chunk to complete with size < MINMATCH. In which case, the chunk becomes a dictionary. No hash was generated and stored, but the chunk is accessible as default position 0 points to dictStart, and position 0 is still within MAX_DISTANCE. Then, next attempt to read 32-bits from position 0 fails. The issue would have been mitigated by starting from index 64 KB, effectively eliminating position 0 as too far away. The proper fix is to eliminate such "dictionary" as too small. Which is what this patch does.
-rw-r--r--lib/lz4.c13
-rw-r--r--lib/lz4frame.c1
-rw-r--r--tests/frametest.c8
3 files changed, 19 insertions, 3 deletions
diff --git a/lib/lz4.c b/lib/lz4.c
index 3860c51..9a922b5 100644
--- a/lib/lz4.c
+++ b/lib/lz4.c
@@ -767,6 +767,7 @@ LZ4_FORCE_INLINE int LZ4_compress_generic(
} else if (dictDirective==usingExtDict) {
if (matchIndex < startIndex) {
DEBUGLOG(7, "extDict candidate: matchIndex=%5u < startIndex=%5u", matchIndex, startIndex);
+ assert(startIndex - matchIndex >= MINMATCH);
match = dictBase + matchIndex;
lowLimit = dictionary;
} else {
@@ -989,6 +990,7 @@ _last_literals:
if (outputLimited == fillOutput) {
*inputConsumed = (int) (((const char*)ip)-source);
}
+ DEBUGLOG(5, "LZ4_compress_generic: compressed %i bytes into %i bytes", inputSize, (int)(((char*)op) - dest));
return (int)(((char*)op) - dest);
}
@@ -1255,12 +1257,19 @@ int LZ4_compress_fast_continue (LZ4_stream_t* LZ4_stream, const char* source, ch
{
const tableType_t tableType = byU32;
LZ4_stream_t_internal* streamPtr = &LZ4_stream->internal_donotuse;
- const BYTE* const dictEnd = streamPtr->dictionary + streamPtr->dictSize;
+ const BYTE* dictEnd = streamPtr->dictionary + streamPtr->dictSize;
if (streamPtr->initCheck) return 0; /* Uninitialized structure detected */
LZ4_renormDictT(streamPtr, inputSize); /* avoid index overflow */
if (acceleration < 1) acceleration = ACCELERATION_DEFAULT;
+ /* invalidate tiny dictionaries */
+ if (streamPtr->dictSize < 4) {
+ streamPtr->dictSize = 0;
+ streamPtr->dictionary = (const BYTE*)source;
+ dictEnd = (const BYTE*)source;
+ }
+
/* Check overlapping input/dictionary space */
{ const BYTE* sourceEnd = (const BYTE*) source + inputSize;
if ((sourceEnd > streamPtr->dictionary) && (sourceEnd < dictEnd)) {
@@ -1402,6 +1411,8 @@ LZ4_FORCE_INLINE int LZ4_decompress_generic(
const BYTE* const shortiend = iend - (endOnInput ? 14 : 8) /*maxLL*/ - 2 /*offset*/;
const BYTE* const shortoend = oend - (endOnInput ? 14 : 8) /*maxLL*/ - 18 /*maxML*/;
+ DEBUGLOG(5, "LZ4_decompress_generic (srcSize:%i)", srcSize);
+
/* Special cases */
if ((partialDecoding) && (oexit > oend-MFLIMIT)) oexit = oend-MFLIMIT; /* targetOutputSize too high => just decode everything */
if ((endOnInput) && (unlikely(outputSize==0))) return ((srcSize==1) && (*ip==0)) ? 0 : -1; /* Empty output buffer */
diff --git a/lib/lz4frame.c b/lib/lz4frame.c
index aa7889b..4e40816 100644
--- a/lib/lz4frame.c
+++ b/lib/lz4frame.c
@@ -808,6 +808,7 @@ size_t LZ4F_compressUpdate(LZ4F_cctx* cctxPtr,
LZ4F_lastBlockStatus lastBlockCompressed = notDone;
compressFunc_t const compress = LZ4F_selectCompression(cctxPtr->prefs.frameInfo.blockMode, cctxPtr->prefs.compressionLevel);
+ DEBUGLOG(4, "LZ4F_compressUpdate (srcSize=%zu)", srcSize);
if (cctxPtr->cStage != 1) return err0r(LZ4F_ERROR_GENERIC);
if (dstCapacity < LZ4F_compressBound_internal(srcSize, &(cctxPtr->prefs), cctxPtr->tmpInSize))
diff --git a/tests/frametest.c b/tests/frametest.c
index 7d69ff7..21d883d 100644
--- a/tests/frametest.c
+++ b/tests/frametest.c
@@ -740,8 +740,9 @@ static void locateBuffDiff(const void* buff1, const void* buff2, size_t size, un
size_t p=0;
const BYTE* b1=(const BYTE*)buff1;
const BYTE* b2=(const BYTE*)buff2;
+ DISPLAY("locateBuffDiff: looking for error position \n");
if (nonContiguous) {
- DISPLAY("Non-contiguous output test (%i bytes)\n", (int)size);
+ DISPLAY("mode %u: non-contiguous output (%zu bytes), cannot search \n", nonContiguous, size);
return;
}
while (p < size && b1[p]==b2[p]) p++;
@@ -838,6 +839,8 @@ int fuzzerTests(U32 seed, unsigned nbTests, unsigned startTest, double compressi
size_t const iSize = MIN(sampleMax, (size_t)(iend-ip));
size_t const oSize = LZ4F_compressBound(iSize, prefsPtr);
cOptions.stableSrc = ((FUZ_rand(&randState) & 3) == 1);
+ DISPLAYLEVEL(6, "Sending %zi bytes to compress (stableSrc:%u) \n",
+ iSize, cOptions.stableSrc);
result = LZ4F_compressUpdate(cCtx, op, oSize, ip, iSize, &cOptions);
CHECK(LZ4F_isError(result), "Compression failed (error %i : %s)", (int)result, LZ4F_getErrorName(result));
@@ -882,7 +885,8 @@ int fuzzerTests(U32 seed, unsigned nbTests, unsigned startTest, double compressi
dOptions.stableDst = FUZ_rand(&randState) & 1;
if (nonContiguousDst==2) dOptions.stableDst = 0; /* overwrite mode */
result = LZ4F_decompress(dCtx, op, &oSize, ip, &iSize, &dOptions);
- if (LZ4F_getErrorCode(result) == LZ4F_ERROR_contentChecksum_invalid) locateBuffDiff(srcStart, decodedBuffer, srcSize, nonContiguousDst);
+ if (LZ4F_getErrorCode(result) == LZ4F_ERROR_contentChecksum_invalid)
+ locateBuffDiff(srcStart, decodedBuffer, srcSize, nonContiguousDst);
CHECK(LZ4F_isError(result), "Decompression failed (error %i:%s)", (int)result, LZ4F_getErrorName(result));
XXH64_update(&xxh64, op, (U32)oSize);
totalOut += oSize;