Fixed Btree update

ZSTD_updateTree() expected to be followed by a Bt match finder, which would update zc->nextToUpdate.
With the new optimal match finder, it's not necessarily the case : a match might be found during repcode or hash3, and stops there because it reaches sufficient_len, without even entering the binary tree.
Previous policy was to nonetheless update zc->nextToUpdate, but the current position would not be inserted, creating "holes" in the btree, aka positions that will no longer be searched.
Now, when current position is not inserted, zc->nextToUpdate is not update, expecting ZSTD_updateTree() to fill the tree later on.

Solution selected is that ZSTD_updateTree() takes care of properly setting zc->nextToUpdate,
so that it no longer depends on a future function to do this job.

It took time to get there, as the issue started with a memory sanitizer error.
The pb would have been easier to spot with a proper `assert()`.
So this patch add a few of them.

Additionnally, I discovered that `make test` does not enable `assert()` during CLI tests.
This patch enables them.

Unfortunately, these `assert()` triggered other (unrelated) bugs during CLI tests, mostly within zstdmt.
So this patch also fixes them.

- Changed packed structure for gcc memory access : memory sanitizer would complain that a read "might" reach out-of-bound position on the ground that the `union` is larger than the type accessed.
  Now, to avoid this issue, each type is independent.
- ZSTD_CCtxParams_setParameter() : @return provides the value of parameter, clamped/fixed appropriately.
- ZSTDMT : changed constant name to ZSTDMT_JOBSIZE_MIN
- ZSTDMT : multithreading is automatically disabled when srcSize <= ZSTDMT_JOBSIZE_MIN, since only one thread will be used in this case (saves memory and runtime).
- ZSTDMT : nbThreads is automatically clamped on setting the value.
This commit is contained in:
Yann Collet
2017-11-16 12:18:56 -08:00
parent dfc14579f5
commit 05dffe43a7
9 changed files with 263 additions and 191 deletions
+34 -22
View File
@@ -18,8 +18,9 @@
/** ZSTD_insertBt1() : add one or multiple positions to tree.
* ip : assumed <= iend-8 .
* @return : nb of positions added */
static U32 ZSTD_insertBt1(ZSTD_CCtx* zc, const BYTE* const ip, const U32 mls, const BYTE* const iend, U32 nbCompares,
U32 extDict)
static U32 ZSTD_insertBt1(ZSTD_CCtx* zc,
const BYTE* const ip, const BYTE* const iend,
U32 nbCompares, U32 const mls, U32 const extDict)
{
U32* const hashTable = zc->hashTable;
U32 const hashLog = zc->appliedParams.cParams.hashLog;
@@ -58,6 +59,7 @@ static U32 ZSTD_insertBt1(ZSTD_CCtx* zc, const BYTE* const ip, const U32 mls, co
while (nbCompares-- && (matchIndex > windowLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
assert(matchIndex < current);
#ifdef ZSTD_C_PREDICT /* note : can create issues when hlog small <= 11 */
const U32* predictPtr = bt + 2*((matchIndex-1) & btMask); /* written this way, as bt is a roll buffer */
@@ -124,6 +126,36 @@ static U32 ZSTD_insertBt1(ZSTD_CCtx* zc, const BYTE* const ip, const U32 mls, co
return 1;
}
FORCE_INLINE_TEMPLATE
void ZSTD_updateTree_internal(ZSTD_CCtx* zc,
const BYTE* const ip, const BYTE* const iend,
const U32 nbCompares, const U32 mls, const U32 extDict)
{
const BYTE* const base = zc->base;
U32 const target = (U32)(ip - base);
U32 idx = zc->nextToUpdate;
DEBUGLOG(7, "ZSTD_updateTree_internal, from %u to %u (extDict:%u)",
idx, target, extDict);
while(idx < target)
idx += ZSTD_insertBt1(zc, base+idx, iend, nbCompares, mls, extDict);
zc->nextToUpdate = target;
}
void ZSTD_updateTree(ZSTD_CCtx* zc,
const BYTE* const ip, const BYTE* const iend,
const U32 nbCompares, const U32 mls)
{
return ZSTD_updateTree_internal(zc, ip, iend, nbCompares, mls, 0 /*extDict*/);
}
void ZSTD_updateTree_extDict(ZSTD_CCtx* zc,
const BYTE* const ip, const BYTE* const iend,
const U32 nbCompares, const U32 mls)
{
return ZSTD_updateTree_internal(zc, ip, iend, nbCompares, mls, 1 /*extDict*/);
}
static size_t ZSTD_insertBtAndFindBestMatch (
ZSTD_CCtx* zc,
@@ -206,16 +238,6 @@ static size_t ZSTD_insertBtAndFindBestMatch (
}
void ZSTD_updateTree(ZSTD_CCtx* zc, const BYTE* const ip, const BYTE* const iend, const U32 nbCompares, const U32 mls)
{
const BYTE* const base = zc->base;
const U32 target = (U32)(ip - base);
U32 idx = zc->nextToUpdate;
while(idx < target)
idx += ZSTD_insertBt1(zc, base+idx, mls, iend, nbCompares, 0);
}
/** ZSTD_BtFindBestMatch() : Tree updater, providing best match */
static size_t ZSTD_BtFindBestMatch (
ZSTD_CCtx* zc,
@@ -246,16 +268,6 @@ static size_t ZSTD_BtFindBestMatch_selectMLS (
}
void ZSTD_updateTree_extDict(ZSTD_CCtx* zc, const BYTE* const ip, const BYTE* const iend, const U32 nbCompares, const U32 mls)
{
const BYTE* const base = zc->base;
const U32 target = (U32)(ip - base);
U32 idx = zc->nextToUpdate;
while (idx < target) idx += ZSTD_insertBt1(zc, base+idx, mls, iend, nbCompares, 1);
}
/** Tree updater, providing best match */
static size_t ZSTD_BtFindBestMatch_extDict (
ZSTD_CCtx* zc,