/*
* StrategyGetBuffer
*
* Called by the bufmgr to get the next candidate buffer to use in
* BufferAlloc(). The only hard requirement BufferAlloc() has is that
* the selected buffer must not currently be pinned by anyone.
*
* To ensure that no one else can pin the buffer before we do, we must
* return the buffer with the buffer header spinlock still held. That
* means that we return with the BufFreelistLock still held, as well;
* the caller must release that lock once the spinlock is dropped.
*/
volatile BufferDesc *
StrategyGetBuffer(void)
{
volatile BufferDesc *buf;
int trycounter;
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
/*
* Try to get a buffer from the freelist. Note that the freeNext fields
* are considered to be protected by the BufFreelistLock not the
* individual buffer spinlocks, so it's OK to manipulate them without
* holding the spinlock.
*/
while (StrategyControl->firstFreeBuffer >= 0)
{
buf = &BufferDescriptors[StrategyControl->firstFreeBuffer];
Assert(buf->freeNext != FREENEXT_NOT_IN_LIST);
/* Unconditionally remove buffer from freelist */
StrategyControl->firstFreeBuffer = buf->freeNext;
buf->freeNext = FREENEXT_NOT_IN_LIST;
buf->freePre = FREEPRE_NOT_IN_LIST;
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; discard it and retry. (This can only happen if VACUUM put a
* valid buffer in the freelist and then someone else used it before
* we got to it.)
*/
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count == 0)
{
elog(LOG, "Get buf %d\n", buf->buf_id);
return buf;
}
UnlockBufHdr(buf);
}
/* Nothing on the freelist, so run the "clock sweep" algorithm */
/* this part is removed to remove "clock sweep" if nothing found in
* freelist, nothing available.
*/
/* not reached */
return NULL;
}
/*
* GetBufferFromRing -- returns a buffer from the ring, or NULL if the
* ring is empty.
*
* The bufhdr spin lock is held on the returned buffer.
*/
static volatile BufferDesc *
GetBufferFromRing(BufferAccessStrategy strategy)
{
volatile BufferDesc *buf;
Buffer bufnum;
/* Advance to next ring slot */
if (++strategy->current >= strategy->ring_size)
strategy->current = 0;
/*
* If the slot hasn't been filled yet, tell the caller to allocate a new
* buffer with the normal allocation strategy. He will then fill this
* slot by calling AddBufferToRing with the new buffer.
*/
bufnum = strategy->buffers[strategy->current];
if (bufnum == InvalidBuffer)
{
strategy->current_was_in_ring = false;
return NULL;
}
/*
* If the buffer is pinned we cannot use it under any circumstances.
*
* If usage_count is 0 or 1 then the buffer is fair game (we expect 1,
* since our own previous usage of the ring element would have left it
* there, but it might've been decremented by clock sweep since then). A
* higher usage_count indicates someone else has touched the buffer, so we
* shouldn't re-use it.
*/
buf = &BufferDescriptors[bufnum - 1];
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count <= 1)
{
strategy->current_was_in_ring = true;
return buf;
}
UnlockBufHdr(buf);
/*
* Tell caller to allocate a new buffer with the normal allocation
* strategy. He'll then replace this ring element via AddBufferToRing.
*/
strategy->current_was_in_ring = false;
return NULL;
}
/*
* StrategyGetBuffer
*
* Called by the bufmgr to get the next candidate buffer to use in
* BufferAlloc(). The only hard requirement BufferAlloc() has is that
* the selected buffer must not currently be pinned by anyone.
*
* strategy is a BufferAccessStrategy object, or NULL for default strategy.
*
* To ensure that no one else can pin the buffer before we do, we must
* return the buffer with the buffer header spinlock still held. If
* *lock_held is set on exit, we have returned with the BufFreelistLock
* still held, as well; the caller must release that lock once the spinlock
* is dropped. We do it that way because releasing the BufFreelistLock
* might awaken other processes, and it would be bad to do the associated
* kernel calls while holding the buffer header spinlock.
*/
volatile BufferDesc *
StrategyGetBuffer(BufferAccessStrategy strategy, bool *lock_held)
{
volatile BufferDesc *buf;
Latch *bgwriterLatch;
int trycounter;
/*
* If given a strategy object, see whether it can select a buffer. We
* assume strategy objects don't need the BufFreelistLock.
*/
if (strategy != NULL)
{
buf = GetBufferFromRing(strategy);
if (buf != NULL)
{
*lock_held = false;
return buf;
}
}
/* Nope, so lock the freelist */
*lock_held = true;
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
/*
* We count buffer allocation requests so that the bgwriter can estimate
* the rate of buffer consumption. Note that buffers recycled by a
* strategy object are intentionally not counted here.
*/
StrategyControl->numBufferAllocs++;
/*
* If bgwriterLatch is set, we need to waken the bgwriter, but we should
* not do so while holding BufFreelistLock; so release and re-grab. This
* is annoyingly tedious, but it happens at most once per bgwriter cycle,
* so the performance hit is minimal.
*/
bgwriterLatch = StrategyControl->bgwriterLatch;
if (bgwriterLatch)
{
StrategyControl->bgwriterLatch = NULL;
LWLockRelease(BufFreelistLock);
SetLatch(bgwriterLatch);
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
}
/*
* Try to get a buffer from the freelist. Note that the freeNext fields
* are considered to be protected by the BufFreelistLock not the
* individual buffer spinlocks, so it's OK to manipulate them without
* holding the spinlock.
*/
while (StrategyControl->firstFreeBuffer >= 0)
{
buf = &BufferDescriptors[StrategyControl->firstFreeBuffer];
Assert(buf->freeNext != FREENEXT_NOT_IN_LIST);
/* Unconditionally remove buffer from freelist */
StrategyControl->firstFreeBuffer = buf->freeNext;
buf->freeNext = FREENEXT_NOT_IN_LIST;
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; discard it and retry. (This can only happen if VACUUM put a
* valid buffer in the freelist and then someone else used it before
* we got to it. It's probably impossible altogether as of 8.3, but
* we'd better check anyway.)
*/
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count == 0)
{
if (strategy != NULL)
AddBufferToRing(strategy, buf);
return buf;
}
UnlockBufHdr(buf);
}
if(true)
{
/* Now using LRU buffer replacement policy */
int currMinTimeStamp = INT_MAX;
BufferStrategyControl *currTargetBuffer;
for (StrategyControl->nextVictimBuffer=0;StrategyControl->nextVictimBuffer<NBuffers;StrategyControl->nextVictimBuffer++)
{
buf = &BufferDescriptors[StrategyControl->nextVictimBuffer];
if(StrategyControl->nextVictimBuffer == NBuffers-1 && currMinTimeStamp == INT_MAX)
{
StrategyControl ->completePasses++;
elog(ERROR,"No available buffer frame");
}
LockBufHdr(buf);
if(buf->refcount==0 && buf->timer<currMinTimeStamp)
{
currMinTimeStamp = buf->timer;
currTargetBuffer = buf;
}
UnlockBufHdr(buf);
}
return currTargetBuffer;
}
else
{
/* Nothing on the freelist, so run the "clock sweep" algorithm */
trycounter = NBuffers;
for (;;)
{
buf = &BufferDescriptors[StrategyControl->nextVictimBuffer];
if (++StrategyControl->nextVictimBuffer >= NBuffers)
{
StrategyControl->nextVictimBuffer = 0;
StrategyControl->completePasses++;
}
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; decrement the usage_count (unless pinned) and keep scanning.
*/
LockBufHdr(buf);
if (buf->refcount == 0)
{
if (buf->usage_count > 0)
{
buf->usage_count--;
trycounter = NBuffers;
}
else
{
/* Found a usable buffer */
if (strategy != NULL)
AddBufferToRing(strategy, buf);
return buf;
}
}
else if (--trycounter == 0)
{
/*
* We've scanned all the buffers without making any state changes,
* so all the buffers are pinned (or were when we looked at them).
* We could hope that someone will free one eventually, but it's
* probably better to fail than to risk getting stuck in an
* infinite loop.
*/
UnlockBufHdr(buf);
elog(ERROR, "no unpinned buffers available");
}
UnlockBufHdr(buf);
}
/* not reached */
return NULL;
}
}
/*
* StrategyGetBuffer
*
* Called by the bufmgr to get the next candidate buffer to use in
* BufferAlloc(). The only hard requirement BufferAlloc() has is that
* the selected buffer must not currently be pinned by anyone.
*
* strategy is a BufferAccessStrategy object, or NULL for default strategy.
*
* To ensure that no one else can pin the buffer before we do, we must
* return the buffer with the buffer header spinlock still held.
*/
volatile BufferDesc *
StrategyGetBuffer(BufferAccessStrategy strategy)
{
volatile BufferDesc *buf;
int bgwprocno;
int trycounter;
/*
* If given a strategy object, see whether it can select a buffer. We
* assume strategy objects don't need buffer_strategy_lock.
*/
if (strategy != NULL)
{
buf = GetBufferFromRing(strategy);
if (buf != NULL)
return buf;
}
/*
* If asked, we need to waken the bgwriter. Since we don't want to rely on
* a spinlock for this we force a read from shared memory once, and then
* set the latch based on that value. We need to go through that length
* because otherwise bgprocno might be reset while/after we check because
* the compiler might just reread from memory.
*
* This can possibly set the latch of the wrong process if the bgwriter
* dies in the wrong moment. But since PGPROC->procLatch is never
* deallocated the worst consequence of that is that we set the latch of
* some arbitrary process.
*/
bgwprocno = INT_ACCESS_ONCE(StrategyControl->bgwprocno);
if (bgwprocno != -1)
{
/* reset bgwprocno first, before setting the latch */
StrategyControl->bgwprocno = -1;
/*
* Not acquiring ProcArrayLock here which is slightly icky. It's
* actually fine because procLatch isn't ever freed, so we just can
* potentially set the wrong process' (or no process') latch.
*/
SetLatch(&ProcGlobal->allProcs[bgwprocno].procLatch);
}
/*
* We count buffer allocation requests so that the bgwriter can estimate
* the rate of buffer consumption. Note that buffers recycled by a
* strategy object are intentionally not counted here.
*/
pg_atomic_fetch_add_u32(&StrategyControl->numBufferAllocs, 1);
/*
* First check, without acquiring the lock, whether there's buffers in the
* freelist. Since we otherwise don't require the spinlock in every
* StrategyGetBuffer() invocation, it'd be sad to acquire it here -
* uselessly in most cases. That obviously leaves a race where a buffer is
* put on the freelist but we don't see the store yet - but that's pretty
* harmless, it'll just get used during the next buffer acquisition.
*
* If there's buffers on the freelist, acquire the spinlock to pop one
* buffer of the freelist. Then check whether that buffer is usable and
* repeat if not.
*
* Note that the freeNext fields are considered to be protected by the
* buffer_strategy_lock not the individual buffer spinlocks, so it's OK to
* manipulate them without holding the spinlock.
*/
if (StrategyControl->firstFreeBuffer >= 0)
{
while (true)
{
/* Acquire the spinlock to remove element from the freelist */
SpinLockAcquire(&StrategyControl->buffer_strategy_lock);
if (StrategyControl->firstFreeBuffer < 0)
{
SpinLockRelease(&StrategyControl->buffer_strategy_lock);
break;
}
buf = GetBufferDescriptor(StrategyControl->firstFreeBuffer);
Assert(buf->freeNext != FREENEXT_NOT_IN_LIST);
/* Unconditionally remove buffer from freelist */
StrategyControl->firstFreeBuffer = buf->freeNext;
buf->freeNext = FREENEXT_NOT_IN_LIST;
/*
* Release the lock so someone else can access the freelist while
* we check out this buffer.
*/
SpinLockRelease(&StrategyControl->buffer_strategy_lock);
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot
* use it; discard it and retry. (This can only happen if VACUUM
* put a valid buffer in the freelist and then someone else used
* it before we got to it. It's probably impossible altogether as
* of 8.3, but we'd better check anyway.)
*/
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count == 0)
{
if (strategy != NULL)
AddBufferToRing(strategy, buf);
return buf;
}
UnlockBufHdr(buf);
}
}
/* Nothing on the freelist, so run the "clock sweep" algorithm */
trycounter = NBuffers;
for (;;)
{
buf = GetBufferDescriptor(ClockSweepTick());
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; decrement the usage_count (unless pinned) and keep scanning.
*/
LockBufHdr(buf);
if (buf->refcount == 0)
{
if (buf->usage_count > 0)
{
buf->usage_count--;
trycounter = NBuffers;
}
else
{
/* Found a usable buffer */
if (strategy != NULL)
AddBufferToRing(strategy, buf);
return buf;
}
}
else if (--trycounter == 0)
{
/*
* We've scanned all the buffers without making any state changes,
* so all the buffers are pinned (or were when we looked at them).
* We could hope that someone will free one eventually, but it's
* probably better to fail than to risk getting stuck in an
* infinite loop.
*/
UnlockBufHdr(buf);
elog(ERROR, "no unpinned buffers available");
}
UnlockBufHdr(buf);
}
}
/*
* StrategyGetBuffer
*
* Called by the bufmgr to get the next candidate buffer to use in
* BufferAlloc(). The only hard requirement BufferAlloc() has is that
* the selected buffer must not currently be pinned by anyone.
*
* strategy is a BufferAccessStrategy object, or NULL for default strategy.
*
* To ensure that no one else can pin the buffer before we do, we must
* return the buffer with the buffer header spinlock still held. If
* *lock_held is set on exit, we have returned with the BufFreelistLock
* still held, as well; the caller must release that lock once the spinlock
* is dropped. We do it that way because releasing the BufFreelistLock
* might awaken other processes, and it would be bad to do the associated
* kernel calls while holding the buffer header spinlock.
*/
volatile BufferDesc *
StrategyGetBuffer(BufferAccessStrategy strategy, bool *lock_held)
{
volatile BufferDesc *buf;
volatile int bufIndex = -1;
volatile int resultIndex = -1;
int trycounter;
/* Lock the freelist */
*lock_held = true;
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
/*
* We count buffer allocation requests so that the bgwriter can estimate
* the rate of buffer consumption.
*/
StrategyControl->numBufferAllocs++;
/*
* Try to get a buffer from the freelist. Note that the freeNext fields
* are considered to be protected by the BufFreelistLock not the
* individual buffer spinlocks, so it's OK to manipulate them without
* holding the spinlock.
*/
while (StrategyControl->firstFreeBuffer >= 0)
{
bufIndex = StrategyControl->firstFreeBuffer;
buf = &BufferDescriptors[bufIndex];
Assert(buf->freeNext != FREENEXT_NOT_IN_LIST);
/* Unconditionally remove buffer from freelist */
StrategyControl->firstFreeBuffer = buf->freeNext;
buf->freeNext = FREENEXT_NOT_IN_LIST;
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; discard it and retry. (This can only happen if VACUUM put a
* valid buffer in the freelist and then someone else used it before
* we got to it. It's probably impossible altogether as of 8.3, but
* we'd better check anyway.)
*/
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count == 0)
{
resultIndex = bufIndex;
break;
}
UnlockBufHdr(buf);
}
/*
* Nothing on the freelist, so use the buffer replacement policy
* to select a buffer to evict.
*/
if (resultIndex == -1)
{
if (BufferReplacementPolicy == POLICY_CLOCK)
{
/* Run the "clock sweep" algorithm */
trycounter = NBuffers;
for (;;)
{
bufIndex = StrategyControl->nextVictimBuffer;
buf = &BufferDescriptors[bufIndex];
/*
* If the clock sweep hand has reached the end of the
* buffer pool, start back at the beginning.
*/
if (++StrategyControl->nextVictimBuffer >= NBuffers)
{
StrategyControl->nextVictimBuffer = 0;
StrategyControl->completePasses++;
}
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; decrement the usage_count (unless pinned) and keep scanning.
*/
LockBufHdr(buf);
if (buf->refcount == 0)
{
if (buf->usage_count > 0)
{
buf->usage_count--;
trycounter = NBuffers;
}
else
{
/* Found a usable buffer */
resultIndex = bufIndex;
break;
}
}
else if (--trycounter == 0)
{
/*
* We've scanned all the buffers without making any state changes,
* so all the buffers are pinned (or were when we looked at them).
* We could hope that someone will free one eventually, but it's
* probably better to fail than to risk getting stuck in an
* infinite loop.
*/
UnlockBufHdr(buf);
elog(ERROR, "no unpinned buffers available");
}
UnlockBufHdr(buf);
}
}
/*
* CS186 TODO: Add code here to implement the LRU, MRU and 2Q buffer
* replacement policies. Once you've selected a buffer to
* evict, assign its index in the BufferDescriptors array to
* "resultIndex". You can model your code on the CLOCK code
* above.
*/
else if (BufferReplacementPolicy == POLICY_LRU)
{
resultIndex = StrategyHelper(&buf, &(StrategyControl->head), &(StrategyControl->tail));
if(resultIndex == -1)
elog(ERROR, "no unpinned buffers available");
}
else if (BufferReplacementPolicy == POLICY_MRU)
{
buf = StrategyControl->head;
if(buf) {
// Deleting the tail ie find refcount == 0 on first try
if(buf->refcount == 0) {
resultIndex = buf->buf_id;
if (StrategyControl->head == StrategyControl->tail) {
StrategyControl->head = NULL;
StrategyControl->tail = NULL;
}
else {
buf->prev->next = buf->next;
buf->next->prev = buf->prev;
StrategyControl->head = buf->next;
}
buf->prev = NULL;
buf->next = NULL;
buf->queueTag = 0;
}
else {
buf = StrategyControl->head->next;
while (buf != StrategyControl->head) {
if(buf->refcount == 0) {
resultIndex = buf->buf_id;
buf->prev->next = buf->next;
buf->next->prev = buf->prev;
if (buf == StrategyControl->tail) {
StrategyControl->tail = buf->prev;
}
buf->prev = NULL;
buf->next = NULL;
buf->queueTag = 0;
break;
}
buf = buf->next;
}
}
}
if(resultIndex == -1)
elog(ERROR, "no unpinned buffers available");
}
else if (BufferReplacementPolicy == POLICY_2Q)
{
int threshhold;
threshhold = NBuffers/2 ;
if (size_of_list(&(StrategyControl->head2Q)) >= threshhold || !(StrategyControl->head)) {
resultIndex = StrategyHelper(&buf, &(StrategyControl->head2Q), &(StrategyControl->tail2Q));
}
else {
resultIndex = StrategyHelper(&buf, &(StrategyControl->head), &(StrategyControl->tail));
}
if (resultIndex == -1) {
elog(ERROR, "no unpinned buffers available");
}
}
else
{
elog(ERROR, "invalid buffer pool replacement policy %d", BufferReplacementPolicy);
}
/*
* CS186 Grading LOG - DON'T TOUCH
* Don't output logs starting with "GRADING" by yourself; they are for grading purposes only.
*/
elog(LOG, "GRADING: EVICT %2d", resultIndex);
}
if (resultIndex == -1)
elog(ERROR, "reached end of StrategyGetBuffer() without selecting a buffer");
return &BufferDescriptors[resultIndex];
}
/*
* StrategyGetBuffer
*
* Called by the bufmgr to get the next candidate buffer to use in
* BufferAlloc(). The only hard requirement BufferAlloc() has is that
* the selected buffer must not currently be pinned by anyone.
*
* To ensure that no one else can pin the buffer before we do, we must
* return the buffer with the buffer header spinlock still held. That
* means that we return with the BufFreelistLock still held, as well;
* the caller must release that lock once the spinlock is dropped.
*/
volatile BufferDesc *
StrategyGetBuffer(void)
{
volatile BufferDesc *buf;
int trycounter;
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
/*
* Try to get a buffer from the freelist. Note that the freeNext fields
* are considered to be protected by the BufFreelistLock not the
* individual buffer spinlocks, so it's OK to manipulate them without
* holding the spinlock.
*/
while (StrategyControl->firstFreeBuffer >= 0)
{
buf = &BufferDescriptors[StrategyControl->firstFreeBuffer];
Assert(buf->freeNext != FREENEXT_NOT_IN_LIST);
/* Unconditionally remove buffer from freelist */
StrategyControl->firstFreeBuffer = buf->freeNext;
buf->freeNext = FREENEXT_NOT_IN_LIST;
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; discard it and retry. (This can only happen if VACUUM put a
* valid buffer in the freelist and then someone else used it before
* we got to it.)
*/
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count == 0)
return buf;
UnlockBufHdr(buf);
}
/* Nothing on the freelist, so run the "clock sweep" algorithm */
trycounter = NBuffers;
for (;;)
{
buf = &BufferDescriptors[StrategyControl->nextVictimBuffer];
if (++StrategyControl->nextVictimBuffer >= NBuffers)
StrategyControl->nextVictimBuffer = 0;
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; decrement the usage_count and keep scanning.
*/
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count == 0)
return buf;
if (buf->usage_count > 0)
{
buf->usage_count--;
trycounter = NBuffers;
}
else if (--trycounter == 0)
{
/*
* We've scanned all the buffers without making any state changes,
* so all the buffers are pinned (or were when we looked at them).
* We could hope that someone will free one eventually, but it's
* probably better to fail than to risk getting stuck in an
* infinite loop.
*/
UnlockBufHdr(buf);
elog(ERROR, "no unpinned buffers available");
}
UnlockBufHdr(buf);
}
/* not reached */
return NULL;
}
Datum
pg_buffercache_pages(PG_FUNCTION_ARGS)
{
FuncCallContext *funcctx;
Datum result;
MemoryContext oldcontext;
BufferCachePagesContext *fctx; /* User function context. */
TupleDesc tupledesc;
HeapTuple tuple;
if (SRF_IS_FIRSTCALL())
{
int i;
volatile BufferDesc *bufHdr;
funcctx = SRF_FIRSTCALL_INIT();
/* Switch context when allocating stuff to be used in later calls */
oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);
/* Create a user function context for cross-call persistence */
fctx = (BufferCachePagesContext *) palloc(sizeof(BufferCachePagesContext));
/* Construct a tuple descriptor for the result rows. */
tupledesc = CreateTemplateTupleDesc(NUM_BUFFERCACHE_PAGES_ELEM, false);
TupleDescInitEntry(tupledesc, (AttrNumber) 1, "bufferid",
INT4OID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 2, "relfilenode",
OIDOID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 3, "reltablespace",
OIDOID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 4, "reldatabase",
OIDOID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 5, "relforknumber",
INT2OID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 6, "relblocknumber",
INT8OID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 7, "isdirty",
BOOLOID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 8, "usage_count",
INT2OID, -1, 0);
fctx->tupdesc = BlessTupleDesc(tupledesc);
/* Allocate NBuffers worth of BufferCachePagesRec records. */
fctx->record = (BufferCachePagesRec *) palloc(sizeof(BufferCachePagesRec) * NBuffers);
/* Set max calls and remember the user function context. */
funcctx->max_calls = NBuffers;
funcctx->user_fctx = fctx;
/* Return to original context when allocating transient memory */
MemoryContextSwitchTo(oldcontext);
/*
* To get a consistent picture of the buffer state, we must lock all
* partitions of the buffer map. Needless to say, this is horrible
* for concurrency. Must grab locks in increasing order to avoid
* possible deadlocks.
*/
for (i = 0; i < NUM_BUFFER_PARTITIONS; i++)
LWLockAcquire(FirstBufMappingLock + i, LW_SHARED);
/*
* Scan though all the buffers, saving the relevant fields in the
* fctx->record structure.
*/
for (i = 0, bufHdr = BufferDescriptors; i < NBuffers; i++, bufHdr++)
{
/* Lock each buffer header before inspecting. */
LockBufHdr(bufHdr);
fctx->record[i].bufferid = BufferDescriptorGetBuffer(bufHdr);
fctx->record[i].relfilenode = bufHdr->tag.rnode.relNode;
fctx->record[i].reltablespace = bufHdr->tag.rnode.spcNode;
fctx->record[i].reldatabase = bufHdr->tag.rnode.dbNode;
fctx->record[i].forknum = bufHdr->tag.forkNum;
fctx->record[i].blocknum = bufHdr->tag.blockNum;
fctx->record[i].usagecount = bufHdr->usage_count;
if (bufHdr->flags & BM_DIRTY)
fctx->record[i].isdirty = true;
else
fctx->record[i].isdirty = false;
/* Note if the buffer is valid, and has storage created */
if ((bufHdr->flags & BM_VALID) && (bufHdr->flags & BM_TAG_VALID))
fctx->record[i].isvalid = true;
else
fctx->record[i].isvalid = false;
UnlockBufHdr(bufHdr);
}
/*
* And release locks. We do this in reverse order for two reasons:
* (1) Anyone else who needs more than one of the locks will be trying
* to lock them in increasing order; we don't want to release the
* other process until it can get all the locks it needs. (2) This
* avoids O(N^2) behavior inside LWLockRelease.
*/
for (i = NUM_BUFFER_PARTITIONS; --i >= 0;)
LWLockRelease(FirstBufMappingLock + i);
}
funcctx = SRF_PERCALL_SETUP();
/* Get the saved state */
fctx = funcctx->user_fctx;
if (funcctx->call_cntr < funcctx->max_calls)
{
uint32 i = funcctx->call_cntr;
Datum values[NUM_BUFFERCACHE_PAGES_ELEM];
bool nulls[NUM_BUFFERCACHE_PAGES_ELEM];
values[0] = Int32GetDatum(fctx->record[i].bufferid);
nulls[0] = false;
/*
* Set all fields except the bufferid to null if the buffer is unused
* or not valid.
*/
if (fctx->record[i].blocknum == InvalidBlockNumber ||
fctx->record[i].isvalid == false)
{
nulls[1] = true;
nulls[2] = true;
nulls[3] = true;
nulls[4] = true;
nulls[5] = true;
nulls[6] = true;
nulls[7] = true;
}
else
{
values[1] = ObjectIdGetDatum(fctx->record[i].relfilenode);
nulls[1] = false;
values[2] = ObjectIdGetDatum(fctx->record[i].reltablespace);
nulls[2] = false;
values[3] = ObjectIdGetDatum(fctx->record[i].reldatabase);
nulls[3] = false;
values[4] = ObjectIdGetDatum(fctx->record[i].forknum);
nulls[4] = false;
values[5] = Int64GetDatum((int64) fctx->record[i].blocknum);
nulls[5] = false;
values[6] = BoolGetDatum(fctx->record[i].isdirty);
nulls[6] = false;
values[7] = Int16GetDatum(fctx->record[i].usagecount);
nulls[7] = false;
}
/* Build and return the tuple. */
tuple = heap_form_tuple(fctx->tupdesc, values, nulls);
result = HeapTupleGetDatum(tuple);
SRF_RETURN_NEXT(funcctx, result);
}
else
SRF_RETURN_DONE(funcctx);
}
Datum
pg_buffercache_pages(PG_FUNCTION_ARGS)
{
FuncCallContext *funcctx;
Datum result;
MemoryContext oldcontext;
BufferCachePagesContext *fctx; /* User function context. */
TupleDesc tupledesc;
TupleDesc expected_tupledesc;
HeapTuple tuple;
if (SRF_IS_FIRSTCALL())
{
int i;
funcctx = SRF_FIRSTCALL_INIT();
/* Switch context when allocating stuff to be used in later calls */
oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);
/* Create a user function context for cross-call persistence */
fctx = (BufferCachePagesContext *) palloc(sizeof(BufferCachePagesContext));
/*
* To smoothly support upgrades from version 1.0 of this extension
* transparently handle the (non-)existence of the pinning_backends
* column. We unfortunately have to get the result type for that... -
* we can't use the result type determined by the function definition
* without potentially crashing when somebody uses the old (or even
* wrong) function definition though.
*/
if (get_call_result_type(fcinfo, NULL, &expected_tupledesc) != TYPEFUNC_COMPOSITE)
elog(ERROR, "return type must be a row type");
if (expected_tupledesc->natts < NUM_BUFFERCACHE_PAGES_MIN_ELEM ||
expected_tupledesc->natts > NUM_BUFFERCACHE_PAGES_ELEM)
elog(ERROR, "incorrect number of output arguments");
/* Construct a tuple descriptor for the result rows. */
tupledesc = CreateTemplateTupleDesc(expected_tupledesc->natts, false);
TupleDescInitEntry(tupledesc, (AttrNumber) 1, "bufferid",
INT4OID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 2, "relfilenode",
OIDOID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 3, "reltablespace",
OIDOID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 4, "reldatabase",
OIDOID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 5, "relforknumber",
INT2OID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 6, "relblocknumber",
INT8OID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 7, "isdirty",
BOOLOID, -1, 0);
TupleDescInitEntry(tupledesc, (AttrNumber) 8, "usage_count",
INT2OID, -1, 0);
if (expected_tupledesc->natts == NUM_BUFFERCACHE_PAGES_ELEM)
TupleDescInitEntry(tupledesc, (AttrNumber) 9, "pinning_backends",
INT4OID, -1, 0);
fctx->tupdesc = BlessTupleDesc(tupledesc);
/* Allocate NBuffers worth of BufferCachePagesRec records. */
fctx->record = (BufferCachePagesRec *)
MemoryContextAllocHuge(CurrentMemoryContext,
sizeof(BufferCachePagesRec) * NBuffers);
/* Set max calls and remember the user function context. */
funcctx->max_calls = NBuffers;
funcctx->user_fctx = fctx;
/* Return to original context when allocating transient memory */
MemoryContextSwitchTo(oldcontext);
/*
* Scan through all the buffers, saving the relevant fields in the
* fctx->record structure.
*
* We don't hold the partition locks, so we don't get a consistent
* snapshot across all buffers, but we do grab the buffer header
* locks, so the information of each buffer is self-consistent.
*/
for (i = 0; i < NBuffers; i++)
{
BufferDesc *bufHdr;
uint32 buf_state;
bufHdr = GetBufferDescriptor(i);
/* Lock each buffer header before inspecting. */
buf_state = LockBufHdr(bufHdr);
fctx->record[i].bufferid = BufferDescriptorGetBuffer(bufHdr);
fctx->record[i].relfilenode = bufHdr->tag.rnode.relNode;
fctx->record[i].reltablespace = bufHdr->tag.rnode.spcNode;
fctx->record[i].reldatabase = bufHdr->tag.rnode.dbNode;
fctx->record[i].forknum = bufHdr->tag.forkNum;
fctx->record[i].blocknum = bufHdr->tag.blockNum;
fctx->record[i].usagecount = BUF_STATE_GET_USAGECOUNT(buf_state);
fctx->record[i].pinning_backends = BUF_STATE_GET_REFCOUNT(buf_state);
if (buf_state & BM_DIRTY)
fctx->record[i].isdirty = true;
else
fctx->record[i].isdirty = false;
/* Note if the buffer is valid, and has storage created */
if ((buf_state & BM_VALID) && (buf_state & BM_TAG_VALID))
fctx->record[i].isvalid = true;
else
fctx->record[i].isvalid = false;
UnlockBufHdr(bufHdr, buf_state);
}
}
funcctx = SRF_PERCALL_SETUP();
/* Get the saved state */
fctx = funcctx->user_fctx;
if (funcctx->call_cntr < funcctx->max_calls)
{
uint32 i = funcctx->call_cntr;
Datum values[NUM_BUFFERCACHE_PAGES_ELEM];
bool nulls[NUM_BUFFERCACHE_PAGES_ELEM];
values[0] = Int32GetDatum(fctx->record[i].bufferid);
nulls[0] = false;
/*
* Set all fields except the bufferid to null if the buffer is unused
* or not valid.
*/
if (fctx->record[i].blocknum == InvalidBlockNumber ||
fctx->record[i].isvalid == false)
{
nulls[1] = true;
nulls[2] = true;
nulls[3] = true;
nulls[4] = true;
nulls[5] = true;
nulls[6] = true;
nulls[7] = true;
/* unused for v1.0 callers, but the array is always long enough */
nulls[8] = true;
}
else
{
values[1] = ObjectIdGetDatum(fctx->record[i].relfilenode);
nulls[1] = false;
values[2] = ObjectIdGetDatum(fctx->record[i].reltablespace);
nulls[2] = false;
values[3] = ObjectIdGetDatum(fctx->record[i].reldatabase);
nulls[3] = false;
values[4] = ObjectIdGetDatum(fctx->record[i].forknum);
nulls[4] = false;
values[5] = Int64GetDatum((int64) fctx->record[i].blocknum);
nulls[5] = false;
values[6] = BoolGetDatum(fctx->record[i].isdirty);
nulls[6] = false;
values[7] = Int16GetDatum(fctx->record[i].usagecount);
nulls[7] = false;
/* unused for v1.0 callers, but the array is always long enough */
values[8] = Int32GetDatum(fctx->record[i].pinning_backends);
nulls[8] = false;
}
/* Build and return the tuple. */
tuple = heap_form_tuple(fctx->tupdesc, values, nulls);
result = HeapTupleGetDatum(tuple);
SRF_RETURN_NEXT(funcctx, result);
}
else
SRF_RETURN_DONE(funcctx);
}
static void
SaveBuffers(void)
{
int i;
int num_buffers;
int log_level = DEBUG3;
SavedBuffer *saved_buffers;
volatile BufferDesc *bufHdr; // XXX: Do we really need volatile here?
FILE *file = NULL;
int database_counter= 0;
Oid prev_database = InvalidOid;
Oid prev_filenode = InvalidOid;
ForkNumber prev_forknum = InvalidForkNumber;
BlockNumber prev_blocknum = InvalidBlockNumber;
BlockNumber range_counter = 0;
const char *savefile_path;
/*
* XXX: If the memory request fails, ask for a smaller memory chunk, and use
* it to create chunks of save-files, and make the workers read those chunks.
*
* This is not a concern as of now, so deferred; there's at least one other
* place that allocates (NBuffers * (much_bigger_struct)), so this seems to
* be an acceptable practice.
*/
saved_buffers = (SavedBuffer *) palloc(sizeof(SavedBuffer) * NBuffers);
/* Lock the buffer partitions for reading. */
for (i = 0; i < NUM_BUFFER_PARTITIONS; ++i)
LWLockAcquire(FirstBufMappingLock + i, LW_SHARED);
/* Scan and save a list of valid buffers. */
for (num_buffers = 0, i = 0, bufHdr = BufferDescriptors; i < NBuffers; ++i, ++bufHdr)
{
/* Lock each buffer header before inspecting. */
LockBufHdr(bufHdr);
/* Skip invalid buffers */
if ((bufHdr->flags & BM_VALID) && (bufHdr->flags & BM_TAG_VALID))
{
saved_buffers[num_buffers].database = bufHdr->tag.rnode.dbNode;
saved_buffers[num_buffers].filenode = bufHdr->tag.rnode.relNode;
saved_buffers[num_buffers].forknum = bufHdr->tag.forkNum;
saved_buffers[num_buffers].blocknum = bufHdr->tag.blockNum;
++num_buffers;
}
UnlockBufHdr(bufHdr);
}
/* Unlock the buffer partitions in reverse order, to avoid a deadlock. */
for (i = NUM_BUFFER_PARTITIONS - 1; i >= 0; --i)
LWLockRelease(FirstBufMappingLock + i);
/*
* Sort the list, so that we can optimize the storage of these buffers.
*
* The side-effect of this storage optimization is that when reading the
* blocks back from relation forks, it leads to sequential reads, which
* improve the restore speeds quite considerably as compared to random reads
* from different blocks all over the data directory.
*/
pg_qsort(saved_buffers, num_buffers, sizeof(SavedBuffer), SavedBufferCmp);
/* Connect to the database and start a transaction for database name lookups. */
BackgroundWorkerInitializeConnection(guc_default_database, NULL);
SetCurrentStatementStartTimestamp();
StartTransactionCommand();
PushActiveSnapshot(GetTransactionSnapshot());
pgstat_report_activity(STATE_RUNNING, "saving buffers");
for (i = 0; i < num_buffers; ++i)
{
int j;
SavedBuffer *buf = &saved_buffers[i];
if (i == 0)
{
/*
* Special case for global objects. The sort brings them to the
* front of the list.
*/
/* Make sure the first buffer we save belongs to global object. */
Assert(buf->database == InvalidOid);
/*
* Database number (and save-file name) 1 is reserverd for storing
* list of buffers of global objects.
*/
database_counter = 1;
savefile_path = getSavefileName(database_counter);
file = fileOpen(savefile_path, PG_BINARY_W);
writeDBName("", file, savefile_path);
prev_database = buf->database;
}
if (buf->database != prev_database)
{
char *dbname;
/*
* We are beginning to process a different database than the
* previous one; close the save-file of previous database, and open
* a new one.
*/
++database_counter;
dbname = get_database_name(buf->database);
Assert(dbname != NULL);
if (file != NULL)
fileClose(file, savefile_path);
savefile_path = getSavefileName(database_counter);
file = fileOpen(savefile_path, PG_BINARY_W);
writeDBName(dbname, file, savefile_path);
pfree(dbname);
/* Reset trackers appropriately */
prev_database = buf->database;
prev_filenode = InvalidOid;
prev_forknum = InvalidForkNumber;
prev_blocknum = InvalidBlockNumber;
range_counter = 0;
}
if (buf->filenode != prev_filenode)
{
/* We're beginning to process a new relation; emit a record for it. */
fileWrite("r", 1, file, savefile_path);
fileWrite(&(buf->filenode), sizeof(Oid), file, savefile_path);
/* Reset trackers appropriately */
prev_filenode = buf->filenode;
prev_forknum = InvalidForkNumber;
prev_blocknum = InvalidBlockNumber;
range_counter = 0;
}
if (buf->forknum != prev_forknum)
{
/*
* We're beginning to process a new fork of this relation; add a
* record for it.
*/
fileWrite("f", 1, file, savefile_path);
fileWrite(&(buf->forknum), sizeof(ForkNumber), file, savefile_path);
/* Reset trackers appropriately */
prev_forknum = buf->forknum;
prev_blocknum = InvalidBlockNumber;
range_counter = 0;
}
ereport(log_level,
(errmsg("writer: writing block db %d filenode %d forknum %d blocknum %d",
database_counter, prev_filenode, prev_forknum, buf->blocknum)));
fileWrite("b", 1, file, savefile_path);
fileWrite(&(buf->blocknum), sizeof(BlockNumber), file, savefile_path);
prev_blocknum = buf->blocknum;
/*
* If a continuous range of blocks follows this block, then emit one
* entry for the range, instead of one for each block.
*/
range_counter = 0;
for ( j = i+1; j < num_buffers; ++j)
{
SavedBuffer *tmp = &saved_buffers[j];
if (tmp->database == prev_database
&& tmp->filenode == prev_filenode
&& tmp->forknum == prev_forknum
&& tmp->blocknum == (prev_blocknum + range_counter + 1))
{
++range_counter;
}
}
if (range_counter != 0)
{
ereport(log_level,
(errmsg("writer: writing range db %d filenode %d forknum %d blocknum %d range %d",
database_counter, prev_filenode, prev_forknum, prev_blocknum, range_counter)));
fileWrite("N", 1, file, savefile_path);
fileWrite(&range_counter, sizeof(range_counter), file, savefile_path);
i += range_counter;
}
}
ereport(LOG,
(errmsg("Buffer Saver: saved metadata of %d blocks", num_buffers)));
Assert(file != NULL);
fileClose(file, savefile_path);
pfree(saved_buffers);
PopActiveSnapshot();
CommitTransactionCommand();
pgstat_report_activity(STATE_IDLE, NULL);
}
/*
* StrategyGetBuffer
*
* Called by the bufmgr to get the next candidate buffer to use in
* BufferAlloc(). The only hard requirement BufferAlloc() has is that
* the selected buffer must not currently be pinned by anyone.
*
* strategy is a BufferAccessStrategy object, or NULL for default strategy.
*
* To ensure that no one else can pin the buffer before we do, we must
* return the buffer with the buffer header spinlock still held. If
* *lock_held is set on exit, we have returned with the BufFreelistLock
* still held, as well; the caller must release that lock once the spinlock
* is dropped. We do it that way because releasing the BufFreelistLock
* might awaken other processes, and it would be bad to do the associated
* kernel calls while holding the buffer header spinlock.
*/
volatile BufferDesc *
StrategyGetBuffer(BufferAccessStrategy strategy, bool *lock_held)
{
volatile BufferDesc *buf;
Latch *bgwriterLatch;
int trycounter;
elog(LOG, "in StrategyGetBuffer");
// fprintf(stderr, "StrategyGetBuffer\n");
/*
* If given a strategy object, see whether it can select a buffer. We
* assume strategy objects don't need the BufFreelistLock.
*/
/* Nope, so lock the freelist */
*lock_held = true;
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
LWLockAcquire(BufDllLock, LW_EXCLUSIVE);
/*
* We count buffer allocation requests so that the bgwriter can estimate
* the rate of buffer consumption. Note that buffers recycled by a
* strategy object are intentionally not counted here.
*/
StrategyControl->numBufferAllocs++;
/*
* If bgwriterLatch is set, we need to waken the bgwriter, but we should
* not do so while holding BufFreelistLock; so release and re-grab. This
* is annoyingly tedious, but it happens at most once per bgwriter cycle,
* so the performance hit is minimal.
*/
bgwriterLatch = StrategyControl->bgwriterLatch;
if (bgwriterLatch)
{
StrategyControl->bgwriterLatch = NULL;
LWLockRelease(BufFreelistLock);
SetLatch(bgwriterLatch);
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
}
/*
* Try to get a buffer from the freelist. Note that the freeNext fields
* are considered to be protected by the BufFreelistLock not the
* individual buffer spinlocks, so it's OK to manipulate them without
* holding the spinlock.
*/
while (StrategyControl->firstFreeBuffer >= 0)
{
buf = &BufferDescriptors[StrategyControl->firstFreeBuffer];
Assert(buf->freeNext != FREENEXT_NOT_IN_LIST);
/* Unconditionally remove buffer from freelist */
StrategyControl->firstFreeBuffer = buf->freeNext;
buf->freeNext = FREENEXT_NOT_IN_LIST;
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; discard it and retry. (This can only happen if VACUUM put a
* valid buffer in the freelist and then someone else used it before
* we got to it. It's probably impossible altogether as of 8.3, but
* we'd better check anyway.)
*/
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count == 0)
{
BufNodes[StrategyControl->firstFreeBuffer] =
dllInsertInt(BufDLL, StrategyControl->firstFreeBuffer, TAIL);
LWLockRelease(BufDllLock);
// elog(ERROR, "Successfully got buf from list");
return buf;
}
UnlockBufHdr(buf);
}
/* Nothing on the freelist, so run the "clock sweep" algorithm */
DllNode *d = BufDLL->head;
for (; d; d = d->next)
{
buf = &BufferDescriptors[d->data];
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; decrement the usage_count (unless pinned) and keep scanning.
*/
LockBufHdr(buf);
if (buf->refcount == 0)
{
/* Found a usable buffer */
dllMove(BufDLL, buf, TAIL);
LWLockRelease(BufDllLock);
elog(LOG, "Successfully got buf from list");
return buf;
}
UnlockBufHdr(buf);
}
/*
* We've scanned all the buffers without making any state changes,
* so all the buffers are pinned (or were when we looked at them).
* We could hope that someone will free one eventually, but it's
* probably better to fail than to risk getting stuck in an
* infinite loop.
*/
UnlockBufHdr(buf);
elog(ERROR, "no unpinned buffers available");
/* not reached */
LWLockRelease(BufDllLock);
return NULL;
}
/*
* StrategyGetBuffer
*
* Called by the bufmgr to get the next candidate buffer to use in
* BufferAlloc(). The only hard requirement BufferAlloc() has is that
* the selected buffer must not currently be pinned by anyone.
*
* strategy is a BufferAccessStrategy object, or NULL for default strategy.
*
* To ensure that no one else can pin the buffer before we do, we must
* return the buffer with the buffer header spinlock still held. If
* *lock_held is set on exit, we have returned with the BufFreelistLock
* still held, as well; the caller must release that lock once the spinlock
* is dropped. We do it that way because releasing the BufFreelistLock
* might awaken other processes, and it would be bad to do the associated
* kernel calls while holding the buffer header spinlock.
*/
volatile BufferDesc *
StrategyGetBuffer(BufferAccessStrategy strategy, bool *lock_held)
{
volatile BufferDesc *buf;
Latch *bgwriterLatch;
int trycounter;
volatile int bufIndex = -1;
volatile int resultIndex = -1;
volatile BufferDesc *next;
volatile BufferDesc *previous;
/*
* If given a strategy object, see whether it can select a buffer. We
* assume strategy objects don't need the BufFreelistLock.
*/
if (strategy != NULL)
{
buf = GetBufferFromRing(strategy);
if (buf != NULL)
{
*lock_held = false;
return buf;
}
}
/* Nope, so lock the freelist */
*lock_held = true;
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
/*
* We count buffer allocation requests so that the bgwriter can estimate
* the rate of buffer consumption. Note that buffers recycled by a
* strategy object are intentionally not counted here.
*/
StrategyControl->numBufferAllocs++;
/*
* If bgwriterLatch is set, we need to waken the bgwriter, but we should
* not do so while holding BufFreelistLock; so release and re-grab. This
* is annoyingly tedious, but it happens at most once per bgwriter cycle,
* so the performance hit is minimal.
*/
bgwriterLatch = StrategyControl->bgwriterLatch;
if (bgwriterLatch)
{
StrategyControl->bgwriterLatch = NULL;
LWLockRelease(BufFreelistLock);
SetLatch(bgwriterLatch);
LWLockAcquire(BufFreelistLock, LW_EXCLUSIVE);
}
/*
* Try to get a buffer from the freelist. Note that the freeNext fields
* are considered to be protected by the BufFreelistLock not the
* individual buffer spinlocks, so it's OK to manipulate them without
* holding the spinlock.
*/
while (StrategyControl->firstFreeBuffer >= 0)
{
buf = &BufferDescriptors[StrategyControl->firstFreeBuffer];
Assert(buf->freeNext != FREENEXT_NOT_IN_LIST);
/* Unconditionally remove buffer from freelist */
StrategyControl->firstFreeBuffer = buf->freeNext;
buf->freeNext = FREENEXT_NOT_IN_LIST;
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; discard it and retry. (This can only happen if VACUUM put a
* valid buffer in the freelist and then someone else used it before
* we got to it. It's probably impossible altogether as of 8.3, but
* we'd better check anyway.)
*/
LockBufHdr(buf);
if (buf->refcount == 0 && buf->usage_count == 0)
{
if (strategy != NULL)
AddBufferToRing(strategy, buf);
return buf;
}
UnlockBufHdr(buf);
}
/* Nothing on the freelist, so run the algorithm defined in the
* BufferReplacementPolicy variable*/
if (resultIndex == -1)
{
if (BufferReplacementPolicy == POLICY_CLOCK)
{
//Running the Clock Sweep Algorithm (Default postgres Algo.)
trycounter = NBuffers;
for (;;)
{
buf = &BufferDescriptors[StrategyControl->nextVictimBuffer];
/*
* If the clock sweep hand has reached the end of the
* buffer pool, start back at the beginning.
*/
if (++StrategyControl->nextVictimBuffer >= NBuffers)
{
StrategyControl->nextVictimBuffer = 0;
StrategyControl->completePasses++;
}
/*
* If the buffer is pinned or has a nonzero usage_count, we cannot use
* it; decrement the usage_count (unless pinned) and keep scanning.
*/
LockBufHdr(buf);
if (buf->refcount == 0)
{
if (buf->usage_count > 0)
{
buf->usage_count--;
trycounter = NBuffers;
}
else
{
/* Found a usable buffer */
if (strategy != NULL)
AddBufferToRing(strategy, buf);
return buf;
}
}
else if (--trycounter == 0)
{
/*
* We've scanned all the buffers without making any state changes,
* so all the buffers are pinned (or were when we looked at them).
* We could hope that someone will free one eventually, but it's
* probably better to fail than to risk getting stuck in an
* infinite loop.
*/
UnlockBufHdr(buf);
elog(ERROR, "no unpinned buffers available");
}
UnlockBufHdr(buf);
}
}
/* Implementation of LRU, MRU and 2Q Algorithms.
* Once we've selected a buffer to evict, its index in
* the BufferDescriptors array is stored in "resultIndex" */
else if (BufferReplacementPolicy == POLICY_LRU)
{
buf = StrategyControl->firstUnpinned;
while (buf != NULL) {
LockBufHdr(buf);
if (buf->refcount == 0) {
resultIndex = buf->buf_id;
break;
} else {
UnlockBufHdr(buf);
buf = buf->next;
}
}
/*
* We've scanned all the buffers without making any state changes,
* so all the buffers are pinned (or were when we looked at them).
* We could hope that someone will free one eventually, but it's
* probably better to fail than to risk getting stuck in an
* infinite loop.
*/
if (buf == NULL) {
UnlockBufHdr(buf); //p added 10/24
elog(ERROR, "no unpinned buffers available");
}
}
else if (BufferReplacementPolicy == POLICY_MRU)
{
buf = StrategyControl->lastUnpinned;
while (buf != NULL) {
LockBufHdr(buf);
if (buf->refcount == 0) {
resultIndex = buf->buf_id;
break;
} else {
UnlockBufHdr(buf);
buf = buf->previous;
}
}
/*
* We've scanned all the buffers without making any state changes,
* so all the buffers are pinned (or were when we looked at them).
* We could hope that someone will free one eventually, but it's
* probably better to fail than to risk getting stuck in an
* infinite loop.
*/
if (buf == NULL) {
UnlockBufHdr(buf); //p added 10/24
elog(ERROR, "no unpinned buffers available");
}
}
else if (BufferReplacementPolicy == POLICY_2Q)
{
int thres = NBuffers/2;
int sizeA1 = 0;
volatile BufferDesc *head = StrategyControl->a1Head;
while (head != NULL) {
head = head->next;
sizeA1++;
}
if (sizeA1 >= thres || StrategyControl->lastUnpinned == NULL) {
buf = StrategyControl->a1Head;
while (buf != NULL) {
if (buf->refcount == 0) {
resultIndex = buf->buf_id;
next = buf->next;
previous = buf->previous;
//adjust neighbors
if (next != NULL) {
if (previous != NULL) { //next and prev != null, buf is already in middle of list
previous->next = next;
next->previous = previous;
} else { //next != null, prev == null, buf is at beginning of list
next->previous = NULL;
StrategyControl->a1Head = next;
}
} else if (previous == NULL) { //next == NULL, prev == null, buf is only item in list
StrategyControl->a1Head = NULL;
StrategyControl->a1Tail = NULL;
} else { //buf is last item in list, next == null, prev != null
StrategyControl->a1Tail = previous;
previous->next = NULL;
}
buf->next = NULL;
buf->previous = NULL;
break;
} else {
buf = buf->next;
}
}
if (buf == NULL) {
elog(ERROR, "no unpinned buffers available");
}
} else { // delete from the head of AM
buf = StrategyControl->firstUnpinned;
while (buf != NULL) {
// LockBufHdr(buf);
if (buf->refcount == 0) {
resultIndex = buf->buf_id;
next = buf->next;
previous = buf->previous;
//adjust neighbors
if (next != NULL) {
if (previous != NULL) { //next and prev != null, buf is already in middle of list
previous->next = next;
next->previous = previous;
} else { //next != null, prev == null, buf is at beginning of list
next->previous = NULL;
StrategyControl->firstUnpinned = next;
}
} else if (previous == NULL) { //next == NULL, prev == null, buf is new to list
StrategyControl->firstUnpinned = NULL;
StrategyControl->lastUnpinned = NULL;
} else {
previous->next = NULL;
StrategyControl->lastUnpinned = previous;
}
buf->next = NULL;
buf->previous = NULL;
break;
} else {
// UnlockBufHdr(buf);
buf = buf->next;
}
// UnlockBufHdr(buf); //phoebe 10/24
}
if (buf == NULL) {
// UnlockBufHdr(buf);
elog(ERROR, "no unpinned buffers available");
}
}
}
else
{
elog(ERROR, "invalid buffer pool replacement policy %d", BufferReplacementPolicy);
}
}
if (resultIndex == -1)
{
elog(ERROR, "reached end of StrategyGetBuffer() without selecting a buffer");
}
return &BufferDescriptors[resultIndex];
}