[haiku-commits] r33807 - haiku/trunk/src/system/kernel/vm
- From: axeld@xxxxxxxxxxxxxxxx
- To: haiku-commits@xxxxxxxxxxxxx
- Date: Wed, 28 Oct 2009 11:44:30 +0100 (CET)
Author: axeld
Date: 2009-10-28 11:44:29 +0100 (Wed, 28 Oct 2009)
New Revision: 33807
Changeset: http://dev.haiku-os.org/changeset/33807/haiku
Modified:
haiku/trunk/src/system/kernel/vm/vm_page.cpp
Log:
* There is no need for the sPageLock to be a spinlock - a mutex should do. This
should improve the kernel latencies, as things like
vm_page_allocate_page_run() is very expensive.
Modified: haiku/trunk/src/system/kernel/vm/vm_page.cpp
===================================================================
--- haiku/trunk/src/system/kernel/vm/vm_page.cpp 2009-10-28 10:15:57 UTC
(rev 33806)
+++ haiku/trunk/src/system/kernel/vm/vm_page.cpp 2009-10-28 10:44:29 UTC
(rev 33807)
@@ -1,5 +1,5 @@
/*
- * Copyright 2002-2008, Axel Dörfler, axeld@xxxxxxxxxxxxxxxxx
+ * Copyright 2002-2009, Axel Dörfler, axeld@xxxxxxxxxxxxxxxxx
* Distributed under the terms of the MIT License.
*
* Copyright 2001-2002, Travis Geiselbrecht. All rights reserved.
@@ -77,7 +77,7 @@
static size_t sModifiedTemporaryPages;
static ConditionVariable sFreePageCondition;
-static spinlock sPageLock;
+static mutex sPageLock = MUTEX_INITIALIZER("pages");
static sem_id sWriterWaitSem;
@@ -851,7 +851,7 @@
if (sFreePageQueue.count == 0)
continue;
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
// Since we temporarily remove pages from the free pages
reserve,
// we must make sure we don't cause a violation of the page
@@ -933,7 +933,7 @@
return;
}
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
remove_page_from_queue(queue, &marker);
marker.state = PAGE_STATE_UNUSED;
@@ -943,7 +943,7 @@
static vm_page *
next_modified_page(struct vm_page &marker)
{
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
vm_page *page;
if (marker.state == PAGE_STATE_MODIFIED) {
@@ -1327,7 +1327,7 @@
fWrappers[checkIndex++].CheckRemoveFromShrunkenCache();
}
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
for (uint32 j = 0; j < transfer.PageCount(); j++)
fWrappers[wrapperIndex++].Done(transfer.Status());
@@ -1461,7 +1461,7 @@
continue;
}
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
// state might have changed while we were locking the
cache
if (page->state != PAGE_STATE_MODIFIED) {
@@ -1518,7 +1518,7 @@
static vm_page *
find_page_candidate(struct vm_page &marker, bool stealActive)
{
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
page_queue *queue;
vm_page *page;
@@ -1605,7 +1605,7 @@
page->cache->RemovePage(page);
- InterruptsSpinLocker _(sPageLock);
+ MutexLocker _(sPageLock);
remove_page_from_queue(page->state == PAGE_STATE_ACTIVE
? &sActivePageQueue : &sInactivePageQueue, page);
return true;
@@ -1633,7 +1633,7 @@
if (steal_page(page, false)) {
if (reserve || stolen >= maxCount) {
- InterruptsSpinLocker _(sPageLock);
+ MutexLocker _(sPageLock);
enqueue_page(&sFreePageQueue, page);
page->state = PAGE_STATE_FREE;
@@ -1649,7 +1649,7 @@
remove_page_marker(marker);
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
if ((reserve && sReservedPages <= free_page_queue_count())
|| count == 0
@@ -1743,7 +1743,7 @@
bool dequeuedPage = false;
if (page != NULL) {
if (page->state == PAGE_STATE_MODIFIED) {
- InterruptsSpinLocker locker(&sPageLock);
+ MutexLocker locker(&sPageLock);
remove_page_from_queue(&sModifiedPageQueue,
page);
dequeuedPage = true;
} else if (page->state == PAGE_STATE_BUSY
@@ -1787,7 +1787,7 @@
wrappers[i]->CheckRemoveFromShrunkenCache();
}
- InterruptsSpinLocker locker(&sPageLock);
+ MutexLocker locker(&sPageLock);
for (int32 i = 0; i < usedWrappers; i++)
wrappers[i]->Done(status);
@@ -2001,8 +2001,7 @@
return B_BAD_VALUE;
}
- cpu_status state = disable_interrupts();
- acquire_spinlock(&sPageLock);
+ MutexLocker _(sPageLock);
for (addr_t i = 0; i < length; i++) {
vm_page *page = &sPages[startPage + i];
@@ -2026,9 +2025,6 @@
}
}
- release_spinlock(&sPageLock);
- restore_interrupts(state);
-
return B_OK;
}
@@ -2044,7 +2040,7 @@
if (count == 0)
return;
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
ASSERT(sReservedPages >= count);
T(UnreservePages(count));
@@ -2067,7 +2063,7 @@
if (count == 0)
return;
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
T(ReservePages(count));
@@ -2091,7 +2087,7 @@
if (count == 0)
return true;
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
T(ReservePages(count));
@@ -2125,7 +2121,7 @@
return NULL; // invalid
}
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
T(AllocatePage(reserved));
@@ -2210,7 +2206,7 @@
vm_page *firstPage = NULL;
uint32 start = base >> PAGE_SHIFT;
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
if (free_page_queue_count() - sReservedPages < length) {
// TODO: add more tries, ie. free some inactive, ...
@@ -2266,7 +2262,7 @@
vm_page *
vm_page_allocate_page_run_no_base(int pageState, addr_t count)
{
- InterruptsSpinLocker locker(sPageLock);
+ MutexLocker locker(sPageLock);
if (free_page_queue_count() - sReservedPages < count) {
// TODO: add more tries, ie. free some inactive, ...
@@ -2370,7 +2366,7 @@
void
vm_page_free(vm_cache *cache, vm_page *page)
{
- InterruptsSpinLocker _(sPageLock);
+ MutexLocker _(sPageLock);
if (page->cache == NULL && page->state == PAGE_STATE_MODIFIED
&& cache->temporary) {
@@ -2384,7 +2380,7 @@
status_t
vm_page_set_state(vm_page *page, int pageState)
{
- InterruptsSpinLocker _(sPageLock);
+ MutexLocker _(sPageLock);
return set_page_state_nolock(page, pageState);
}
@@ -2396,7 +2392,7 @@
void
vm_page_requeue(struct vm_page *page, bool tail)
{
- InterruptsSpinLocker _(sPageLock);
+ MutexLocker _(sPageLock);
page_queue *queue = NULL;
switch (page->state) {
Other related posts:
- » [haiku-commits] r33807 - haiku/trunk/src/system/kernel/vm - axeld
