[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
#pragma once
|
|
|
|
|
|
|
|
|
|
#include <internal/arch/mm.h>
|
|
|
|
|
|
|
|
|
|
/* TYPES *********************************************************************/
|
|
|
|
|
|
2012-03-28 19:41:40 +00:00
|
|
|
#define PFN_FROM_SSE(E) ((PFN_NUMBER)((E) >> PAGE_SHIFT))
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
#define IS_SWAP_FROM_SSE(E) ((E) & 0x00000001)
|
|
|
|
|
#define MM_IS_WAIT_PTE(E) \
|
2012-03-28 18:39:21 +00:00
|
|
|
(IS_SWAP_FROM_SSE(E) && SWAPENTRY_FROM_SSE(E) == MM_WAIT_ENTRY)
|
2012-03-28 19:41:40 +00:00
|
|
|
#define MAKE_PFN_SSE(P) ((ULONG_PTR)((P) << PAGE_SHIFT))
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
#define SWAPENTRY_FROM_SSE(E) ((E) >> 1)
|
2012-03-28 19:41:40 +00:00
|
|
|
#define MAKE_SWAP_SSE(S) (((ULONG_PTR)(S) << 1) | 0x1)
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
#define DIRTY_SSE(E) ((E) | 2)
|
|
|
|
|
#define CLEAN_SSE(E) ((E) & ~2)
|
|
|
|
|
#define IS_DIRTY_SSE(E) ((E) & 2)
|
2012-03-28 19:41:40 +00:00
|
|
|
#define PAGE_FROM_SSE(E) ((E) & 0xFFFFF000)
|
|
|
|
|
#define SHARE_COUNT_FROM_SSE(E) (((E) & 0x00000FFC) >> 2)
|
|
|
|
|
#define MAX_SHARE_COUNT 0x3FF
|
|
|
|
|
#define MAKE_SSE(P, C) ((ULONG_PTR)((P) | ((C) << 2)))
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
#define MM_SEGMENT_FINALIZE (0x40000000)
|
|
|
|
|
|
2012-02-05 16:42:22 +00:00
|
|
|
#define RMAP_SEGMENT_MASK ~((ULONG_PTR)0xff)
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
#define RMAP_IS_SEGMENT(x) (((ULONG_PTR)(x) & RMAP_SEGMENT_MASK) == RMAP_SEGMENT_MASK)
|
|
|
|
|
|
|
|
|
|
#define MIN(x,y) (((x)<(y))?(x):(y))
|
|
|
|
|
#define MAX(x,y) (((x)>(y))?(x):(y))
|
|
|
|
|
|
|
|
|
|
/* Determine what's needed to make paged pool fit in this category.
|
|
|
|
|
* it seems that something more is required to satisfy arm3. */
|
|
|
|
|
#define BALANCER_CAN_EVICT(Consumer) \
|
2012-03-28 18:39:21 +00:00
|
|
|
(((Consumer) == MC_USER) || \
|
|
|
|
|
((Consumer) == MC_CACHE))
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
#define SEC_CACHE (0x20000000)
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
2012-03-28 18:39:21 +00:00
|
|
|
#define MiWaitForPageEvent(Process,Address) do { \
|
2013-08-31 16:02:13 +00:00
|
|
|
DPRINT("MiWaitForPageEvent %p:%p #\n", Process, Address); \
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
KeWaitForSingleObject(&MmWaitPageEvent, 0, KernelMode, FALSE, NULL); \
|
|
|
|
|
} while(0)
|
|
|
|
|
|
2012-03-28 18:39:21 +00:00
|
|
|
#define MiSetPageEvent(Process,Address) do { \
|
2013-02-16 17:37:17 +00:00
|
|
|
DPRINT("MiSetPageEvent %p:%p #\n",Process, (PVOID)(Address)); \
|
2012-03-28 18:39:21 +00:00
|
|
|
KeSetEvent(&MmWaitPageEvent, IO_NO_INCREMENT, FALSE); \
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
} while(0)
|
|
|
|
|
|
|
|
|
|
/* We store 8 bits of location with a page association */
|
|
|
|
|
#define ENTRIES_PER_ELEMENT 256
|
|
|
|
|
|
|
|
|
|
extern KEVENT MmWaitPageEvent;
|
|
|
|
|
|
2012-03-28 18:39:21 +00:00
|
|
|
typedef struct _CACHE_SECTION_PAGE_TABLE
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
{
|
|
|
|
|
LARGE_INTEGER FileOffset;
|
2012-03-28 18:39:21 +00:00
|
|
|
PMM_SECTION_SEGMENT Segment;
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
ULONG Refcount;
|
2012-03-28 19:41:40 +00:00
|
|
|
ULONG_PTR PageEntries[ENTRIES_PER_ELEMENT];
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
} CACHE_SECTION_PAGE_TABLE, *PCACHE_SECTION_PAGE_TABLE;
|
|
|
|
|
|
|
|
|
|
struct _MM_REQUIRED_RESOURCES;
|
|
|
|
|
|
2012-03-28 18:39:21 +00:00
|
|
|
typedef NTSTATUS (NTAPI * AcquireResource)(
|
|
|
|
|
PMMSUPPORT AddressSpace,
|
|
|
|
|
struct _MEMORY_AREA *MemoryArea,
|
|
|
|
|
struct _MM_REQUIRED_RESOURCES *Required);
|
|
|
|
|
|
|
|
|
|
typedef NTSTATUS (NTAPI * NotPresentFaultHandler)(
|
|
|
|
|
PMMSUPPORT AddressSpace,
|
|
|
|
|
struct _MEMORY_AREA *MemoryArea,
|
|
|
|
|
PVOID Address,
|
|
|
|
|
BOOLEAN Locked,
|
|
|
|
|
struct _MM_REQUIRED_RESOURCES *Required);
|
|
|
|
|
|
|
|
|
|
typedef NTSTATUS (NTAPI * FaultHandler)(
|
|
|
|
|
PMMSUPPORT AddressSpace,
|
|
|
|
|
struct _MEMORY_AREA *MemoryArea,
|
|
|
|
|
PVOID Address,
|
|
|
|
|
struct _MM_REQUIRED_RESOURCES *Required);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
typedef struct _MM_REQUIRED_RESOURCES
|
|
|
|
|
{
|
2012-03-28 18:39:21 +00:00
|
|
|
ULONG Consumer;
|
|
|
|
|
ULONG Amount;
|
|
|
|
|
ULONG Offset;
|
|
|
|
|
ULONG State;
|
|
|
|
|
PVOID Context;
|
|
|
|
|
LARGE_INTEGER FileOffset;
|
|
|
|
|
AcquireResource DoAcquisition;
|
|
|
|
|
PFN_NUMBER Page[2];
|
|
|
|
|
PVOID Buffer[2];
|
|
|
|
|
SWAPENTRY SwapEntry;
|
|
|
|
|
const char *File;
|
|
|
|
|
int Line;
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
} MM_REQUIRED_RESOURCES, *PMM_REQUIRED_RESOURCES;
|
|
|
|
|
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmCreateCacheSection(PROS_SECTION_OBJECT *SectionObject,
|
|
|
|
|
ACCESS_MASK DesiredAccess,
|
|
|
|
|
POBJECT_ATTRIBUTES ObjectAttributes,
|
|
|
|
|
PLARGE_INTEGER UMaximumSize,
|
|
|
|
|
ULONG SectionPageProtection,
|
|
|
|
|
ULONG AllocationAttributes,
|
|
|
|
|
PFILE_OBJECT FileObject);
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
PFN_NUMBER
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmWithdrawSectionPage(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
PLARGE_INTEGER FileOffset,
|
|
|
|
|
BOOLEAN *Dirty);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmFinalizeSectionPageOut(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
PLARGE_INTEGER FileOffset,
|
|
|
|
|
PFN_NUMBER Page,
|
|
|
|
|
BOOLEAN Dirty);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
/* sptab.c *******************************************************************/
|
|
|
|
|
|
|
|
|
|
VOID
|
|
|
|
|
NTAPI
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
MiInitializeSectionPageTable(PMM_SECTION_SEGMENT Segment);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MmSetPageEntrySectionSegment(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
PLARGE_INTEGER Offset,
|
2012-03-28 19:41:40 +00:00
|
|
|
ULONG_PTR Entry,
|
2012-03-28 18:39:21 +00:00
|
|
|
const char *file,
|
|
|
|
|
int line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
2012-03-28 19:41:40 +00:00
|
|
|
ULONG_PTR
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MmGetPageEntrySectionSegment(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
PLARGE_INTEGER Offset,
|
|
|
|
|
const char *file,
|
|
|
|
|
int line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
#define MmSetPageEntrySectionSegment(S,O,E) _MmSetPageEntrySectionSegment(S,O,E,__FILE__,__LINE__)
|
2012-03-28 18:39:21 +00:00
|
|
|
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
#define MmGetPageEntrySectionSegment(S,O) _MmGetPageEntrySectionSegment(S,O,__FILE__,__LINE__)
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
2012-03-28 18:39:21 +00:00
|
|
|
typedef VOID (NTAPI *FREE_SECTION_PAGE_FUN)(
|
|
|
|
|
PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
PLARGE_INTEGER Offset);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
VOID
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmFreePageTablesSectionSegment(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
FREE_SECTION_PAGE_FUN FreePage);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
/* Yields a lock */
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
PMM_SECTION_SEGMENT
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmGetSectionAssociation(PFN_NUMBER Page,
|
|
|
|
|
PLARGE_INTEGER Offset);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmSetSectionAssociation(PFN_NUMBER Page,
|
|
|
|
|
PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
PLARGE_INTEGER Offset);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
VOID
|
|
|
|
|
NTAPI
|
|
|
|
|
MmDeleteSectionAssociation(PFN_NUMBER Page);
|
|
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
|
|
|
|
MmpPageOutPhysicalAddress(PFN_NUMBER Page);
|
|
|
|
|
|
|
|
|
|
/* io.c **********************************************************************/
|
|
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
MmspWaitForFileLock(PFILE_OBJECT File);
|
|
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiSimpleRead(PFILE_OBJECT FileObject,
|
|
|
|
|
PLARGE_INTEGER FileOffset,
|
|
|
|
|
PVOID Buffer,
|
|
|
|
|
ULONG Length,
|
|
|
|
|
BOOLEAN Paging,
|
|
|
|
|
PIO_STATUS_BLOCK ReadStatus);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MiSimpleWrite(PFILE_OBJECT FileObject,
|
|
|
|
|
PLARGE_INTEGER FileOffset,
|
|
|
|
|
PVOID Buffer,
|
|
|
|
|
ULONG Length,
|
|
|
|
|
PIO_STATUS_BLOCK ReadStatus,
|
|
|
|
|
const char *file,
|
|
|
|
|
int line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
#define MiSimpleWrite(F,O,B,L,R) _MiSimpleWrite(F,O,B,L,R,__FILE__,__LINE__)
|
|
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MiWriteBackPage(PFILE_OBJECT FileObject,
|
|
|
|
|
PLARGE_INTEGER Offset,
|
|
|
|
|
ULONG Length,
|
|
|
|
|
PFN_NUMBER Page,
|
|
|
|
|
const char *File,
|
|
|
|
|
int Line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
#define MiWriteBackPage(F,O,L,P) _MiWriteBackPage(F,O,L,P,__FILE__,__LINE__)
|
|
|
|
|
|
|
|
|
|
/* section.c *****************************************************************/
|
|
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmAccessFaultCacheSection(KPROCESSOR_MODE Mode,
|
|
|
|
|
ULONG_PTR Address,
|
|
|
|
|
BOOLEAN FromMdl);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiReadFilePage(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMEMORY_AREA MemoryArea,
|
|
|
|
|
PMM_REQUIRED_RESOURCES RequiredResources);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiGetOnePage(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMEMORY_AREA MemoryArea,
|
|
|
|
|
PMM_REQUIRED_RESOURCES RequiredResources);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiSwapInPage(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMEMORY_AREA MemoryArea,
|
|
|
|
|
PMM_REQUIRED_RESOURCES RequiredResources);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiWriteSwapPage(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMEMORY_AREA MemoryArea,
|
|
|
|
|
PMM_REQUIRED_RESOURCES Resources);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiWriteFilePage(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMEMORY_AREA MemoryArea,
|
|
|
|
|
PMM_REQUIRED_RESOURCES Resources);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
VOID
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiFreeSegmentPage(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
PLARGE_INTEGER FileOffset);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
2012-12-19 23:49:13 +00:00
|
|
|
_Success_(1)
|
|
|
|
|
_When_(return==STATUS_MORE_PROCESSING_REQUIRED, _At_(Required->DoAcquisition, _Post_notnull_))
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-12-19 23:49:13 +00:00
|
|
|
MiCowCacheSectionPage (
|
|
|
|
|
_In_ PMMSUPPORT AddressSpace,
|
|
|
|
|
_In_ PMEMORY_AREA MemoryArea,
|
|
|
|
|
_In_ PVOID Address,
|
|
|
|
|
_In_ BOOLEAN Locked,
|
|
|
|
|
_Inout_ PMM_REQUIRED_RESOURCES Required);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
VOID
|
2012-03-28 18:39:21 +00:00
|
|
|
MmPageOutDeleteMapping(PVOID Context,
|
|
|
|
|
PEPROCESS Process,
|
|
|
|
|
PVOID Address);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
VOID
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MmLockSectionSegment(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
const char *file,
|
|
|
|
|
int line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
#define MmLockSectionSegment(x) _MmLockSectionSegment(x,__FILE__,__LINE__)
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
VOID
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MmUnlockSectionSegment(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
const char *file,
|
|
|
|
|
int line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
#define MmUnlockSectionSegment(x) _MmUnlockSectionSegment(x,__FILE__,__LINE__)
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
VOID
|
2012-03-28 18:39:21 +00:00
|
|
|
MmFreeCacheSectionPage(PVOID Context,
|
|
|
|
|
MEMORY_AREA* MemoryArea,
|
|
|
|
|
PVOID Address,
|
|
|
|
|
PFN_NUMBER Page,
|
|
|
|
|
SWAPENTRY SwapEntry,
|
|
|
|
|
BOOLEAN Dirty);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MiFlushMappedSection(PVOID BaseAddress,
|
|
|
|
|
PLARGE_INTEGER BaseOffset,
|
|
|
|
|
PLARGE_INTEGER FileSize,
|
|
|
|
|
BOOLEAN Dirty,
|
|
|
|
|
const char *File,
|
|
|
|
|
int Line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
#define MiFlushMappedSection(A,O,S,D) _MiFlushMappedSection(A,O,S,D,__FILE__,__LINE__)
|
|
|
|
|
|
|
|
|
|
VOID
|
|
|
|
|
NTAPI
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
MmFinalizeSegment(PMM_SECTION_SEGMENT Segment);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
VOID
|
|
|
|
|
NTAPI
|
|
|
|
|
MmFreeSectionSegments(PFILE_OBJECT FileObject);
|
|
|
|
|
|
2012-03-28 18:39:21 +00:00
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
|
|
|
|
MmMapCacheViewInSystemSpaceAtOffset(IN PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
OUT PVOID* MappedBase,
|
|
|
|
|
IN PLARGE_INTEGER ViewOffset,
|
|
|
|
|
IN OUT PULONG ViewSize);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MiMapViewOfSegment(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
PVOID* BaseAddress,
|
|
|
|
|
SIZE_T ViewSize,
|
|
|
|
|
ULONG Protect,
|
|
|
|
|
PLARGE_INTEGER ViewOffset,
|
|
|
|
|
ULONG AllocationType,
|
|
|
|
|
const char *file,
|
|
|
|
|
int line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
2012-03-28 18:39:21 +00:00
|
|
|
#define MiMapViewOfSegment(AddressSpace,Segment,BaseAddress,ViewSize,Protect,ViewOffset,AllocationType) \
|
|
|
|
|
_MiMapViewOfSegment(AddressSpace,Segment,BaseAddress,ViewSize,Protect,ViewOffset,AllocationType,__FILE__,__LINE__)
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmUnmapViewOfCacheSegment(PMMSUPPORT AddressSpace,
|
|
|
|
|
PVOID BaseAddress);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
|
|
|
|
MmUnmapCacheViewInSystemSpace(PVOID Address);
|
|
|
|
|
|
2012-12-19 23:49:13 +00:00
|
|
|
_Success_(1)
|
|
|
|
|
_When_(return==STATUS_MORE_PROCESSING_REQUIRED, _At_(Required->DoAcquisition, _Post_notnull_))
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-12-19 23:49:13 +00:00
|
|
|
MmNotPresentFaultCachePage (
|
|
|
|
|
_In_ PMMSUPPORT AddressSpace,
|
|
|
|
|
_In_ MEMORY_AREA* MemoryArea,
|
|
|
|
|
_In_ PVOID Address,
|
|
|
|
|
_In_ BOOLEAN Locked,
|
|
|
|
|
_Inout_ PMM_REQUIRED_RESOURCES Required);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmPageOutPageFileView(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMEMORY_AREA MemoryArea,
|
|
|
|
|
PVOID Address,
|
|
|
|
|
PMM_REQUIRED_RESOURCES Required);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
FORCEINLINE
|
|
|
|
|
BOOLEAN
|
2012-03-28 18:39:21 +00:00
|
|
|
_MmTryToLockAddressSpace(IN PMMSUPPORT AddressSpace,
|
|
|
|
|
const char *file,
|
|
|
|
|
int line)
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
{
|
2012-03-28 18:39:21 +00:00
|
|
|
BOOLEAN Result = KeTryToAcquireGuardedMutex(&CONTAINING_RECORD(AddressSpace, EPROCESS, Vm)->AddressCreationLock);
|
|
|
|
|
//DbgPrint("(%s:%d) Try Lock Address Space %x -> %s\n", file, line, AddressSpace, Result ? "true" : "false");
|
|
|
|
|
return Result;
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#define MmTryToLockAddressSpace(x) _MmTryToLockAddressSpace(x,__FILE__,__LINE__)
|
|
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiWidenSegment(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMEMORY_AREA MemoryArea,
|
|
|
|
|
PMM_REQUIRED_RESOURCES RequiredResources);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiSwapInSectionPage(PMMSUPPORT AddressSpace,
|
|
|
|
|
PMEMORY_AREA MemoryArea,
|
|
|
|
|
PMM_REQUIRED_RESOURCES RequiredResources);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmExtendCacheSection(PROS_SECTION_OBJECT Section,
|
|
|
|
|
PLARGE_INTEGER NewSize,
|
|
|
|
|
BOOLEAN ExtendFile);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
_MiFlushMappedSection(PVOID BaseAddress,
|
|
|
|
|
PLARGE_INTEGER BaseOffset,
|
|
|
|
|
PLARGE_INTEGER FileSize,
|
|
|
|
|
BOOLEAN Dirty,
|
|
|
|
|
const char *File,
|
|
|
|
|
int Line);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
#define MiFlushMappedSection(A,O,S,D) _MiFlushMappedSection(A,O,S,D,__FILE__,__LINE__)
|
|
|
|
|
|
|
|
|
|
PVOID
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmGetSegmentRmap(PFN_NUMBER Page,
|
|
|
|
|
PULONG RawOffset);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MmNotPresentFaultCacheSection(KPROCESSOR_MODE Mode,
|
|
|
|
|
ULONG_PTR Address,
|
|
|
|
|
BOOLEAN FromMdl);
|
[CACHE]
The cache manager rewrite I started years ago has finally appeared in
ReactOS' trunk and although at this point it's not quite perfectly
integrated, it's enough to boot up the bootcd or livecd. To check out
the more mature original, check out arty-newcc-reactos, branch
arty-newcc on bitbucket.org . Amine Khaldi encouraged me quite a bit
to not give up on it, and was able to reach out and be an advocate
when i really wasn't able to. Others agree that the time has come to
begin removing the old cache manager. I expect the remaining problems
in the version going to trunk will be taken care of relatively
quickly.
The motivation for this effort lies in the particularly hairy
relationship between ReactOS' cache manager and data sections. This
code completely removes page sharing between cache manager and section
and reimagines cache manager as being a facility layered on the memory
manager, not really caring about individual pages, but simply managing
data section objects where caching might occur.
It took me about 2 years to do the first pass of this rewrite and most
of this year to fix some lingering issues, properly implement demand
paging in ReactOS (code which didn't come with this patch in a
recognizable form), and finish getting the PrivateCacheMap and
SharedCacheMap relationship correct.
Currently, the new ntoskrnl/cache directory contains an own
implementation of data file sections. After things have settled down,
we can begin to deprecate and remove the parts of ReactOS' section
implementation that depend on a close relationship with cache
manager. Eventually, I think that the extra code added to
ntoskrnl/cache/section will be removed and ReactOS' own sections will
replace the use of the special MM_CACHE_SECTION_SEGMENT in the cache
path.
Note also, that this makes all cache manager (and new section parts)
use wide file offsets. If my section code were to take over other
parts of the ReactOS memory manager, they would also benefit from
these improvements.
I invite anyone who wants to to peek at this code and fix whatever
bugs can be found.
svn path=/trunk/; revision=49423
2010-11-02 02:32:39 +00:00
|
|
|
|
|
|
|
|
ULONG
|
|
|
|
|
NTAPI
|
2012-03-28 18:39:21 +00:00
|
|
|
MiCacheEvictPages(PMM_SECTION_SEGMENT Segment,
|
|
|
|
|
ULONG Target);
|
[NEWCC]
A reintegration checkpoint for the NewCC branch, brought to you by Team NewCC.
Differences with current ReactOS trunk:
* A new memory area type, MEMORY_AREA_CACHE, is added, which represents a mapped region of a file. In NEWCC mode, user sections are MEMORY_AREA_CACHE type as well, and obey new semantics. In non-NEWCC mode, they aren't used.
* A way of claiming a page entry for a specific thread's work is added. Placing the special SWAPENTRY value MM_WAIT_ENTRY in a page table, or in a section page table should indicate that memory management code is intended to wait for another thread to make some status change before checking the state of the page entry again. In code that uses this convention, a return value of STATUS_SUCCESS + 1 is used to indicate that the caller should use the MiWaitForPageEvent macro to wait until somebody has change the state of a wait entry before checking again. This is a lighter weight mechanism than PAGEOPs.
* A way of asking the caller to perform some blocking operation without locks held is provided. This replaces some spaghettified code in which locks are repeatedly taken and broken by code that performs various blocking operations. Using this mechanism, it is possible to do a small amount of non-blocking work, fill in a request, then return STATUS_MORE_PROCESSING_REQUIRED to request that locks be dropped and the blocking operation be carried out. A MM_REQUIRED_RESOURCES structure is provided to consumers of this contract to use to accumulate state across many blocking operations. Several functions wrapping blocking operations are provided in ntoskrnl/cache/reqtools.c.
* Image section pages are no longer direct mapped. This is done to simplify consolidation of ownership of pages under the data section system. At a later time, it may be possible to make data pages directly available to image sections for the same file. This is likely the only direct performance impact this code makes on non-NEWCC mode.
RMAPs:
* A new type of RMAP entry is introduced, distinguished by RMAP_IS_SEGMENT(Address) of the rmap entry. This kind of entry contains a pointer to a section page table node in the Process pointer, which in turn links back to the MM_SECTION_SEGMENT it belongs to. Therefore, a page belonging only to a segment (that is, a segment page that isn't mapped) can exist and be evicted using the normal page eviction mechanism in balance.c. Each of the rmap function has been modified to deal with segment rmaps.
* The low 8 bits of the Address field in a segment rmap denote the entry number in the generic table node pointed to by Process that points to the page the rmap belongs to. By combining them, you can determine the file offset the page belongs to.
* In NEWCC mode, MmSharePageEntry/UnsharePageEntry are not used, and instead the page reference count is used to keep track of the number of mappings of a page, allowing the last reference expiring to allow the page to be recycled without much intervention. These are still used in non-NEWCC mode. One change has been made, the count fields have been narrowed by 1 bit to make room for a dirty bit in SSE entries, needed when a page is present but unmapped.
Section page tables:
* The section page tables are now implemented using RtlGenericTables. This enables a fairly compact representation of section page tables without having the existence of a section object imply 4k of fake PDEs. In addition, each node in the generic table has a wide file offset that is a multiple of 256 pages, or 1 megabyte total. Besides needing wide file offsets, the only other visible change caused by the switch to generic tables for section page tables is the need to lock the section segment before interacting with the section page table.
Eviction:
* Page eviction in cache sections is accomplished by MmpPageOutPhysicalAddress. In the case of a shared page, it tries to remove all mappings of the indicated page. If this process fails at any point, the page will simply be drawn back into the target address spaces. After succeeding at this, if TRUE has been accumulated into the page's dirty bit in the section page table, it is written back, and then permanently removed.
NewCC mode:
* NEWCC mode is introduced, which rewrites the file cache to a set of cache stripes actively mapped, along with unmapped section data.
* NewCC is more authentic in its interpretation of the external interface to the windows cache than the current cache manager, implementing each of the cache manager functions according to the documented interface with no preconceived ideas about how anything should be implemented internally. Cache stripes are implemented on top of section objects, using the same memory manager paths, and therefore economizing code and complexity. This replaces a rather complicated system in which pages can be owned by the cache manager and the memory manager simultaneously and they must cooperate in a fairly sophisticated way to manage them. Since they're quite interdependent in the current code, modifying either is very difficult. In NEWCC, they have a clear division of labor and thus can be worked on independently.
* Several third party filesystems that use the kernel Cc interface work properly using NEWCC, including matt wu's ext3 driver.
* In contrast with code that tries to make CcInitializeCacheMap and CcUninitializeCacheMap into a pair that supports reference counting, NEWCC lazily initializes the shared and private cache maps as needed and uses the presence of a PrivateCacheMap on at least one file pointing to the SharedCacheMap as an indication that the FILE_OBJECT reference in the SharedCacheMap should still be held. When the last PrivateCacheMap is discarded, that's the appropriate time to tear down caching for a specific file, as the SharedCacheMap data is allowed to be saved and reused. We honor this by making the SharedCacheMap into a depot for keeping track of the PrivateCacheMap objects associated with views of a file.
svn path=/trunk/; revision=55833
2012-02-23 12:03:06 +00:00
|
|
|
|
|
|
|
|
NTSTATUS
|
2012-03-28 18:39:21 +00:00
|
|
|
MiRosTrimCache(ULONG Target,
|
|
|
|
|
ULONG Priority,
|
|
|
|
|
PULONG NrFreed);
|