(the time of the last access), as what we really want is a frequency/last
access time scoring, and accessed alone is useless for that.
* put_cached_block() no longer frees any unused blocks.
* The low memory handler will now only lock the cache if there is something
to do. Also, it did not take address space warnings into account.
* Even when memory is critical, we don't free all unused blocks anymore - if
the number of blocks we free now (10000) is not sufficient to get out of the
critical condition, chances are good that we will be called again :-)
* block_notifier_and_writer() now tries to make sure that the total block cache
memory consumption grows not much larger than half of the available RAM.
* This should all help to limit the block cache usage a bit better. Hopefully,
a checkfs run will no longer run out of memory here (couldn't test yet).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35733 a95241bf-73f2-0310-859d-f6bbb57e9c96
remove the block from an unlocked cache and the cache->lock would be unlocked
twice when the calling function unlocked. It panics in that case anyway, but
this should make it continuable and is more correct.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35648 a95241bf-73f2-0310-859d-f6bbb57e9c96
they were never freed unless the cache was destroyed (I just wondered why
my system would bury >1G in the magazines).
* Made the magazine capacity variable per cache, ie. for larger objects, it's
not a good idea to have 64*CPU buffers lying around in the worst case.
* Furthermore, the create_object_cache_etc()/object_depot_init() now have
arguments for the magazine capacity as well as the maximum number of full
unused magazines.
* By default, you might want to initialize both to zero, as then some hopefully
usable defaults are computed. Otherwise (the only current example is the
vm_page_mapping cache) you can just put in the values you'd want there.
The page mapping cache uses larger values, as its objects are usually
allocated and deleted in larger chunks.
* Beware, though, I couldn't test these changes yet as Qemu didn't like to run
today. I'll test these changes on another machine now.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35601 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Since the same block can be in up to two transactions, it's very well possible
that one shouldn't write all transactions in a single run.
* Forgot to pass on the iterator from BlockWriter::Add(transaction) to
Add(block).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35565 a95241bf-73f2-0310-859d-f6bbb57e9c96
whether the pages where mapped. Was introduced already in r22731.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35532 a95241bf-73f2-0310-859d-f6bbb57e9c96
therefore does not unlock anymore while iterating over the transactions.
This gave other threads the opportunity to finish a transaction, causing
bug #5412. Also, the BlockWriter will now always close transactions on its
own, and you need to pass the transaction hash iterator to Add().
* Also, transactions that contain blocks that are currently written back will
be ignored by the block writer, as well as cache_sync_transaction(). This
fixes bug #5415.
* Improved error handling if BlockWriter fails to write back blocks. Most
notably, they are no longer left busy_writing, and the functions calling
it do proper error reporting (besides block_cache_discard() that does not
return any erro code; I've added a TODO note there for now).
* The BlockWriter now starts with a larger array once it has to allocate one.
* One can now limit the number of blocks that go into a BlockWriter. This is
used by the block writer thread, that shouldn't always write back everything
every two seconds.
* Also, the fixed array is larger now (leaving enough space such that the
block writer/notifier does not need to allocate anything).
* And finally, if allocating the array fails, the BlockWriter falls back to the
synchronous write back used previously. IOW it will never write back less
blocks than you ask for.
* Added static BlockWriter::WriteBlock() method replacing write_cached_block().
* Forgot to rename block_cache::busy_count to busy_reading_count.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35510 a95241bf-73f2-0310-859d-f6bbb57e9c96
particular it wouldn't set the flag when writing something to a page, but
only move it to the modified queue. Since mapping the page would move it to
another queue, the information that the page was modified would be lost and
it would never be written to disk. Was well reproducible with a Haiku image
build and limited amount of memory.
Fixes the hopefully last remaining cause for #5374.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35500 a95241bf-73f2-0310-859d-f6bbb57e9c96
block_cache_discard() use the BlockWriter directly as well.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35496 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added new KDL command "cached_block" that dumps a cached block.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35493 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Renamed the "busy" stuff to "busy_reading", and added a "busy_writing"
concept.
* This now allows reading a block (and other blocks), while blocks are written
back. This should speed all operations needing to write back blocks, like
unzipping or compiling.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35464 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added vm_clear_page_mapping_accessed_flags() and
vm_remove_all_page_mappings_if_unaccessed(), which combine the functionality
of vm_test_map_activation(), vm_clear_map_flags(), and
vm_remove_all_page_mappings(), thus saving lots of calls to translation map
methods. The backend is the new method
VMTranslationMap::ClearAccessedAndModified().
* Started to make use of the cached page queue and changed the meaning of the
other non-free queues slightly:
- Active queue: Contains mapped pages that have been used recently.
- Inactive queue: Contains mapped pages that have not been used recently. Also
contains unmapped temporary pages.
- Modified queue: Contains unmapped modified pages.
- Cached queue: Contains unmapped unmodified pages (LRU sorted).
Unless we're actually low on memory and actively do paging, modified and
cached queues only contain non-temporary pages. Cached pages are considered
quasi free. They still belong to a cache, but since they are unmodified and
unmapped, they can be freed immediately. And this is what
vm_page_[try_]reserve_pages() do now when there are no more actually free
pages at hand. Essentially this means that pages storing cached file data,
unless mmap()ped, no longer are considered used and don't contribute to page
pressure. Paging will not happen as long there are enough free + cached pages
available.
* Reimplemented the page daemon. It no longer scans all pages, but instead works
the page queues. As long as the free pages situation is harmless, it only
iterates through the active queue and deactivates pages that have not been
used recently. When paging occurs it additionally scans the inactive queue and
frees pages that have not been used recently.
* Changed the page reservation/allocation interface:
vm_page_[try_]reserve_pages(), vm_page_unreserve_pages(), and
vm_page_allocate_page() now take a vm_page_reservation structure pointer.
The reservation functions initialize the structure -- currently consisting
only of a count member for the number of still reserved pages.
vm_page_allocate_page() decrements the count and vm_page_unreserve_pages()
unreserves the remaining pages (if any). Advantages are that reservation/
unreservation mismatches cannot occur anymore, that vm_page_allocate_page()
can verify that the caller has indeed a reserved page left, and that there's
no unnecessary pressure on the free page pool anymore. The only disadvantage
is that the vm_page_reservation object needs to be passed around a bit.
* Reworked the page reservation implementation:
- Got rid of sSystemReservedPages and sPageDeficit. Instead
sUnreservedFreePages now actually contains the number of free pages that
have not yet been reserved (it cannot become negative anymore) and the new
sUnsatisfiedPageReservations contains the number of pages that are still
needed for reservation.
- Threads waiting for reservations do now add themselves to a waiter queue,
which is ordered by descending priority (VM priority and thread priority).
High priority waiters are served first when pages become available.
Fixes#5328.
* cache_prefetch_vnode(): Would reserve one less page than allocated later, if
the size wasn't page aligned.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35393 a95241bf-73f2-0310-859d-f6bbb57e9c96
that does not have a transaction.
* This should fix#5340.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35390 a95241bf-73f2-0310-859d-f6bbb57e9c96
general "flags" parameter. It encodes the target state of the page -- so
that the page isn't unnecessarily put in the wrong page queue first -- a
flag whether the page should be cleared, and one to indicate whether the
page should be marked busy.
* Added page state PAGE_STATE_CACHED. Not used yet.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35333 a95241bf-73f2-0310-859d-f6bbb57e9c96
flag. The obvious advantage is that one can still see what state a page is in
and even move it between states while being marked busy.
* Removed the vm_page::is_dummy flag. Instead we mark marker pages busy, which
in all cases has the same effect. Introduced a vm_page_is_dummy() that can
still check whether a given page is a dummy page.
* vm_page_unreserve_pages(): Before adding to the system reserve make sure
sUnreservedFreePages is non-negative. Otherwise we'd make nonexisting pages
available for allocation. steal_pages() still has the same problem and it
can't be solved that easily.
* map_page(): No longer changes the page state/mark the page unbusy. That's the
caller's responsibility.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35331 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added flags to avoid notifying the busy condition variable unnecessarily.
* get_writable_cached_block(): Unlock the cache while memcpy()ing/memset()ing
the block's data. The idea is to reduce lock contention. Less effective
than I hoped, though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35328 a95241bf-73f2-0310-859d-f6bbb57e9c96
argument. They replace the previous special-purpose allocation functions
(malloc_nogrow(), vip_io_request_malloc()).
* Moved the I/O VIP heap to heap.cpp accordingly.
* Added quite a bit of passing around of allocation flags in the VM,
particularly in the VM*AddressSpace classes.
* Fixed IOBuffer::GetNextVirtualVec(): It was ignoring the VIP flag and always
allocated on the normal heap.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35316 a95241bf-73f2-0310-859d-f6bbb57e9c96
on the fly, clearing and writing it each time, we now use an iovec with 32
identical entries pointing to a clear page that we prepare once at
initialization. This speeds up clear_image in low memory situations
dramatically.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35304 a95241bf-73f2-0310-859d-f6bbb57e9c96
beginning whether to pass the cache by really doesn't help when
reading/writing a huge amount of data, since a low memory situation is likely
to occur at some point during the operation. This should fix the main issue
of #3768.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35299 a95241bf-73f2-0310-859d-f6bbb57e9c96
memory and page reservation functions have a new "priority" parameter that
indicates how deep the function may tap into that reserve. The currently
existing priority levels are "user", "system", and "VIP". The idea is that
user programs should never be able to cause a state that gets the kernel into
trouble due to heavy battling for memory. The "VIP" level (not really used
yet) is intended for allocations that are required to free memory eventually
(in the page writer). More levels are thinkable in the future, like "user real
time" or "user system server".
* Added "priority" parameters to several VMCache methods.
* Replaced the map_backing_store() "unmapAddressRange" parameter by a "flags"
parameter.
* Added area creation flag CREATE_AREA_PRIORITY_VIP and slab allocator flag
CACHE_PRIORITY_VIP indicating the importance of the request.
* Changed most code to pass the right priorities/flags.
These changes already significantly improve the behavior in low memory
situations. I've tested a bit with 64 MB (virtual) RAM and, while not
particularly fast and responsive, the system remains at least usable under high
memory pressure.
As a side effect the slab allocator can now be used as general memory allocator.
Not done by default yet, though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35295 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Implemented a more elaborated raw memory allocation backend (MemoryManager).
We allocate 8 MB areas whose pages we allocate and map when needed. An area is
divided into equally-sized chunks which form the basic units of allocation. We
have areas with three possible chunk sizes (small, medium, large), which is
basically what the ObjectCache implementations were using anyway.
* Added "uint32 flags" parameter to several of the slab allocator's object
cache and object depot functions. E.g. object_depot_store() potentially wants
to allocate memory for a magazine. But also in pure freeing functions it
might eventually become useful to have those flags, since they could end up
deleting an area, which might not be allowable in all situations. We should
introduce specific flags to indicate that.
* Reworked the block allocator. Since the MemoryManager allocates block-aligned
areas, maintains a hash table for lookup, and maps chunks to object caches,
we can quickly find out which object cache a to be freed allocation belongs
to and thus don't need the boundary tags anymore.
* Reworked the slab boot strap process. We allocate from the initial area only
when really necessary, i.e. when the object cache for the respective
allocation size has not been created yet. A single page is thus sufficient.
other:
* vm_allocate_early(): Added boolean "blockAlign" parameter. If true, the
semantics is the same as for B_ANY_KERNEL_BLOCK_ADDRESS.
* Use an object cache for page mappings. This significantly reduces the
contention on the heap bin locks.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35232 a95241bf-73f2-0310-859d-f6bbb57e9c96
access to a vm_page. It is basically an atomically accessed thread ID field
in the vm_page structure, which is explicitly set by macros marking the
critical sections. As a first positive effect I had to review quite a bit of
code and found several issues.
* Added several TODOs and comments. Some harmless ones, but also a few
troublesome ones in vm.cpp regarding page unmapping.
* file_cache: PrecacheIO::Prepare()/read_into_cache: Removed superfluous
vm_page_allocate_page() return value checks. It cannot fail anymore.
* Removed the heavily contended "pages" lock. We use different policies now:
- sModifiedTemporaryPages is accessed atomically.
- sPageDeficitLock and sFreePageCondition are protected by a new mutex.
- The page queues have individual locks (mutexes).
- Renamed set_page_state_nolock() to set_page_state(). Unless the caller says
otherwise, it does now lock the affected pages queues itself. Also changed
the return value to void -- we panic() anyway.
* set_page_state(): Add free/clear pages to the beginning of their respective
queues as this is more cache-friendly.
* Pages with the states PAGE_STATE_WIRED or PAGE_STATE_UNUSED are no longer
in any queue. They were in the "active" queue, but there's no good reason
to have them there. In case we decide to let the page daemon work the queues
(like FreeBSD) they would just be in the way.
* Pulled the common part of vm_page_allocate_page_run[_no_base]() into a helper
function. Also fixed a bug I introduced previously: The functions must not
vm_page_unreserve_pages() on success, since they remove the pages from the
free/clear queue without decrementing sUnreservedFreePages.
* vm_page_set_state(): Changed return type to void. The function cannot really
fail and no-one was checking it anyway.
* vm_page_free(), vm_page_set_state(): Added assertion: The page must not be
free/clear before. This is implied by the policy that no-one is allowed to
access free/clear pages without holding the respective queue's lock, which is
not the case at this point. This found the bug fixed in r34912.
* vm_page_requeue(): Added general assertions. panic() when requeuing of
free/clear pages is requested. Same reason as above.
* vm_clone_area(), B_FULL_LOCK case: Don't map busy pages. The implementation is
still not correct, though.
My usual -j8 Haiku build test runs another 10% faster, now. The total kernel
time drops about 18%. As hoped the new locks have only a fraction of the old
"pages" lock contention. Other locks lead the "most wanted list" now.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34933 a95241bf-73f2-0310-859d-f6bbb57e9c96
improve the reliability as long as our slab implementation is a PITA.
* Removed an assertion that will no longer work (due to the DoublyLinkedList
changes).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34877 a95241bf-73f2-0310-859d-f6bbb57e9c96
destruction and VMVnodeCache::AcquireUnreferencedStoreRef(). Solved by
adding a flag to VMVnodeCache and letting AcquireUnreferencedStoreRef()
fail, if set.
* Added TODO regarding replacing the snooze() waiting for busy vnodes.
* get_vnode(): Unlock sVnodeMutex while calling the put_vnode() hook on
error.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34841 a95241bf-73f2-0310-859d-f6bbb57e9c96
unlocked, allowing for more parallel access.
* Writing is still done synchronously, though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34611 a95241bf-73f2-0310-859d-f6bbb57e9c96
waits for certain events on a given page, NotifyPageEvents() wakes up
waiting threads respectively.
* Used the new feature instead of condition variables for waiting on busy
pages. We save publishing and unpublishing of a condition variable whenever
a page is marked busy. There's only something to do, if there's at least
one thread waiting in the list of the respective cache. The general
assumption is that this is only rarely the case and even if it happens,
there should be only very few threads.
* Added an apparently missing notification in cache_io(). At least I didn't
see the reason for it not being there.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34537 a95241bf-73f2-0310-859d-f6bbb57e9c96
are required to be physically contiguos, which should be reworked to put them
into seperate iovecs. Still this manages to combine a great deal of page
writes into larger bursts already. Reduces the amount of IO requests being
scheduled (and greatly benefits media where page wise writes are slow when
they are accessed through a non-IOScheduler path, i.e. USB mass storage until
that is properly implemented).
* Abstracted per page page writing tasks into a PageWriteWrapper class.
* Abstracted per transfer page writing tasks into PageWriteTransfer class which
formerly was the PageWriterCallback.
* Use both classes from the PageWriterRun and from
vm_page_write_modified_page_range to remove code duplication.
* Adjusted synchronous VMAnonymousCache::Write() to cope correctly with larger
iovecs and more than one iovec. It assumed that there was exactly one page per
vector previously.
* Introduced MaxPagesPerWrite() and MaxPagesPerAsyncWrite() to VMCache to allow
a cache to specify restricitions. VMAnonymousCache does restrict the max pages
to 1 for WriteAsync right now as I didn't feel like reworking that one to cope
with non single page writes just yet.
* Pulled out PageWriteTransfer methods for better readability.
* Some typo fixes.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@33507 a95241bf-73f2-0310-859d-f6bbb57e9c96
cannot assume that the file is that long - we need to check for the actual
file size specifically, and then clear the extraneous bytes based on that one.
* This fixes seeing old file left-overs in the space beyond the file size in
mmapped file. Thanks to Oliver who provided me with a nice test program to
reproduce this problem (this should also fix gcc's fix_includes app (although
I wonder what it does if a file actually ends on a page boundary).
* Also fixed a bug in precaching that would cause the last page of the cache
to be discarded, causing it to be re-read later on.
* Moved partial page clearing into its own function
read_pages_and_clear_partial(), even though it's currently only used once.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32001 a95241bf-73f2-0310-859d-f6bbb57e9c96
UI freezes (ActivityMonitor and ProcessController both use get_system_info() a
lot), although this is only the symptom of another problem.
* The downside is that the block cache usage information isn't as up to date as
it was previously - it's updated by the block write/notifier thread now (worst
case every 2 seconds).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@31812 a95241bf-73f2-0310-859d-f6bbb57e9c96
fixes#4106 for good.
* Also removed a bit of dead code in that function.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@31732 a95241bf-73f2-0310-859d-f6bbb57e9c96
