vm_object_deallocate(), replacing the assertion GIANT_REQUIRED.
o Remove GIANT_REQUIRED from vm_map_protect() and vm_map_simplify_entry().
o Acquire and release Giant around vm_map_protect()'s call to pmap_protect().
Altogether, these changes eliminate the need for mprotect() to acquire
and release Giant.
mallochash. Mallochash is going to go away as soon as I introduce the
kfree/kmalloc api and partially overhaul the malloc wrapper. This can't happen
until all users of the malloc api that expect memory to be aligned on the size
of the allocation are fixed.
Implement the following checks on freed memory in the bucket path:
- Slab membership
- Alignment
- Duplicate free
This previously was only done if we skipped the buckets. This code will slow
down INVARIANTS a bit, but it is smp safe. The checks were moved out of the
normal path and into hooks supplied in uma_dbg.
0xdeadc0de and then check for it just before memory is handed off as part
of a new request. This will catch any post free/pre alloc modification of
memory, as well as introduce errors for anything that tries to dereference
it as a pointer.
This code takes the form of special init, fini, ctor and dtor routines that
are specificly used by malloc. It is in a seperate file because additional
debugging aids will want to live here as well.
mutex class. Currently this is only used for kmapentzone because kmapents
are are potentially allocated when freeing memory. This is not dangerous
though because no other allocations will be done while holding the
kmapentzone lock.
i386/ia64/alpha - catch up to sparc64/ppc:
- replace pmap_kernel() with refs to kernel_pmap
- change kernel_pmap pointer to (&kernel_pmap_store)
(this is a speedup since ld can set these at compile/link time)
all platforms (as suggested by jake):
- gc unused pmap_reference
- gc unused pmap_destroy
- gc unused struct pmap.pm_count
(we never used pm_count - we track address space sharing at the vmspace)
in the same style as sys/proc.h.
o Undo the de-inlining of several trivial, MPSAFE methods on the vm_map.
(Contrary to the commit message for vm_map.h revision 1.66 and vm_map.c
revision 1.206, de-inlining these methods increased the kernel's size.)
due to conditions that suggest the possible need for stack growth.
This has two beneficial effects: (1) we can
now remove calls to vm_map_growstack() from the MD trap handlers and (2)
simple page faults are faster because we no longer unnecessarily perform
vm_map_growstack() on every page fault.
o Remove vm_map_growstack() from the i386's trap_pfault().
o Remove the acquisition and release of Giant from i386's trap_pfault().
(vm_fault() still acquires it.)
statclock can access it in the tail end of statclock_process() at an
unfortunate time. This bit me several times on an SMP alpha (UP2000)
and the problem went away with this change. I'm not sure why it doesn't
break x86 as well. Maybe it's because the clocks are much faster
on alpha (HZ=1024 by default).
and pmap_copy_page(). This gets rid of a couple more physical addresses
in upper layers, with the eventual aim of supporting PAE and dealing with
the physical addressing mostly within pmap. (We will need either 64 bit
physical addresses or page indexes, possibly both depending on the
circumstances. Leaving this to pmap itself gives more flexibilitly.)
Reviewed by: jake
Tested on: i386, ia64 and (I believe) sparc64. (my alpha was hosed)
hash while holding the lock on a zone. Fix this by doing the allocation
seperately from the actual hash expansion.
The lock is dropped before the allocation and reacquired before the expansion.
The expansion code checks to see if we lost the race and frees the new hash
if we do. We really never will lose this race because the hash expansion is
single threaded via the timeout mechanism.
Fortunately we have no large zones with maximums specified yet, so it wasn't
breaking anything.
Implement blocking when a zone exceeds the maximum and M_WAITOK is specified.
Previously this just failed like the old zone allocator did. The old zone
allocator didn't support WAITOK/NOWAIT though so we should do what we
advertise.
While I was in there I cleaned up some more zalloc logic to further simplify
that code path and reduce redundant code. This was needed to make the blocking
work properly anyway.
didn't like the wait argument and that if you were removing a zone it had
better be empty.
Also, I broke out part of hash_expand and made a seperate hash_free() for use
in uma_zdestroy.
never held across blocking operations. Also, fix two other lock order
reversals that were exposed by jhb's witness change.
The free path previously had a bug that would cause it to skip the free bucket
list in some cases and go straight to allocating a new bucket. This has been
fixed as well.
These changes made the bucket handling code much cleaner and removed quite a
few lock operations. This should be marginally faster now.
It is now possible to call malloc w/o Giant and avoid any witness warnings.
This still isn't entirely safe though because malloc_type statistics are not
protected by any lock.
relating to extreme low memory situations occured. This was only ever seen on
the port build cluster, so many thanks to kris for helping me debug this.
Tested by: kris
PCPU_LAZY_INC() which increments elements in it for cases where we
can afford the occassional inaccuracy. Use of per-cpu stats counters
avoids significant cache stalls in various critical paths that would
otherwise severely limit our cpu scaleability.
Adjust all sysctl's accessing cnt.* elements to now use a procedure
which aggregates the requested field for all cpus and for the global
vmmeter.
The global vmmeter is retained, since some stats counters, like v_free_min,
cannot be made per-cpu. Also, this allows us to convert counters from
the global vmmeter to the per-cpu vmmeter in a piecemeal fashion, so
have at it!
most cases NULL is passed, but in some cases such as network driver locks
(which use the MTX_NETWORK_LOCK macro) and UMA zone locks, a name is used.
Tested on: i386, alpha, sparc64
memory in phys_avail will fit in 'int', use vm_size_t. This fixes booting
on sparc64 machines with more than 2 gigs of ram.
Thanks to Jan Chrillesen for providing me with access to a 4 gig machine.
general cleanup of the API. The entire API now consists of two functions
similar to the pre-KSE API. The suser() function takes a thread pointer
as its only argument. The td_ucred member of this thread must be valid
so the only valid thread pointers are curthread and a few kernel threads
such as thread0. The suser_cred() function takes a pointer to a struct
ucred as its first argument and an integer flag as its second argument.
The flag is currently only used for the PRISON_ROOT flag.
Discussed on: smp@
with this flag. Remove the dup_list and dup_ok code from subr_witness. Now
we just check for the flag instead of doing string compares.
Also, switch the process lock, process group lock, and uma per cpu locks over
to this interface. The original mechanism did not work well for uma because
per cpu lock names are unique to each zone.
Approved by: jhb
where a sysctl within 20 seconds of a cache_drain could yield negative "USED"
counts.
Also, grab the uma_mtx while in the sysctl handler. This hadn't caused
problems yet because Giant is held all the time.
Reported by: kkenn
best path forward now is likely to change the lockmgr locks to simple
sleep mutexes, then see if any extra contention it generates is greater
than removed overhead of managing local locking state information,
cost of extra calls into lockmgr, etc.
Additionally, making the vm_map lock a mutex and respecting it properly
will put us much closer to not needing Giant magic in vm.
the motivation for saving and restoring the map->hint in useracc() is gone.
(The same tests that motivated this change in revision 1.57 now show that
there is no performance loss from removing it.) This was really a hack and
some day we would have had to add new synchronization here on map->hint
to maintain it.
pmap_qremove. pmap_kenter is not safe to use in MI code because it is not
guaranteed to flush the mapping from the tlb on all cpus. If the process
in question is preempted and migrates cpus between the call to pmap_kenter
and pmap_kremove, the original cpu will be left with stale mappings in its
tlb. This is currently not a problem for i386 because we do not use PG_G on
SMP, and thus all mappings are flushed from the tlb on context switches, not
just user mappings. This is not the case on all architectures, and if PG_G
is to be used with SMP on i386 it will be a problem. This was committed by
peter earlier as part of his fine grained tlb shootdown work for i386, which
was backed out for other reasons.
Reviewed by: peter
the bio and buffer structures to have daddr64_t bio_pblkno,
b_blkno, and b_lblkno fields which allows access to disks
larger than a Terabyte in size. This change also requires
that the VOP_BMAP vnode operation accept and return daddr64_t
blocks. This delta should not affect system operation in
any way. It merely sets up the necessary interfaces to allow
the development of disk drivers that work with these larger
disk block addresses. It also allows for the development of
UFS2 which will use 64-bit block addresses.
While doing this, move it earlier in the sysinit boot process so that the
VM system can use it.
After that, the system is now able to use sx locks instead of lockmgr
locks in the VM system. To accomplish this, some of the more
questionable uses of the locks (such as testing whether they are
owned or not, as well as allowing shared+exclusive recursion) are
removed, and simpler logic throughout is used so locks should also be
easier to understand.
This has been tested on my laptop for months, and has not shown any
problems on SMP systems, either, so appears quite safe. One more
user of lockmgr down, many more to go :)
style(9)
- Minor space adjustment in cases where we have "( ", " )", if(), return(),
while(), for(), etc.
- Add /* SYMBOL */ after a few #endifs.
Reviewed by: alc
the revived code.
vm pages newly allocated are marked busy (PG_BUSY), thus calling
vm_page_delete before the pages has been freed or unbusied will
cause a deadlock since vm_page_object_page_remove will wait for the
busy flag to be cleared. This can be triggered by calling malloc
with size > PAGE_SIZE and the M_NOWAIT flag on systems low on
physical free memory.
A kernel module that reproduces the problem, written by Logan Gabriel
<logan@mail.2cactus.com>, can be found in the freebsd-hackers mail
archive (12 Apr 2001). The problem was recently noticed again by
Archie Cobbs <archie@dellroad.org>.
Reviewed by: dillon
moderately improves msync's and VM object flushing for objects containing
randomly dirtied pages (fsync(), msync(), filesystem update daemon),
and improves cpu use for small-ranged sequential msync()s in the face of
very large mmap()ings from O(N) to O(1) as might be performed by a database.
A sysctl, vm.msync_flush_flag, has been added and defaults to 3 (the two
committed optimizations are turned on by default). 0 will turn off both
optimizations.
This code has already been tested under stable and is one in a series of
memq / vp->v_dirtyblkhd / fsync optimizations to remove O(N^2) restart
conditions that will be coming down the pipe.
MFC after: 3 days
and again in vm_page.c and vm_pageq.c.
o Delete unusused prototypes. (Mainly a result of the earlier renaming
of various functions from vm_page_*() to vm_pageq_*().)
and isn't strictly required. However, it lowers the number of false
positives found when grep'ing the kernel sources for p_ucred to ensure
proper locking.
In order to determine what to page out, the vm_daemon checks
reference bits on all pages belonging to all processes. Unfortunately,
the algorithm used reacted badly with shared pages; each shared page
would be checked once per process sharing it; this caused an O(N^2)
growth of tlb invalidations. The algorithm has been changed so that
each page will be checked only 16 times.
Prior to this change, a fork/sleepbomb of 1300 processes could cause
the vm_daemon to take over 60 seconds to complete, effectively
freezing the system for that time period. With this change
in place, the vm_daemon completes in less than a second. Any system
with hundreds of processes sharing pages should benefit from this change.
Note that the vm_daemon is only run when the system is under extreme
memory pressure. It is likely that many people with loaded systems saw
no symptoms of this problem until they reached the point where swapping
began.
Special thanks go to dillon, peter, and Chuck Cranor, who helped me
get up to speed with vm internals.
PR: 33542, 20393
Reviewed by: dillon
MFC after: 1 week
There is some unresolved badness that has been eluding me, particularly
affecting uniprocessor kernels. Turning off PG_G helped (which is a bad
sign) but didn't solve it entirely. Userland programs still crashed.
shootdowns in a couple of key places. Do the same for i386. This also
hides some physical addresses from higher levels and has it use the
generic vm_page_t's instead. This will help for PAE down the road.
Obtained from: jake (MI code, suggestions for MD part)
count that would otherwise be on one of the free queues. This eliminates a
panic when broken programs unmap memory that still has pending IO from raw
devices.
Reviewed by: dillon, alc
- Allow the OOM killer to target processes currently locked in
memory. These very often are the ones doing the memory hogging.
- Drop the wakeup priority of processes currently sleeping while
waiting for their page fault to complete. In order for the OOM
killer to work well, the killed process and other system processes
waiting on memory must be allowed to wakeup first.
Reviewed by: dillon
MFC after: 1 week
this is a low-functionality change that changes the kernel to access the main
thread of a process via the linked list of threads rather than
assuming that it is embedded in the process. It IS still embeded there
but remove all teh code that assumes that in preparation for the next commit
which will actually move it out.
Reviewed by: peter@freebsd.org, gallatin@cs.duke.edu, benno rice,
shared.
Also introduce vm_endcopy instead of using pointer tricks when
initializing new vmspaces.
The race occured because of how the reference was utilized:
test vmspace reference,
possibly block,
decrement reference
When sharing a vmspace between multiple processes it was possible
for two processes exiting at the same time to test the reference
count, possibly block and neither one free because they wouldn't
see the other's update.
Submitted by: green
Seigo Tanimura (tanimura) posted the initial delta.
I've polished it quite a bit reducing the need for locking and
adapting it for KSE.
Locks:
1 mutex in each filedesc
protects all the fields.
protects "struct file" initialization, while a struct file
is being changed from &badfileops -> &pipeops or something
the filedesc should be locked.
1 mutex in each struct file
protects the refcount fields.
doesn't protect anything else.
the flags used for garbage collection have been moved to
f_gcflag which was the FILLER short, this doesn't need
locking because the garbage collection is a single threaded
container.
could likely be made to use a pool mutex.
1 sx lock for the global filelist.
struct file * fhold(struct file *fp);
/* increments reference count on a file */
struct file * fhold_locked(struct file *fp);
/* like fhold but expects file to locked */
struct file * ffind_hold(struct thread *, int fd);
/* finds the struct file in thread, adds one reference and
returns it unlocked */
struct file * ffind_lock(struct thread *, int fd);
/* ffind_hold, but returns file locked */
I still have to smp-safe the fget cruft, I'll get to that asap.
mutex releases to not require flags for the cases when preemption is
not allowed:
The purpose of the MTX_NOSWITCH and SWI_NOSWITCH flags is to prevent
switching to a higher priority thread on mutex releease and swi schedule,
respectively when that switch is not safe. Now that the critical section
API maintains a per-thread nesting count, the kernel can easily check
whether or not it should switch without relying on flags from the
programmer. This fixes a few bugs in that all current callers of
swi_sched() used SWI_NOSWITCH, when in fact, only the ones called from
fast interrupt handlers and the swi_sched of softclock needed this flag.
Note that to ensure that swi_sched()'s in clock and fast interrupt
handlers do not switch, these handlers have to be explicitly wrapped
in critical_enter/exit pairs. Presently, just wrapping the handlers is
sufficient, but in the future with the fully preemptive kernel, the
interrupt must be EOI'd before critical_exit() is called. (critical_exit()
can switch due to a deferred preemption in a fully preemptive kernel.)
I've tested the changes to the interrupt code on i386 and alpha. I have
not tested ia64, but the interrupt code is almost identical to the alpha
code, so I expect it will work fine. PowerPC and ARM do not yet have
interrupt code in the tree so they shouldn't be broken. Sparc64 is
broken, but that's been ok'd by jake and tmm who will be fixing the
interrupt code for sparc64 shortly.
Reviewed by: peter
Tested on: i386, alpha
against VM_WAIT in the pageout code. Both fixes involve adjusting
the lockmgr's timeout capability so locks obtained with timeouts do not
interfere with locks obtained without a timeout.
Hopefully MFC: before the 4.5 release
commit by Kirk also fixed a softupdates bug that could easily be triggered
by server side NFS.
* An edge case with shared R+W mmap()'s and truncate whereby
the system would inappropriately clear the dirty bits on
still-dirty data. (applicable to all filesystems)
THIS FIX TEMPORARILY DISABLED PENDING FURTHER TESTING.
see vm/vm_page.c line 1641
* The straddle case for VM pages and buffer cache buffers when
truncating. (applicable to NFS client side)
* Possible SMP database corruption due to vm_pager_unmap_page()
not clearing the TLB for the other cpu's. (applicable to NFS
client side but could effect all filesystems). Note: not
considered serious since the corruption occurs beyond the file
EOF.
* When flusing a dirty buffer due to B_CACHE getting cleared,
we were accidently setting B_CACHE again (that is, bwrite() sets
B_CACHE), when we really want it to stay clear after the write
is complete. This resulted in a corrupt buffer. (applicable
to all filesystems but probably only triggered by NFS)
* We have to call vtruncbuf() when ftruncate()ing to remove
any buffer cache buffers. This is still tentitive, I may
be able to remove it due to the second bug fix. (applicable
to NFS client side)
* vnode_pager_setsize() race against nfs_vinvalbuf()... we have
to set n_size before calling nfs_vinvalbuf or the NFS code
may recursively vnode_pager_setsize() to the original value
before the truncate. This is what was causing the user mmap
bus faults in the nfs tester program. (applicable to NFS
client side)
* Fix to softupdates (see ufs/ffs/ffs_inode.c 1.73, commit made
by Kirk).
Testing program written by: Avadis Tevanian, Jr.
Testing program supplied by: jkh / Apple (see Dec2001 posting to freebsd-hackers with Subject 'NFS: How to make FreeBS fall on its face in one easy step')
MFC after: 1 week
you run out of mbuf address space.
kern/subr_mbuf.c: print a warning message when mb_alloc fails, again
rate-limited to at most once per second. This covers other
cases of mbuf allocation failures. Probably it also overlaps the
one handled in vm/vm_kern.c, so maybe the latter should go away.
This warning will let us gradually remove the printf that are scattered
across most network drivers to report mbuf allocation failures.
Those are potentially dangerous, in that they are not rate-limited and
can easily cause systems to panic.
Unless there is disagreement (which does not seem to be the case
judging from the discussion on -net so far), and because this is
sort of a safety bugfix, I plan to commit a similar change to STABLE
during the weekend (it affects kern/uipc_mbuf.c there).
Discussed-with: jlemon, silby and -net
a page boundary, since we've already allocated all our contiguous kva
space up front. This eliminates some memory wastage, and allows us to
actually reach the # of objects were specified in the zinit() call.
Reviewed by: peter, dillon
on a vnode-backed object must be incremented *after* obtaining the vnode
lock. If it is bumped before obtaining the vnode lock we can deadlock
against vtruncbuf().
Submitted by: peter, ps
MFC after: 3 days
file EOF. This works around a bug in the ISOFS (CDRom) BMAP code which
returns bogus values for requests beyond the file EOF rather then returning
an error, resulting in either corrupt data being mmap()'d beyond the file EOF
or resulting in a seg-fault on the last page of a mmap()'d file (mmap()s of
CDRom files).
Reported by: peter / Yahoo
MFC after: 3 days
(allocating pv entries w/ zalloci) when called in a loop due to
an madvise(). It is possible to completely exhaust the free page list and
cause a system panic when an expected allocation fails.
real effect.
Optimize vfs_msync(). Avoid having to continually drop and re-obtain
mutexes when scanning the vnode list. Improves looping case by 500%.
Optimize ffs_sync(). Avoid having to continually drop and re-obtain
mutexes when scanning the vnode list. This makes a couple of assumptions,
which I believe are ok, in regards to vnode stability when the mount list
mutex is held. Improves looping case by 500%.
(more optimization work is needed on top of these fixes)
MFC after: 1 week
to kern_synch.c in preparation for adding some jitter to the
inter-sample time.
Note that the "vm.loadavg" sysctl still lives in vm_meter.c which
isn't the right place, but it is appropriate for the current (bad)
name of that sysctl.
Suggested by: jhb (some time ago)
Reviewed by: bde
- vm map entries are not valid after the map has been unlocked.
- An exclusive lock on the map is needed before calling
vm_map_simplify_entry().
Fix cleanup after page wiring failure to unwire all pages that had been
successfully wired before the failure was detected.
Reviewed by: dillon
kernel map and object in a manner that contigfree() is actually able to
free. Previously contigfree() freed up the KVA space but could not
unwire & free the underlying VM pages due to mismatched pageability between
the map entry and the VM pages.
Submitted by: Thomas Moestl <tmoestl@gmx.net>
Testing by: mark tinguely <tinguely@web.cs.ndsu.nodak.edu>
MFC after: 3 days
would sometimes prevent a dirty page from being cleaned, even when synced,
resulting in the dirty page being re-flushed to disk every 30-60 seconds or
so, forever. The problem is that when the filesystem flushes a page to
its backing file it typically does not clear dirty bits representing areas
of the page that are beyond the file EOF. If the file is also mmap()'d and
a fault is taken, vm_fault (properly, is required to) set the vm_page_t->dirty
bits to VM_PAGE_BITS_ALL. This combination could leave us with an uncleanable,
unfreeable page.
The solution is to have the vnode_pager detect the edge case and manually
clear the dirty bits representing areas beyond the file EOF. The filesystem
does the rest and the page comes up clean after the write completes.
MFC after: 3 days
- crhold() returns a reference to the ucred whose refcount it bumps.
- crcopy() now simply copies the credentials from one credential to
another and has no return value.
- a new crshared() primitive is added which returns true if a ucred's
refcount is > 1 and false (0) otherwise.
been a no-op for as long as our CVS history goes back. Processes in
state SSLEEP could only be counted if p_slptime == 0, but immediately
before loadav() is called, schedcpu() has just incremented p_slptime
on all SSLEEP processes.
Note ALL MODULES MUST BE RECOMPILED
make the kernel aware that there are smaller units of scheduling than the
process. (but only allow one thread per process at this time).
This is functionally equivalent to teh previousl -current except
that there is a thread associated with each process.
Sorry john! (your next MFC will be a doosie!)
Reviewed by: peter@freebsd.org, dillon@freebsd.org
X-MFC after: ha ha ha ha
it to the MI area. KSE touched cpu_wait() which had the same change
replicated five ways for each platform. Now it can just do it once.
The only MD parts seemed to be dealing with fpu state cleanup and things
like vm86 cleanup on x86. The rest was identical.
XXX: ia64 and powerpc did not have cpu_throw(), so I've put a functional
stub in place.
Reviewed by: jake, tmm, dillon
on and off since John Dyson left his work-in-progress.
It is off by default for now. sysctl vm.zeroidle_enable=1 to turn it on.
There are some hacks here to deal with the present lack of preemption - we
yield after doing a small number of pages since we wont preempt otherwise.
This is basically Matt's algorithm [with hysteresis] with an idle process
to call it in a similar way it used to be called from the idle loop.
I cleaned up the includes a fair bit here too.
timeout callwheel and buffer cache, out of the platform specific areas
and into the machine independant area. i386 and alpha adjusted here.
Other cpus can be fixed piecemeal.
Reviewed by: freebsd-smp, jake
information. The default limits only effect machines with > 1GB of ram
and can be overriden with two new kernel conf variables VM_SWZONE_SIZE_MAX
and VM_BCACHE_SIZE_MAX, or with loader variables kern.maxswzone and
kern.maxbcache. This has the effect of leaving more KVM available for
sizing NMBCLUSTERS and 'maxusers' and should avoid tripups where a sysad
adds memory to a machine and then sees the kernel panic on boot due to
running out of KVM.
Also change the default swap-meta auto-sizing calculation to allocate half
of what it was previously allocating. The prior defaults were way too high.
Note that we cannot afford to run out of swap-meta structures so we still
stay somewhat conservative here.
- Callers of asleep() and await() have been converted to calling tsleep().
The only caller outside of M_ASLEEP was the ata driver, which called both
asleep() and await() with spl-raised, so there was no need for the
asleep() and await() pair. M_ASLEEP was unused.
Reviewed by: jasone, peter
- Callers of asleep() and await() have been converted to calling tsleep().
The only caller outside of M_ASLEEP was the ata driver, which called both
asleep() and await() with spl-raised, so there was no need for the
asleep() and await() pair. M_ASLEEP was unused.
Reviewed by: jasone, peter
1) allocate fewer buckets
2) when failing to allocate swap zone, keep reducing the zone by
a third rather than a half in order to reduce the chance of
allocating way too little.
I also moved around some code for readability.
Suggested by: dillon
Reviewed by: dillon
already allow this for NFS swap configured via BOOTP, so it is
known to work fine.
For many diskless configurations is is more flexible to have the
client set up swapping itself; it can recreate a sparse swap file
to save on server space for example, and it works with a non-NFS
root filesystem such as an in-kernel filesystem image.
Also removed some spl's and added some VM mutexes, but they are not actually
used yet, so this commit does not really make any operational changes
to the system.
vm_page.c relates to vm_page_t manipulation, including high level deactivation,
activation, etc... vm_pageq.c relates to finding free pages and aquiring
exclusive access to a page queue (exclusivity part not yet implemented).
And the world still builds... :-)
most of these inlines had been bloated in -current far beyond their
original intent. Normalize prototypes and function declarations to be ANSI
only (half already were). And do some general cleanup.
(kernel size also reduced by 50-100K, but that isn't the prime intent)
(this commit is just the first stage). Also add various GIANT_ macros to
formalize the removal of Giant, making it easy to test in a more piecemeal
fashion. These macros will allow us to test fine-grained locks to a degree
before removing Giant, and also after, and to remove Giant in a piecemeal
fashion via sysctl's on those subsystems which the authors believe can
operate without Giant.
introduce a modified allocation mechanism for mbufs and mbuf clusters; one
which can scale under SMP and which offers the possibility of resource
reclamation to be implemented in the future. Notable advantages:
o Reduce contention for SMP by offering per-CPU pools and locks.
o Better use of data cache due to per-CPU pools.
o Much less code cache pollution due to excessively large allocation macros.
o Framework for `grouping' objects from same page together so as to be able
to possibly free wired-down pages back to the system if they are no longer
needed by the network stacks.
Additional things changed with this addition:
- Moved some mbuf specific declarations and initializations from
sys/conf/param.c into mbuf-specific code where they belong.
- m_getclr() has been renamed to m_get_clrd() because the old name is really
confusing. m_getclr() HAS been preserved though and is defined to the new
name. No tree sweep has been done "to change the interface," as the old
name will continue to be supported and is not depracated. The change was
merely done because m_getclr() sounds too much like "m_get a cluster."
- TEMPORARILY disabled mbtypes statistics displaying in netstat(1) and
systat(1) (see TODO below).
- Fixed systat(1) to display number of "free mbufs" based on new per-CPU
stat structures.
- Fixed netstat(1) to display new per-CPU stats based on sysctl-exported
per-CPU stat structures. All infos are fetched via sysctl.
TODO (in order of priority):
- Re-enable mbtypes statistics in both netstat(1) and systat(1) after
introducing an SMP friendly way to collect the mbtypes stats under the
already introduced per-CPU locks (i.e. hopefully don't use atomic() - it
seems too costly for a mere stat update, especially when other locks are
already present).
- Optionally have systat(1) display not only "total free mbufs" but also
"total free mbufs per CPU pool."
- Fix minor length-fetching issues in netstat(1) related to recently
re-enabled option to read mbuf stats from a core file.
- Move reference counters at least for mbuf clusters into an unused portion
of the cluster itself, to save space and need to allocate a counter.
- Look into introducing resource freeing possibly from a kproc.
Reviewed by (in parts): jlemon, jake, silby, terry
Tested by: jlemon (Intel & Alpha), mjacob (Intel & Alpha)
Preliminary performance measurements: jlemon (and me, obviously)
URL: http://people.freebsd.org/~bmilekic/mb_alloc/
processes a little earlier to avoid a deadlock. Second, when calculating
the 'largest process' do not just count RSS. Instead count the RSS + SWAP
used by the process. Without this the code tended to kill small
inconsequential processes like, oh, sshd, rather then one of the many
'eatmem 200MB' I run on a whim :-). This fix has been extensively tested on
-stable and somewhat tested on -current and will be MFCd in a few days.
Shamed into fixing this by: ps
canonical: define a versioned struct xswdev, and add a sysctl node
handler that allows the user to get this structure for a certain device
index by specifying this index as last element of the MIB.
This new node handler, vm.swap_info, replaces the old vm.nswapdev
and vm.swapdevX.* (where X was the index) sysctls.
- move the sysctl code to kern_intr.c
- do not use INTRCNT_COUNT, but rather eintrcnt - intrcnt to determine
the length of the intrcnt array
- move the declarations of intrnames, eintrnames, intrcnt and eintrcnt
from machine-dependent include files to sys/interrupt.h
- remove the hw.nintr sysctl, it is not needed.
- fix various style bugs
Requested by: bde
Reviewed by: bde (some time ago)
Tor created a while ago, removes the raw I/O piece (that has cache coherency
problems), and adds a buffer cache / VM freeing piece.
Essentially this patch causes O_DIRECT I/O to not be left in the cache, but
does not prevent it from going through the cache, hence the 80%. For
the last 20% we need a method by which the I/O can be issued directly to
buffer supplied by the user process and bypass the buffer cache entirely,
but still maintain cache coherency.
I also have the code working under -stable but the changes made to sys/file.h
may not be MFCable, so an MFC is not on the table yet.
Submitted by: tegge, dillon
- Assert Giant in vm_pageout_scan() for the vnode hacking that it does.
- Don't hold vm_mtx around vget() or vput().
- Lock Giant when calling vm_pageout_scan() from the pagedaemon. Also,
lock curproc while setting the P_BUFEXHAUST flag.
- For now we still hold Giant for all of the vm_daemon. When process
limits are locked we will be only need Giant for swapout_procs().
- Restore the previous order of setting up a new vm_object. The previous
had a small bug where we zero'd out the flags after we set the
OBJ_ONEMAPPING flag.
- Add several asserts of vm_mtx.
- Assert Giant is held rather than locking and unlocking it in a few
places.
- Add in some #ifdef objlocks code to lock individual vm objects when
vm objects each have their own lock someday.
- Don't bother acquiring the allproc lock for a ddb command. If DDB
blocked on the lock, that would be worse than having an inconsistent
allproc list.
- Add a few KTR tracepoints to track the addition and removal of
vm_map_entry's and the creation adn free'ing of vmspace's.
- Adjust a few portions of code so that we update the process' vmspace
pointer to its new vmspace before freeing the old vmspace.
- Don't lock Giant in the scheduler() function except for when calling
faultin().
- In swapout_procs(), lock the VM before the proccess to avoid a lock order
violation.
- In swapout_procs(), release the allproc lock before calling swapout().
We restart the process scan after swapping out a process.
- In swapout_procs(), un #if 0 the code to bump the vmspace reference count
and lock the process' vm structures. This bug was introduced by me and
could result in the vmspace being free'd out from under a running
process.
- Fix an old bug where the vmspace reference count was not free'd if we
failed the swap_idle_threshold2 test.
acquired.
- Assert Giant is held in the strategy, getpages, and putpages methods and
the getchainbuf, flushchainbuf, and waitchainbuf functions.
- Always call flushchainbuf() w/o the VM lock.
vnodes.
- Fix an old bug that would leak a reference to a fd if the vnode being
mmap'd wasn't of type VREG or VCHR.
- Lock Giant in vm_mmap() around calls into the VM that can call into
pager routines that need Giant or into other VM routines that need
Giant.
- Replace code that used a goto to jump around the else branch of a test
to use an else branch instead.
vm_mtx does not recurse and is required for most low level
vm operations.
faults can not be taken without holding Giant.
Memory subsystems can now call the base page allocators safely.
Almost all atomic ops were removed as they are covered under the
vm mutex.
Alpha and ia64 now need to catch up to i386's trap handlers.
FFS and NFS have been tested, other filesystems will need minor
changes (grabbing the vm lock when twiddling page properties).
Reviewed (partially) by: jake, jhb
wakeup proc0 by hand to enforce the timeout.
- When swapping out a process, keep the process locked via the proc lock
from the first checks up until we clear PS_INMEM and set PS_SWAPPING in
swapout(). The swapout() function now must be called with the proc lock
held and releases it before returning.
- Comment out the code to attempt to lock a process' VM structures before
swapping out. It is broken in that it releases the lock after obtaining
it. If it does grab the lock, it needs to hand it off to swapout()
instead of releasing it. This can be revisisted when the VM is locked
as this is a valid test to perform. It also causes a lock order reversal
for the time being, which is the immediate cause for temporarily
disabling it.
the process in question locked as soon as we find it and determine it to
be eligible until we actually kill it. To avoid deadlock, we don't block
on the process lock but skip any process that is already locked during our
search.
- Don't hold Giant in the swapper daemon while we walk the list of
processes looking for a process to swap back in.
- Don't bother grabbing the sched_lock while checking a process' sleep
time in swapout_procs() to ensure that a process has been idle for at
least swap_idle_threshold2 before swapping it out. If we lose the race
we just let a process stay in memory until the next call of
swapout_procs().
- Remove some unneeded spl's, sched_lock does all the locking needed in
this case.
other "system" header files.
Also help the deprecation of lockmgr.h by making it a sub-include of
sys/lock.h and removing sys/lockmgr.h form kernel .c files.
Sort sys/*.h includes where possible in affected files.
OK'ed by: bde (with reservations)
The zone allocator's locks should be leaflocks, meaning that they
should never be held when entering into another subsystem, however
the sysctl grabs the zone global mutex and individual zone mutexes
while holding the lock it calls SYSCTL_OUT which recurses into the
VM subsystem in order to wire user memory to do a safe copy. This
can block and cause lock order reversals.
To fix this:
lock zone global.
get a count of the number of zones.
unlock global.
allocate temporary storage.
format and SYSCTL_OUT the banner.
lock global.
traverse list.
make sure we haven't looped more than the initial count taken
to avoid overflowing the allocated buffer.
lock each nodes.
read values and format into buffer.
unlock individual node.
unlock global.
format and SYSCTL_OUT the rest of the data.
free storage.
return.
Other problems included not checking for errors when doing sysctl out
of the column header. Fixed.
Inconsistant termination of the copied string. Fixed.
Objected to by: des (for not using sbuf)
Since the output is not variable length and I'm actually over
allocating signifigantly and I'd like to get this fixed now, I'll
work on the sbuf convertion at a later date. I would not object
to someone else taking it upon themselves to convert it to sbuf.
I hold no MAINTIANER rights to this code (for now).
Protect pager object list manipulation with a mutex.
It doesn't look possible to combine them under a single sx lock because
creation may block and we can't have the object list manipulation block
on anything other than a mutex because of interrupt requests.
VOP_BWRITE() was a hack which made it possible for NFS client
side to use struct buf with non-bio backing.
This patch takes a more general approach and adds a bp->b_op
vector where more methods can be added.
The success of this patch depends on bp->b_op being initialized
all relevant places for some value of "relevant" which is not
easy to determine. For now the buffers have grown a b_magic
element which will make such issues a tiny bit easier to debug.
programs. There is a case during a fork() which can cause a deadlock.
From Tor -
The workaround that consists of setting a flag in the vm map that
indicates that a fork is in progress and using that mark in the page
fault handling to force a revalidation failure. That change will only
affect (pessimize) page fault handling during fork for threaded
(linuxthreads style) applications and applications using aio_*().
Submited by: tegge
call is correct, but it interferes with the massive hack called
vm_map_growstack(). The call will be returned after our stack handling
code is fixed.
Reported by: tegge
reference count was transferred to the new object, but both the
new and the old map entries had pointers to the new object.
Correct this by transferring the second reference.
This fixes a panic that can occur when mmap(2) is used with the
MAP_INHERIT flag.
PR: i386/25603
Reviewed by: dillon, alc
supported architectures such as the alpha. This allows us to save
on kernel virtual address space, TLB entries, and (on the ia64) VHPT
entries. pmap_map() now modifies the passed in virtual address on
architectures that do not support direct-mapped segments to point to
the next available virtual address. It also returns the actual
address that the request was mapped to.
- On the IA64 don't use a special zone of PV entries needed for early
calls to pmap_kenter() during pmap_init(). This gets us in trouble
because we end up trying to use the zone allocator before it is
initialized. Instead, with the pmap_map() change, the number of needed
PV entries is small enough that we can get by with a static pool that is
used until pmap_init() is complete.
Submitted by: dfr
Debugging help: peter
Tested by: me
of memory, rather than from the start.
This fixes problems allocating bouncebuffers on alphas where there is only
1 chunk of memory (unlike PCs where there is generally at least one small
chunk and a large chunk). Having 1 chunk had been fatal, because these
structures take over 13MB on a machine with 1GB of ram. This doesn't leave
much room for other structures and bounce buffers if they're at the front.
Reviewed by: dfr, anderson@cs.duke.edu, silence on -arch
Tested by: Yoriaki FUJIMORI <fujimori@grafin.fujimori.cache.waseda.ac.jp>
make sure that PG_NOSYNC is properly set. Previously we only set it
for a write-fault, but this can occur on a read-fault too.
(will be MFCd prior to 4.3 freeze)
this information via the vm.nswapdev sysctl (number of swap areas)
and vm.swapdevX nodes (where X is the device), which contain the MIBs
dev, blocks, used, and flags. These changes are required to allow
top and other userland swap-monitoring utilities to run without
setgid kmem.
Submitted by: Thomas Moestl <tmoestl@gmx.net>
Reviewed by: freebsd-audit
- All processes go into the same array of queues, with different
scheduling classes using different portions of the array. This
allows user processes to have their priorities propogated up into
interrupt thread range if need be.
- I chose 64 run queues as an arbitrary number that is greater than
32. We used to have 4 separate arrays of 32 queues each, so this
may not be optimal. The new run queue code was written with this
in mind; changing the number of run queues only requires changing
constants in runq.h and adjusting the priority levels.
- The new run queue code takes the run queue as a parameter. This
is intended to be used to create per-cpu run queues. Implement
wrappers for compatibility with the old interface which pass in
the global run queue structure.
- Group the priority level, user priority, native priority (before
propogation) and the scheduling class into a struct priority.
- Change any hard coded priority levels that I found to use
symbolic constants (TTIPRI and TTOPRI).
- Remove the curpriority global variable and use that of curproc.
This was used to detect when a process' priority had lowered and
it should yield. We now effectively yield on every interrupt.
- Activate propogate_priority(). It should now have the desired
effect without needing to also propogate the scheduling class.
- Temporarily comment out the call to vm_page_zero_idle() in the
idle loop. It interfered with propogate_priority() because
the idle process needed to do a non-blocking acquire of Giant
and then other processes would try to propogate their priority
onto it. The idle process should not do anything except idle.
vm_page_zero_idle() will return in the form of an idle priority
kernel thread which is woken up at apprioriate times by the vm
system.
- Update struct kinfo_proc to the new priority interface. Deliberately
change its size by adjusting the spare fields. It remained the same
size, but the layout has changed, so userland processes that use it
would parse the data incorrectly. The size constraint should really
be changed to an arbitrary version number. Also add a debug.sizeof
sysctl node for struct kinfo_proc.
mtx_enter(lock, type) becomes:
mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks)
mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized)
similarily, for releasing a lock, we now have:
mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN.
We change the caller interface for the two different types of locks
because the semantics are entirely different for each case, and this
makes it explicitly clear and, at the same time, it rids us of the
extra `type' argument.
The enter->lock and exit->unlock change has been made with the idea
that we're "locking data" and not "entering locked code" in mind.
Further, remove all additional "flags" previously passed to the
lock acquire/release routines with the exception of two:
MTX_QUIET and MTX_NOSWITCH
The functionality of these flags is preserved and they can be passed
to the lock/unlock routines by calling the corresponding wrappers:
mtx_{lock, unlock}_flags(lock, flag(s)) and
mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN
locks, respectively.
Re-inline some lock acq/rel code; in the sleep lock case, we only
inline the _obtain_lock()s in order to ensure that the inlined code
fits into a cache line. In the spin lock case, we inline recursion and
actually only perform a function call if we need to spin. This change
has been made with the idea that we generally tend to avoid spin locks
and that also the spin locks that we do have and are heavily used
(i.e. sched_lock) do recurse, and therefore in an effort to reduce
function call overhead for some architectures (such as alpha), we
inline recursion for this case.
Create a new malloc type for the witness code and retire from using
the M_DEV type. The new type is called M_WITNESS and is only declared
if WITNESS is enabled.
Begin cleaning up some machdep/mutex.h code - specifically updated the
"optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN
and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently
need those.
Finally, caught up to the interface changes in all sys code.
Contributors: jake, jhb, jasone (in no particular order)
by myself. It solves a serious vm_map corruption problem that can occur
with the buffer cache when block sizes > 64K are used. This code has been
heavily tested in -stable but only tested somewhat on -current. An MFC
will occur in a few days. My additions include the vm_map_simplify_entry()
and minor buffer cache boundry case fix.
Make the buffer cache use a system map for buffer cache KVM rather then a
normal map.
Ensure that VM objects are not allocated for system maps. There were cases
where a buffer map could wind up with a backing VM object -- normally
harmless, but this could also result in the buffer cache blocking in places
where it assumes no blocking will occur, possibly resulting in corrupted
maps.
Fix a minor boundry case in the buffer cache size limit is reached that
could result in non-optimal code.
Add vm_map_simplify_entry() calls to prevent 'creeping proliferation'
of vm_map_entry's in the buffer cache's vm_map. Previously only a simple
linear optimization was made. (The buffer vm_map typically has only a
handful of vm_map_entry's. This stabilizes it at that level permanently).
PR: 20609
Submitted by: (Tor Egge) tegge
- Proc locking in a few places.
- faultin() now must be called with the proc lock held.
- Split up swappable() into a couple of tests so that it can be locke in
swapout_procs().
- Use queue macros.
- replace the simplelock in struct vm_zone with a mutex.
- use a proper SLIST rather than a hand-rolled job for the zone list.
- add a subsystem lock that protects the zone list and the statistics
counters.
- merge _zalloc() into zalloc() and _zfree() into zfree(), and
move them below _zget() so there's no need for a prototype.
- add two initialization functions: one which initializes the
subsystem mutex and the zone list, and one that currently doesn't
do anything.
- zap zerror(); use KASSERTs instead.
- dike out half of sysctl_vm_zone(), which was mostly trying to do
manually what the snprintf() call could do better.
Reviewed by: jhb, jasone
in 4.2-REL which I ripped out in -stable and -current when implementing the
low-memory handling solution. However, maxlaunder turns out to be the saving
grace in certain very heavily loaded systems (e.g. newsreader box). The new
algorithm limits the number of pages laundered in the first pageout daemon
pass. If that is not sufficient then suceessive will be run without any
limit.
Write I/O is now pipelined using two sysctls, vfs.lorunningspace and
vfs.hirunningspace. This prevents excessive buffered writes in the
disk queues which cause long (multi-second) delays for reads. It leads
to more stable (less jerky) and generally faster I/O streaming to disk
by allowing required read ops (e.g. for indirect blocks and such) to occur
without interrupting the write stream, amoung other things.
NOTE: eventually, filesystem write I/O pipelining needs to be done on a
per-device basis. At the moment it is globalized.
The fix works by reverting the ordering of free memory so that the
chances of contig_malloc() succeeding increases.
PR: 23291
Submitted by: Andrew Atrens <atrens@nortel.ca>
struct swblock entries by dividing the number of the entries by 2
until the swap metadata fits.
- Reject swapon(2) upon failure of swap_zone allocation.
This is just a temporary fix. Better solutions include:
(suggested by: dillon)
o reserving swap in SWAP_META_PAGES chunks, and
o swapping the swblock structures themselves.
Reviewed by: alfred, dillon
of explicit calls to lockmgr. Also provides macros for the flags
pased to specify shared, exclusive or release which map to the
lockmgr flags. This is so that the use of lockmgr can be easily
replaced with optimized reader-writer locks.
- Add some locking that I missed the first time.
cases with file fragments and read-write mmap's can lead to a situation
where a VM page has odd dirty bits, e.g. 0xFC - due to being dirtied by
an mmap and only the fragment (representing a non-page-aligned end of
file) synced via a filesystem buffer. A correct solution that
guarentees consistent m->dirty for the file EOF case is being
worked on. In the mean time we can't be so conservative in the
KASSERT.
Backout the previous delta (rev 1.4), it didn't make any difference.
If the requested handle is NULL then don't add it to the list of
objects, to be found by handle.
The problem is that when asking for a NULL handle you are implying
you want a new object. Because objects with NULL handles were
being added to the list, any further requests for phys backed
objects with NULL handles would return a reference to the initial
NULL handle object after finding it on the list.
Basically one couldn't have more than one phys backed object without
a handle in the entire system without this fix. If you did more
than one shared memory allocation using the phys pager it would
give you your initial allocation again.
