to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
insert the replacement page into the same position
as the original page on the object memq so that
genfs_putpages (and lfs) won't be confused.
noted by Stephan Uphoff (PR/24328)
the latter is not a appropriate place to do so and it broke vfork.
- deactivate pmap before calling cpu_exit() to keep a balance of
pmap_activate/deactivate.
which is zero by default.
perform rbtree sanity checks only when it isn't zero
because the check is very heavy weight especially when
there're many entries.
in the case that there's no cached entries,
if kmem_map is already up, allocate a entry from it
so that we won't try to vm_map_lock recursively.
XXX assuming usage pattern of kmem_map.
and uncontrolled growth.
The key fix is from Dan Carasone, who noticed that buf_canfree() was
counting in _bytes_ but freeing in _buffers_, which caused the instant
drop to lowater observed by some users.
We now control the rate of growth; the probability of getting a new
allocation is inversely proportional to the current size of the
cache. This idea is from a long-ago conversation with Kirk McKusick
and, if memory serves, was used for the file-system cache in some
other BSD variant at some point in history.
With growth and shrinkage more or less dealt with, we return the
default maximum cache size to 15%. The default _minimum_ cache size
is raised from 1/16 of the maximum cache size to 1/8, since 1/16 was
chosen when the maximum size was 30% of memory.
Finally, after observing the behaviour of the pagedaemon and the
buffer cache drainer under pathological workloads (e.g. a benchmark
that steps through 75% of available memory backwards) I have moved
the call to buf_drain() to the beginning of the pagedaemon from the
end; if the pagedaemon bogs down, it still won't get run as often
as it should, but at least this way it will see the state of the
free count and free target _before_ the scan step does its thing.
- for in-kernel maps, disable map entry merging so that
unmap operations won't block. (workaround for PR/24039)
- for in-kernel maps, allocate kva for vm_map_entry from
the map itsself and eliminate MAX_KMAPENT and
uvm_map_entry_kmem_pool.
VOP_STRATEGY(bp) is replaced by one of two new functions:
- VOP_STRATEGY(vp, bp) Call the strategy routine of vp for bp.
- DEV_STRATEGY(bp) Call the d_strategy routine of bp->b_dev for bp.
DEV_STRATEGY(bp) is used only for block-to-block device situations.
This avoids problems with the kernel grovelling vmstat -u/-U when
using LOCKDEBUG, which changes the size of struct simplelock.
Replaced the original location of the simplelock with "int unused"
so that binary compatibility will be retained with old vmstat.
process context ('reaper').
From within the exiting process context:
* deactivate pmap and free vmspace while we can still block
* introduce MD cpu_lwp_free() - this cleans all MD-specific context (such
as FPU state), and is the last potentially blocking operation;
all of cpu_wait(), and most of cpu_exit(), is now folded into cpu_lwp_free()
* process is now immediatelly marked as zombie and made available for pickup
by parent; the remaining last lwp continues the exit as fully detached
* MI (rather than MD) code bumps uvmexp.swtch, cpu_exit() is now same
for both 'process' and 'lwp' exit
uvm_lwp_exit() is modified to never block; the u-area memory is now
always just linked to the list of available u-areas. Introduce (blocking)
uvm_uarea_drain(), which is called to release the excessive u-area memory;
this is called by parent within wait4(), or by pagedaemon on memory shortage.
uvm_uarea_free() is now private function within uvm_glue.c.
MD process/lwp exit code now always calls lwp_exit2() immediatelly after
switching away from the exiting lwp.
g/c now unneeded routines and variables, including the reaper kernel thread
virtual memory reservation and a private pool of memory pages -- by a scheme
based on memory pools.
This allows better utilization of memory because buffers can now be allocated
with a granularity finer than the system's native page size (useful for
filesystems with e.g. 1k or 2k fragment sizes). It also avoids fragmentation
of virtual to physical memory mappings (due to the former fixed virtual
address reservation) resulting in better utilization of MMU resources on some
platforms. Finally, the scheme is more flexible by allowing run-time decisions
on the amount of memory to be used for buffers.
On the other hand, the effectiveness of the LRU queue for buffer recycling
may be somewhat reduced compared to the traditional method since, due to the
nature of the pool based memory allocation, the actual least recently used
buffer may release its memory to a pool different from the one needed by a
newly allocated buffer. However, this effect will kick in only if the
system is under memory pressure.
uvm_km_valloc1(), and use it to express all of
uvm_km_valloc()
uvm_km_valloc_wait()
uvm_km_valloc_prefer()
uvm_km_valloc_prefer_wait()
uvm_km_valloc_align()
in terms of it by macro expansion.
Gone are the old kern_sysctl(), cpu_sysctl(), hw_sysctl(),
vfs_sysctl(), etc, routines, along with sysctl_int() et al. Now all
nodes are registered with the tree, and nodes can be added (or
removed) easily, and I/O to and from the tree is handled generically.
Since the nodes are registered with the tree, the mapping from name to
number (and back again) can now be discovered, instead of having to be
hard coded. Adding new nodes to the tree is likewise much simpler --
the new infrastructure handles almost all the work for simple types,
and just about anything else can be done with a small helper function.
All existing nodes are where they were before (numerically speaking),
so all existing consumers of sysctl information should notice no
difference.
PS - I'm sorry, but there's a distinct lack of documentation at the
moment. I'm working on sysctl(3/8/9) right now, and I promise to
watch out for buses.
copyin() or copyout().
uvm_useracc() tells us whether the mapping permissions allow access to
the desired part of an address space, and many callers assume that
this is the same as knowing whether an attempt to access that part of
the address space will succeed. however, access to user space can
fail for reasons other than insufficient permission, most notably that
paging in any non-resident data can fail due to i/o errors. most of
the callers of uvm_useracc() make the above incorrect assumption. the
rest are all misguided optimizations, which optimize for the case
where an operation will fail. we'd rather optimize for operations
succeeding, in which case we should just attempt the access and handle
failures due to insufficient permissions the same way we handle i/o
errors. since there appear to be no good uses of uvm_useracc(), we'll
just remove it.
(1) split the single list of pages allocated to a pool into three lists:
completely full, partially full, and completely empty.
there is no longer any need to traverse any list looking for a
certain type of page.
(2) replace the 8-element hash table for out-of-page page headers
with a splay tree.
these two changes (together with the recent enhancements to the wait code)
give us linear scaling for a fork+exit microbenchmark.
to improve scalability of operations on the map.
originally done by Niels Provos for OpenBSD.
tweaked for NetBSD by me with some advices from enami tsugutomo.
discussed on tech-kern@ and tech-perform@.
- after sleeping for memory, re-check if we have a page.
- put the allocated page to pageq to appease UVM_PAGE_TRKOWN.
- dequeue the page when doing ->K loan.
The case where l_stat == LSONPROC and l_cpu == curcpu cannot happen
because the pagedaemon is the LWP on curcpu and the pagedaemon is a
kernel thread and the code is only used by the pagedaemon.
See also updated patch in PR kern/23095, which I ment to checkin
originally.
uvm_swapout_threads will swapout LWPs which are running on another CPU:
- uvm_swapout_threads considers LWPs running on another CPU for swapout
if their l_swtime is high
- uvm_swapout_threads considers LWPs on the runqueue for swapout if their
l_swtime is high but these LWPs might be running by the time uvm_swapout
is called
symptoms of failure: panic in setrunqueue
fixes PR kern/23095
with the KVA of the newly-wired uarea.
This is useful on some architectures (e.g. xscale) where the uarea mapping
can be tweaked to use the mini-data cache instead of the main cache.
it may return space already in use as free space under some condition.
The symptom of the bug is that exec fails if stack is unlimited on
topdown VM kernel.
uvm_swap_free() being called with a zero slot; this might have been
the reason for crashes with sysvshm and heavy swapping.
(PR kern/22752 by Tom Spindler)
Confirmed by Chuck Silvers.
the `# swap page in use' and `# swap page only' counters. However, at the
time of swap device removal we can no longer figure out how many of the
bad swap pages are actually also `swap only' pages.
So, on swap I/O errors arrange things to not include the bad swap pages in
the `swpgonly' counter as follows: uvm_swap_markbad() decrements `swpgonly'
by the number of bad pages, and the various VM object deallocation routines
do not decrement `swpgonly' for swap slots marked as SWSLOT_BAD.
and make the stack and heap non-executable by default. the changes
fall into two basic catagories:
- pmap and trap-handler changes. these are all MD:
= alpha: we already track per-page execute permission with the (software)
PG_EXEC bit, so just have the trap handler pay attention to it.
= i386: use a new GDT segment for %cs for processes that have no
executable mappings above a certain threshold (currently the
bottom of the stack). track per-page execute permission with
the last unused PTE bit.
= powerpc/ibm4xx: just use the hardware exec bit.
= powerpc/oea: we already track per-page exec bits, but the hardware only
implements non-exec mappings at the segment level. so track the
number of executable mappings in each segment and turn on the no-exec
segment bit iff the count is 0. adjust the trap handler to deal.
= sparc (sun4m): fix our use of the hardware protection bits.
fix the trap handler to recognize text faults.
= sparc64: split the existing unified TSB into data and instruction TSBs,
and only load TTEs into the appropriate TSB(s) for the permissions.
fix the trap handler to check for execute permission.
= not yet implemented: amd64, hppa, sh5
- changes in all the emulations that put a signal trampoline on the stack.
instead, we now put the trampoline into a uvm_aobj and map that into
the process separately.
originally from openbsd, adapted for netbsd by me.
* Remove the "lwp *" argument that was added to vget(). Turns out
that nothing actually used it!
* Remove the "lwp *" arguments that were added to VFS_ROOT(), VFS_VGET(),
and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted
above, didn't use it).
* Remove all of the "lwp *" arguments to internal functions that were added
just to appease the above.
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
http://mail-index.netbsd.org/source-changes/2003/05/08/0068.html
There were some side-effects that I didn't anticipate, and fixing them
is proving to be more difficult than I thought, do just eject for now.
Maybe one day we can look at this again.
Fixes PR kern/21517.
space is advertised to UVM by making virtual_avail and virtual_end
first-class exported variables by UVM. Machine-dependent code is
responsible for initializing them before main() is called. Anything
that steals KVA must adjust these variables accordingly.
This reduces the number of instances of this info from 3 to 1, and
simplifies the pmap(9) interface by removing the pmap_virtual_space()
function call, and removing two arguments from pmap_steal_memory().
This also eliminates some kludges such as having to burn kernel_map
entries on space used by the kernel and stolen KVA.
This also eliminates use of VM_{MIN,MAX}_KERNEL_ADDRESS from MI code,
this giving MD code greater flexibility over the bounds of the managed
kernel virtual address space if a given port's specific platforms can
vary in this regard (this is especially true of the evb* ports).
first step towards per-device MAXPHYS, and has the beneficial side effect
of allowing clustering to MAXPHYS even on systems that need to run with
a reduced MAXBSIZE to get more metadata buffers.
* Remove DEFAULT_PAGE_SIZE. We don't use PAGE_SIZE the way Mach did.
* In uvm_setpagesize(), if we are called with uvmexp.pagesize == 0,
then assert that PAGE_SIZE != 0 (i.e. a constant), and set uvmexp.pagesize
accordingly.
* Provide defaults for MIN_PAGE_SIZE and MAX_PAGE_SIZE if not defined
by <machine/vmparam.h>. If PAGE_SIZE is not a constant, MIN_PAGE_SIZE
and MAX_PAGE_SIZE must be provided.
* If MIN_PAGE_SIZE and MAX_PAGE_SIZE are not equal (i.e. PAGE_SIZE may
not be a constant in all configurations), then ensure that PAGE_SIZE
and friends expand to variable references for LKMs.
* User allocates ZFOD region, but does not actually touch the buffer
to fault in the pages.
* In a loop, user writes this buffer to a network socket, triggering
sosend_loan().
* uvm_loan() calls uvm_loanzero() once for each page in the loaned
region (since the pages have not yet faulted in). This causes a
page to be allocated and zero'd. The result is the kernel spends
a lot of time allocating and zero'ing pages.
This fixes creates a special object which owns a single zero'd page.
This single zero'd page is used to satisfy all loans of non-resident
ZFOD mappings.
Thanks to Allen Briggs for discovering the problem and for providing
an initial patch.
previous entry. (not if the current entry starts at the end of the new
space; that case doesn't take into account if the new space had a specified
alignment).
means that the dynamic linker gets mapped in at the top of available
user virtual memory (typically just below the stack), shared libraries
get mapped downwards from that point, and calls to mmap() that don't
specify a preferred address will get mapped in below those.
This means that the heap and the mmap()ed allocations will grow
towards each other, allowing one or the other to grow larger than
before. Previously, the heap was limited to MAXDSIZ by the placement
of the dynamic linker (and the process's rlimits) and the space
available to mmap was hobbled by this reservation.
This is currently only enabled via an *option* for the i386 platform
(though other platforms are expected to follow). Add "options
USE_TOPDOWN_VM" to your kernel config file, rerun config, and rebuild
your kernel to take advantage of this.
Note that the pmap_prefer() interface has not yet been modified to
play nicely with this, so those platforms require a bit more work
(most notably the sparc) before they can use this new memory
arrangement.
This change also introduces a VM_DEFAULT_ADDRESS() macro that picks
the appropriate default address based on the size of the allocation or
the size of the process's text segment accordingly. Several drivers
and the SYSV SHM address assignment were changed to use this instead
of each one picking their own "default".
(there are still some details to work out) but expect that to go
away soon. To support these basic changes (creation of lfs_putpages,
lfs_gop_write, mods to lfs_balloc) several other changes were made, to
wit:
* Create a writer daemon kernel thread whose purpose is to handle page
writes for the pagedaemon, but which also takes over some of the
functions of lfs_check(). This thread is started the first time an
LFS is mounted.
* Add a "flags" parameter to GOP_SIZE. Current values are
GOP_SIZE_READ, meaning that the call should return the size of the
in-core version of the file, and GOP_SIZE_WRITE, meaning that it
should return the on-disk size. One of GOP_SIZE_READ or
GOP_SIZE_WRITE must be specified.
* Instead of using malloc(...M_WAITOK) for everything, reserve enough
resources to get by and use malloc(...M_NOWAIT), using the reserves if
necessary. Use the pool subsystem for structures small enough that
this is feasible. This also obsoletes LFS_THROTTLE.
And a few that are not strictly necessary:
* Moves the LFS inode extensions off onto a separately allocated
structure; getting closer to LFS as an LKM. "Welcome to 1.6O."
* Unified GOP_ALLOC between FFS and LFS.
* Update LFS copyright headers to correct values.
* Actually cast to unsigned in lfs_shellsort, like the comment says.
* Keep track of which segments were empty before the previous
checkpoint; any segments that pass two checkpoints both dirty and
empty can be summarily cleaned. Do this. Right now lfs_segclean
still works, but this should be turned into an effectless
compatibility syscall.
we read-lock the map and call uvm_map_lookup_entry() instead of simply
walking from the header to the next and to the next, etc.
Dumping from sparsely populated amaps could cause faults that would
result in amaps being split, which (in turn) resulted in the core
dumping routines dumping some regions of memory twice. This makes the
core file too large, the headers not match, gdb not work properly,
and so on.
Addresses PR 19260.
malloc types into a structure, a pointer to which is passed around,
instead of an int constant. Allow the limit to be adjusted when the
malloc type is defined, or with a function call, as suggested by
Jonathan Stone.
request may contain PGO_DONTCARE and nfs_getpages may unbusy them on error.
Fix is provided in PR#20028 by YAMAMOTO Takashi. (and same one is approved
by chuq while ago in private mail). It was my fault to forget to commit.
allocated ppref data to zero in the case of an amap that has empty
space at the front.
Don't set anything in the ppref array if "len" is zero.
Many thanks to Sami Kantoluoto for providing gdb access to a machine
that would reliably crash with problems related to the above, and to
Stephan Thesing for corroborating that the patch properly addressed
the problem.
Note that the ar_pageoff (and related variables) types must be changed
soon. The use of "int" here is not theoretically sufficient.
to sleep. Define UVM_KMF_NOWAIT in terms of UVM_FLAG_NOWAIT.
From Manuel Bouyer. Fixes a problem where any mapping with
read protection was created in a "nowait" context, causing
spurious failures.
uvm_map(). Change uvm_map() to honnor UVM_KMF_NOWAIT. For this, change
amap_extend() to take a flags parameter instead of just boolean for
direction, and introduce AMAP_EXTEND_FORWARDS and AMAP_EXTEND_NOWAIT flags
(AMAP_EXTEND_BACKWARDS is still defined as 0x0, to keep the code easier to
read).
Add a flag parameter to uvm_mapent_alloc().
This solves a problem a pool_get(PR_NOWAIT) could trigger a pool_get(PR_WAITOK)
in uvm_mapent_alloc().
Thanks to Chuck Silvers, enami tsugutomo, Andrew Brown and Jason R Thorpe
for feedback.
backed by physical pages (ie. because it reused a previously-freed one),
so that we can skip a bunch of useless work in that case.
this fixes the underlying problem behind PR 18543, and also speeds up fork()
quite a bit (eg. 7% on my pc, 1% on my ultra2) when we get a cache hit.
delay freeing the old am_ppref so that if we bail early due to
malloc() failures, valid ppref data hasn't been freed for no reason.
Based on comments from enami.
with:
Case #1 -- adjust offset: The slot offset in the aref can be
decremented to cover the required size addition.
Case #2 -- move pages and adjust offset: The slot offset is not large
enough, but the amap contains enough inactive space *after* the mapped
pages to make up the difference, so active slots are slid to the "end"
of the amap, and the slot offset is, again, adjusted to cover the
required size addition. This optimizes for hitting case #1 again on
the next small extension.
Case #3 -- reallocate, move pages, and adjust offset: There is not
enough inactive space in the amap, so the arrays are reallocated, and
the active pages are copied again to the "end" of the amap, and the
slot offset is adjusted to cover the required size. This also
optimizes for hitting case #1 on the next backwards extension.
This provides the missing piece in the "forward extension of
vm_map_entries" logic, so the merge failure counters have been
removed.
Not many applications will make any use of this at this time (except
for jvms and perhaps gcc3), but a "top-down" memory allocator will use
it extensively.
backwards and forwards) if the previous entry was backed by an amap.
Fixes pr kern/18789, where netscape 7 + a java applet actually manage
to incur forward and bimerges in userspace.
Code reviewed by fvdl and thorpej.
kqueue provides a stateful and efficient event notification framework
currently supported events include socket, file, directory, fifo,
pipe, tty and device changes, and monitoring of processes and signals
kqueue is supported by all writable filesystems in NetBSD tree
(with exception of Coda) and all device drivers supporting poll(2)
based on work done by Jonathan Lemon for FreeBSD
initial NetBSD port done by Luke Mewburn and Jason Thorpe
allocations can be merged either forwards or backwards, meaning no new
entries will be added to the list, and some can even be merged in both
directions, resulting in a surplus entry.
This code typically reduces the number of map entries in the
kernel_map by an order of magnitude or more. It also makes possible
recovery from the pathological case of "5000 processes created and
then killed", which leaves behind a large number of map entries.
The only forward merge case not covered is the instance of an amap
that has to be extended backwards (WIP). Note that this only affects
processes, not the kernel (the kernel doesn't use amaps), and that
merge opportunities like this come up *very* rarely, if at all. Eg,
after being up for eight days, I see only three failures in this
regard, and even those are most likely due to programs I'm developing
to exercise this case.
Code reviewed by thorpej, matt, christos, mrg, chuq, chuck, perry,
tls, and probably others. I'd like to thank my mother, the Hollywood
Foreign Press...
tearing down a vm_map. use this to skip the pmap_update()
at the end of all the removes, which allows pmaps to optimize
pmap tear-down. also, use the new pmap_remove_all() hook to
let the pmap implemenation know what we're up to.
return failure if swap is full and there are no free physical pages.
have malloc() use this flag if M_CANFAIL is passed to it.
use M_CANFAIL to allow amap_extend() to fail when memory is scarce.
this should prevent most of the remaining hangs in low-memory situations.
This merge changes the device switch tables from static array to
dynamically generated by config(8).
- All device switches is defined as a constant structure in device drivers.
- The new grammer ``device-major'' is introduced to ``files''.
device-major <prefix> char <num> [block <num>] [<rules>]
- All device major numbers must be listed up in port dependent majors.<arch>
by using this grammer.
- Added the new naming convention.
The name of the device switch must be <prefix>_[bc]devsw for auto-generation
of device switch tables.
- The backward compatibility of loading block/character device
switch by LKM framework is broken. This is necessary to convert
from block/character device major to device name in runtime and vice versa.
- The restriction to assign device major by LKM is completely removed.
We don't need to reserve LKM entries for dynamic loading of device switch.
- In compile time, device major numbers list is packed into the kernel and
the LKM framework will refer it to assign device major number dynamically.
the page is still loaned to an anon, we should put the page back on a
paging queue. this is because while pages loaned to the kernel really
do need to stay resident (since the kernel is accessing the physical
memory directly), pages loaned to anons can be paged out just fine.
(the page will be paged out twice, first to the object and then again
to the anon, but after that the page can be reused.)
-pass vm_physseg* instead of physseg index, and PFN (int) instead
of physical address (could be done even more)
-simplify detection of boundary crossing and behave more intelligently
in this case
-take stuff out of the inner loops, or put into "#ifdef DEBUG"
(because we move along physsegs we don't need to check that the
pages are physically contigous)
-make the "simple" and "contigous" branches look more uniform; at
least the outer loops might coalesce one day
Makoto Fujiwara <makoto@ki.nu> and Manuel Bouyer <bouyer@netbsd.org>.
Help from Allen Briggs, Jason Thorpe, and Matt Thomas.
We need to call cpu_cache_probe() early in boot (machdep.c).
Add 603 info for completeness, and use NBPG not PAGESIZE, as the
latter relies on uvm being setup (cpu_subr.c).
Let uvm_page_recolor() be called before uvm has been set up; just
note the page coloring value (uvm_page.c).
obey the preferences expressed by freelist assignment,
to avoid wasting valuable "low memory" to devices which
don't really need it.
comments:
-I'm not sure searching the physsegs within a freelist
beginning with the biggest is the right thing. This is
what the "memory steal" code in uvm_page.c does, so
keep it consistent.
-There seems to be some confusion whether the upper
address limit passed is inclusive or not. Stays on
the save side, possibly leaving one page out.
-The boundary/pagemask check can be simplified, also some
arguments passed are only used for diagnostic checks.
-Integration with UVM_PAGE_TRKOWN???
no alignment / boundary / nsegs restrictions apply.
This one doesn't insist in a contigous range, and it honours the "waitok"
flag, thus succeeds in situations which were hopeless with the existing one.
(A solution which searches for a minimum number of contiguous ranges using
some best-fit or so algorithm would be expensive to implement; I believe the
"either-or" done here does reflect the current use by bus_dma quite well.)
Now agp memory allocation is robust for me. (tested on i810)
we can't simply reuse the pointor to the page. Instead, we need to
acquire it again. So, rearrange the loop like genfs_putpages() does.
Reviewed by chuq.
This makes `tail -<N> <FILE> | cat > file' correctly, where <FILE> is
a regular file larger than 10Mbytes (makes tail to map part of file)
and <N> is big enough to produce output larger than 8kbytes (makes pipe
to use page loan facility). Problem reported by FUKAUMI Naoki on japanese
local mailing list.
uvm_swap_stats(). This is done in order to allow COMPAT_* swapctl()
emulation to use it directly without going through sys_swapctl().
The problem with using sys_swapctl() there is that it involves
copying the swapent array to the stackgap, and this array's size
is not known at build time. Hence it would not be possible to
ensure it would fit in the stackgap in any case.
deal with shortages of the VM maps where the backing pages are mapped
(usually kmem_map). Try to deal with this:
* Group all information about the backend allocator for a pool in a
separate structure. The pool references this structure, rather than
the individual fields.
* Change the pool_init() API accordingly, and adjust all callers.
* Link all pools using the same backend allocator on a list.
* The backend allocator is responsible for waiting for physical memory
to become available, but will still fail if it cannot callocate KVA
space for the pages. If this happens, carefully drain all pools using
the same backend allocator, so that some KVA space can be freed.
* Change pool_reclaim() to indicate if it actually succeeded in freeing
some pages, and use that information to make draining easier and more
efficient.
* Get rid of PR_URGENT. There was only one use of it, and it could be
dealt with by the caller.
From art@openbsd.org.
just skip that page. this situation can arise legitimately when a file
with a wired mapping is truncated so that a wired page is no longer
part of the file.
from VM_FAULT_WIRE in that when the pages being wired are faulted in,
the simulated fault is at the maximum protection allowed for the mapping
instead of the current protection. use this in uvm_map_pageable{,_all}()
to fix the problem where writing via ptrace() to shared libraries that
are also mapped with wired mappings in another process causes a
diagnostic panic when the wired mapping is removed.
this is a really obscure problem so it deserves some more explanation.
ptrace() writing to another process ends up down in uvm_map_extract(),
which for MAP_PRIVATE mappings (such as shared libraries) will cause
the amap to be copied or created. then the amap is made shared
(ie. the AMAP_SHARED flag is set) between the kernel and the ptrace()d
process so that the kernel can modify pages in the amap and have the
ptrace()d process see the changes. then when the page being modified
is actually faulted on, the object pages (from the shared library vnode)
is copied to a new anon page and inserted into the shared amap.
to make all the processes sharing the amap actually see the new anon
page instead of the vnode page that was there before, we need to
invalidate all the pmap-level mappings of the vnode page in the pmaps
of the processes sharing the amap, but we don't have a good way of
doing this. the amap doesn't keep track of the vm_maps which map it.
so all we can do at this point is to remove all the mappings of the
page with pmap_page_protect(), but this has the unfortunate side-effect
of removing wired mappings as well. removing wired mappings with
pmap_page_protect() is a legitimate operation, it can happen when a file
with a wired mapping is truncated. so the pmap has no way of knowing
whether a request to remove a wired mapping is normal or when it's due to
this weird situation. so the pmap has to remove the weird mapping.
the process being ptrace()d goes away and life continues. then,
much later when we go to unwire or remove the wired vm_map mapping,
we discover that the pmap mapping has been removed when it should
still be there, and we panic.
so where did we go wrong? the problem is that we don't have any way
to update just the pmap mappings that need to be updated in this
scenario. we could invent a mechanism to do this, but that is much
more complicated than this change and it doesn't seem like the right
way to go in the long run either.
the real underlying problem here is that wired pmap mappings just
aren't a good concept. one of the original properties of the pmap
design was supposed to be that all the information in the pmap could
be thrown away at any time and the VM system could regenerate it all
through fault processing, but wired pmap mappings don't allow that.
a better design for UVM would not require wired pmap mappings,
and Chuck C. and I are talking about this, but it won't be done
anytime soon, so this change will do for now.
this change has the effect of causing MAP_PRIVATE mappings to be
copied to anonymous memory when they are mlock()d, so that uvm_fault()
doesn't need to copy these pages later when called from ptrace(), thus
avoiding the call to pmap_page_protect() and the panic that results
from this when the mlock()d region is unlocked or freed. note that
this change doesn't help the case where the wired mapping is MAP_SHARED.
discussed at great length with Chuck Cranor.
fixes PRs 10363, 12554, 12604, 13041, 13487, 14580 and 14853.
we need to make sure that vnode pages are written to disk at least once,
otherwise processes could gain access to whatever data was previously stored
in disk blocks which are freshly allocated to a file.
uobject and uanon pointers rather than at the PQ_ANON flag to determine
which lock to hold, since PQ_ANON can be clear even when the anon's lock
is the one which we should hold (if the page was loaned from an object
and then freed by the object).
if the vec pointer is valid rather than using uvm_useracc().
uvm_useracc() just tells you if the permissions of a user mapping allow
the desired access, not whether faulting on that mapping will succeed.
will be allocated for the respective usage types when there is contention
for memory.
replace "vnode" and "vtext" with "file" and "exec" in uvmexp field names
and sysctl names.
- fix the loaned case in uvm_pagefree().
- redo uvmexp.swpgonly accounting to work with page loaning.
add an assertion before each place we adjust uvmexp.swpgonly.
- fix uvm_km_pgremove() to always free any swap space associated with
the range being removed.
- get rid of UVM_LOAN_WIRED flag. instead, we just make sure that
pages loaned to the kernel are never on the page queues.
this allows us to assert that pages are not loaned and wired
at the same time.
- add yet more assertions.
(either the current protection or the max protection) that reference
vnodes associated with a file system mounted with the NOEXEC option.
uvm_mmap(): Don't allow PROT_EXEC mappings to be established of vnodes
which are associated with a file system mounted with the NOEXEC option.
executable mappings. Stop overloading VTEXT for this purpose (VTEXT
also has another meaning).
- Rename vn_marktext() to vn_markexec(), and use it when executable
mappings of a vnode are established.
- In places where we want to set VTEXT, set it in v_flag directly, rather
than making a function call to do this (it no longer makes sense to
use a function call, since we no longer overload VTEXT with VEXECMAP's
meaning).
VEXECMAP suggested by Chuq Silvers.
are only wired if this flag is present (i.e. they are not wired by default now)
loaned pages are unloaned via new uvm_unloan(), uvm_unloananon() and
uvm_unloanpage() are no longer exported
adjust uvm_unloanpage() to unwire the pages if UVM_LOAN_WIRED is specified
mark uvm_loanuobj() and uvm_loanzero() static also in function implementation
kern/sys_pipe.c: uvm_unloanpage() --> uvm_unloan()
- remove special treatment of pager_map mappings in pmaps. this is
required now, since I've removed the globals that expose the address range.
pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's
no longer any need to special-case it.
- eliminate struct uvm_vnode by moving its fields into struct vnode.
- rewrite the pageout path. the pager is now responsible for handling the
high-level requests instead of only getting control after a bunch of work
has already been done on its behalf. this will allow us to UBCify LFS,
which needs tighter control over its pages than other filesystems do.
writing a page to disk no longer requires making it read-only, which
allows us to write wired pages without causing all kinds of havoc.
- use a new PG_PAGEOUT flag to indicate that a page should be freed
on behalf of the pagedaemon when it's unlocked. this flag is very similar
to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the
pageout fails due to eg. an indirect-block buffer being locked.
this allows us to remove the "version" field from struct vm_page,
and together with shrinking "loan_count" from 32 bits to 16,
struct vm_page is now 4 bytes smaller.
- no longer use PG_RELEASED for swap-backed pages. if the page is busy
because it's being paged out, we can't release the swap slot to be
reallocated until that write is complete, but unlike with vnodes we
don't keep a count of in-progress writes so there's no good way to
know when the write is done. instead, when we need to free a busy
swap-backed page, just sleep until we can get it busy ourselves.
- implement a fast-path for extending writes which allows us to avoid
zeroing new pages. this substantially reduces cpu usage.
- encapsulate the data used by the genfs code in a struct genfs_node,
which must be the first element of the filesystem-specific vnode data
for filesystems which use genfs_{get,put}pages().
- eliminate many of the UVM pagerops, since they aren't needed anymore
now that the pager "put" operation is a higher-level operation.
- enhance the genfs code to allow NFS to use the genfs_{get,put}pages
instead of a modified copy.
- clean up struct vnode by removing all the fields that used to be used by
the vfs_cluster.c code (which we don't use anymore with UBC).
- remove kmem_object and mb_object since they were useless.
instead of allocating pages to these objects, we now just allocate
pages with no object. such pages are mapped in the kernel until they
are freed, so we can use the mapping to find the page to free it.
this allows us to remove splvm() protection in several places.
The sum of all these changes improves write throughput on my
decstation 5000/200 to within 1% of the rate of NetBSD 1.5
and reduces the elapsed time for "make release" of a NetBSD 1.5
source tree on my 128MB pc to 10% less than a 1.5 kernel took.
This will allow improvements to the pmaps so that they can more easily defer expensive operations, eg tlb/cache flush, til the last possible moment.
Currently this is a no-op on most platforms, so they should see no difference.
Reviewed by Jason.
kernel_map. use this instead of the static map entries when allocating
map entries for kernel_map. this greatly reduces the number of static
map entries used and should eliminate the problems with running out.
loop returns 0. loanentry was returning >0, but was unlocking the maps
(because of the zero). reworked to avoid this. problem reported by
chuck silvers. also clarify a comment that jdolecek asked about.
provides us more flexibility with pageq-locked fields, and clarifies the
locking semantics for platforms which cannot address shorts.
From Ross Harvey.
an spl-protected "interrupt safe map" list, simply require that callers
of uvm_fault() never call us in interrupt context (MD code must make
the assertion), and check for interrupt-safe maps in uvmfault_lookup()
before we lock the map.
between creation of a file descriptor and close(2) when using kernel
assisted threads. What we do is stick descriptors in the table, but
mark them as "larval". This causes essentially everything to treat
it as a non-existent descriptor, except for fdalloc(), which sees a
filled slot so that it won't (incorrectly) allocate it again. When
a descriptor is fully constructed, the code that has constructed it
marks it as "mature" (which actually clears the "larval" flag), and
things continue to work as normal.
While here, gather all the code that gets a descriptor from the table
into a fd_getfile() function, and call it, rather than having the
same (sometimes incorrect) code copied all over the place.
some time ago. The mistake was to check that the page was not
referenced since the last active scan before moving it to inactive.
Now we just clear reference and move it to inacive (which is where
the second clock hand sweep occurs).
dynamically re-coloring pages; as machine-dependent code discovers
the size of the system's caches, it may call uvm_page_recolor() with
the new number of colors to use. If the new mumber of colors is
smaller (or equal to) the current number of colors, then uvm_page_recolor()
is a no-op.
The system defaults to one bucket if machine-dependent code does not
initialize uvmexp.ncolors before uvm_page_init() is called.
Note that the number of color bins should be initialized to something
reasonable as early as possible -- for many early memory allocations,
we live with the consequences of the page choice for the lifetime of
the boot.
each vm_page structure. Add a VM_MDPAGE_INIT() macro to init this
data when pages are initialized by UVM. These macros are mandatory,
but ports may #define them to nothing if they are not needed/used.
This deprecates struct pmap_physseg. As a transitional measure,
allow a port to #define PMAP_PHYSSEG so that it can continue to
use it until its pmap is converted to use VM_MDPAGE_MEMBERS.
Use all this stuff to eliminate a lot of extra work in the Alpha
pmap module (it's smaller and faster now). Changes to other pmap
modules will follow.
algorithm (Solaris calls this "Bin Hopping").
This implementation currently relies on MD code to define a
constant defining the number of buckets. This will change
reasonably soon (MD code will be able to dynamically size
the bucket array).
- pmap_enter()
- pmap_remove()
- pmap_protect()
- pmap_kenter_pa()
- pmap_kremove()
as described in pmap(9).
These calls are relatively conservative. It may be possible to
optimize these a little more.
which have pmap_steal_memory(). This is to reduce the API differences
between pmaps that implement pmap_steal_memory() and pmaps which do
not.
Note that pmap_steal_memory() needs to adjust *vstartp and/or
*vendp only if it used addresses within the range provided to UVM
via the pmap_virtual_space() call. I.e. it is not necessary to do
so in any current pmap_steal_memory() implementation.
on locking rules to make code easier to understand. locking in
uvm_loananon still needs some work on fringe cases where anon's page
is actually on loan from a uobj.
if uvm_loanentry() returned 0; otherwise, the unlocking would already
have been done by uvmfault_unlockall() call in uvm_loanentry().
Okay'ed by Chuck Silvers
pk