We added _NORECLAIM to request that kmem_alloc_contig_pages() not spend
time scanning physical memory for candidates to reclaim. In some
situations the scanning can induce large amounts of undesirable latency,
and it's less important that the request be satisfied than it is that we
not spend many milliseconds scanning.
The problem extends to vm_reserv_reclaim_contig(), which unlike
vm_reserv_reclaim() may have to scan the entire list of partially
populated reservations. Use VM_ALLOC_NORECLAIM to request that this
scan not be executed.[1]
As a side effect, this fixes a regression in 02fb0585e7 ("vm_page:
Drop handling of VM_ALLOC_NOOBJ in vm_page_alloc_contig_domain()")
where VM_ALLOC_CONTIG was not included in VPAC_FLAGS or VPANC_FLAGS even
though it is not masked by kmem_alloc_contig_pages().[2]
Reported by: gallatin [1], glebius [2]
Reviewed by: alc, glebius, kib
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32899
Kegs with no items reserved have uk_reserve = 0. So the check
keg->uk_reserve >= dom->ud_free_items will be true once all slabs are
depleted. Then, rather than go and allocate a fresh slab, we return to
the cache layer.
The intent was to do this only when the keg actually has a reserve, so
modify the check to verify this first. Another approach would be to
make uk_reserve signed and set it to -1 until uma_zone_reserve() is
called, but this requires a few casts elsewhere.
Fixes: 1b2dcc8c54 ("uma: Avoid depleting keg reserves when filling a bucket")
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32516
M_USE_RESERVE is used in a couple of places in the VM to avoid unbounded
recursion when the direct map is not available, as is the case on 32-bit
platforms or when certain kernel sanitizers (KASAN and KMSAN) are
enabled. For example, to allocate KVA, the kernel might allocate a
kernel map entry, which might require a new slab, which requires KVA.
For these zones, we use uma_prealloc() to populate a reserve of items,
and then in certain serialized contexts M_USE_RESERVE can be used to
guarantee a successful allocation. uma_prealloc() allocates the
requested number of items, distributing them evenly among NUMA domains.
Thus, in a first-touch zone, to satisfy an M_USE_RESERVE allocation we
might have to check the slab lists of other domains than the current one
to provide the semantics expected by consumers.
So, try harder to find an item if M_USE_RESERVE is specified and the keg
doesn't have anything for current (first-touch) domain. Specifically,
fall back to a round-robin slab allocation. This change fixes boot-time
panics on NUMA systems with KASAN or KMSAN enabled.[1]
Alternately we could have uma_prealloc() allocate the requested number
of items for each domain, but for some existing consumers this would be
quite wasteful. In general I think keg_fetch_slab() should try harder
to find free slabs in other domains before trying to allocate fresh
ones, but let's limit this to M_USE_RESERVE for now.
Also fix a separate problem that I noticed: in a non-round-robin slab
allocation with M_WAITOK, rather than sleeping after a failed slab
allocation we simply try again. Call vm_wait_domain() before retrying.
Reported by: mjg, tuexen [1]
Reviewed by: alc
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32515
For making the call faster, do not count active/inactive object queues,
and do not report vnode info if any (for tmpfs).
Reviewed by: markj
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D31163
vm_reserv_reclaim_*() will release pages to the default freepool, not
the direct freepool from which noobj allocations are drawn. But if both
pools are empty, the noobj allocator variants must break reservations to
make progress.
Reported by: cy
Reviewed by: kib (previous version)
Fixes: b498f71bc5 ("vm_page: Add a new page allocator interface for unnamed pages")
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32592
Now vm_page_alloc() and friends will unconditionally preserve PG_ZERO,
so there is no point in setting this flag.
Eliminate a local variable and add a comment explaining why we
prioritize the allocation when the process is doomed.
No functional change intended.
Reviewed by: kib, alc
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32036
As in vm_page_alloc_domain_after(), unconditionally preserve PG_ZERO.
Implement vm_page_alloc_noobj_contig_domain().
Reviewed by: kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32034
This makes the allocator simpler since it can assume object != NULL.
Also modify the function to unconditionally preserve PG_ZERO, so
VM_ALLOC_ZERO is effectively ignored (and still must be implemented by
the caller for now).
Reviewed by: alc, kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32033
This is the same as vm_page_alloc_noobj(), but allocates physically
contiguous runs of memory. For now it is implemented in terms of
vm_page_alloc_contig(), with the difference that
vm_page_alloc_noobj_contig() implements VM_ALLOC_ZERO by zeroing the
page.
Reviewed by: alc, kib
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32005
Remove page zeroing code from consumers and stop specifying
VM_ALLOC_NOOBJ. In a few places, also convert an allocation loop to
simply use VM_ALLOC_WAITOK.
Similarly, convert vm_page_alloc_domain() callers.
Note that callers are now responsible for assigning the pindex.
Reviewed by: alc, hselasky, kib
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D31986
The diff adds vm_page_alloc_noobj() and vm_page_alloc_noobj_domain().
These mostly correspond to vm_page_alloc() and vm_page_alloc_domain()
when no VM object is specified, with the exception that they handle
VM_ALLOC_ZERO by zeroing the page, rather than by preserving PG_ZERO.
This simplifies callers and will permit simplification of the
vm_page_alloc_domain() definition.
Since the new allocator variant is similar to vm_page_alloc_freelist(),
implement both of them using a common backend allocator function. No
functional change intended.
Reviewed by: alc, kib
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D31985
Make vmdaemon timeout configurable, so that one can adjust
how often it runs.
Here's a trick: set this to 1, then run 'limits -m 0 sh',
then run whatever you want with 'ktrace -it XXX', and observe
how the working set changes over time.
Reviewed By: kib
Sponsored By: EPSRC
Differential Revision: https://reviews.freebsd.org/D22038
Calling setrlimit with stack gap enabled and with low values of stack
resource limit often caused the program to abort immediately after
exiting the syscall. This happened due to the fact that the resource
limit was calculated assuming that the stack started at sv_usrstack,
while with stack gap enabled the stack is moved by a random number
of bytes.
Save information about stack size in struct vmspace and adjust the
rlim_cur value. If the rlim_cur and stack gap is bigger than rlim_max,
then the value is truncated to rlim_max.
PR: 253208
Reviewed by: kib
Obtained from: Semihalf
Sponsored by: Stormshield
MFC after: 1 month
Differential Revision: https://reviews.freebsd.org/D31516
sys/sysctl.h moved struct thread forward declaration under #ifdef
_KERNEL and so this header fails when included from userland. Add a
forward declaration here.
Fixes: 99eefc727e
Sponsored by: Netflix
Wakeup in vm_waitpfault() does not mean that the thread would get the
page on the next vm_page_alloc() call, other thread might steal the free
page we were waiting for. On the other hand, this wakeup might come much
earlier than just vm_pfault_oom_wait seconds, if the rate of the page
reclamation is high enough.
If wakeups come fast and we loose the allocation race enough times, OOM
could be undeservably triggered much earlier than vm_pfault_oom_attempts
x vm_pfault_oom_wait seconds. Fix it by not counting the number of sleeps,
but measuring the time to th first allocation failure, and triggering OOM
when it was older than oom_attempts x oom_wait seconds.
Reviewed by: markj
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D32287
The function is identical in each minidump implementation, so move it to
vm_phys.c. The only slight exception is powerpc where the function was
public, for use in moea64_scan_pmap().
Reviewed by: kib, markj, imp (earlier version)
MFC after: 2 weeks
Sponsored by: Juniper Networks, Inc.
Sponsored by: Klara, Inc.
Differential Revision: https://reviews.freebsd.org/D31884
GEOM_ELI needs to know the value, cause it will soon have special
memory handling for IO operations associated with swap.
Move initialization to swap_pager_init(), which is executed at
SI_SUB_VM, unlike swap_pager_swap_init(), which would be executed
only when a swap is configured. GEOM_ELI might need the value at
SI_SUB_DRIVERS, when disks are tasted by GEOM.
Reviewed by: kib
Differential Revision: https://reviews.freebsd.org/D24400
This is useful for measuring the number of pages that could be freed
from a NOFREE zone under memory pressure.
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
If phys_avail[] segment only intersect with some vm_phys segment, add
pages from it to the free list that belong to the given vm_phys_seg,
instead of dropping them.
The vm_phys segments are generally result of subdivision of phys_avail
segments, for instance DMA32 or LOWMEM boundaries split them. On
amd64, after UEFI in-place kernel activation (copy_staging disable)
was enabled, we typically have a large phys_avail[] segment below 4G
which crosses LOWMEM (1M) boundary. With the current way of requiring
phys_avail[] fully fit into vm_phys_seg, this memory was ignored.
Reported by: madpilot
Reviewed by: markj
Discussed with: alc
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D31958
This interface is used solely by md(4) when the MD_RESERVE flag is
specified, as in `mdconfig -a -t swap -s 1G -o reserve`. It
pre-allocates swap blocks for the entire object.
The number of blocks to be reserved is specified as a vm_size_t, but
swp_pager_getswapspace() can allocate at most INT_MAX blocks. vm_size_t
also seems like the incorrect type to use here it refers only to the
size of the VM object, not the size of a mapping. So:
- change the type of "size" in swap_pager_reserve() to vm_pindex_t, and
- clamp the requested number of blocks for a single
swp_pager_getswapspace() call to INT_MAX.
Reported by: syzkaller
Reviewed by: dougm, alc, kib
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D31875
In fee2a2fa39 the KASSERTs in
vm_page_unwire_noq() changed from "vm_page_unwire" to "vm_page_unref".
While the former no longer was part of that function the latter does
not exist as a function and is highly confusing when hit when using
tools to lookup the functions and not doing a full-text search.
Use %s __func__ for printing the function name, as that will do the
right thing as code moves around and functions get renamed.
Hit: while debugging a wired page leak with linuxkpi/iwlwifi
Sponsored by: The FreeBSD Foundation
Reviewed by: markj
Differential Revision: https://reviews.freebsd.org/D31635
For now, just hook the allocation path: upon allocation, items are
marked as initialized (absent M_ZERO). Some zones are exempted from
this when it would otherwise raise false positives.
Use kmsan_orig() to update the origin map for UMA and malloc(9)
allocations. This allows KMSAN to print the return address when an
uninitialized UMA item is implicated in a report. For example:
panic: MSan: Uninitialized UMA memory from m_getm2+0x7fe
Sponsored by: The FreeBSD Foundation
- During boot, allocate PDP pages for the shadow maps. The region above
KERNBASE is currently not shadowed.
- Create a dummy shadow for the vm page array. For now, this array is
not protected by the shadow map to help reduce kernel memory usage.
- Grow shadows when growing the kernel map.
- Increase the default kernel stack size when KMSAN is enabled. As with
KASAN, sanitizer instrumentation appears to create stack frames large
enough that the default value is not sufficient.
- Disable UMA's use of the direct map when KMSAN is configured. KMSAN
cannot validate the direct map.
- Disable unmapped I/O when KMSAN configured.
- Lower the limit on paging buffers when KMSAN is configured. Each
buffer has a static MAXPHYS-sized allocation of KVA, which in turn
eats 2*MAXPHYS of space in the shadow map.
Reviewed by: alc, kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D31295
This KPI is created in addition to the existing vnode_pager_setsize(9)
KPI. The KPI is intended for file systems that are able to turn a range
of file into sparse range, also known as hole-punching.
Sponsored by: The FreeBSD Foundation
Reviewed by: kib
Differential Revision: https://reviews.freebsd.org/D27194
and remove repetetive code that calculates vnode locking type for write.
Reviewed by: khng, markj
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D31405
which is the place to put MD asserts about allocated pages.
On amd64, verify that allocated page does not belong to the kernel
(text, data) or early allocated pages.
Reviewed by: markj
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D31121
redzone(9) does some munging of the allocation to insert redzones before
and after a valid memory buffer, but KASAN does not know about this and
will raise false positives if both are configured. Until this is fixed,
do not allow both to be configured. Note that KASAN provides similar
checking on its own but currently does not force the creation of
redzones for all UMA allocations; this should be addressed as well.
Sponsored by: The FreeBSD Foundation
- Ensure that all items returned by UMA are aligned to
KASAN_SHADOW_SCALE (8). This was true in practice since smaller
alignments are not used by any consumers, but we should enforce it
anyway.
- Use a non-zero code for marking redzones that appear naturally in
items that are not a multiple of the scale factor in size. Currently
we do not modify keg layouts to force the creation of redzones.
- Use a non-zero code for marking freed per-CPU items, otherwise
accesses of freed per-CPU items are not detected by the runtime.
Sponsored by: The FreeBSD Foundation
Remove OBJT_SWAP_TMPFS. Move tmpfs-specific swap pager bits into
tmpfs_subr.c.
There is no longer any code to directly support tmpfs in sys/vm, most
tmpfs knowledge is shared by non-anon swap object type implementation.
The tmpfs-specific methods are provided by registered tmpfs pager, which
inherits from the swap pager.
Reviewed by: markj
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D30168
Pager is allowed to inherit part of its implementation from the existing
pager, which is done by copying non-NULL virtual method slots.
Reviewed by: markj
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D30168
Mostly in cases where OBJ_SWAP flag works as well, or by reversing the
condition so that object types can be listed.
Reviewed by: markj
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D30168
This avoids the need to know all existing object types in advance, by the
cost of loosing the assert that unknown object type is handled in a sane
manner.
Reviewed by: markj
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D30168
This is OBJT_SWAP pager, specialized for tmpfs. Right now, both swap pager
and generic vm code have to explicitly handle swap objects which are tmpfs
vnode v_object, in the special ways. Replace (almost) all such places with
proper methods.
Since VM still needs a notion of the 'swap object', regardless of its
use, add yet another type-classification flag OBJ_SWAP. Set it in
vm_object_allocate() where other type-class flags are set.
This change almost completely eliminates the knowledge of tmpfs from VM,
and opens a way to make OBJT_SWAP_TMPFS loadable from tmpfs.ko.
Reviewed by: markj
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D30070
