Nishanth Aravamudan
d1c3fb1f8f
hugetlb: introduce nr_overcommit_hugepages sysctl ...
hugetlb: introduce nr_overcommit_hugepages sysctl
While examining the code to support /proc/sys/vm/hugetlb_dynamic_pool, I
became convinced that having a boolean sysctl was insufficient:
1) To support per-node control of hugepages, I have previously submitted
patches to add a sysfs attribute related to nr_hugepages. However, with
a boolean global value and per-mount quota enforcement constraining the
dynamic pool, adding corresponding control of the dynamic pool on a
per-node basis seems inconsistent to me.
2) Administration of the hugetlb dynamic pool with multiple hugetlbfs
mount points is, arguably, more arduous than it needs to be. Each quota
would need to be set separately, and the sum would need to be monitored.
To ease the administration, and to help make the way for per-node
control of the static & dynamic hugepage pool, I added a separate
sysctl, nr_overcommit_hugepages. This value serves as a high watermark
for the overall hugepage pool, while nr_hugepages serves as a low
watermark. The boolean sysctl can then be removed, as the condition
nr_overcommit_hugepages > 0
indicates the same administrative setting as
hugetlb_dynamic_pool == 1
Quotas still serve as local enforcement of the size of the pool on a
per-mount basis.
A few caveats:
1) There is a race whereby the global surplus huge page counter is
incremented before a hugepage has allocated. Another process could then
try grow the pool, and fail to convert a surplus huge page to a normal
huge page and instead allocate a fresh huge page. I believe this is
benign, as no memory is leaked (the actual pages are still tracked
correctly) and the counters won't go out of sync.
2) Shrinking the static pool while a surplus is in effect will allow the
number of surplus huge pages to exceed the overcommit value. As long as
this condition holds, however, no more surplus huge pages will be
allowed on the system until one of the two sysctls are increased
sufficiently, or the surplus huge pages go out of use and are freed.
Successfully tested on x86_64 with the current libhugetlbfs snapshot,
modified to use the new sysctl.
Signed-off-by: Nishanth Aravamudan <nacc@us.ibm.com>
Acked-by: Adam Litke <agl@us.ibm.com>
Cc: William Lee Irwin III <wli@holomorphy.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-12-17 19:28:17 -08:00
net: libertas sdio driver
2007-10-17 22:51:13 +02:00
[NETFILTER]: Sort matches/targets in Kbuild file
2007-11-07 04:08:21 -08:00
[NETFILTER]: Sort matches/targets in Kbuild file
2007-11-07 04:08:21 -08:00
[NETFILTER]: Sort matches/targets in Kbuild file
2007-11-07 04:08:21 -08:00
fs/nfsd/export.c: make 3 functions static
2007-10-16 09:43:10 -07:00
bitmap.h: remove dead artifacts
2007-10-17 08:43:03 -07:00
Merge branch 'linus' of master.kernel.org:/pub/scm/linux/kernel/git/perex/alsa
2007-10-16 10:13:38 -07:00
SUNRPC: make sunrpc/xprtsock.c:xs_setup_{udp,tcp}() static
2007-11-26 16:24:50 -05:00
ACPI: fix two IRQ8 issues in IOAPIC mode
2007-11-20 01:16:29 -05:00
Remove struct task_struct::io_wait
2007-10-18 14:37:20 -07:00
apm_event{,info}_t are userspace types
2007-12-17 19:28:16 -08:00
libata: handle broken cable reporting
2007-11-05 18:10:28 -05:00
Driver for the Atmel on-chip SSC on AT32AP and AT91
2007-10-17 08:42:47 -07:00
[PATCH] audit: watching subtrees
2007-10-21 02:37:45 -04:00
increase AT_VECTOR_SIZE to terminate saved_auxv properly
2007-10-17 08:43:00 -07:00
mm: per device dirty threshold
2007-10-17 08:42:45 -07:00
Implement file posix capabilities
2007-10-17 08:43:07 -07:00
bit_spin_lock: use lock bitops
2007-10-18 14:37:29 -07:00
remove BITS_TO_TYPE macro
2007-10-19 11:53:42 -07:00
remove BITS_TO_TYPE macro
2007-10-19 11:53:42 -07:00
Add UNPLUG traces to all appropriate places
2007-11-09 13:41:32 +01:00
Revert "x86_64: allocate sparsemem memmap above 4G"
2007-10-29 14:05:37 -07:00
capabilities: clean up file capability reading
2007-10-22 08:13:18 -07:00
Typo: depricated -> deprecated
2007-10-20 03:10:57 +02:00
Add cgroupstats
2007-10-19 11:53:36 -07:00
sched: cpu accounting controller (V2)
2007-12-02 20:04:49 +01:00
Add cgroupstats
2007-10-19 11:53:36 -07:00
Fix discrepancy between VDSO based gettimeofday() and sys_gettimeofday().
2007-10-18 14:37:20 -07:00
Replace __attribute_pure__ with __pure
2007-10-18 14:37:32 -07:00
Force erroneous inclusions of compiler-*.h files to be errors
2007-10-17 08:42:47 -07:00
Force erroneous inclusions of compiler-*.h files to be errors
2007-10-17 08:42:47 -07:00
Force erroneous inclusions of compiler-*.h files to be errors
2007-10-17 08:42:47 -07:00
Permit silencing of __deprecated warnings.
2007-10-25 15:10:17 -07:00
sched: fix fastcall mismatch in completion APIs
2007-10-24 18:23:48 +02:00
uvesafb: add connector entries
2007-10-16 09:43:13 -07:00
serial: turn serial console suspend a boot rather than compile time option
2007-10-18 14:37:19 -07:00
vt/vgacon: Check if screen resize request comes from userspace
2007-10-16 09:43:20 -07:00
unicode diacritics support
2007-10-17 08:42:52 -07:00
Redefine {un}register_hotcpu_notifier() !HOTPLUG_CPU stubs
2007-10-18 14:37:28 -07:00
cpuidle: fix HP nx6125 regression
2007-11-19 21:43:22 -05:00
hotplug cpu: migrate a task within its cpuset
2007-10-19 11:53:44 -07:00
cramfs: error message about endianess
2007-10-17 08:42:53 -07:00
crypto: convert crypto.h to UTF-8
2007-10-19 23:07:36 +02:00
Char: cyclades, fix some -W warnings
2007-10-18 14:37:26 -07:00
DCA: Add Direct Cache Access driver
2007-10-16 09:43:09 -07:00
[PATCH] audit: watching subtrees
2007-10-21 02:37:45 -04:00
[DCCP]: Convert Reset code into socket error number
2007-10-24 10:27:48 -02:00
Add cgroupstats
2007-10-19 11:53:36 -07:00
dm: bounce_pfn limit added
2007-11-02 08:47:25 +01:00
Revert "Driver core: remove class_device_*_bin_file"
2007-10-31 12:51:29 -07:00
dm io:ctl use constant struct size
2007-10-20 02:00:58 +01:00
stop using DMA_xxBIT_MASK
2007-10-18 14:37:21 -07:00
Intel IOMMU: DMAR fault handling support
2007-10-22 08:13:19 -07:00
Make asm-x86/bootparam.h includable from userspace.
2007-10-23 15:49:47 +10:00
kexec: add BSS to resource tree
2007-10-22 08:13:19 -07:00
efs: new export ops
2007-10-22 08:13:20 -07:00
increase AT_VECTOR_SIZE to terminate saved_auxv properly
2007-10-17 08:43:00 -07:00
Add linux/elfcore-compat.h
2007-10-17 08:42:51 -07:00
x86 merge fallout: uml
2007-10-29 07:41:32 -07:00
exportfs: update documentation
2007-10-22 08:13:21 -07:00
fix up ext2_fs.h for userspace after reservations backport
2007-11-29 09:24:53 -08:00
ext2 reservations
2007-10-17 08:43:02 -07:00
Fix 64KB blocksize in ext3 directories
2007-11-14 18:45:43 -08:00
ext3: show all mount options
2007-10-17 08:42:48 -07:00
ext4: Convert s_r_blocks_count and s_free_blocks_count
2007-10-17 18:50:02 -04:00
ext4: Convert ext4_extent_idx.ei_leaf to ext4_extent_idx.ei_leaf_lo
2007-10-17 18:50:03 -04:00
ext4: Remove (partial, never completed) fragment support
2007-10-17 18:49:59 -04:00
ext4: Remove (partial, never completed) fragment support
2007-10-17 18:49:59 -04:00
jbd2: JBD_XXX to JBD2_XXX naming cleanup
2007-10-17 18:49:58 -04:00
hwmon: (f75375s) Allow setting up fans with platform_data
2007-11-08 08:42:46 -05:00
fb: move and rename extern declaration for global_mode_option
2007-10-16 09:43:22 -07:00
F_DUPFD_CLOEXEC implementation
2007-10-17 08:43:01 -07:00
r/o bind mounts: filesystem helpers for custom 'struct file's
2007-10-17 08:43:04 -07:00
[NET]: Introduce the sk_detach_filter() call
2007-10-17 21:21:26 -07:00
firewire: adopt read cycle timer ABI from raw1394
2007-10-17 00:00:08 +02:00
freezer: introduce freezer-friendly waiting macros
2007-10-18 14:37:19 -07:00
exportfs: make struct export_operations const
2007-10-22 08:13:21 -07:00
[PATCH] pass dentry to audit_inode()/audit_inode_child()
2007-10-21 02:37:18 -04:00
fuse: pass open flags to read and write
2007-11-29 09:24:54 -08:00
kernel/futex.c: make 3 functions static
2007-11-05 21:53:46 +11:00
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
2007-10-16 09:43:00 -07:00
get rid of input BIT* duplicate defines
2007-10-19 11:53:42 -07:00
hrtimer: Rework hrtimer_nanosleep to make sys_compat_nanosleep easier
2007-10-18 22:54:18 +02:00
hugetlb: introduce nr_overcommit_hugepages sysctl
2007-12-17 19:28:17 -08:00
i2c: Make i2c_check_addr static
2007-11-15 19:24:02 +01:00
Input: i8042 - export i8042_command()
2007-10-22 00:56:52 -04:00
ide: DMA reporting and validity checking fixes (take 3)
2007-12-12 23:31:58 +01:00
bonding: Add new layer2+3 hash for xor/802.3ad modes
2007-12-07 15:00:32 -05:00
include linux/types.h in if_fddi.h
2007-10-17 08:42:52 -07:00
[LRO]: fix lro_gen_skb() alignment
2007-12-05 05:37:32 -08:00
store __setup_str_* in a more compact way
2007-10-17 08:42:56 -07:00
Isolate the explicit usage of signal->pgrp
2007-10-19 11:53:43 -07:00
[PATCH] new helper - inotify_evict_watch()
2007-10-21 02:37:38 -04:00
Input: adds the context menu key (HUT GenDesc 0x84)
2007-11-27 00:47:04 -05:00
Completely remove deprecated IRQ flags (SA_*)
2007-10-17 08:43:00 -07:00
memory unplug: memory hotplug cleanup
2007-10-16 09:43:01 -07:00
virtualization of sysv msg queues is incomplete
2007-10-19 11:53:48 -07:00
IPMI: add polled interface
2007-10-18 14:37:32 -07:00
IPMI: fix comparison in demangle_device_id
2007-10-30 08:06:55 -07:00
use mutex instead of semaphore in isdn subsystem common functions
2007-10-16 09:43:10 -07:00
jbd: Fix assertion failure in fs/jbd/checkpoint.c
2007-12-05 09:21:20 -08:00
jbd2: JBD_XXX to JBD2_XXX naming cleanup
2007-10-17 18:49:58 -04:00
Missing include file in kallsyms.h
2007-11-05 15:12:32 -08:00
unicode diacritics support
2007-10-17 08:42:52 -07:00
fix headers_install
2007-12-17 19:28:15 -08:00
Fix <linux/kd.h> usage in userspace
2007-11-29 09:24:52 -08:00
printk: add KERN_CONT annotation
2007-10-17 08:43:01 -07:00
Add scaled time to taskstats based process accounting
2007-10-18 14:37:28 -07:00
Extended crashkernel command line
2007-10-19 11:53:49 -07:00
KEYS: Make request_key() and co fundamentally asynchronous
2007-10-17 08:42:57 -07:00
KEYS: Make request_key() and co fundamentally asynchronous
2007-10-17 08:42:57 -07:00
Console keyboard events and accessibility
2007-10-19 11:53:33 -07:00
kprobes: support kretprobe blacklist
2007-10-16 09:43:10 -07:00
leds: Fix led trigger locking bugs
2007-12-07 09:06:53 +00:00
lguest: documentation update
2007-10-25 15:02:50 +10:00
lguest: documentation update
2007-10-25 15:02:50 +10:00
libata: workaround DRQ=1 ERR=1 for ATAPI tape drives
2007-11-19 12:28:11 +09:00
add WEAK() for creating weak asm labels
2007-10-22 08:13:17 -07:00
kernel-api docbook: fix content problems
2007-10-19 11:53:35 -07:00
workqueue: debug flushing deadlocks with lockdep
2007-10-19 11:53:38 -07:00
Add a "rounddown_pow_of_two" routine to log2.h
2007-10-17 08:42:56 -07:00
Task Control Groups: basic task cgroup framework
2007-10-19 11:53:36 -07:00
Linux Kernel Markers - Samples
2007-10-19 11:53:55 -07:00
rtc: fall back to requesting only the ports we actually use
2007-11-14 18:45:41 -08:00
memory hotplug: make kmem_cache_node for SLUB on memory online avoid panic
2007-10-22 08:13:17 -07:00
fix memory hot remove not configured case.
2007-10-16 09:43:02 -07:00
Task Control Groups: make cpusets a client of cgroups
2007-10-19 11:53:36 -07:00
Security: round mmap hint address above mmap_min_addr
2007-12-06 00:25:10 +11:00
increase AT_VECTOR_SIZE to terminate saved_auxv properly
2007-10-17 08:43:00 -07:00
mm: test and set zone reclaim lock before starting reclaim
2007-10-17 08:42:46 -07:00
Virtio interface
2007-10-23 15:49:54 +10:00
Linux Kernel Markers
2007-10-19 11:53:54 -07:00
constify string/array kparam tracking structures
2007-10-17 08:42:56 -07:00
[IPV4]: Clean the ip_sockglue.c from some ugly ifdefs
2007-11-07 04:08:55 -08:00
ipc: store ipcs into IDRs
2007-10-19 11:53:44 -07:00
Mutex documentation is unclear about software interrupts, tasklets and timers
2007-10-17 08:42:57 -07:00
mv643xx_eth: Move ethernet register definitions into private header
2007-10-23 08:23:00 -07:00
mv643xx_eth: Fix MV643XX_ETH offsets used by Pegasos 2
2007-10-30 14:32:16 -04:00
partially fix up the lookup_one_noperm mess
2007-10-17 08:42:44 -07:00
NBD: allow hung network I/O to be cancelled
2007-10-17 08:42:55 -07:00
[NET]: Add the helper kernel_sock_shutdown()
2007-11-12 18:10:39 -08:00
[NET]: docbook fixes for netif_ functions
2007-11-01 02:21:47 -07:00
[NETFILTER]: bridge: fix missing link layer headers on outgoing routed packets
2007-12-14 13:54:39 -08:00
[NETLINK]: Fix unicast timeouts
2007-11-07 04:15:12 -08:00
NFS: make nfs_wb_page_priority() static
2007-11-26 16:24:48 -05:00
add consts where appropriate in fs/nls/*
2007-10-17 08:42:58 -07:00
Memoryless nodes: Add N_CPU node state
2007-10-16 09:42:58 -07:00
Console events and accessibility
2007-10-19 11:53:34 -07:00
Make access to task's nsproxy lighter
2007-10-19 11:53:37 -07:00
remove asm/bitops.h includes
2007-10-19 11:53:41 -07:00
[POWERPC] Move of_device allocation into of_device.[ch]
2007-10-17 22:30:07 +10:00
[SPARC/64]: Consolidate of_register_driver
2007-10-17 21:17:42 -07:00
oom: prevent including sched.h in header file
2007-10-17 08:42:46 -07:00
memory unplug: page isolation
2007-10-16 09:43:02 -07:00
memory unplug: page isolation
2007-10-16 09:43:02 -07:00
[PARPORT] Remove unused 'irq' argument from parport irq functions
2007-10-23 19:53:16 -04:00
PCI: Add Kconfig option to disable deprecated pci_find_* API
2007-11-05 13:35:17 -08:00
forcedeth: new mcp79 pci ids
2007-11-23 20:54:01 -05:00
lib: percpu_counter_init_irq
2007-10-17 08:42:44 -07:00
Misc: phantom, improved data passing
2007-10-19 11:53:41 -07:00
PHY: Add the phy_device_release device method.
2007-12-04 15:06:33 -05:00
Uninline find_pid etc set of functions
2007-10-19 11:53:41 -07:00
pidns: Place under CONFIG_EXPERIMENTAL
2007-11-14 18:45:43 -08:00
PM: Move definition of struct pm_ops to suspend.h
2007-10-18 14:37:18 -07:00
PNP: increase the maximum number of resources
2007-11-29 09:24:52 -08:00
jbd2: JBD_XXX to JBD2_XXX naming cleanup
2007-10-17 18:49:58 -04:00
x86 & generic: change to __builtin_prefetch()
2007-10-19 20:35:04 +02:00
Fix cpusets update_cpumask
2007-10-19 11:53:41 -07:00
proc: fix proc_dir_entry refcounting
2007-12-05 09:21:20 -08:00
make kernel/profile.c:time_hook static
2007-10-17 08:42:55 -07:00
lib: floating proportions
2007-10-17 08:42:45 -07:00
quota: send messages via netlink
2007-10-17 08:42:56 -07:00
Immunize rcu_dereference() against crazy compiler writers
2007-10-17 08:42:46 -07:00
reiserfs: new export ops
2007-10-22 08:13:20 -07:00
reiserfs: remove first_zero_hint
2007-10-19 11:53:35 -07:00
RTC: convert mutex to bitfield
2007-11-29 09:24:54 -08:00
[IPV6]: Add ifindex field to ND user option messages.
2007-11-12 17:58:35 -08:00
scatterlist: add more safeguards
2007-11-27 09:30:39 +01:00
sched: clean up, move __sched_text_start/end to sched.h
2007-11-28 15:52:56 +01:00
x86_64 EFI boot support: EFI frame buffer driver
2007-11-29 09:24:54 -08:00
fix up security_socket_getpeersec_* documentation
2007-10-20 00:53:30 +02:00
vt: Fix warnings in selection.h
2007-10-16 09:43:17 -07:00
SELinux: return EOPNOTSUPP not ENOTSUPP
2007-11-17 10:38:16 +11:00
ipc: store ipcs into IDRs
2007-10-19 11:53:44 -07:00
Fix f_version type: should be u64 instead of unsigned long
2007-10-17 08:42:53 -07:00
IP22ZILOG: fix lockup and sysrq
2007-11-29 09:24:53 -08:00
ipc: store ipcs into IDRs
2007-10-19 11:53:44 -07:00
rename signalfd_siginfo fields
2007-10-17 08:43:01 -07:00
[SKBUFF]: Free old skb properly in skb_morph
2007-11-26 23:11:19 +08:00
Slab API: remove useless ctor parameter and reorder parameters
2007-10-17 08:42:45 -07:00
Slab API: remove useless ctor parameter and reorder parameters
2007-10-17 08:42:45 -07:00
sm501fb: Ensure panel interface is not tristated when setup
2007-10-16 09:43:14 -07:00
KVM: fix !SMP build error
2007-11-09 22:39:38 +01:00
[Bluetooth] Add constant for Bluetooth socket options level
2007-10-22 02:59:48 -07:00
[TIPC]: Fix headercheck wrt. tipc_config.h
2007-10-30 21:44:00 -07:00
Hibernation: Check if ACPI is enabled during restore in the right place
2007-10-18 14:37:20 -07:00
oom: move prototypes to appropriate header file
2007-10-17 08:42:45 -07:00
[S390] appldata: remove unused binary sysctls.
2007-11-20 11:13:45 +01:00
add-scaled-time-to-taskstats-based-process-accounting fix
2007-10-18 14:37:28 -07:00
futex: fix for futex_wait signal stack corruption
2007-12-05 15:46:09 +01:00
kernel/time/timekeeping.c: cleanups
2007-10-17 08:42:53 -07:00
time: add ADJ_OFFSET_SS_READ
2007-11-26 20:42:19 +01:00
[TTY]: Fix network driver interactions with TCGET/SET calls.
2007-11-07 04:14:19 -08:00
fix abuses of ptrdiff_t
2007-10-29 07:41:33 -07:00
ufs: move non-layout parts of ufs_fs.h to fs/ufs/
2007-10-17 08:42:51 -07:00
get rid of input BIT* duplicate defines
2007-10-19 11:53:42 -07:00
usb.h: fix kernel-doc warning
2007-12-17 10:47:15 -08:00
usb-storage: Fix devices that cannot handle 32k transfers
2007-12-17 10:47:14 -08:00
USB: Fix signr comment in usbdevice_fs.h
2007-11-28 13:58:34 -08:00
V4L/DVB (6320): v4l core: remove the unused .hardware V4L1 field
2007-10-22 12:01:24 -02:00
V4L/DVB (6321): Remove obsolete VIDIOC_S/G_MPEGCOMP ioctls
2007-10-22 12:01:30 -02:00
Virtio interface
2007-10-23 15:49:54 +10:00
9p: add virtio transport
2007-10-23 13:47:31 -05:00
Block driver using virtio.
2007-10-23 15:49:54 +10:00
Virtio interface
2007-10-23 15:49:54 +10:00
Virtio console driver
2007-10-23 15:49:55 +10:00
Net driver using virtio
2007-10-23 15:49:54 +10:00
virtio: Force use of power-of-two for descriptor ring sizes
2007-11-12 13:59:40 +11:00
Console events and accessibility
2007-10-19 11:53:34 -07:00
add CONFIG_VT_UNICODE
2007-10-17 08:42:56 -07:00
workqueue: debug flushing deadlocks with lockdep
2007-10-19 11:53:38 -07:00
introduce I_SYNC
2007-10-17 08:43:02 -07:00
d1c3fb1f8f