Johannes Weiner [Thu, 8 Dec 2011 04:42:40 +0000 (15:42 +1100)]
mm: vmscan: distinguish between memcg triggering reclaim and memcg being scanned
Memory cgroup hierarchies are currently handled completely outside of the
traditional reclaim code, which is invoked with a single memory cgroup as
an argument for the whole call stack.
Subsequent patches will switch this code to do hierarchical reclaim, so
there needs to be a distinction between a) the memory cgroup that is
triggering reclaim due to hitting its limit and b) the memory cgroup that
is being scanned as a child of a).
This patch introduces a struct mem_cgroup_zone that contains the
combination of the memory cgroup and the zone being scanned, which is then
passed down the stack instead of the zone argument.
Signed-off-by: Johannes Weiner <jweiner@redhat.com> Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Reviewed-by: Michal Hocko <mhocko@suse.cz> Reviewed-by: Kirill A. Shutemov <kirill@shutemov.name> Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Ying Han <yinghan@google.com> Cc: Greg Thelen <gthelen@google.com> Cc: Michel Lespinasse <walken@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Hugh Dickins <hughd@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Johannes Weiner [Thu, 8 Dec 2011 04:42:39 +0000 (15:42 +1100)]
mm: vmscan: distinguish global reclaim from global LRU scanning
The traditional zone reclaim code is scanning the per-zone LRU lists
during direct reclaim and kswapd, and the per-zone per-memory cgroup LRU
lists when reclaiming on behalf of a memory cgroup limit.
Subsequent patches will convert the traditional reclaim code to reclaim
exclusively from the per-memory cgroup LRU lists. As a result, using the
predicate for which LRU list is scanned will no longer be appropriate to
tell global reclaim from limit reclaim.
This patch adds a global_reclaim() predicate to tell direct/kswapd reclaim
from memory cgroup limit reclaim and substitutes it in all places where
currently scanning_global_lru() is used for that.
Signed-off-by: Johannes Weiner <jweiner@redhat.com> Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Reviewed-by: Michal Hocko <mhocko@suse.cz> Reviewed-by: Kirill A. Shutemov <kirill@shutemov.name> Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Ying Han <yinghan@google.com> Cc: Greg Thelen <gthelen@google.com> Cc: Michel Lespinasse <walken@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Hugh Dickins <hughd@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Memory control groups are currently bolted onto the side of
traditional memory management in places where better integration would
be preferrable. To reclaim memory, for example, memory control groups
maintain their own LRU list and reclaim strategy aside from the global
per-zone LRU list reclaim. But an extra list head for each existing
page frame is expensive and maintaining it requires additional code.
This patchset disables the global per-zone LRU lists on memory cgroup
configurations and converts all its users to operate on the per-memory
cgroup lists instead. As LRU pages are then exclusively on one list,
this saves two list pointers for each page frame in the system:
page_cgroup array size with 4G physical memory
vanilla: [ 0.000000] allocated 31457280 bytes of page_cgroup
patched: [ 0.000000] allocated 15728640 bytes of page_cgroup
At the same time, system performance for various workloads is
unaffected:
100G sparse file cat, 4G physical memory, 10 runs, to test for code
bloat in the traditional LRU handling and kswapd & direct reclaim
paths, without/with the memory controller configured in
4 unlimited memcgs running kbuild -j32 each, 4G physical memory, 500M
swap on SSD, 10 runs, to test for regressions in kswapd & direct
reclaim using per-memcg LRU lists with multiple memcgs and multiple
allocators within each memcg
The task counter subsystem has been written assuming that
can_attach_task/attach_task/cancel_attach_task calls are serialized per
task. This is true when we attach only one task but not when we attach a
whole thread group, in which case the sequence is:
for each thread
if (can_attach_task() < 0)
goto rollback
for each_thread
attach_task()
rollback:
for each thread
cancel_attach_task()
The common ancestor, searched on task_counter_attach_task(), can thus
change between each of these calls for a given task. This breaks if some
tasks in the thread group are not in the same cgroup origin. The uncharge
made in attach_task() or the rollback in cancel_attach_task() there would
have an erroneous propagation.
This can even break seriously is some scenario. For example there
with $PID beeing the pid of a multithread process:
On the last move, attach_task() is called on the thread leader with
the wrong common_ancestor, leading to a crash because we uncharge
a res_counter that doesn't exist:
To solve this, keep the original cgroup of each thread in the thread
group cached in the flex array and pass it to can_attach_task()/attach_task()
and cancel_attach_task() so that the correct common ancestor between the old
and new cgroup can be safely retrieved for each task.
This is inspired by a previous patch from Li Zefan:
"[PATCH] cgroups: don't cache common ancestor in task counter subsys".
Reported-by: Ben Blum <bblum@andrew.cmu.edu> Reported-by: Li Zefan <lizf@cn.fujitsu.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Menage <paul@paulmenage.org> Cc: Tim Hockin <thockin@hockin.org> Cc: Tejun Heo <htejun@gmail.com> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add a new subsystem to limit the number of running tasks, similar to the
NR_PROC rlimit but in the scope of a cgroup.
The user can set an upper bound limit that is checked every time a task
forks in a cgroup or is moved into a cgroup with that subsystem binded.
The primary goal is to protect against forkbombs that explode inside a
container. The traditional NR_PROC rlimit is not efficient in that case
because if we run containers in parallel under the same user, one of these
could starve all the others by spawning a high number of tasks close to
the user wide limit.
This is a prevention against forkbombs, so it's not deemed to cure the
effects of a forkbomb when the system is in a state where it's not
responsive. It's aimed at preventing from ever reaching that state and
stop the spreading of tasks early. While defining the limit on the
allowed number of tasks, it's up to the user to find the right balance
between the resource its containers may need and what it can afford to
provide.
As it's totally dissociated from the rlimit NR_PROC, both can be
complementary: the cgroup task counter can set an upper bound per
container and the rlmit can be an upper bound on the overall set of
containers.
Also this subsystem can be used to kill all the tasks in a cgroup without
races against concurrent forks, by setting the limit of tasks to 0, any
further forks can be rejected. This is a good way to kill a forkbomb in a
container, or simply kill any container without the need to retry an
unbound number of times.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Menage <paul@paulmenage.org> Reviewed-by: Li Zefan <lizf@cn.fujitsu.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Aditya Kali <adityakali@google.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Tim Hockin <thockin@hockin.org> Cc: Tejun Heo <htejun@gmail.com> Acked-by: Kirill A. Shutemov <kirill@shutemov.name> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Let the subsystem's fork callback return an error value so that they can
cancel a fork. This is going to be used by the task counter subsystem to
implement the limit.
Suggested-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Menage <paul@paulmenage.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Aditya Kali <adityakali@google.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Tim Hockin <thockin@hockin.org> Cc: Tejun Heo <htejun@gmail.com> Acked-by: Kirill A. Shutemov <kirill@shutemov.name> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
cgroups: pull up res counter charge failure interpretation to caller
res_counter_charge() always returns -ENOMEM when the limit is reached and
the charge thus can't happen.
However it's up to the caller to interpret this failure and return the
appropriate error value. The task counter subsystem will need to report
the user that a fork() has been cancelled because of some limit reached,
not because we are too short on memory.
Fix this by returning -1 when res_counter_charge() fails.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Menage <paul@paulmenage.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Aditya Kali <adityakali@google.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Tim Hockin <thockin@hockin.org> Cc: Tejun Heo <htejun@gmail.com> Acked-by: Kirill A. Shutemov <kirill@shutemov.name> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
cgroups: ability to stop res charge propagation on bounded ancestor
Moving a task from a cgroup to another may require to substract its
resource charge from the old cgroup and add it to the new one.
For this to happen, the uncharge/charge propagation can just stop when we
reach the common ancestor for the two cgroups. Further the performance
reasons, we also want to avoid to temporarily overload the common
ancestors with a non-accurate resource counter usage if we charge first
the new cgroup and uncharge the old one thereafter. This is going to be a
requirement for the coming max number of task subsystem.
To solve this, provide a pair of new API that can charge/uncharge a
resource counter until we reach a given ancestor.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Acked-by: Paul Menage <paul@paulmenage.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Aditya Kali <adityakali@google.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Tim Hockin <thockin@hockin.org> Cc: Tejun Heo <htejun@gmail.com> Acked-by: Kirill A. Shutemov <kirill@shutemov.name> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
cgroups: new cancel_attach_task() subsystem callback
To cancel a process attachment on a subsystem, we only call the
cancel_attach() callback once on the leader but we have no way to cancel
the attachment individually for each member of the process group.
This is going to be needed for the max number of tasks susbystem that is
coming.
To prepare for this integration, call a new cancel_attach_task() callback
on each task of the group until we reach the member that failed to attach.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Acked-by: Paul Menage <paul@paulmenage.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Aditya Kali <adityakali@google.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Tim Hockin <thockin@hockin.org> Cc: Tejun Heo <htejun@gmail.com> Acked-by: Kirill A. Shutemov <kirill@shutemov.name> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Provide an API to inherit a counter value from a parent. This can be
useful to implement cgroup.clone_children on a resource counter.
Still the resources of the children are limited by those of the parent, so
this is only to provide a default setting behaviour when clone_children is
set.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Menage <paul@paulmenage.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Aditya Kali <adityakali@google.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Tim Hockin <thockin@hockin.org> Cc: Tejun Heo <htejun@gmail.com> Acked-by: Kirill A. Shutemov <kirill@shutemov.name> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Extend the resource counter API with a mirror of res_counter_read_u64() to
make it handy to update a resource counter value from a cgroup subsystem
u64 value file.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Acked-by: Paul Menage <paul@paulmenage.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Aditya Kali <adityakali@google.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Tim Hockin <thockin@hockin.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Kirill A. Shutemov <kirill@shutemov.name> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Mike Galbraith [Thu, 8 Dec 2011 04:42:34 +0000 (15:42 +1100)]
cpusets, cgroups: disallow attaching kthreadd
Allowing kthreadd to be moved to a non-root group makes no sense, it being
a global resource, and needlessly leads unsuspecting users toward trouble.
1. An RT workqueue worker thread spawned in a task group with no
rt_runtime allocated is not schedulable. Simple user error, but
harmful to the box.
2. A worker thread which acquires PF_THREAD_BOUND can never leave a
cpuset, rendering the cpuset immortal.
Save the user some unexpected trouble, just say no.
Signed-off-by: Mike Galbraith <efault@gmx.de> Acked-by: David Rientjes <rientjes@google.com> Acked-by: Paul Menage <paul@paulmenage.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Nothing requires that we lock the filesystem until the root inode is
provided.
Also iget5_locked() triggers a warning because we are holding the
filesystem lock while allocating the inode, which result in a lockdep
suspicion that we have a lock inversion against the reclaim path:
The deadlock shouldn't happen since we are doing that allocation in the
mount path, the filesystem is not available for any reclaim. Still the
warning is annoying.
To solve this, acquire the lock later only where we need it, right before
calling reiserfs_read_locked_inode() that wants to lock to walk the tree.
Reported-by: Knut Petersen <Knut_Petersen@t-online.de> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Christoph Hellwig <hch@lst.de> Cc: Jeff Mahoney <jeffm@suse.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
journal_init() doesn't need the lock since no operation on the filesystem
is involved there. journal_read() and get_list_bitmap() have yet to be
reviewed carefully though before removing the lock there. Just keep the
it around these two calls for safety.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Christoph Hellwig <hch@lst.de> Cc: Jeff Mahoney <jeffm@suse.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
reiserfs: delay reiserfs lock until journal initialization
In the mount path, transactions that are made before journal
initialization don't involve the filesystem. We can delay the reiserfs
lock until we play with the journal.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Christoph Hellwig <hch@lst.de> Cc: Jeff Mahoney <jeffm@suse.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Mark Brown [Thu, 8 Dec 2011 04:42:32 +0000 (15:42 +1100)]
drivers/rtc/rtc-wm831x.c: remove unused period IRQ handler
Due to changes in the RTC core the period interrupt is now unused so
delete the code managing it.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Cc: Alessandro Zummo <a.zummo@towertech.it> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
ChangeLog v2->v3:
- back to square 1. 0x80 is not allowed because the representation
is not two's complement but bit 7 is a sign bit, thus 0x80 is
just another way to say "zero". Sorry for the mess, clarified this
with a comment in the code.
Acked-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> Signed-off-by: Mark Godfrey <mark.godfrey@stericsson.com> Signed-off-by: Linus Walleij <linus.walleij@stericsson.com> Cc: Alessandro Zummo <a.zummo@towertech.it> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Alessandro Zummo <a.zummo@towertech.it>
WARNING: line over 80 characters
#48: FILE: drivers/rtc/rtc-ab8500.c:268:
+ * Check that the calibration value (which is in units of 0.5 parts-per-million)
ERROR: need consistent spacing around '-' (ctx:WxV)
#64: FILE: drivers/rtc/rtc-ab8500.c:284:
+ rtccal = ~(calibration -1) | 0x80;
^
total: 1 errors, 1 warnings, 139 lines checked
./patches/rtc-ab8500-add-calibration-attribute-to-ab8500-rtc.patch has style problems, please review.
If any of these errors are false positives, please report
them to the maintainer, see CHECKPATCH in MAINTAINERS.
Please run checkpatch prior to sending patches
Cc: Alessandro Zummo <a.zummo@towertech.it> Cc: Linus Walleij <linus.walleij@stericsson.com> Cc: Mark Godfrey <mark.godfrey@stericsson.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Mark Godfrey [Thu, 8 Dec 2011 04:42:31 +0000 (15:42 +1100)]
rtc/ab8500: add calibration attribute to AB8500 RTC
The rtc_calibration attribute allows user-space to get and set the
AB8500's RtcCalibration register. The AB8500 will then use the value in
this register to compensate for RTC drift every 60 seconds.
Signed-off-by: Mark Godfrey <mark.godfrey@stericsson.com> Signed-off-by: Linus Walleij <linus.walleij@stericsson.com> Acked-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> Cc: Alessandro Zummo <a.zummo@towertech.it> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Linus Walleij [Thu, 8 Dec 2011 04:42:30 +0000 (15:42 +1100)]
drivers/rtc/rtc-ab8500.c: change msleep() to usleep_range()
The resolution of msleep is related to HZ, so with HZ set to 100 any
msleep of less than 10ms will become ~10ms. This is not what we want.
Use the hrtimer-based usleep_range() and allow for some slack in the
non-critical path so we have more control of what is happening here.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Cc: Jonas Aaberg <jonas.aberg@stericsson.com> Cc: Alessandro Zummo <a.zummo@towertech.it> Cc: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Yauhen Kharuzhy [Thu, 8 Dec 2011 04:42:28 +0000 (15:42 +1100)]
drivers/rtc/rtc-mxc.c: fix setting time for MX1 SoC
There is no way to track year in the i.MX1 RTC: Days Counter register is
9-bit wide only. Attempt to save date after 1970-01-01 plus 512 days
causes endless loop in mxc_rtc_set_mmss(). Fix this by resetting year to
1970.
Signed-off-by: Yauhen Kharuzhy <jekhor@gmail.com> Cc: Daniel Mack <daniel@caiaq.de> Cc: Alessandro Zummo <a.zummo@towertech.it> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
YanHong [Thu, 8 Dec 2011 04:42:27 +0000 (15:42 +1100)]
init/do_mounts.c: create /root if it does not exist
If someone supplies an initramfs without /root in it, and we fail to
execute rdinit, we will try to mount root device and fail, for the mount
point does not exits.
But we get error message "VFS: Cannot open root device". It's confusing.
We can give a more detailed error message, or we can go further: if /root
does not exit, create it.
Signed-off-by: YanHong <tempname2@hotmail.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Woody Suwalski <terraluna977@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
David Daney [Thu, 8 Dec 2011 04:42:27 +0000 (15:42 +1100)]
MIPS: randomize PIE load address
... by selecting ARCH_BINFMT_ELF_RANDOMIZE_PIE
Signed-off-by: David Daney <david.daney@cavium.com> Cc: Russell King <linux@arm.linux.org.uk> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
David Daney [Thu, 8 Dec 2011 04:42:27 +0000 (15:42 +1100)]
fs: binfmt_elf: create Kconfig variable for PIE randomization
Randomization of PIE load address is hard coded in binfmt_elf.c for X86
and ARM. Create a new Kconfig variable
(CONFIG_ARCH_BINFMT_ELF_RANDOMIZE_PIE) for this and use it instead. Thus
architecture specific policy is pushed out of the generic binfmt_elf.c and
into the architecture Kconfig files.
X86 and ARM Kconfigs are modified to select the new variable so there is
no change in behavior. A follow on patch will select it for MIPS too.
Signed-off-by: David Daney <david.daney@cavium.com> Cc: Russell King <linux@arm.linux.org.uk> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Acked-by: H. Peter Anvin <hpa@zytor.com> Cc: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Jason Baron [Thu, 8 Dec 2011 04:42:26 +0000 (15:42 +1100)]
epoll: limit paths
The current epoll code can be tickled to run basically indefinitely in
both loop detection path check (on ep_insert()), and in the wakeup paths.
The programs that tickle this behavior set up deeply linked networks of
epoll file descriptors that cause the epoll algorithms to traverse them
indefinitely. A couple of these sample programs have been previously
posted in this thread: https://lkml.org/lkml/2011/2/25/297.
To fix the loop detection path check algorithms, I simply keep track of
the epoll nodes that have been already visited. Thus, the loop detection
becomes proportional to the number of epoll file descriptor and links.
This dramatically decreases the run-time of the loop check algorithm. In
one diabolical case I tried it reduced the run-time from 15 mintues (all
in kernel time) to .3 seconds.
Fixing the wakeup paths could be done at wakeup time in a similar manner
by keeping track of nodes that have already been visited, but the
complexity is harder, since there can be multiple wakeups on different
cpus...Thus, I've opted to limit the number of possible wakeup paths when
the paths are created.
This is accomplished, by noting that the end file descriptor points that
are found during the loop detection pass (from the newly added link), are
actually the sources for wakeup events. I keep a list of these file
descriptors and limit the number and length of these paths that emanate
from these 'source file descriptors'. In the current implemetation I
allow 1000 paths of length 1, 500 of length 2, 100 of length 3, 50 of
length 4 and 10 of length 5. Note that it is sufficient to check the
'source file descriptors' reachable from the newly added link, since no
other 'source file descriptors' will have newly added links. This allows
us to check only the wakeup paths that may have gotten too long, and not
re-check all possible wakeup paths on the system.
In terms of the path limit selection, I think its first worth noting that
the most common case for epoll, is probably the model where you have 1
epoll file descriptor that is monitoring n number of 'source file
descriptors'. In this case, each 'source file descriptor' has a 1 path of
length 1. Thus, I believe that the limits I'm proposing are quite
reasonable and in fact may be too generous. Thus, I'm hoping that the
proposed limits will not prevent any workloads that currently work to
fail.
In terms of locking, I have extended the use of the 'epmutex' to all
epoll_ctl add and remove operations. Currently its only used in a subset
of the add paths. I need to hold the epmutex, so that we can correctly
traverse a coherent graph, to check the number of paths. I believe that
this additional locking is probably ok, since its in the setup/teardown
paths, and doesn't affect the running paths, but it certainly is going to
add some extra overhead. Also, worth noting is that the epmuex was
recently added to the ep_ctl add operations in the initial path loop
detection code using the argument that it was not on a critical path.
Another thing to note here, is the length of epoll chains that is allowed.
Currently, eventpoll.c defines:
/* Maximum number of nesting allowed inside epoll sets */
#define EP_MAX_NESTS 4
This basically means that I am limited to a graph depth of 5 (EP_MAX_NESTS
+ 1). However, this limit is currently only enforced during the loop
check detection code, and only when the epoll file descriptors are added
in a certain order. Thus, this limit is currently easily bypassed. The
newly added check for wakeup paths, stricly limits the wakeup paths to a
length of 5, regardless of the order in which ep's are linked together.
Thus, a side-effect of the new code is a more consistent enforcement of
the graph depth.
Thus far, I've tested this, using the sample programs previously
mentioned, which now either return quickly or return -EINVAL. I've also
testing using the piptest.c epoll tester, which showed no difference in
performance. I've also created a number of different epoll networks and
tested that they behave as expectded.
I believe this solves the original diabolical test cases, while still
preserving the sane epoll nesting.
Signed-off-by: Jason Baron <jbaron@redhat.com> Cc: Nelson Elhage <nelhage@ksplice.com> Cc: Davide Libenzi <davidel@xmailserver.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Taking a pointer reference to each row in the crc table matrix, one can
reduce the inner loop with a few insn's
Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se> Cc: Bob Pearson <rpearson@systemfabricworks.com> Cc: Frank Zago <fzago@systemfabricworks.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Andy Whitcroft [Thu, 8 Dec 2011 04:42:25 +0000 (15:42 +1100)]
checkpatch: catch all occurances of type and cast spacing errors per line
Fix up type and cast spacing checks such that all occurances on a line are
examined and reported. For example the line below has a valid cast and a
bad type, but currently we check the cast first which is good and stop:
u16* bar = (u16 *)baz;
We will also only report one of the errors in this example:
u16* bar = (u16*)bad;
Move to iterating across all casts and all types, reporting any failure.
Signed-off-by: Andy Whitcroft <apw@canonical.com> Cc: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Andy Whitcroft [Thu, 8 Dec 2011 04:42:23 +0000 (15:42 +1100)]
checkpatch: only apply kconfig help checks for options which prompt
The intent of this check is to catch the options which the user will see
and ensure they are properly described. It is also common for internal
only options to have a brief description. Allow this form.
Reported-by: Steven Rostedt <rostedt@goodmis.org> Tested-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andy Whitcroft <apw@canonical.com> Cc: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Andy Whitcroft [Thu, 8 Dec 2011 04:42:23 +0000 (15:42 +1100)]
checkpatch: optimise statement scanner when mid-statement
In the middle of a long definition or similar, there is no possibility of
finding a smaller sub-statement. Optimise this case by skipping statement
aquirey where there are no starts of statement (open brace '{' or
semi-colon ';'). We are likely to scan slightly more than needed still
but this is safest.
Signed-off-by: Andy Whitcroft <apw@canonical.com> Cc: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Andy Whitcroft [Thu, 8 Dec 2011 04:42:22 +0000 (15:42 +1100)]
checkpatch: ## is not a valid modifier
Inserting a # into the modifiers list will incorrectly add the null string
to the modifiers list, leading to an infinite loop. As neither of these
is a valid modifier form simply ignore them.
Signed-off-by: Andy Whitcroft <apw@canonical.com> Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Joe Perches [Thu, 8 Dec 2011 04:42:22 +0000 (15:42 +1100)]
checkpatch: improve memset and min/max with cast checking
Improve the checking of arguments to memset and min/max tests.
Move the checking of min/max to statement blocks instead of single line.
Change $Constant to allow any case type 0x initiator and trailing ul
specifier. Add $FuncArg type as any function argument with or without a
cast. Print the whole statement when showing memset or min/max messages.
Improve the memset with 0 as 3rd argument error message.
There are still weaknesses in the $FuncArg and $Constant code as arbitrary
parentheses and negative signs are not generically supported.
Signed-off-by: Joe Perches <joe@perches.com> Acked-by: Andy Whitcroft <apw@canonical.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Andy Whitcroft [Thu, 8 Dec 2011 04:42:21 +0000 (15:42 +1100)]
checkpatch: check for common memset parameter issues against statments
Move the memset checks over to work against the statement. Also add
checks for 0 and 1 used as lengths. Generally these indicate badly
ordered parameters.
Signed-off-by: Andy Whitcroft <apw@canonical.com> Cc: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Andy Whitcroft [Thu, 8 Dec 2011 04:42:21 +0000 (15:42 +1100)]
checkpatch: correctly track the end of preprocessor commands in context
When looking for a statement we currently run on through preprocessor
commands. This means that a header file with just definitions is parsed
over and over again combining all of the lines from the current line to
the end of file leading to severe performance issues.
Fix up context accumulation to track preprocessor commands and stop when
reaching the end of them. At the same time vastly simplify the #define
handling.
Signed-off-by: Andy Whitcroft <apw@canonical.com> Cc: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Joe Perches [Thu, 8 Dec 2011 04:42:20 +0000 (15:42 +1100)]
checkpatch: update signature "might be better as" warning
email header lines can look like signature tags. It's valid to have
multiple email recipients on a single line but not valid to have multiple
signatures on a single line.
Validate signatures only when not in the email headers.
Clear the $in_commit_log flag when the patch filename appears.
Add '-' to the valid chars in a message header for headers
like "Message-Id:" and "In-Reply-To:".
Signed-off-by: Joe Perches <joe@perches.com> Reported-by: Julia Lawall <julia.lawall@lip6.fr> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Axel Lin [Thu, 8 Dec 2011 04:42:19 +0000 (15:42 +1100)]
drivers/leds/leds-netxbig.c: use gpio_request_one()
Use gpio_request_one() instead of multiple gpiolib calls. This also
simplifies error handling a bit.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Cc: Simon Guinot <sguinot@lacie.com> Cc: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Axel Lin [Thu, 8 Dec 2011 04:42:19 +0000 (15:42 +1100)]
drivers/leds/leds-bd2802.c: use gpio_request_one()
Use gpio_request_one() instead of multiple gpiolib calls.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Cc: Kim Kyuwon <q1.kim@samsung.com> Cc: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Axel Lin <axel.lin@gmail.com> Cc: Samu Onkalo <samu.p.onkalo@nokia.com> Cc: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Axel Lin [Thu, 8 Dec 2011 04:42:18 +0000 (15:42 +1100)]
leds: convert leds-dac124s085 to module_spi_driver
Factor out some boilerplate code for spi driver registration into
module_spi_driver.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Cc: Haojian Zhuang <hzhuang1@marvell.com> Cc: Mark Brown <broonie@opensource.wolfsonmicro.com> Cc: Richard Purdie <rpurdie@rpsys.net> Cc: Michael Hennerich <hennerich@blackfin.uclinux.org> Cc: Mike Rapoport <mike@compulab.co.il> Acked-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Axel Lin [Thu, 8 Dec 2011 04:42:18 +0000 (15:42 +1100)]
leds: convert led i2c drivers to module_i2c_driver
Factor out some boilerplate code for i2c driver registration
into module_i2c_driver.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Cc: Haojian Zhuang <hzhuang1@marvell.com> Cc: Mark Brown <broonie@opensource.wolfsonmicro.com> Cc: Richard Purdie <rpurdie@rpsys.net> Cc: Michael Hennerich <hennerich@blackfin.uclinux.org> Cc: Mike Rapoport <mike@compulab.co.il> Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Axel Lin [Thu, 8 Dec 2011 04:42:17 +0000 (15:42 +1100)]
leds: convert led platform drivers to module_platform_driver
Factor out some boilerplate code for platform driver registration into
module_platform_driver.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Acked-by: Haojian Zhuang <hzhuang1@marvell.com> [led-88pm860x.c] Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Cc: Richard Purdie <rpurdie@rpsys.net> Cc: Michael Hennerich <hennerich@blackfin.uclinux.org> Cc: Mike Rapoport <mike@compulab.co.il> Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Axel Lin [Thu, 8 Dec 2011 04:42:15 +0000 (15:42 +1100)]
backlight: convert drivers/video/backlight/* to use module_platform_driver()
Convert the drivers in drivers/video/backlight/* to use the
module_platform_driver() macro which makes the code smaller and a bit
simpler.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> Acked-by: H Hartley Sweeten <hsweeten@visionengravers.com> [ep93xx_bl.c] Cc: Mike Rapoport <mike@compulab.co.il> Cc: Richard Purdie <rpurdie@rpsys.net> Acked-by: Michael Hennerich <michael.hennerich@analog.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Paul Bolle [Thu, 8 Dec 2011 04:41:57 +0000 (15:41 +1100)]
backlight: remove ADX backlight device support
Support for the Avionic Design Xanthos backlight device got added in
commit 3b96ea9ef8 ("backlight: Add support for the Avionic Design Xanthos
backlight device."). That support depends on ARCH_PXA_ADX. The code that
should have provided that Kconfig symbol never got submitted. It has
never been possible to even build this driver. Remove it.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl> Acked-by: Thierry Reding <thierry.reding@avionic-design.de> Cc: Richard Purdie <rpurdie@rpsys.net> Cc: Wim Van Sebroeck <wim@iguana.be> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Ian Campbell [Thu, 8 Dec 2011 04:41:56 +0000 (15:41 +1100)]
get_maintainers.pl: follow renames when looking up commit signers
I happen to have had a commit to various network drivers since the big
renaming/reorg which happened to drivers/net recently. This means that I
now appear to be in the top few commit signers (by %age) for many of them
so am getting sent all sorts of stuff and people who are involved with the
driver are not. e.g. (to pick one at random):
$ ./scripts/get_maintainer.pl -f drivers/net/ethernet/nvidia/forcedeth.c
"David S. Miller" <davem@davemloft.net> (commit_signer:5/7=71%)
Ian Campbell <ian.campbell@citrix.com> (commit_signer:2/7=29%)
Eric Dumazet <eric.dumazet@gmail.com> (commit_signer:1/7=14%)
Jeff Kirsher <jeffrey.t.kirsher@intel.com> (commit_signer:1/7=14%)
Jiri Pirko <jpirko@redhat.com> (commit_signer:1/7=14%)
netdev@vger.kernel.org (open list:NETWORKING DRIVERS)
linux-kernel@vger.kernel.org (open list)
With the following patch the renames are followed and the result appears
much more sensible:
1 is a power of two, therefore rounddown_pow_of_two(1) should return 1.
It does in case the argument is a variable but in case it's a constant it
behaves wrong and returns 0. Probably nobody ever did it so this was
never noticed, however net/drivers/vmxnet3 with latest GCC does and breaks
on unicpu systems.
This is similar to Rolf's patch to roundup_pow_of_two(1).
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de> Reviewed-by: Jesper Juhl <jj@chaosbits.net> Signed-off-by: Andrei Warkentin <andreiw@vmware.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Andi Kleen [Thu, 8 Dec 2011 04:41:55 +0000 (15:41 +1100)]
brlocks/lglocks: clean up code
lglocks and brlocks are currently generated with some complicated macros
in lglock.h. But there's no reason I can see to not just use common
utility functions that get pointers to the lglock.
Since there are at least two users it makes sense to share this code in a
library.
This will also make it later possible to dynamically allocate lglocks.
In general the users now look more like normal function calls with
pointers, not magic macros.
The patch is rather large because I move over all users in one go to keep
it bisectable. This impacts the VFS somewhat in terms of lines changed.
But no actual behaviour change.
Signed-off-by: Andi Kleen <ak@linux.intel.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Nick Piggin <npiggin@kernel.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Jesper Juhl [Thu, 8 Dec 2011 04:41:55 +0000 (15:41 +1100)]
audit: always follow va_copy() with va_end()
A call to va_copy() should always be followed by a call to va_end() in the
same function. In kernel/autit.c::audit_log_vformat() this is not always
done. This patch makes sure va_end() is always called.
Signed-off-by: Jesper Juhl <jj@chaosbits.net> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Eric Paris <eparis@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Heiko Carstens [Thu, 8 Dec 2011 04:41:54 +0000 (15:41 +1100)]
mm,slub,x86: decouple size of struct page from CONFIG_CMPXCHG_LOCAL
While implementing cmpxchg_double() on s390 I realized that we don't set
CONFIG_CMPXCHG_LOCAL besides the fact that we have support for it.
However setting that option will increase the size of struct page by eight
bytes on 64 bit, which we certainly do not want. Also, it doesn't make
sense that a present cpu feature should increase the size of struct page.
Besides that it looks like the dependency to CMPXCHG_LOCAL is wrong and
that it should depend on CMPXCHG_DOUBLE instead.
This patch:
If an architecture supports CMPXCHG_LOCAL this shouldn't result
automatically in larger struct pages if the SLUB allocator is used.
Instead introduce a new config option "HAVE_ALIGNED_STRUCT_PAGE" which can
be selected if a double word aligned struct page is required. Also update
x86 Kconfig so that it should work as before.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com> Acked-by: Christoph Lameter <cl@linux.com> Cc: Pekka Enberg <penberg@kernel.org> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
WARNING: please, no spaces at the start of a line
#57: FILE: arch/m68k/amiga/config.c:515:
+ __noreturn;$
total: 0 errors, 1 warnings, 106 lines checked
./patches/treewide-convert-uses-of-attrib_noreturn-to-__noreturn.patch has style problems, please review.
If any of these errors are false positives, please report
them to the maintainer, see CHECKPATCH in MAINTAINERS.
Please run checkpatch prior to sending patches
Cc: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Shaohua Li [Thu, 8 Dec 2011 04:41:51 +0000 (15:41 +1100)]
intel_idle: fix API misuse
smp_call_function() only lets all other CPUs execute a specific function,
while we expect all CPUs do in intel_idle. Without the fix, we could have
one cpu which has auto_demotion enabled or has no boradcast timer setup.
Usually we don't see impact because auto demotion just harms power and the
intel_idle init is called in CPU 0, where boradcast timer delivers
interrupt, but this still could be a problem.
Signed-off-by: Shaohua Li <shaohua.li@intel.com> Cc: Len Brown <lenb@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Magnus Lynch [Thu, 8 Dec 2011 04:41:51 +0000 (15:41 +1100)]
hpet: factor timer allocate from open
The current implementation of the /dev/hpet driver couples opening the
device with allocating one of the (scarce) timers (aka comparators). This
is a limitation in that the main counter may be valuable to applications
seeking a high-resolution timer who have no use for the interrupt
generating functionality of the comparators.
This patch alters the open semantics so that when the device is opened, no
timer is allocated. Operations that depend on a timer being in context
implicitly attempt allocating a timer, to maintain backward compatibility.
There is also an IOCTL (HPET_ALLOC_TIMER _IO) added so that the
allocation may be done explicitly. (I prefer the explicit open then
allocate pattern but don't know how practical it would be to require all
existing code to be changed.)
/dev/hpet is accessed via mmap(). This is the only interface of /dev/hpet
that is actually used in practice.
[akpm@linux-foundation.org: coding-style tweaks]
[arnd@arndb.de: fix build] Signed-off-by: Magnus Lynch <maglyx@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: john stultz <johnstul@us.ibm.com> Acked-by: Clemens Ladisch <clemens@ladisch.de> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Michal Hocko [Thu, 8 Dec 2011 04:41:50 +0000 (15:41 +1100)]
mm: fix off-by-two in __zone_watermark_ok()
88f5acf8 ("mm: page allocator: adjust the per-cpu counter threshold when
memory is low") changed the form how free_pages is calculated but it
forgot that we used to do free_pages - ((1 << order) - 1) so we ended up
with off-by-two when calculating free_pages.
Reported-by: Wang Sheng-Hui <shhuiw@gmail.com> Signed-off-by: Michal Hocko <mhocko@suse.cz> Acked-by: Mel Gorman <mgorman@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Hillf Danton [Thu, 8 Dec 2011 04:41:50 +0000 (15:41 +1100)]
mm: compaction: push isolate search base of compact control one pfn ahead
After isolated the current pfn will no longer be scanned and isolated if
the next round is necessary, so push the isolate_migratepages search base
of the given compact_control one step ahead.
Signed-off-by: Hillf Danton <dhillf@gmail.com> Reviewed-by: Andrea Arcangeli <aarcange@redhat.com> Cc: Mel Gorman <mgorman@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Johannes Weiner [Thu, 8 Dec 2011 04:41:50 +0000 (15:41 +1100)]
Btrfs: pass __GFP_WRITE for buffered write page allocations
Tell the page allocator that pages allocated for a buffered write are
expected to become dirty soon.
Signed-off-by: Johannes Weiner <jweiner@redhat.com> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Michal Hocko <mhocko@suse.cz> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Dave Chinner <david@fromorbit.com> Cc: Jan Kara <jack@suse.cz> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Chris Mason <chris.mason@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Johannes Weiner [Thu, 8 Dec 2011 04:41:49 +0000 (15:41 +1100)]
mm: filemap: pass __GFP_WRITE from grab_cache_page_write_begin()
Tell the page allocator that pages allocated through
grab_cache_page_write_begin() are expected to become dirty soon.
Signed-off-by: Johannes Weiner <jweiner@redhat.com> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Reviewed-by: Michal Hocko <mhocko@suse.cz> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Dave Chinner <david@fromorbit.com> Cc: Jan Kara <jack@suse.cz> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Chris Mason <chris.mason@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Johannes Weiner [Thu, 8 Dec 2011 04:41:49 +0000 (15:41 +1100)]
mm: try to distribute dirty pages fairly across zones
The maximum number of dirty pages that exist in the system at any time is
determined by a number of pages considered dirtyable and a user-configured
percentage of those, or an absolute number in bytes.
This number of dirtyable pages is the sum of memory provided by all the
zones in the system minus their lowmem reserves and high watermarks, so
that the system can retain a healthy number of free pages without having
to reclaim dirty pages.
But there is a flaw in that we have a zoned page allocator which does not
care about the global state but rather the state of individual memory
zones. And right now there is nothing that prevents one zone from filling
up with dirty pages while other zones are spared, which frequently leads
to situations where kswapd, in order to restore the watermark of free
pages, does indeed have to write pages from that zone's LRU list. This
can interfere so badly with IO from the flusher threads that major
filesystems (btrfs, xfs, ext4) mostly ignore write requests from reclaim
already, taking away the VM's only possibility to keep such a zone
balanced, aside from hoping the flushers will soon clean pages from that
zone.
Enter per-zone dirty limits. They are to a zone's dirtyable memory what
the global limit is to the global amount of dirtyable memory, and try to
make sure that no single zone receives more than its fair share of the
globally allowed dirty pages in the first place. As the number of pages
considered dirtyable excludes the zones' lowmem reserves and high
watermarks, the maximum number of dirty pages in a zone is such that the
zone can always be balanced without requiring page cleaning.
As this is a placement decision in the page allocator and pages are
dirtied only after the allocation, this patch allows allocators to pass
__GFP_WRITE when they know in advance that the page will be written to and
become dirty soon. The page allocator will then attempt to allocate from
the first zone of the zonelist - which on NUMA is determined by the task's
NUMA memory policy - that has not exceeded its dirty limit.
At first glance, it would appear that the diversion to lower zones can
increase pressure on them, but this is not the case. With a full high
zone, allocations will be diverted to lower zones eventually, so it is
more of a shift in timing of the lower zone allocations. Workloads that
previously could fit their dirty pages completely in the higher zone may
be forced to allocate from lower zones, but the amount of pages that
"spill over" are limited themselves by the lower zones' dirty constraints,
and thus unlikely to become a problem.
For now, the problem of unfair dirty page distribution remains for NUMA
configurations where the zones allowed for allocation are in sum not big
enough to trigger the global dirty limits, wake up the flusher threads and
remedy the situation. Because of this, an allocation that could not
succeed on any of the considered zones is allowed to ignore the dirty
limits before going into direct reclaim or even failing the allocation,
until a future patch changes the global dirty throttling and flusher
thread activation so that they take individual zone states into account.
Test results
15M DMA + 3246M DMA32 + 504 Normal = 3765M memory
40% dirty ratio
16G USB thumb drive
10 runs of dd if=/dev/zero of=disk/zeroes bs=32k count=$((10 << 15))
Signed-off-by: Johannes Weiner <jweiner@redhat.com> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Acked-by: Mel Gorman <mgorman@suse.de> Reviewed-by: Michal Hocko <mhocko@suse.cz> Tested-by: Wu Fengguang <fengguang.wu@intel.com> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Jan Kara <jack@suse.cz> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Rik van Riel <riel@redhat.com> Cc: Chris Mason <chris.mason@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Johannes Weiner [Thu, 8 Dec 2011 04:41:49 +0000 (15:41 +1100)]
mm: writeback: cleanups in preparation for per-zone dirty limits
The next patch will introduce per-zone dirty limiting functions in
addition to the traditional global dirty limiting.
Rename determine_dirtyable_memory() to global_dirtyable_memory() before
adding the zone-specific version, and fix up its documentation.
Also, move the functions to determine the dirtyable memory and the
function to calculate the dirty limit based on that together so that their
relationship is more apparent and that they can be commented on as a
group.
Signed-off-by: Johannes Weiner <jweiner@redhat.com> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Acked-by: Mel Gorman <mel@suse.de> Reviewed-by: Michal Hocko <mhocko@suse.cz> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Dave Chinner <david@fromorbit.com> Cc: Jan Kara <jack@suse.cz> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Rik van Riel <riel@redhat.com> Cc: Chris Mason <chris.mason@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Chris Mason <chris.mason@oracle.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Jan Kara <jack@suse.cz> Cc: Johannes Weiner <jweiner@redhat.com> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Mel Gorman <mgorman@suse.de> Cc: Michal Hocko <mhocko@suse.cz> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Rik van Riel <riel@redhat.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Johannes Weiner [Thu, 8 Dec 2011 04:41:48 +0000 (15:41 +1100)]
mm: exclude reserved pages from dirtyable memory
Per-zone dirty limits try to distribute page cache pages allocated for
writing across zones in proportion to the individual zone sizes, to reduce
the likelihood of reclaim having to write back individual pages from the
LRU lists in order to make progress.
This patch:
The amount of dirtyable pages should not include the full number of free
pages: there is a number of reserved pages that the page allocator and
kswapd always try to keep free.
The closer (reclaimable pages - dirty pages) is to the number of reserved
pages, the more likely it becomes for reclaim to run into dirty pages:
This patch introduces a per-zone dirty reserve that takes both the lowmem
reserve as well as the high watermark of the zone into account, and a
global sum of those per-zone values that is subtracted from the global
amount of dirtyable pages. The lowmem reserve is unavailable to page
cache allocations and kswapd tries to keep the high watermark free. We
don't want to end up in a situation where reclaim has to clean pages in
order to balance zones.
Not treating reserved pages as dirtyable on a global level is only a
conceptual fix. In reality, dirty pages are not distributed equally
across zones and reclaim runs into dirty pages on a regular basis.
But it is important to get this right before tackling the problem on a
per-zone level, where the distance between reclaim and the dirty pages is
mostly much smaller in absolute numbers.
Signed-off-by: Johannes Weiner <jweiner@redhat.com> Reviewed-by: Rik van Riel <riel@redhat.com> Reviewed-by: Michal Hocko <mhocko@suse.cz> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Dave Chinner <david@fromorbit.com> Cc: Jan Kara <jack@suse.cz> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Chris Mason <chris.mason@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
David Rientjes [Thu, 8 Dec 2011 04:41:47 +0000 (15:41 +1100)]
mm, debug: test for online nid when allocating on single node
Calling alloc_pages_exact_node() means the allocation only passes the
zonelist of a single node into the page allocator. If that node isn't
online, it's zonelist may never have been initialized causing a strange
oops that may not immediately be clear.
I recently debugged an issue where node 0 wasn't online and an allocator
was passing 0 to alloc_pages_exact_node() and it resulted in a NULL
pointer on zonelist->_zoneref. If CONFIG_DEBUG_VM is enabled, though, it
would be nice to catch this a bit earlier.
Signed-off-by: David Rientjes <rientjes@google.com> Acked-by: Mel Gorman <mgorman@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Shawn Bohrer [Thu, 8 Dec 2011 04:41:47 +0000 (15:41 +1100)]
fadvise: only initiate writeback for specified range with FADV_DONTNEED
Previously POSIX_FADV_DONTNEED would start writeback for the entire file
when the bdi was not write congested. This negatively impacts performance
if the file contians dirty pages outside of the requested range. This
change uses __filemap_fdatawrite_range() to only initiate writeback for
the requested range.
Signed-off-by: Shawn Bohrer <sbohrer@rgmadvisors.com> Acked-by: Johannes Weiner <jweiner@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Disable slub debug facilities and allocate slabs at minimal order when
debug_guardpage_minorder > 0 to increase probability to catch random
memory corruption by cpu exception.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Christoph Lameter <cl@linux.com> Cc: Mel Gorman <mgorman@suse.de> Cc: Stanislaw Gruszka <sgruszka@redhat.com> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
When debugging with CONFIG_DEBUG_PAGEALLOC and debug_guardpage_minorder >
0, we have lot of free pages that are not marked so. Snapshot code
account them as savable, what cause hibernate memory preallocation
failure.
It is pretty hard to make hibernate allocation succeed with
debug_guardpage_minorder=1. This change at least make it possible when
system has relatively big amount of RAM.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com> Acked-by: Rafael J. Wysocki <rjw@sisk.pl> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Christoph Lameter <cl@linux-foundation.org> Cc: Mel Gorman <mgorman@suse.de> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
With CONFIG_DEBUG_PAGEALLOC configured, the CPU will generate an exception
on access (read,write) to an unallocated page, which permits us to catch
code which corrupts memory. However the kernel is trying to maximise
memory usage, hence there are usually few free pages in the system and
buggy code usually corrupts some crucial data.
This patch changes the buddy allocator to keep more free/protected pages
and to interlace free/protected and allocated pages to increase the
probability of catching corruption.
When the kernel is compiled with CONFIG_DEBUG_PAGEALLOC,
debug_guardpage_minorder defines the minimum order used by the page
allocator to grant a request. The requested size will be returned with
the remaining pages used as guard pages.
The default value of debug_guardpage_minorder is zero: no change from
current behaviour.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com> Cc: Mel Gorman <mgorman@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Christoph Lameter <cl@linux-foundation.org> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
David Daney [Thu, 8 Dec 2011 04:41:45 +0000 (15:41 +1100)]
hugetlb: replace BUG() with BUILD_BUG() for dummy definitions
The file linux/hugetlb.h has many places where dummy symbols were defined
so that the main source code would contain fewer:
#ifdef CONFIG_HUGETLBFS
or
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
If there were any misuse of these symbols, the only symptom would be an
OOPS at runtime. Change the BUG() to BUILD_BUG() to catch any such abuse
at compile time instead.
Signed-off-by: David Daney <david.daney@cavium.com> Cc: David Rientjes <rientjes@google.com> Cc: DM <dm.n9107@gmail.com> Cc: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
David Daney [Thu, 8 Dec 2011 04:41:45 +0000 (15:41 +1100)]
kernel.h: Add BUILD_BUG() macro.
We can place this in definitions that we expect the compiler to remove
by dead code elimination. If this assertion fails, we get a nice
error message at build time.
The GCC function attribute error("message") was added in version 4.3,
so we define a new macro __linktime_error(message) to expand to this
for GCC-4.3 and later. This will give us an error diagnostic from the
compiler on the line that fails. For other compilers
__linktime_error(message) expands to nothing, and we have to be
content with a link time error, but at least we will still get a build
error.
BUILD_BUG() expands to the undefined function __build_bug_failed() and
will fail at link time if the compiler ever emits code for it. On
GCC-4.3 and later, attribute((error())) is used so that the failure
will be noted at compile time instead.
Acked-by: David Howells <dhowells@redhat.com> Acked-by: David Rientjes <rientjes@google.com> Signed-off-by: David Daney <david.daney@cavium.com> Cc: David Rientjes <rientjes@google.com> Cc: DM <dm.n9107@gmail.com> Cc: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
projects / karo-tx-linux.git / log