page-writeback accounting is presently performed in the page-flags macros.
This is inconsistent and a bit ugly and makes it awkward to implement
per-backing_dev under-writeback page accounting.
So move this accounting down to the callsite(s).
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
__acquire
|
lock _____
| \
| __contended
| |
| wait
| _______/
|/
|
__acquired
|
__release
|
unlock
We measure acquisition and contention bouncing.
This is done by recording a cpu stamp in each lock instance.
Contention bouncing requires the cpu stamp to be set on acquisition. Hence we
move __acquired into the generic path.
__acquired is then used to measure acquisition bouncing by comparing the
current cpu with the old stamp before replacing it.
__contended is used to measure contention bouncing (only useful for preemptable
locks)
[akpm@linux-foundation.org: cleanups]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
- update the copyright notices
- use the default hash function
- fix a thinko in a BUILD_BUG_ON
- add a WARN_ON to spot inconsitent naming
- fix a termination issue in /proc/lock_stat
[akpm@linux-foundation.org: cleanups]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Introduce the core lock statistics code.
Lock statistics provides lock wait-time and hold-time (as well as the count
of corresponding contention and acquisitions events). Also, the first few
call-sites that encounter contention are tracked.
Lock wait-time is the time spent waiting on the lock. This provides insight
into the locking scheme, that is, a heavily contended lock is indicative of
a too coarse locking scheme.
Lock hold-time is the duration the lock was held, this provides a reference for
the wait-time numbers, so they can be put into perspective.
1)
lock
2)
... do stuff ..
unlock
3)
The time between 1 and 2 is the wait-time. The time between 2 and 3 is the
hold-time.
The lockdep held-lock tracking code is reused, because it already collects locks
into meaningful groups (classes), and because it is an existing infrastructure
for lock instrumentation.
Currently lockdep tracks lock acquisition with two hooks:
lock()
lock_acquire()
_lock()
... code protected by lock ...
unlock()
lock_release()
_unlock()
We need to extend this with two more hooks, in order to measure contention.
lock_contended() - used to measure contention events
lock_acquired() - completion of the contention
These are then placed the following way:
lock()
lock_acquire()
if (!_try_lock())
lock_contended()
_lock()
lock_acquired()
... do locked stuff ...
unlock()
lock_release()
_unlock()
(Note: the try_lock() 'trick' is used to avoid instrumenting all platform
dependent lock primitive implementations.)
It is also possible to toggle the two lockdep features at runtime using:
/proc/sys/kernel/prove_locking
/proc/sys/kernel/lock_stat
(esp. turning off the O(n^2) prove_locking functionaliy can help)
[akpm@linux-foundation.org: build fixes]
[akpm@linux-foundation.org: nuke unneeded ifdefs]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Jason Baron <jbaron@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use the lockdep infrastructure to track lock contention and other lock
statistics.
It tracks lock contention events, and the first four unique call-sites that
encountered contention.
It also measures lock wait-time and hold-time in nanoseconds. The minimum and
maximum times are tracked, as well as a total (which together with the number
of event can give the avg).
All statistics are done per lock class, per write (exclusive state) and per read
(shared state).
The statistics are collected per-cpu, so that the collection overhead is
minimized via having no global cachemisses.
This new lock statistics feature is independent of the lock dependency checking
traditionally done by lockdep; it just shares the lock tracking code. It is
also possible to enable both and runtime disabled either component - thereby
avoiding the O(n^2) lock chain walks for instance.
This patch:
raw_spinlock_t should not use lockdep (and doesn't) since lockdep itself
relies on it.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We ignore signals for about 30 seconds to give userspace a chance to see the
upcall. As we did not block signals we ended up in a busy loop for the
remainder of the period when a signal is received.
Signed-off-by: Jan Harkes <jaharkes@cs.cmu.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch adds an interface to set/reset flags which determines each memory
segment should be dumped or not when a core file is generated.
/proc/<pid>/coredump_filter file is provided to access the flags. You can
change the flag status for a particular process by writing to or reading from
the file.
The flag status is inherited to the child process when it is created.
Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: David Howells <dhowells@redhat.com>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch changes mm_struct.dumpable to a pair of bit flags.
set_dumpable() converts three-value dumpable to two flags and stores it into
lower two bits of mm_struct.flags instead of mm_struct.dumpable.
get_dumpable() behaves in the opposite way.
[akpm@linux-foundation.org: export set_dumpable]
Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: David Howells <dhowells@redhat.com>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Stackable file systems, among others, frequently need to lookup paths or
path components starting from an arbitrary point in the namespace
(identified by a dentry and a vfsmount). Currently, such file systems use
lookup_one_len, which is frowned upon [1] as it does not pass the lookup
intent along; not passing a lookup intent, for example, can trigger BUG_ON's
when stacking on top of NFSv4.
The first patch introduces a new lookup function to allow lookup starting
from an arbitrary point in the namespace. This approach has been suggested
by Christoph Hellwig [2].
The second patch changes sunrpc to use vfs_path_lookup.
The third patch changes nfsctl.c to use vfs_path_lookup.
The fourth patch marks link_path_walk static.
The fifth, and last patch, unexports path_walk because it is no longer
unnecessary to call it directly, and using the new vfs_path_lookup is
cleaner.
For example, the following snippet of code, looks up "some/path/component"
in a directory pointed to by parent_{dentry,vfsmnt}:
err = vfs_path_lookup(parent_dentry, parent_vfsmnt,
"some/path/component", 0, &nd);
if (!err) {
/* exits */
...
/* once done, release the references */
path_release(&nd);
} else if (err == -ENOENT) {
/* doesn't exist */
} else {
/* other error */
}
VFS functions such as lookup_create can be used on the nameidata structure
to pass the create intent to the file system.
Signed-off-by: Josef 'Jeff' Sipek <jsipek@cs.sunysb.edu>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Christoph Hellwig <hch@lst.de>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: Neil Brown <neilb@suse.de>
Cc: Michael Halcrow <mhalcrow@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove the arg+env limit of MAX_ARG_PAGES by copying the strings directly from
the old mm into the new mm.
We create the new mm before the binfmt code runs, and place the new stack at
the very top of the address space. Once the binfmt code runs and figures out
where the stack should be, we move it downwards.
It is a bit peculiar in that we have one task with two mm's, one of which is
inactive.
[a.p.zijlstra@chello.nl: limit stack size]
Signed-off-by: Ollie Wild <aaw@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: <linux-arch@vger.kernel.org>
Cc: Hugh Dickins <hugh@veritas.com>
[bunk@stusta.de: unexport bprm_mm_init]
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The purpose of audit_bprm() is to log the argv array to a userspace daemon at
the end of the execve system call. Since user-space hasn't had time to run,
this array is still in pristine state on the process' stack; so no need to
copy it, we can just grab it from there.
In order to minimize the damage to audit_log_*() copy each string into a
temporary kernel buffer first.
Currently the audit code requires that the full argument vector fits in a
single packet. So currently it does clip the argv size to a (sysctl) limit,
but only when execve auditing is enabled.
If the audit protocol gets extended to allow for multiple packets this check
can be removed.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ollie Wild <aaw@google.com>
Cc: <linux-audit@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
I realise jprobes are a razor-blades-included type of interface, but that
doesn't mean we can't try and make them safer to use. This guy I know once
wrote code like this:
struct jprobe jp = { .kp.symbol_name = "foo", .entry = "jprobe_foo" };
And then his kernel exploded. Oops.
This patch adds an arch hook, arch_deref_entry_point() (I don't like it
either) which takes the void * in a struct jprobe, and gives back the text
address that it represents.
We can then use that in register_jprobe() to check that the entry point we're
passed is actually in the kernel text, rather than just some random value.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Cc: Prasanna S Panchamukhi <prasanna@in.ibm.com>
Acked-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently jprobe.entry is a kprobe_opcode_t *, but that's a lie. On some
platforms it doesn't point to an opcode at all, it points to a function
descriptor.
It's really a pointer to something that the arch code can turn into a function
entry point. And that's what actually happens, none of the generic code ever
looks at jprobe.entry, it's only ever dereferenced by arch code.
So just make it a void *.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Cc: Prasanna S Panchamukhi <prasanna@in.ibm.com>
Acked-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Split ondemand readahead interface into two functions. I think this makes it
a little clearer for non-readahead experts (like Rusty).
Internally they both call ondemand_readahead(), but the page argument is
changed to an obvious boolean flag.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Share the same page flag bit for PG_readahead and PG_reclaim.
One is used only on file reads, another is only for emergency writes. One
is used mostly for fresh/young pages, another is for old pages.
Combinations of possible interactions are:
a) clear PG_reclaim => implicit clear of PG_readahead
it will delay an asynchronous readahead into a synchronous one
it actually does _good_ for readahead:
the pages will be reclaimed soon, it's readahead thrashing!
in this case, synchronous readahead makes more sense.
b) clear PG_readahead => implicit clear of PG_reclaim
one(and only one) page will not be reclaimed in time
it can be avoided by checking PageWriteback(page) in readahead first
c) set PG_reclaim => implicit set of PG_readahead
will confuse readahead and make it restart the size rampup process
it's a trivial problem, and can mostly be avoided by checking
PageWriteback(page) first in readahead
d) set PG_readahead => implicit set of PG_reclaim
PG_readahead will never be set on already cached pages.
PG_reclaim will always be cleared on dirtying a page.
so not a problem.
In summary,
a) we get better behavior
b,d) possible interactions can be avoided
c) racy condition exists that might affect readahead, but the chance
is _really_ low, and the hurt on readahead is trivial.
Compound pages also use PG_reclaim, but for now they do not interact with
reclaim/readahead code.
Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is a minimal readahead algorithm that aims to replace the current one.
It is more flexible and reliable, while maintaining almost the same behavior
and performance. Also it is full integrated with adaptive readahead.
It is designed to be called on demand:
- on a missing page, to do synchronous readahead
- on a lookahead page, to do asynchronous readahead
In this way it eliminated the awkward workarounds for cache hit/miss,
readahead thrashing, retried read, and unaligned read. It also adopts the
data structure introduced by adaptive readahead, parameterizes readahead
pipelining with `lookahead_index', and reduces the current/ahead windows to
one single window.
HEURISTICS
The logic deals with four cases:
- sequential-next
found a consistent readahead window, so push it forward
- random
standalone small read, so read as is
- sequential-first
create a new readahead window for a sequential/oversize request
- lookahead-clueless
hit a lookahead page not associated with the readahead window,
so create a new readahead window and ramp it up
In each case, three parameters are determined:
- readahead index: where the next readahead begins
- readahead size: how much to readahead
- lookahead size: when to do the next readahead (for pipelining)
BEHAVIORS
The old behaviors are maximally preserved for trivial sequential/random reads.
Notable changes are:
- It no longer imposes strict sequential checks.
It might help some interleaved cases, and clustered random reads.
It does introduce risks of a random lookahead hit triggering an
unexpected readahead. But in general it is more likely to do good
than to do evil.
- Interleaved reads are supported in a minimal way.
Their chances of being detected and proper handled are still low.
- Readahead thrashings are better handled.
The current readahead leads to tiny average I/O sizes, because it
never turn back for the thrashed pages. They have to be fault in
by do_generic_mapping_read() one by one. Whereas the on-demand
readahead will redo readahead for them.
OVERHEADS
The new code reduced the overheads of
- excessively calling the readahead routine on small sized reads
(the current readahead code insists on seeing all requests)
- doing a lot of pointless page-cache lookups for small cached files
(the current readahead only turns itself off after 256 cache hits,
unfortunately most files are < 1MB, so never see that chance)
That accounts for speedup of
- 0.3% on 1-page sequential reads on sparse file
- 1.2% on 1-page cache hot sequential reads
- 3.2% on 256-page cache hot sequential reads
- 1.3% on cache hot `tar /lib`
However, it does introduce one extra page-cache lookup per cache miss, which
impacts random reads slightly. That's 1% overheads for 1-page random reads on
sparse file.
PERFORMANCE
The basic benchmark setup is
- 2.6.20 kernel with on-demand readahead
- 1MB max readahead size
- 2.9GHz Intel Core 2 CPU
- 2GB memory
- 160G/8M Hitachi SATA II 7200 RPM disk
The benchmarks show that
- it maintains the same performance for trivial sequential/random reads
- sysbench/OLTP performance on MySQL gains up to 8%
- performance on readahead thrashing gains up to 3 times
iozone throughput (KB/s): roughly the same
==========================================
iozone -c -t1 -s 4096m -r 64k
2.6.20 on-demand gain
first run
" Initial write " 61437.27 64521.53 +5.0%
" Rewrite " 47893.02 48335.20 +0.9%
" Read " 62111.84 62141.49 +0.0%
" Re-read " 62242.66 62193.17 -0.1%
" Reverse Read " 50031.46 49989.79 -0.1%
" Stride read " 8657.61 8652.81 -0.1%
" Random read " 13914.28 13898.23 -0.1%
" Mixed workload " 19069.27 19033.32 -0.2%
" Random write " 14849.80 14104.38 -5.0%
" Pwrite " 62955.30 65701.57 +4.4%
" Pread " 62209.99 62256.26 +0.1%
second run
" Initial write " 60810.31 66258.69 +9.0%
" Rewrite " 49373.89 57833.66 +17.1%
" Read " 62059.39 62251.28 +0.3%
" Re-read " 62264.32 62256.82 -0.0%
" Reverse Read " 49970.96 50565.72 +1.2%
" Stride read " 8654.81 8638.45 -0.2%
" Random read " 13901.44 13949.91 +0.3%
" Mixed workload " 19041.32 19092.04 +0.3%
" Random write " 14019.99 14161.72 +1.0%
" Pwrite " 64121.67 68224.17 +6.4%
" Pread " 62225.08 62274.28 +0.1%
In summary, writes are unstable, reads are pretty close on average:
access pattern 2.6.20 on-demand gain
Read 62085.61 62196.38 +0.2%
Re-read 62253.49 62224.99 -0.0%
Reverse Read 50001.21 50277.75 +0.6%
Stride read 8656.21 8645.63 -0.1%
Random read 13907.86 13924.07 +0.1%
Mixed workload 19055.29 19062.68 +0.0%
Pread 62217.53 62265.27 +0.1%
aio-stress: roughly the same
============================
aio-stress -l -s4096 -r128 -t1 -o1 knoppix511-dvd-cn.iso
aio-stress -l -s4096 -r128 -t1 -o3 knoppix511-dvd-cn.iso
2.6.20 on-demand delta
sequential 92.57s 92.54s -0.0%
random 311.87s 312.15s +0.1%
sysbench fileio: roughly the same
=================================
sysbench --test=fileio --file-io-mode=async --file-test-mode=rndrw \
--file-total-size=4G --file-block-size=64K \
--num-threads=001 --max-requests=10000 --max-time=900 run
threads 2.6.20 on-demand delta
first run
1 59.1974s 59.2262s +0.0%
2 58.0575s 58.2269s +0.3%
4 48.0545s 47.1164s -2.0%
8 41.0684s 41.2229s +0.4%
16 35.8817s 36.4448s +1.6%
32 32.6614s 32.8240s +0.5%
64 23.7601s 24.1481s +1.6%
128 24.3719s 23.8225s -2.3%
256 23.2366s 22.0488s -5.1%
second run
1 59.6720s 59.5671s -0.2%
8 41.5158s 41.9541s +1.1%
64 25.0200s 23.9634s -4.2%
256 22.5491s 20.9486s -7.1%
Note that the numbers are not very stable because of the writes.
The overall performance is close when we sum all seconds up:
sum all up 495.046s 491.514s -0.7%
sysbench oltp (trans/sec): up to 8% gain
========================================
sysbench --test=oltp --oltp-table-size=10000000 --oltp-read-only \
--mysql-socket=/var/run/mysqld/mysqld.sock \
--mysql-user=root --mysql-password=readahead \
--num-threads=064 --max-requests=10000 --max-time=900 run
10000-transactions run
threads 2.6.20 on-demand gain
1 62.81 64.56 +2.8%
2 67.97 70.93 +4.4%
4 81.81 85.87 +5.0%
8 94.60 97.89 +3.5%
16 99.07 104.68 +5.7%
32 95.93 104.28 +8.7%
64 96.48 103.68 +7.5%
5000-transactions run
1 48.21 48.65 +0.9%
8 68.60 70.19 +2.3%
64 70.57 74.72 +5.9%
2000-transactions run
1 37.57 38.04 +1.3%
2 38.43 38.99 +1.5%
4 45.39 46.45 +2.3%
8 51.64 52.36 +1.4%
16 54.39 55.18 +1.5%
32 52.13 54.49 +4.5%
64 54.13 54.61 +0.9%
That's interesting results. Some investigations show that
- MySQL is accessing the db file non-uniformly: some parts are
more hot than others
- It is mostly doing 4-page random reads, and sometimes doing two
reads in a row, the latter one triggers a 16-page readahead.
- The on-demand readahead leaves many lookahead pages (flagged
PG_readahead) there. Many of them will be hit, and trigger
more readahead pages. Which might save more seeks.
- Naturally, the readahead windows tend to lie in hot areas,
and the lookahead pages in hot areas is more likely to be hit.
- The more overall read density, the more possible gain.
That also explains the adaptive readahead tricks for clustered random reads.
readahead thrashing: 3 times better
===================================
We boot kernel with "mem=128m single", and start a 100KB/s stream on every
second, until reaching 200 streams.
max throughput min avg I/O size
2.6.20: 5MB/s 16KB
on-demand: 15MB/s 140KB
Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn>
Cc: Steven Pratt <slpratt@austin.ibm.com>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
