From: Andi Kleen <andi@firstfloor.org>
To: linux-kernel@vger.kernel.org
Cc: torvalds@linux-foundation.org, akpm@linux-foundation.org,
x86@kernel.org, Andi Kleen <ak@linux.intel.com>
Subject: [PATCH 29/29] tsx: Add documentation for lock-elision
Date: Fri, 22 Mar 2013 18:25:23 -0700 [thread overview]
Message-ID: <1364001923-10796-30-git-send-email-andi@firstfloor.org> (raw)
In-Reply-To: <1364001923-10796-1-git-send-email-andi@firstfloor.org>
From: Andi Kleen <ak@linux.intel.com>
Document the tunables and the statistics in Documentation/lock-elision.txt
Signed-off-by: Andi Kleen <ak@linux.intel.com>
---
Documentation/lock-elision.txt | 94 ++++++++++++++++++++++++++++++++++++++++
1 files changed, 94 insertions(+), 0 deletions(-)
create mode 100644 Documentation/lock-elision.txt
diff --git a/Documentation/lock-elision.txt b/Documentation/lock-elision.txt
new file mode 100644
index 0000000..bc5a5ac
--- /dev/null
+++ b/Documentation/lock-elision.txt
@@ -0,0 +1,94 @@
+
+The Linux kernel uses Intel TSX lock elision for most of its locking primitives,
+when the hardware supports TSX.
+
+This allows to run lock regions that follow specific conditions to run
+in parallel without explicit blocking in a memory transaction. If the
+transaction fails the code falls back to the normal lock.
+
+The lock elision is implemented using RTM.
+
+To measure transaction success the TSX perf extensions should be used.
+(At the time of this writing this requires unmerged perf patches,
+as in the full Haswell PMU patch series in hsw/pmu* in
+git://git.kernel.org/pub/scm/linux/kernel/git/ak/linux-misc.git)
+
+perf stat -T ...
+perf record -e tx-aborts ...
+
+The elision code is mostly adaptive, that is the lock predicts whether
+it should elide or not based on the history.
+
+There are various tunables available in /sys/module/rtm_locks/parameters.
+The tunables are grouped by lock types.
+
+Elision can be enabled by lock types:
+
+rwsem_elision Enable elision for read-write semaphores
+mutex_elision Enable elision for mutexes
+spinlock_elision Enable elision for spinlocks
+rwlock_elision Enable elision for read-write spinlocks
+bitlock_elision Enable elision for bit spinlocks
+
+Some lock types use adaptive algorithms. The adaptive algorithm
+has a number of tunables which can be tuned. Each tunable
+is named LOCKTYPE_TUNABLE. For example to tune the conflict retries
+for mutexes do
+
+echo NUMBER > /sys/module/rtm_locks/parameters/mutex_conflict_retry
+
+Valid adaptive lock types are:
+mutex, readlock, writelock, readsem, writesem, spinlock
+
+The current tunables are (subject to change):
+
+In general increasing the skip counts makes lock elision more conservative,
+while increasing retries or timeout makes it more aggressive.
+
+*_elision Global elision enable for the lock type.
+*_conflict_abort_skip Number of elisions to skip when a transaction
+ failed due to a number of retried conflicts.
+*_conflict_retry Number of retries on memory conflicts
+*_internal_abort_skip Number of elision to skip when a transaction
+ failed due to a reason inside the transaction
+ (that is not caused by another CPU)
+*_lock_busy_retry How often to retry wen the lock is busy on
+ elision.
+*_lock_busy_skip How often to skip elision when retries on
+ lock busy failed.
+*_other_abort_skip How often to skip elision when the transaction
+ failed for other reasons (e.g. capacity overflow)
+*_retry_timeout How often to spin while waiting for a lock to free
+ before retrying
+
+Note these tunables do not present an ABI and may change as the internals
+evolve.
+
+Additional statistics:
+
+lock_el_skip Number of times a lock didn't elide due to skipping
+lock_el_skip_start Number of times a lock started skipping
+
+The average skip is lock_el_skip / lock_el_start_skip
+Additional statistics can be gotten using perf PMU TSX events.
+
+There is also a elision:elision_skip_start trace point that triggers every time
+a lock starts skipping elision due to non success.
+
+References:
+"Adding Lock elision to Linux"
+http://halobates.de/adding-lock-elision-to-linux.pdf
+
+"Adding lock elision to the GNU C Library"
+http://lwn.net/Articles/533894/
+
+Full TSX specification:
+http://software.intel.com/file/41417 (chapter 8)
+
+Glossary:
+cache line Aligned 64 byte region
+conflict Another CPU writes to a cache line read in a elided lock region,
+ or reads from a cache line written to in the region.
+abort Rolling back the transaction and undoing its side effect
+
+Andi Kleen
--
1.7.7.6
next prev parent reply other threads:[~2013-03-23 1:26 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-03-23 1:24 RFC: Kernel lock elision for TSX Andi Kleen
2013-03-23 1:24 ` [PATCH 01/29] tsx: Add generic noop macros for RTM intrinsics Andi Kleen
2013-03-25 3:39 ` Michael Neuling
2013-03-25 8:19 ` Andi Kleen
2013-03-25 8:50 ` Michael Neuling
2013-03-23 1:24 ` [PATCH 02/29] x86, tsx: Add " Andi Kleen
2013-03-25 3:40 ` Michael Neuling
2013-03-25 8:15 ` Andi Kleen
2013-03-25 8:54 ` Michael Neuling
2013-03-25 9:32 ` Andi Kleen
2013-03-23 1:24 ` [PATCH 03/29] tsx: Add generic disable_txn macros Andi Kleen
2013-03-23 1:24 ` [PATCH 04/29] tsx: Add generic linux/elide.h macros Andi Kleen
2013-03-23 1:24 ` [PATCH 05/29] x86, tsx: Add a minimal RTM tester at bootup Andi Kleen
2013-03-23 1:25 ` [PATCH 06/29] checkpatch: Don't warn about if ((status = _xbegin()) == _XBEGIN_STARTED) Andi Kleen
2013-03-25 3:39 ` Michael Neuling
2013-03-23 1:25 ` [PATCH 07/29] x86, tsx: Don't abort immediately in __read/write_lock_failed Andi Kleen
2013-03-23 1:25 ` [PATCH 08/29] locking, tsx: Add support for arch_read/write_unlock_irq/flags Andi Kleen
2013-03-23 1:25 ` [PATCH 09/29] x86, xen: Support arch_spin_unlock_irq/flags Andi Kleen
2013-03-23 1:25 ` [PATCH 10/29] locking, tsx: Add support for arch_spin_unlock_irq/flags Andi Kleen
2013-03-23 1:25 ` [PATCH 11/29] x86, paravirt: Add support for arch_spin_unlock_flags/irq Andi Kleen
2013-03-23 1:25 ` [PATCH 12/29] x86, tsx: Add a per thread transaction disable count Andi Kleen
2013-03-23 11:51 ` Borislav Petkov
2013-03-23 13:51 ` Andi Kleen
2013-03-23 15:52 ` Borislav Petkov
2013-03-23 16:25 ` Borislav Petkov
2013-03-23 17:16 ` Linus Torvalds
2013-03-23 17:32 ` Borislav Petkov
2013-03-23 18:01 ` Andi Kleen
2013-03-23 1:25 ` [PATCH 13/29] params: Add a per cpu module param type Andi Kleen
2013-03-23 1:25 ` [PATCH 14/29] params: Add static key module param Andi Kleen
2013-03-23 1:25 ` [PATCH 15/29] x86, tsx: Add TSX lock elision infrastructure Andi Kleen
2013-03-23 1:25 ` [PATCH 16/29] locking, tsx: Allow architecture to control mutex fast path owner field Andi Kleen
2013-03-23 1:25 ` [PATCH 17/29] x86, tsx: Enable lock elision for mutexes Andi Kleen
2013-03-23 1:25 ` [PATCH 18/29] locking, tsx: Abort is mutex_is_locked() Andi Kleen
2013-03-23 1:25 ` [PATCH 19/29] x86, tsx: Add support for rwsem elision Andi Kleen
2013-03-23 1:25 ` [PATCH 20/29] x86, tsx: Enable elision for read write spinlocks Andi Kleen
2013-03-23 1:25 ` [PATCH 21/29] locking, tsx: Protect assert_spin_locked() with _xtest() Andi Kleen
2013-03-23 1:25 ` [PATCH 22/29] locking, tsx: Add a trace point for elision skipping Andi Kleen
2013-03-23 1:25 ` [PATCH 23/29] x86, tsx: Add generic per-lock adaptive lock elision support Andi Kleen
2013-03-23 1:25 ` [PATCH 24/29] x86, tsx: Use adaptive elision for mutexes Andi Kleen
2013-03-23 1:25 ` [PATCH 25/29] x86, tsx: Add adaption support for spinlocks Andi Kleen
2013-03-23 1:25 ` [PATCH 26/29] x86, tsx: Add adaptation support to rw spinlocks Andi Kleen
2013-03-23 1:25 ` [PATCH 27/29] locking, tsx: Add elision to bit spinlocks Andi Kleen
2013-03-23 1:25 ` [PATCH 28/29] x86, tsx: Add adaptive elision for rwsems Andi Kleen
2013-03-23 1:25 ` Andi Kleen [this message]
2013-03-23 17:11 ` RFC: Kernel lock elision for TSX Linus Torvalds
2013-03-23 18:00 ` Andi Kleen
2013-03-23 18:02 ` Andi Kleen
2013-03-24 14:17 ` Benjamin Herrenschmidt
2013-03-25 0:59 ` Michael Neuling
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1364001923-10796-30-git-send-email-andi@firstfloor.org \
--to=andi@firstfloor.org \
--cc=ak@linux.intel.com \
--cc=akpm@linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=torvalds@linux-foundation.org \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®