urcu.git
8 years agoFix: compat futex duplicated lock and completion
Mathieu Desnoyers [Mon, 30 Sep 2013 18:54:22 +0000 (14:54 -0400)] 
Fix: compat futex duplicated lock and completion

compat_futex.c has one instance included in each urcu shared object, as
well as within some of the test applications. However, it is expected
that an entire program interact with the same lock and completion
variables. Therefore, define them as globally visible, but weak, so the
entire program agree on which object should be used.

Reported-by: Vladimir Nikulichev <nvs@tbricks.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: i386 compat code duplicated mutex instances
Mathieu Desnoyers [Mon, 30 Sep 2013 18:40:33 +0000 (14:40 -0400)] 
Fix: i386 compat code duplicated mutex instances

compat_arch_x86.c is linked into many .so and even into test programs.
The basic problem with this is that it contains a statically defined
mutex, which will fail to protect concurrent use of this compat code by
different shared objects.

Fix this by defining both the mutex (now called __urcu_x86_compat_mutex)
and __rcu_cas_avail as weak symbols. Therefore, the first symbol that
gets loaded in a program will by used by everyone.

Reported-by: Vladimir Nikulichev <nvs@tbricks.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: urcu-bp: Bulletproof RCU arena resize bug
Mathieu Desnoyers [Mon, 30 Sep 2013 15:49:32 +0000 (11:49 -0400)] 
Fix: urcu-bp: Bulletproof RCU arena resize bug

> From: "Milosz Tanski" <milosz@adfin.com>

> While trying to use the BP flavor of RCU I ran into random crashes. I
> tracked it down to issues with resizing of the BP RCU memory pool.
>
> The problem is in the urcu-bp.c file in the resize_arena() function.
> On successful allocation / remapping the len member of the
> registry_arena struct is never set anywhere function. On the second
> resize of the arena the code in resize_arena() still thinks the
> previous size is equal to the original mapping size. I've fixed this
> issue locally by just adding the following code at the bottom of
> resize_arena().

Good catch !!

However, I think your fix misses one case: if we happen to re-use the
same region, we want to update the length too.

Reported-by: Milosz Tanski <milosz@adfin.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: test_mutex.c uninitialized mutex
Vladimir Nikulichev [Mon, 30 Sep 2013 14:32:22 +0000 (10:32 -0400)] 
Fix: test_mutex.c uninitialized mutex

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoVersion 0.8.0 v0.8.0
Mathieu Desnoyers [Fri, 6 Sep 2013 11:58:28 +0000 (07:58 -0400)] 
Version 0.8.0

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: hash table growth (for small tables) should be limited
Mathieu Desnoyers [Tue, 27 Aug 2013 21:58:22 +0000 (17:58 -0400)] 
Fix: hash table growth (for small tables) should be limited

Buckets with many entries encountered in a hash table could cause it to
grow to a large size, beyond the scope for which this mechanism is
expected to play a role when node accounting is available. Indeed, when
the hash table grows to larger size, split-counter node accounting is
expected to deal with resize/shrink rather than relying on an heuristic
based on the largest bucket size.

This is fixing an issue where we see hash tables sometimes reaching 65k
entries index (65536*8 = 524288 bytes) for a workload limited to adding
1000 entries and then removing all of them, done in a loop (random
keys).

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: doc/examples cross-build
Mathieu Desnoyers [Mon, 26 Aug 2013 15:16:06 +0000 (11:16 -0400)] 
Fix: doc/examples cross-build

We need to pass automake's CPPFLAGS and LDFLAGS to examples. Add CFLAGS
too for completeness.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoIntroduce URCU_INLINE_SMALL_FUNCTIONS
Mathieu Desnoyers [Sun, 11 Aug 2013 20:29:27 +0000 (16:29 -0400)] 
Introduce URCU_INLINE_SMALL_FUNCTIONS

Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoAdd missing tests/common/Makefile.am
Mathieu Desnoyers [Sat, 10 Aug 2013 05:29:29 +0000 (01:29 -0400)] 
Add missing tests/common/Makefile.am

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoREADME: document make check/regtest/bench
Mathieu Desnoyers [Fri, 9 Aug 2013 22:19:08 +0000 (18:19 -0400)] 
README: document make check/regtest/bench

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agotests: split in check, regtest and bench targets
Mathieu Desnoyers [Fri, 9 Aug 2013 22:13:23 +0000 (18:13 -0400)] 
tests: split in check, regtest and bench targets

Allow make check to run without requiring build machines to use hours of
CPU time.

make check: short unit tests
make regtest: long regression tests
make bench: long benchmarks

Fixes #611

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoCleanup: doc/examples makefile
Mathieu Desnoyers [Sun, 21 Jul 2013 21:57:31 +0000 (17:57 -0400)] 
Cleanup: doc/examples makefile

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: doc/examples VPATH build
Mathieu Desnoyers [Sun, 21 Jul 2013 21:34:50 +0000 (17:34 -0400)] 
Fix: doc/examples VPATH build

Fixes #568

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: Move the LIBS after the OBJECTS in the Makefile
Alexandre Montplaisir [Mon, 8 Jul 2013 16:24:30 +0000 (12:24 -0400)] 
doc/examples: Move the LIBS after the OBJECTS in the Makefile

Fixes #569.

Signed-off-by: Alexandre Montplaisir <alexmonthy@voxpopuli.im>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoDocument build work-around on MacOS X
Mathieu Desnoyers [Mon, 8 Jul 2013 16:22:00 +0000 (12:22 -0400)] 
Document build work-around on MacOS X

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix tests: use of uninitialized variables
Mathieu Desnoyers [Thu, 4 Jul 2013 19:23:22 +0000 (15:23 -0400)] 
Fix tests: use of uninitialized variables

Were working fine by luck, since they were allocated at the start of
newly spawned test programs. Identified by Coverity.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agotest_urcu_hash*: initialize TLS seeds
Mathieu Desnoyers [Wed, 26 Jun 2013 21:01:20 +0000 (17:01 -0400)] 
test_urcu_hash*: initialize TLS seeds

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_lfht_for_each_entry_duplicate
Mathieu Desnoyers [Sun, 23 Jun 2013 19:03:45 +0000 (15:03 -0400)] 
doc/examples: cds_lfht_for_each_entry_duplicate

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_lfht_lookup
Mathieu Desnoyers [Sun, 23 Jun 2013 18:56:44 +0000 (14:56 -0400)] 
doc/examples: cds_lfht_lookup

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_lfht_destroy
Mathieu Desnoyers [Sun, 23 Jun 2013 18:38:10 +0000 (14:38 -0400)] 
doc/examples: cds_lfht_destroy

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_lfht_add_replace
Mathieu Desnoyers [Sun, 23 Jun 2013 18:16:45 +0000 (14:16 -0400)] 
doc/examples: cds_lfht_add_replace

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_lfht_add_unique
Mathieu Desnoyers [Sun, 23 Jun 2013 18:01:01 +0000 (14:01 -0400)] 
doc/examples: cds_lfht_add_unique

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_lfht_add/cds_lfht_del
Mathieu Desnoyers [Sun, 23 Jun 2013 17:40:42 +0000 (13:40 -0400)] 
doc/examples: cds_lfht_add/cds_lfht_del

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: add rculfqueue example
Mathieu Desnoyers [Sun, 23 Jun 2013 15:31:33 +0000 (11:31 -0400)] 
doc/examples: add rculfqueue example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: add synchronize_rcu()
Mathieu Desnoyers [Sat, 22 Jun 2013 18:59:35 +0000 (14:59 -0400)] 
doc/examples: add synchronize_rcu()

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: add bp flavor
Mathieu Desnoyers [Sat, 22 Jun 2013 18:49:06 +0000 (14:49 -0400)] 
doc/examples: add bp flavor

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: add dist toplevel makefile
Mathieu Desnoyers [Sat, 22 Jun 2013 18:39:12 +0000 (14:39 -0400)] 
doc/examples: add dist toplevel makefile

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: add membarrier flavor
Mathieu Desnoyers [Sat, 22 Jun 2013 18:18:22 +0000 (14:18 -0400)] 
doc/examples: add membarrier flavor

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: document call_rcu()
Mathieu Desnoyers [Sat, 22 Jun 2013 18:05:10 +0000 (14:05 -0400)] 
doc/examples: document call_rcu()

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: update qsbr example
Mathieu Desnoyers [Sat, 22 Jun 2013 18:02:30 +0000 (14:02 -0400)] 
doc/examples: update qsbr example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agourcu signal: remove assertion on exit
Mathieu Desnoyers [Sat, 22 Jun 2013 18:00:39 +0000 (14:00 -0400)] 
urcu signal: remove assertion on exit

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: signal flavor
Mathieu Desnoyers [Sat, 22 Jun 2013 17:34:53 +0000 (13:34 -0400)] 
doc/examples: signal flavor

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: add mb flavor
Mathieu Desnoyers [Sat, 22 Jun 2013 17:26:16 +0000 (13:26 -0400)] 
doc/examples: add mb flavor

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: update qsbr
Mathieu Desnoyers [Sat, 22 Jun 2013 17:25:03 +0000 (13:25 -0400)] 
doc/examples: update qsbr

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: introduce urcu-flavors examples directory
Mathieu Desnoyers [Sat, 22 Jun 2013 17:19:36 +0000 (13:19 -0400)] 
doc/examples: introduce urcu-flavors examples directory

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: enhance rcu-flavor-qsbr example
Mathieu Desnoyers [Sat, 22 Jun 2013 17:13:22 +0000 (13:13 -0400)] 
doc/examples: enhance rcu-flavor-qsbr example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: rename qsbr-minimal to rcu-flavor-qsbr
Mathieu Desnoyers [Sat, 22 Jun 2013 17:02:46 +0000 (13:02 -0400)] 
doc/examples: rename qsbr-minimal to rcu-flavor-qsbr

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: automake stop on error
Mathieu Desnoyers [Sat, 22 Jun 2013 16:08:31 +0000 (12:08 -0400)] 
doc/examples: automake stop on error

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: hlist
Mathieu Desnoyers [Sat, 22 Jun 2013 16:06:00 +0000 (12:06 -0400)] 
doc/examples: hlist

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agohlist/rcuhlist update
Mathieu Desnoyers [Sat, 22 Jun 2013 15:50:44 +0000 (11:50 -0400)] 
hlist/rcuhlist update

introduce:

- cds_hlist_for_each()
- cds_hlist_for_each_safe()
- CDS_HLIST_HEAD()
- CDS_HLIST_HEAD_INIT()
- cds_hlist_for_each_entry_2() (takes 3 argument, like the Linux kernel
  API),
- cds_hlist_for_each_entry_safe_2() (takes 4 arguments, like the Linux
  kernel API),
- cds_hlist_for_each_rcu()
- cds_hlist_for_each_entry_rcu_2() (takes 3 arguments, like the Linux
  kernel API).

Left cds_hlist_for_each_entry(), cds_hlist_for_each_entry_safe() and
cds_hlist_for_each_entry_rcu() as-is (different from the ones found in
the Linux kernel) because those APIs were already exposed by Userspace
RCU.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: fix typo in list example
Mathieu Desnoyers [Fri, 21 Jun 2013 21:45:48 +0000 (17:45 -0400)] 
doc/examples: fix typo in list example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agorcuhlist: make pointer stores atomic
Mathieu Desnoyers [Fri, 21 Jun 2013 21:29:42 +0000 (17:29 -0400)] 
rcuhlist: make pointer stores atomic

Use rcu_assign_pointer() to store into head->next in
cds_hlist_add_head_rcu(). This includes the write barrier needed before
publishing the new node.

Use CMM_STORE_SHARED() to store into elem->prev->next in
cds_hlist_del_rcu().

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agohlist, rcuhlist: cleanup coding style
Mathieu Desnoyers [Fri, 21 Jun 2013 21:26:09 +0000 (17:26 -0400)] 
hlist, rcuhlist: cleanup coding style

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: lfstack
Mathieu Desnoyers [Fri, 21 Jun 2013 21:17:32 +0000 (17:17 -0400)] 
doc/examples: lfstack

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: update cds_wfs_pop_all_blocking
Mathieu Desnoyers [Fri, 21 Jun 2013 20:51:14 +0000 (16:51 -0400)] 
doc/examples: update cds_wfs_pop_all_blocking

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_wfs_pop_all_blocking
Mathieu Desnoyers [Fri, 21 Jun 2013 20:46:04 +0000 (16:46 -0400)] 
doc/examples: cds_wfs_pop_all_blocking

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_wfs_pop
Mathieu Desnoyers [Fri, 21 Jun 2013 20:31:49 +0000 (16:31 -0400)] 
doc/examples: cds_wfs_pop

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: add missing Makefile
Mathieu Desnoyers [Fri, 21 Jun 2013 20:21:34 +0000 (16:21 -0400)] 
doc/examples: add missing Makefile

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_wfs_push
Mathieu Desnoyers [Fri, 21 Jun 2013 20:19:01 +0000 (16:19 -0400)] 
doc/examples: cds_wfs_push

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/wfcqueue: cds_wfcq_splice
Mathieu Desnoyers [Fri, 21 Jun 2013 19:57:35 +0000 (15:57 -0400)] 
doc/wfcqueue: cds_wfcq_splice

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: add cds_wfcq_dequeue
Mathieu Desnoyers [Fri, 21 Jun 2013 19:32:14 +0000 (15:32 -0400)] 
doc/examples: add cds_wfcq_dequeue

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: wfcq needs to link against urcu-common
Mathieu Desnoyers [Fri, 21 Jun 2013 19:31:44 +0000 (15:31 -0400)] 
doc/examples: wfcq needs to link against urcu-common

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: update queue comment
Mathieu Desnoyers [Fri, 21 Jun 2013 19:03:48 +0000 (15:03 -0400)] 
doc/examples: update queue comment

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: fix make clean
Mathieu Desnoyers [Fri, 21 Jun 2013 18:56:44 +0000 (14:56 -0400)] 
doc/examples: fix make clean

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agogitignore: add qsbr-minimal
Mathieu Desnoyers [Fri, 21 Jun 2013 18:54:42 +0000 (14:54 -0400)] 
gitignore: add qsbr-minimal

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_wfcq_enqueue
Mathieu Desnoyers [Fri, 21 Jun 2013 18:53:52 +0000 (14:53 -0400)] 
doc/examples: cds_wfcq_enqueue

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: Move LIBS to each makefile
Mathieu Desnoyers [Fri, 21 Jun 2013 18:52:44 +0000 (14:52 -0400)] 
doc/examples: Move LIBS to each makefile

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_list_for_each_rcu
Mathieu Desnoyers [Fri, 21 Jun 2013 18:22:32 +0000 (14:22 -0400)] 
doc/examples: cds_list_for_each_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_list_for_each_entry_rcu
Mathieu Desnoyers [Fri, 21 Jun 2013 18:12:54 +0000 (14:12 -0400)] 
doc/examples: cds_list_for_each_entry_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_list_replace_rcu
Mathieu Desnoyers [Fri, 21 Jun 2013 17:59:09 +0000 (13:59 -0400)] 
doc/examples: cds_list_replace_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_list_add_tail_rcu
Mathieu Desnoyers [Fri, 21 Jun 2013 17:43:43 +0000 (13:43 -0400)] 
doc/examples: cds_list_add_tail_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_list_del_rcu
Mathieu Desnoyers [Fri, 21 Jun 2013 16:01:06 +0000 (12:01 -0400)] 
doc/examples: cds_list_del_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodoc/examples: cds_list_add_rcu
Mathieu Desnoyers [Fri, 21 Jun 2013 15:46:17 +0000 (11:46 -0400)] 
doc/examples: cds_list_add_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agorculist: ensure atomic updates of next pointers
Mathieu Desnoyers [Fri, 21 Jun 2013 16:51:13 +0000 (12:51 -0400)] 
rculist: ensure atomic updates of next pointers

in cds_list_add_rcu, use rcu_assign_pointer to update head->next
atomically and provide the memory barrier before publishing head->next.
Notice that we don't need the wmb() prior to store to prev, because RCU
traversals only go forward, and thus only use "next".

in cds_list_del_rcu, use CMM_STORE_SHARED() to store to elem->prev->next
atomically.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agorculist: implement cds_list_add_tail_rcu
Mathieu Desnoyers [Fri, 21 Jun 2013 16:48:18 +0000 (12:48 -0400)] 
rculist: implement cds_list_add_tail_rcu

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agorculist.h and list.h style cleanup
Mathieu Desnoyers [Fri, 21 Jun 2013 16:27:44 +0000 (12:27 -0400)] 
rculist.h and list.h style cleanup

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoexample makefile: add missing cd ..
Mathieu Desnoyers [Thu, 20 Jun 2013 13:21:03 +0000 (09:21 -0400)] 
example makefile: add missing cd ..

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoUpdate gitignore
Mathieu Desnoyers [Thu, 20 Jun 2013 12:57:57 +0000 (08:57 -0400)] 
Update gitignore

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: examples Makefile on FreeBSD
Mathieu Desnoyers [Thu, 20 Jun 2013 12:50:33 +0000 (08:50 -0400)] 
Fix: examples Makefile on FreeBSD

Support BSD-style make.

Fixes #565

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agohash table test: don't redefine CACHE_LINE_SIZE
Mathieu Desnoyers [Wed, 19 Jun 2013 14:31:57 +0000 (10:31 -0400)] 
hash table test: don't redefine CACHE_LINE_SIZE

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agotests: use thread-id.h wrapper
Mathieu Desnoyers [Wed, 19 Jun 2013 14:31:09 +0000 (10:31 -0400)] 
tests: use thread-id.h wrapper

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoImplement thread-id.h wrapper
Mathieu Desnoyers [Wed, 19 Jun 2013 14:30:58 +0000 (10:30 -0400)] 
Implement thread-id.h wrapper

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agotests: add missing unsigned long casts to pthread_self()
Mathieu Desnoyers [Wed, 19 Jun 2013 13:04:55 +0000 (09:04 -0400)] 
tests: add missing unsigned long casts to pthread_self()

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: don't build examples in static builds
Mathieu Desnoyers [Fri, 14 Jun 2013 22:32:04 +0000 (18:32 -0400)] 
Fix: don't build examples in static builds

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoAdd QSBR minimal example
Mathieu Desnoyers [Mon, 3 Jun 2013 20:18:34 +0000 (16:18 -0400)] 
Add QSBR minimal example

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agocompiler.h: implement CAA_ARRAY_SIZE()
Mathieu Desnoyers [Fri, 14 Jun 2013 21:28:07 +0000 (17:28 -0400)] 
compiler.h: implement CAA_ARRAY_SIZE()

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agodocument rcu barrier
Mathieu Desnoyers [Fri, 14 Jun 2013 12:30:21 +0000 (08:30 -0400)] 
document rcu barrier

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agorcu barrier: handle OOM die urcu_die
Mathieu Desnoyers [Fri, 14 Jun 2013 12:22:50 +0000 (08:22 -0400)] 
rcu barrier: handle OOM die urcu_die

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoImplement rcu_barrier()
Mathieu Desnoyers [Fri, 31 May 2013 15:32:16 +0000 (11:32 -0400)] 
Implement rcu_barrier()

Awaits for all in-flight call_rcu handlers to complete execution before
returning.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agorculfhash: document destroy context limitations
Mathieu Desnoyers [Fri, 14 Jun 2013 12:15:21 +0000 (08:15 -0400)] 
rculfhash: document destroy context limitations

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoAdd MIPS to README
Mathieu Desnoyers [Tue, 11 Jun 2013 15:18:52 +0000 (11:18 -0400)] 
Add MIPS to README

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoUpdate README
Mathieu Desnoyers [Sat, 8 Jun 2013 12:24:23 +0000 (08:24 -0400)] 
Update README

Detail i386/amd64 as tested for FreeBSD.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoUpdate README testing info about FreeBSD
Mathieu Desnoyers [Sat, 8 Jun 2013 11:57:57 +0000 (07:57 -0400)] 
Update README testing info about FreeBSD

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agotest: fix api.h missing if brackets
Mathieu Desnoyers [Wed, 15 May 2013 14:29:44 +0000 (16:29 +0200)] 
test: fix api.h missing if brackets

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agotests: fix incorrect counter
Mathieu Desnoyers [Wed, 15 May 2013 14:26:57 +0000 (16:26 +0200)] 
tests: fix incorrect counter

- not initialized to 0 in 2 tests,
- within inner loop in 3rd test, thus always 0.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: membarrier fallback symbol conflict
Mathieu Desnoyers [Wed, 8 May 2013 13:53:45 +0000 (09:53 -0400)] 
Fix: membarrier fallback symbol conflict

* Lai Jiangshan (laijs@cn.fujitsu.com) wrote:
> Hi, Mathieu,
>
> There is a big compatible problem in URCU which should be fix in next round.
>
> LB: liburcu built on the system which has sys_membarrier().
> LU: liburcu built on the system which does NOT have sys_membarrier().
>
> LBM: liburcu-mb ....
> LUM: liburcu-mb ...
>
> AB: application(-lliburcu) built on the system which has sys_membarrier().
> AU: application(-lliburcu) built on the system which does NOT have
> sys_membarrier().
>
> ABM application(-lliburcu-mb) ...
> AUM application(-lliburcu-mb) ...
>
> AB/AU + LB/LU: 4 combinations
> ABM/AUM + LBM/LUM: 4 combinations
>
> I remember some of the 8 combinations can't works due to symbols are
> miss match.  only LU+AB and LB+AU ?
>
> could you check it?
>
> How to fix it: In LU and AU, keep all the symbol name/ABI as LA and
> AB, but only the behaviors falls back to URCU_MB.

Define membarrier() as -ENOSYS when SYS_membarrier is not found in the
system headers. Check dynamically for membarrier availability to ensure
ABI compatibility between applications and librairies.

Reported-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: Use a filled signal mask to disable all signals
Mathieu Desnoyers [Fri, 10 May 2013 11:30:18 +0000 (07:30 -0400)] 
Fix: Use a filled signal mask to disable all signals

Changelog from David Pelton's original patch:

While using lttng-ust with an application that was calling fork()
with pending signals, I found that all signals were getting unmasked
shortly before the underlying call to fork().  After some
investigation, I found that the rcu_bp_before_fork() function was
unmasking all signals.  Based on the comments for this function, it
should be masking all signals.  Inspection of the rest of the code
in urcu-bp.c revealed the same pattern in two other functions.

This patch changes the code to use a filled signal mask to disable
all signals.  The change to rcu_bp_before_fork() addressed the
problem I was seeing while using lttng-ust.  The changes to the
other two functions appear to fix other instances of the same
problem.

Updates by Mathieu Desnoyers:

- Use SIG_BLOCK instead of SIG_SETMASK when setting a filled mask. This
  has the same behavior in this case (since we're blocking all signals),
  but is semantically neater: if we ever some signals from that mask,
  we'd like to to a union with the signal mask already blocked by the
  application.
- Also fix incorrect signal masking in compat_arch_x86.c.

Reported-by: David Pelton <dpelton@ciena.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agourcu-bp: introduce struct urcu_gp
Mathieu Desnoyers [Mon, 6 May 2013 14:30:57 +0000 (10:30 -0400)] 
urcu-bp: introduce struct urcu_gp

Make urcu-bp similar to urcu-qsbr and other urcu flavors.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoFix: struct urcu_gp broke multiflavor
Mathieu Desnoyers [Mon, 6 May 2013 14:24:14 +0000 (10:24 -0400)] 
Fix: struct urcu_gp broke multiflavor

Add mapping to namespace urcu_gp.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoCleanup test usage printout
Mathieu Desnoyers [Mon, 6 May 2013 14:03:55 +0000 (10:03 -0400)] 
Cleanup test usage printout

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agowfstack tests: use pop "last" state info
Mathieu Desnoyers [Mon, 6 May 2013 13:35:42 +0000 (09:35 -0400)] 
wfstack tests: use pop "last" state info

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agowfstack: return whether pop is popping the last element
Mathieu Desnoyers [Mon, 6 May 2013 13:35:07 +0000 (09:35 -0400)] 
wfstack: return whether pop is popping the last element

Newly introduced "with_state" pop API members return stack state
atomically sampled with the pop operation.

Allow testing behavior of pop with respect to number of push-to-empty
and pop-all-from-non-empty.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agowfcqueue tests: use dequeue empty state
Mathieu Desnoyers [Mon, 6 May 2013 13:34:00 +0000 (09:34 -0400)] 
wfcqueue tests: use dequeue empty state

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agowfcqueue: return whether dequeue is dequeuing last element
Mathieu Desnoyers [Mon, 6 May 2013 13:33:36 +0000 (09:33 -0400)] 
wfcqueue: return whether dequeue is dequeuing last element

Newly introduced "with_state" dequeue API members return queue state
atomically sampled with the dequeue operation.

Allow testing behavior of dequeue with respect to number of
enqueue-to-empty and splice-from-non-empty.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agourcu: avoid false sharing for rcu_gp_ctr
Lai Jiangshan [Mon, 6 May 2013 12:42:27 +0000 (08:42 -0400)] 
urcu: avoid false sharing for rcu_gp_ctr

@rcu_gp_ctr and @registry share the same cache line, it causes
false sharing and slowdown both of the read site and update site.

Fix: Use different cache line for them.

Although rcu_gp_futex is updated less than rcu_gp_ctr, but
they always be accessed at almost the same time, so we also move rcu_gp_futex
to the cacheline of rcu_gp_ctr to reduce the cacheline-usage or cache-missing
of read site.

test: (4X6=24 CPUs)

Before patch:

[root@localhost userspace-rcu]# ./tests/test_urcu_mb 20 1 20
SUMMARY ./tests/test_urcu_mb      testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads   2100285330 nr_writes      3390219 nr_ops   2103675549
[root@localhost userspace-rcu]# ./tests/test_urcu_mb 20 1 20
SUMMARY ./tests/test_urcu_mb      testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads   1619868562 nr_writes      3529478 nr_ops   1623398040
[root@localhost userspace-rcu]# ./tests/test_urcu_mb 20 1 20
SUMMARY ./tests/test_urcu_mb      testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads   1949067038 nr_writes      3469334 nr_ops   1952536372

after patch:

[root@localhost userspace-rcu]# ./tests/test_urcu_mb 20 1 20
SUMMARY ./tests/test_urcu_mb      testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads   3380191848 nr_writes      4903248 nr_ops   3385095096
[root@localhost userspace-rcu]# ./tests/test_urcu_mb 20 1 20
SUMMARY ./tests/test_urcu_mb      testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads   3397637486 nr_writes      4129809 nr_ops   3401767295

Singed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agourcu: make the code of urcu-qsbr as normal urcu
Lai Jiangshan [Mon, 6 May 2013 12:32:02 +0000 (08:32 -0400)] 
urcu: make the code of urcu-qsbr as normal urcu

urcu-qsbr's read site's quiescence is much longer than normal urcu ==>
synchronize_rcu() is much slower ==>
rcu_gp_ctr is updated much less ==>
the whole urcu-qsbr will not be slowed down by false sharing of rcu_gp_ctr.

But this patch makes sense to keep the code of urcu-qsbr like normal urcu,
better readability and maintenance.

Test: (4*6 CPUs)
Before patch:
[root@localhost userspace-rcu]# ./tests/test_urcu_qsbr 20 1 20
SUMMARY ./tests/test_urcu_qsbr    testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads  65498297587 nr_writes      2000665 nr_ops  65500298252
[root@localhost userspace-rcu]# ./tests/test_urcu_qsbr 20 1 20
SUMMARY ./tests/test_urcu_qsbr    testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads  67218079467 nr_writes      1981593 nr_ops  67220061060

After patch
./tests/test_urcu_qsbr 20 1 20
SUMMARY ./tests/test_urcu_qsbr    testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads  67473798999 nr_writes      1999151 nr_ops  67475798150
[root@localhost userspace-rcu]# ./tests/test_urcu_qsbr 20 1 20
SUMMARY ./tests/test_urcu_qsbr    testdur   20 nr_readers  20 rdur      0 wdur      0 nr_writers   1 wdelay      0 nr_reads  67065521397 nr_writes      1993956 nr_ops  67067515353

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agorculfhash: detect if resize/destroy are called within RCU read-side C.S.
Mathieu Desnoyers [Tue, 30 Apr 2013 01:30:17 +0000 (21:30 -0400)] 
rculfhash: detect if resize/destroy are called within RCU read-side C.S.

Report errors.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoDocumentation: rculfhash: cds_lfht_resize not within read-side C.S.
Mathieu Desnoyers [Tue, 30 Apr 2013 00:48:40 +0000 (20:48 -0400)] 
Documentation: rculfhash: cds_lfht_resize not within read-side C.S.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agofix: rculfhash don't change qsbr online state
Mathieu Desnoyers [Tue, 30 Apr 2013 00:28:10 +0000 (20:28 -0400)] 
fix: rculfhash don't change qsbr online state

resize and destroy should not change the QSBR online state. Use the new
rcu_read_ongoing() API for this.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
8 years agoAdd rcu_read_ongoing() API to each urcu flavor
Mathieu Desnoyers [Tue, 30 Apr 2013 00:17:22 +0000 (20:17 -0400)] 
Add rcu_read_ongoing() API to each urcu flavor

This will allow checking whether:

- thread is online (QSBR),
- thread is nested within read-side critical section (other flavors),

This is useful for libraries that need to know if QSBR is online in
order to save the original state temporarily so it can be restored
before returning to the caller.

Eventually, this API can be called by a "debugging" implementation of
rcu_dereference() and other urcu-pointer.h API members to check that no
RCU pointer is read outside of RCU read-side critical sections.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
This page took 0.063067 seconds and 4 git commands to generate.