Mathieu Desnoyers [Tue, 22 Sep 2009 21:09:55 +0000 (17:09 -0400)] 
Add inc/dec x86 atomics
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 22 Sep 2009 20:41:24 +0000 (16:41 -0400)]
update x86 atomic, add test atomic
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 22 Sep 2009 20:28:49 +0000 (16:28 -0400)]
extend x86 atomic operations
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 21 Sep 2009 16:26:16 +0000 (12:26 -0400)]
Add missing files in make clean
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sun, 20 Sep 2009 16:03:37 +0000 (12:03 -0400)]
urcu-defer: remove unnecessary memory barrier
All synchronization between queue producer/consumer is performed by the write to
"head". Before this write, none of the queued data is visible from the consumer
point of view.
Therefore, just a single wmb() is required before writing to head to ensure
correct synchronization. This matches with the rmb() after reading head on the
consumer side.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sun, 20 Sep 2009 15:39:30 +0000 (11:39 -0400)]
urcu-defer: Add fast path for empty queues
defer_barrier does not need to call synchronize_rcu() when all queues are empty.
Skip the G.P. if this is the case.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sun, 20 Sep 2009 15:18:23 +0000 (11:18 -0400)]
urcu-defer: cleanup debug code
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sun, 20 Sep 2009 02:23:11 +0000 (22:23 -0400)]
Deferral test update
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sun, 20 Sep 2009 02:20:16 +0000 (22:20 -0400)]
generic urcu deferral (call_rcu())
Found out a way to encode the queues so the standard scenario is to use a single
pointer per call_rcu().
Uses more space for:
- unaligned functions pointers.
- unaligned data pointers.
- function/data value : -2L. (this is arbitrary)
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sun, 20 Sep 2009 00:05:10 +0000 (20:05 -0400)]
Rename liburcu-reclaim to liburcu-defer
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 19 Sep 2009 21:55:47 +0000 (17:55 -0400)]
urcu-reclaim cleanup
Rename "reader" to "reclaimer", because urcu-reclaim keeps track of "reclaimer"
threads (rcu writers) and has its own reclaiming thread periodically performing
the batch reclamation.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 19 Sep 2009 21:45:41 +0000 (17:45 -0400)]
Update header heading
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 19 Sep 2009 21:28:06 +0000 (17:28 -0400)]
Add rcu-reclaim.so library
Add "automated" RCU memory reclamation.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 19 Sep 2009 21:26:41 +0000 (17:26 -0400)]
add static declarations
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 18 Sep 2009 13:12:35 +0000 (09:12 -0400)]
qsbr: Add write+read thread support to 32-bit QSBR
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 18 Sep 2009 13:05:03 +0000 (09:05 -0400)]
QSBR: Implement 2-phase grace period for 32-bit arch
Ensures we never run in overflow on 32-bit arch.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 18 Sep 2009 12:59:21 +0000 (08:59 -0400)]
Permit both 32 and 64-bit builds
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 18 Sep 2009 12:47:55 +0000 (08:47 -0400)]
Default to architecture size, add Makefile32
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 18 Sep 2009 12:46:24 +0000 (08:46 -0400)]
Update makefile to compiler for 32-bit architectures on x86_64
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 18 Sep 2009 12:32:23 +0000 (08:32 -0400)]
Rename define
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 18 Sep 2009 12:30:03 +0000 (08:30 -0400)]
Fix urcu.c comment
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 18 Sep 2009 12:20:22 +0000 (08:20 -0400)]
qsbr: use defines to clarify code.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 17 Sep 2009 15:11:42 +0000 (11:11 -0400)]
qsbr: portability fixes: use unsigned long for the gp counters.
unsigned wraps properly in C (modulo ULONG_MAX + 1), IOW ULONG_MAX + 2
_is_ defined and is 1.
Rewrite rcu_gp_ongoing test to work in the unsigned case:
a - b < 0
becomes
a - b > ULONG_MAX / 2
This test actually means a - b > 0x7f....f which is exactly what we want
to test for in the first place, but in a portable way.
Signed-off-by: Pierre Habouzit <madcoder@debian.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 17 Sep 2009 12:38:06 +0000 (08:38 -0400)]
qsbr: micro optimization of the gp use.
Follow-up on 1a80f3: since now `0' for a gp counter means "offline", just
be sure that the global GP counter is never zero. To achieve that, let its
first value ever be one, and increment it 2 by 2 like previously done.
Note that it internaly uses the fact that signed integers wraps (which the
previous code already assumed anyways) but this is undefined behaviour in
C. Using unsigned longs would probably be more portable.
This way, setting the cpu online, marking a quiescent state is just a
matter of copying the global gp counter instead of incrementing it by one.
This saves one CPU instruction, and supposedly even register use for
architectures providing memory to memory copy operands.
Signed-off-by: Pierre Habouzit <madcoder@debian.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Pierre Habouzit [Thu, 17 Sep 2009 12:30:19 +0000 (08:30 -0400)] 
qsbr: save two full smp_mb() when the writer is also a reader.
Now that you've put the barriers before taking lock, and after releasing
it, one can expand the _rcu_thread_{on,off}line call, and optimize two
barrier calls away. (this is patch 1/2).
Signed-off-by: Pierre Habouzit <madcoder@debian.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Wed, 16 Sep 2009 16:58:58 +0000 (12:58 -0400)]
rename QSBR force_mb_all_threads to smp_mb
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Wed, 16 Sep 2009 16:44:27 +0000 (12:44 -0400)]
Move mb() outside of the synchronize C.S.
Will be useful for future Q.S. merging. Does not change the RCU g.p. guarantee.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Wed, 16 Sep 2009 16:40:27 +0000 (12:40 -0400)]
Ensure LOAD_SHARED/STORE_SHARED semantic is used in QSBR RCU
Ensures all volatile accesses are indeed volatile.
On architectures without coherent caches, will perform the proper cache flushes
when needed.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Wed, 16 Sep 2009 13:39:20 +0000 (09:39 -0400)]
QSBR offline thread micro-optimization
> That's an interesting paper for sure. I had one micro-optimization in
> mind when I read the paper for QSBR, you can decide that making a CPU
> offline can be done by writing 0 to the per thread counter, instead of
> reading the global counter. It saves a load from a shared variable,
> which probably helps in its tiny way.
>
In wait_for_quiescent_state, the test for Q.S. is :
while (rcu_gp_ongoing(index->rcu_reader_qs_gp) &&
(*index->rcu_reader_qs_gp - urcu_gp_ctr < 0))
cpu_relax();
where :
static inline int rcu_gp_ongoing(long *value)
{
if (value == NULL)
return 0;
return LOAD_SHARED(*value) & 1;
}
Your proposal would work for the rcu_gp_ongoing test, as it only checks
for the parity. Given this test is enough to guarantee that we skip the
reader thread, then yes, it seems to work just fine.
Proposed-by: Pierre Habouzit <madcoder@debian.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Wed, 16 Sep 2009 13:30:22 +0000 (09:30 -0400)]
Expose nop QSBR read lock to non-GPL/LGPL programs
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 23:46:07 +0000 (19:46 -0400)]
Update README (again)
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 23:40:56 +0000 (19:40 -0400)]
update README
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 23:38:36 +0000 (19:38 -0400)]
Rename CONFIG_URCU_AVOID_SIGNALS to URCU_MB, and README updates
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 22:09:36 +0000 (18:09 -0400)]
Revert "qsbr urcu: make it safe to call rcu_sychronize from a registered thread."
This reverts commit
d1238495fdaf1c85447690231bf5d1838b83bf88.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 20:39:35 +0000 (16:39 -0400)]
qsbr urcu: make it safe to call rcu_sychronize from a registered thread
Actually, the proper way to deal with threads registered as readers but also
calling synchronize_rcu() is to mark them as offline while the synchronize is in
progress.
Else, since calling rcu_sychronize will increment urcu_gp_ctr, we will be
stupidly waiting for ourselves and deadlock.
Original-patch-from: Pierre Habouzit <madcoder@debian.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Pierre Habouzit [Tue, 15 Sep 2009 20:17:22 +0000 (16:17 -0400)]
qsbr urcu: make it safe to call rcu_sychronize from a registered thread.
Else, since calling rcu_sychronize will increment urcu_gp_ctr, we will be
stupidly waiting for ourselves and deadlock.
No need to use _rcu_quiescent_state() since we are currently owning the
internal URCU lock, so the memory barriers and ACCESS_ONCE() are overkill.
Signed-off-by: Pierre Habouzit <madcoder@debian.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Pierre Habouzit [Tue, 15 Sep 2009 20:15:27 +0000 (16:15 -0400)]
qsbr urcu: simplifications and fixes.
need_mb is unused in qsbr, get rid of it.
add missing static's
Signed-off-by: Pierre Habouzit <madcoder@debian.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 19:51:34 +0000 (15:51 -0400)]
Build urcu-qsbr.so library, update README
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 18:36:55 +0000 (14:36 -0400)]
urcu qsbr: add DEBUG_RCU self check
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 18:24:34 +0000 (14:24 -0400)]
update README
- Add comment about new threads and signals
- document DEBUG_RCU
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 15 Sep 2009 18:13:09 +0000 (14:13 -0400)]
urcu qsbr: move thread online after add reader
For a matter of symmetry with the unregister, make thread online only after it
has been added as reader.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Pierre Habouzit [Tue, 15 Sep 2009 18:09:59 +0000 (14:09 -0400)]
qsbr rcu: rc_register_thread also make the thread online.
Else, all its RCU reader critical sections aren't safe until its first
_rcu_quiescent_state() call.
Signed-off-by: Pierre Habouzit <madcoder@debian.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 5 Sep 2009 23:52:26 +0000 (19:52 -0400)]
update rcu_dereference comment
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 5 Sep 2009 16:35:46 +0000 (12:35 -0400)]
Document access ordering of rcu_dereference wrt VSS
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sun, 30 Aug 2009 14:50:36 +0000 (10:50 -0400)]
add global include Makefile
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sun, 30 Aug 2009 14:50:04 +0000 (10:50 -0400)]
Move test programs to tests/ subdir
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Pierre-Marc Fournier [Tue, 21 Jul 2009 22:09:47 +0000 (18:09 -0400)] 
urcu.c: declare noop urcu_init() function non-static
The RCU library is sometimes used in constructor context. In these cases,
because it is impossible to predict whether the liburcu constructor has already
been called, an explicit call to the urcu_init function must be made.
Declaring the noop version of the urcu_init() function as static breaks code
that calls urcu_init() explicitly when CONFIG_AVOID_SIGNALS is true. To prevent
this, this patch makes the declaration non-static.
From: Pierre-Marc Fournier <pierre-marc.fournier@polymtl.ca>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Wed, 8 Jul 2009 04:13:13 +0000 (00:13 -0400)]
Move DEBUG_FULL_MB to properly named CONFIG_URCU_AVOID_SIGNALS
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Pierre-Marc Fournier [Tue, 7 Jul 2009 22:14:19 +0000 (18:14 -0400)]
Add SA_RESTART flag to signal handler
With this flag, the kernel restarts automatically some system calls
that were interrupted by the signal. Without this flag, these system
calls would fail with errno = EINTR.
Even with this flag, the kernel cannot restart some system calls, and
developers should verify whether these system calls failed with EINTR
and restart them manually in that case. For the list of system calls
that cannot be restarted automatically, see signal(7).
Because of this, we cannot completely eliminate the side-effects of
liburcu on the application, but I think we should try our best and
set SA_RESTART.
From: Pierre-Marc Fournier <pierre-marc.fournier@polymtl.ca>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 29 Jun 2009 14:59:00 +0000 (10:59 -0400)]
update license
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 29 Jun 2009 14:57:19 +0000 (10:57 -0400)]
Update license files and readme
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Jan Blunck [Mon, 29 Jun 2009 09:46:43 +0000 (11:46 +0200)] 
Re-license arch_atomic_s390.h and arch_s390.h
Mathieu Desnoyers asked for re-license the low-level architecture specific
bits under MIT style license.
Signed-off-by: Jan Blunck <jblunck@suse.de>
Jan Blunck [Wed, 17 Jun 2009 11:49:44 +0000 (13:49 +0200)]
s390: Fix some compiler warnings
Fix warning about _atomic_exchange being defined but unused and explicitly
cast the 2nd argument.
Signed-off-by: Jan Blunck <jblunck@suse.de>
Jan Blunck [Wed, 17 Jun 2009 11:40:38 +0000 (13:40 +0200)]
Add S390 support
This patch adds support for the S390 architecture.
Signed-off-by: Jan Blunck <jblunck@suse.de>
Paul E. McKenney [Tue, 16 Jun 2009 18:26:36 +0000 (11:26 -0700)] 
URCU's api_*.h
On Tue, Jun 16, 2009 at 07:09:11PM +0200, Jan Blunck wrote:
> I already did most of that for libkcompat. That was actually why I was
> asking if I should port that over.
Nevertheless, I offer the following patch to userspace-rcu to provide
a gcc-library version of the api_*.h files and to clean up the build a
bit.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Jan Blunck [Tue, 16 Jun 2009 09:22:13 +0000 (11:22 +0200)]
api.h is only needed by urcutorture
Signed-off-by: Jan Blunck <jblunck@suse.de>
Jan Blunck [Tue, 16 Jun 2009 09:20:55 +0000 (11:20 +0200)]
Add S390 architecture specific headers
This implements atomic exchange and some other trivial operations like
barriers for the S390 architecture. All of this is based on the
"z/Architecture Principles of Operation" as released by IBM.
Note that api_s390.h is missing so urcutortures applications don't build.
Signed-off-by: Jan Blunck <jblunck@suse.de>
Mathieu Desnoyers [Fri, 26 Jun 2009 15:18:07 +0000 (11:18 -0400)]
Add runpaul-phase6.sh
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 26 Jun 2009 15:17:18 +0000 (11:17 -0400)]
Fix phase5 test log collection
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Tue, 23 Jun 2009 15:25:28 +0000 (11:25 -0400)]
Add local reclaim tests
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 22 Jun 2009 18:35:58 +0000 (14:35 -0400)]
Update phase3 batch size
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 22 Jun 2009 18:33:59 +0000 (14:33 -0400)]
Add phase 5 test, educated guess for batch size 32768
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 22 Jun 2009 14:55:39 +0000 (10:55 -0400)]
Update default batch size to 32768, after removal of mutex
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 22 Jun 2009 14:50:53 +0000 (10:50 -0400)]
Remote mutex from RCU write-side
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 22 Jun 2009 01:33:38 +0000 (21:33 -0400)]
Protect shared affinity table with mutex
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Mon, 22 Jun 2009 01:32:05 +0000 (21:32 -0400)]
Deal with kernel affinity bug for tests
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 20 Jun 2009 15:55:38 +0000 (11:55 -0400)]
Test qsbr debug
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 20 Jun 2009 15:32:01 +0000 (11:32 -0400)]
Add verbose debug to test_qsbr
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Pierre-Marc Fournier [Fri, 19 Jun 2009 20:51:23 +0000 (16:51 -0400)]
urcu: add quick start guide to README
Signed-off-by: Pierre-Marc Fournier <pierre-marc.fournier@polymtl.ca>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Fri, 19 Jun 2009 14:05:04 +0000 (10:05 -0400)]
Add test phase 4 for striding
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 20:59:35 +0000 (16:59 -0400)]
Add verif results
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 20:26:17 +0000 (16:26 -0400)]
Add nosched model results (for signal readers)
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 20:21:18 +0000 (16:21 -0400)]
Cleanup verif directory.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 18:11:00 +0000 (14:11 -0400)]
Add phase 3 : scalability run
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 16:47:44 +0000 (12:47 -0400)]
Put back 4096 for runall.sh batch size
Basically, when we get to 16384, it's the error due to non-handled removals that
makes this value "look" faster.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 16:42:06 +0000 (12:42 -0400)]
update runtests.sh
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 16:40:58 +0000 (12:40 -0400)]
Update runall.sh default batch size
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 16:33:47 +0000 (12:33 -0400)]
Make batch test shorter
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 16:25:41 +0000 (12:25 -0400)]
Add 2 phases test
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 16:20:37 +0000 (12:20 -0400)]
Add "runpaul.sh" test script
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 15:52:10 +0000 (11:52 -0400)]
Fix makefile
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 15:33:06 +0000 (11:33 -0400)]
Add more batch sizes to tests, just in case
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 15:09:51 +0000 (11:09 -0400)]
Update default value for batch size
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 15:01:18 +0000 (11:01 -0400)]
Fix gc tests
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 13:36:35 +0000 (09:36 -0400)]
Update summary
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 13:33:51 +0000 (09:33 -0400)]
Update makefile64
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 13:33:12 +0000 (09:33 -0400)]
Update makefile
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Thu, 18 Jun 2009 13:32:46 +0000 (09:32 -0400)]
Add GC tests
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Wed, 17 Jun 2009 16:31:58 +0000 (12:31 -0400)]
Merge branch 'master' of lttng.org:/home/git/userspace-rcu
Mathieu Desnoyers [Wed, 17 Jun 2009 16:31:36 +0000 (12:31 -0400)]
Fix typo in license file
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
compudj [Tue, 16 Jun 2009 21:40:28 +0000 (17:40 -0400)] 
Add cpu time
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
compudj [Tue, 16 Jun 2009 21:29:25 +0000 (17:29 -0400)]
Add output to runall.log
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Paul E. McKenney [Sat, 13 Jun 2009 23:53:05 +0000 (19:53 -0400)]
Fix test_looplen, use double-precition arithmetic
Here you go!!!
[root@perfsqh userspace-rcu]# ./test_looplen
CALIBRATION : 0 cycles per loop
Hmmm... Probably not what you had in mind. Breaking out the
floating-point arithmetic:
[root@perfsqh userspace-rcu]# ./test_looplen
CALIBRATION : 0.386172 cycles per loop
time_tot =
4049312, LOOPS =
1048576, TESTS = 10
More helpful? ;-)
Thanx, Paul
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 13 Jun 2009 23:18:38 +0000 (19:18 -0400)]
Other timing tests, update cache line size
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 13 Jun 2009 23:10:31 +0000 (19:10 -0400)]
Deal with POWER5+ 256B L3 cachefalse sharing for per thread lock
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 13 Jun 2009 21:52:43 +0000 (17:52 -0400)]
Add looplen cycle test
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Paul E. McKenney [Sat, 13 Jun 2009 21:02:36 +0000 (17:02 -0400)]
Fix affinity in test script
You need to change the script!!! See patch below. All prior results
with more than 32 CPUs are bogus because the -a arguments confined the
threads to only 32 CPUs (the even-numbered ones from zero to 62).
Allow 64 CPUs. Really needs to do the calculation based on the
stride and the number of CPUs, but a quick hack for now.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Mathieu Desnoyers [Sat, 13 Jun 2009 19:00:03 +0000 (15:00 -0400)]
Added a few points to the readercs test
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 13 Jun 2009 18:51:02 +0000 (14:51 -0400)]
Update update fraction test
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Mathieu Desnoyers [Sat, 13 Jun 2009 18:40:53 +0000 (14:40 -0400)]
Update readcslen test to use NUM_CPUS readers
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
This page took 0.040301 seconds and 4 git commands to generate.
projects / userspace-rcu.git / log
