10c3b22e48
Merge remote-tracking branch 'upstream/master'
...
Fix regressions for linux systems.
12 years ago
aef827b5a3
regressions: Remove unnecessary include.
12 years ago
81b9313e10
build: Fix regression introduced by MingW32 port.
12 years ago
8f1c31aa2c
Style changes to common.h's alarm macros.
12 years ago
3e469d7fb5
Namespace changes to regressions/common.h (alarm).
...
Add "local_name" to COMMON_ALARM_DECLARE_LOCAL and
COMMON_ALARM_DECLARE_INIT to avoid namespace conflicts.
12 years ago
9c298fc78b
whitespace: Remove unnecessary newline.
12 years ago
1298a6c335
Add support for MinGW32.
...
This primarily involved changing the configure script and adding
several utility functions to regressions/common.h for unit testing.
Signed-off-by: Samy Al Bahra <sbahra@appnexus.com>
12 years ago
9152cc7cea
regressions/ck_hp_fifo_donner: This was submitted from a Windows editor.
...
Switch to using UNIX EOL.
12 years ago
f9a484a10d
build: Disclaimer that regressions are unsupported for out-of-source builds.
12 years ago
cff7b1b21f
regressions/ck_queue: Add coverage for previous regression.
12 years ago
e517126247
regressions: Enable rdtsc for x86-32 targets.
12 years ago
72855320c4
regressions/ck_hs: Do not rely on ck_pr_add_64 for benchmarks.
...
ck_hs now officially targets the 32-bit ports. No advantage
to using ck_pr_*_64 operations here.
12 years ago
5d1fff2e06
regressions/ck_pr: Place double tests under feature flag.
12 years ago
33a9222923
legal: Update Copyright statements.
12 years ago
ca6637a57c
regressions/ck_pr: Utilize _GNU_SOURCE for ck_pr_store unit test.
12 years ago
3293cde2d2
regressions/common: Add missing header files in common.h
12 years ago
cb650246e2
regressions/ck_hp: Do not use empty loop body.
...
Fixes some GCC warnings in 4.7+.
12 years ago
0866206c1f
regressions/ck_ring: Add ck_ring_spmc to fast path test.
13 years ago
fbdc512304
regressions/ck_ring: Use variable for ring size.
13 years ago
0631e396e0
ck_backoff: Replace ck_backoff_eb with ck_backoff_gb.
...
The distinction between additive/exponential implementation
and geometric implementation does little but confuse users.
The terminology used in ck_backoff now reflects terminology
used in literature.
ck_backoff_gb has been removed.
13 years ago
d55e588ee4
regressions/ck_ring: Fix use of uninitialized data.
...
Execution history exists such that first thread will dequeue
from an uninitialized ring buffer. The unit test has been
(un)fortunate in that it would busy-wait during dequeue.
Full barrier semantics will enforce visibility.
Reported by: Maxime Henrion <mhenrion@gma....>.
13 years ago
1f13c75a66
regressions/ck_pr_store: Add coverage for double.
...
It appears GCC doesn't handle IMM constraint gracefully
for double. This isn't an issue with clang.
13 years ago
56bd0b7424
regressions/ck_fifo: cache align MPMC fifo.
13 years ago
3f0a90813a
ck_ring: Unit test coverage for ck_ring_trydequeue.
13 years ago
5c93ee3add
regressions: Do not use ck_error for debug output.
13 years ago
f1be36b6d2
regressions/ck_ring: Fix erroneous usage of ck_error.
13 years ago
6bd946d019
regressions: Add ck_error convenience function.
...
Prints a message and then executes exit(EXIT_FAILURE).
13 years ago
90745e4d60
regressions/ck_ring: Silence GCC false positives.
13 years ago
911909ef31
regressions/ck_ring: Add ck_ring_spmc_template unit test.
13 years ago
b07c7808ce
regressions/ck_ring: Add strict fence around barrier and cache align ring.
13 years ago
d705e02b76
ck_ring: Replace MPMC variant with SPMC variant for starters.
...
Immediate use-case is SPMC. Next is MPSC, then I will generalize
to MPMC (unfortunately, I don't have an algorithm that avoids CAS2
yet for MPMC).
13 years ago
8d90fa28ed
regressions/ck_ring: Add missing header file.
13 years ago
8d8f76845b
regressions/ck_ring: Add MPMC unit test.
13 years ago
68e6752e53
ck_ring: Due to popular request, add MPMC variants of enqueue/dequeue.
...
Fencing semantics are heavier-weight than they need to be.
13 years ago
cb1c72d36b
ck_queue: Add CK_?LIST_MOVE operation.
...
This operation is of format:
CK_S*LIST_MOVE(a, b, linkage) and is equivalent to intializing
a with the contents of b. This is done in a manner that is atomic
with respect to readers. Read-only operations are still valid in
b, but behavior is undefined for write-side operations on b after
a MOVE operation.
13 years ago
533136627f
regressions/ck_stack: Cache align stack.
13 years ago
b0277346b8
regressions/ck_bytelock: Cache-align bytelock.
...
Signed-off-by: Samy Al Bahra <sbahra@repnop.org>
13 years ago
ac50efbeaf
regressions/ck_epoch: GCC 4.6.3 ignores used attribute in this case.
13 years ago
ddfdda3649
regressions/ck_epoch: Force generation of stack dereference in ck_epoch_synchronize.
13 years ago
099f13a24d
regressions/ck_epoch: Force generation of stack pointer.
13 years ago
31040a925d
regressions/ck_ht: Use random get counter.
...
We were erroneously printing reverse get results in serial.
13 years ago
ee2ae5a50f
regressions/ck_epoch: Update check target.
...
Makefile was ignored due to bogus gitignore entry.
13 years ago
690782aa67
ck_hs: Add hash function wrapper to maintain opacity of ck_hs_t.
...
This is in the form of CK_HS_HASH(X, Y, Z) which expands to
Y(Z, X.seed).
13 years ago
43b24d9f28
regressions/ck_hs: Initialize n_state.
13 years ago
44b845a53a
regressions/ck_hs: Use (char *) for debug output.
13 years ago
bc7a3e5606
regressions/ck_bag: Add status around sleep.
13 years ago
ece2895e9e
ck_hs: Drop ck_hs_hash_t abstraction.
...
No need for opaque type anymore.
13 years ago
adaf20b31c
ck_ht: Add ck_ht_stat function.
...
This can be used to expose certain performance metrics.
13 years ago
a40b724270
regressions/ck_hs: Add Makefile for benchmarks.
13 years ago
0f5e540afa
ck_hs: Lock-free set loosely modeled after ck_ht.
...
Currently only SPMC, but MPMC transformation is relatively trivial.
Documentation will come in a follow-up commit.
13 years ago
67895cb87e
regressions/ck_epoch: Add support for multiple writers in poll unit test.
13 years ago
375a9eb816
regressions/ck_epoch: Drop debug flags.
13 years ago
034759c3b9
regressions: Update callers to ck_epoch_call.
13 years ago
657c50f823
regressions/ck_epoch: Multiple writers for synchronize unit test.
13 years ago
5293ad2348
regressions/ck_epoch: Reduce number of poll iterations, change peak calculation in synchronize.
13 years ago
57104fcde7
ck_epoch: Change to epoch semantics, bump grace period to 4.
13 years ago
ce7278140c
regressions/ck_epoch: Add missing header files.
13 years ago
c274e8bc54
ck_epoch: Major redesign and rewrite.
...
I had the pleasure of spending a significant amount of time at the most
recent LPC with Mathieu Desnoyers and Paul McKenney. In discussing
RCU semantics in relation to epoch reclamation, it was argued that
epoch reclamation is a specialisation of RCU (rather than a generalization).
In light of this discussion, I thought it would make more sense to not expose
write-side synchronization semantics aside from ck_epoch_call (similar to
RCU call), ck_epoch_poll (identical to tick), ck_epoch_barrier and
ck_epoch_synchronization (similar to ck_epoch_synchronization). Writers will
now longer have to use write-side epoch sections but can instead rely on
epoch_barrier/synchronization for blocking semantics and ck_epoch_poll
for old tick semantics.
One advantage of this is we can avoid write-side recursion for certain workloads.
Additionally, for infrequent writes, epoch_barrier and epoch_synchronization both
allow for blocking semantics to be used so you don't have to pay the cost of
epoch_entry for non-blocking dispatch.
Example usage:
e = stack_pop(mystack);
ck_epoch_synchronize(...);
free(e);
read_begin and read_end has been replaced with ck_epoch_begin and ck_epoch_end.
If multiple writers need SMR guarantees, then they can also use ck_epoch_begin
and ck_epoch_end. Any dispatch in presence of multiple writers should be done
with-in an epoch section (for now).
There are some follow-up commits to come.
13 years ago
b220e50f54
regressions/ck_pr: Disable rest of width checks on bi-endian platforms.
13 years ago
6961547dfe
regressions/ck_pr: Remove unnecessary assignment in Makefile.
13 years ago
98660bf560
regressions/ck_pr: Disable width checks on bi-endian architectures.
13 years ago
7ebbcf55b1
regressions/ck_spinlock: Disable linux_spinlock on unsupported platforms.
13 years ago
d72ca391f6
ck_bytelock: Switch to RMO model.
13 years ago
bac1c1d26a
regressions/ck_stack: Disable various TRYMPMC benchmarks on unsupported platforms.
13 years ago
3e4aa8239a
regressions: Add _GNU_SOURCE to ck_backoff. Don't apply to all.
13 years ago
f4e50d16d8
regressions: Use _GNU_SOURCE by default on Linux.
13 years ago
8043f52130
ck_pr: Add ck_pr_barrier for compiler barrier.
...
CK_CC_BARRIER isn't idiomatic, roll this into PR memory model
instead.
13 years ago
7e9a0d8751
regressions/ck_brlock: Reduce number of iterations.
13 years ago
f8c19a89b4
regressions/ck_hp: LLVM's optimizer keeps getting the best of us.
...
If I see more regressions in LLVM, I may end up enforcing compiler
barriers in all inlined code.
13 years ago
f567513f8c
regressions/ck_hp: Add Donner's unit test.
13 years ago
3f6a3660eb
regressions/ck_stack: Add trypush unit tests to check target.
...
Submitted by Andrew Schorr <aschorr@telemetry-inves....>.
13 years ago
d454f729e0
regressions/ck_fifo: Add dependency to ck_fifo.h.
13 years ago
290add2e50
build/regressions: Add CORES option, add "check" target.
...
The check target will attempt to run the unit tests with some
sane defaults.
13 years ago
1a8b3db453
ck_ht: Various performance improvements.
...
Though a new implementation is in the works, roll in
some performance improvements in the mean time.
The probe routines have been broken out into separate
reader/writer variants. These variants are much less
branch-intensive (and don't involve predict to stall
in many cases).
New implementation is attempting to deal with
interface-induced overheads.
13 years ago
e1ec55819e
regressions/common. Move edx/eax to x86_64 block.
13 years ago
a56cac3a38
regressions/ck_sequence: Work-around LLVM bug.
...
For some reason, load operations were being re-ordered with respect
to read begin operation (even with explicit compiler barriers).
13 years ago
59158c824b
ck_ht: Do not re-hash on growth for non-PP case.
...
We already store the hash value in the hash table entry
if pointer packing is not enabled.
13 years ago
72a8adb599
ck_ht: Add support for per-hash-table allocator.
...
Documentation and regressions tests have been updated to reflect this.
This functionality allows for individual hash tables use to different
allocation functions. Thanks to Wez Furlong for pointing out the necessary
documentation update for ck_ht.
13 years ago
af7b877414
regressions/ck_ht: Turn off CK_MD_RDTSCP by default.
13 years ago
cc23781512
regressions/ck_ht: Switch to a format that is more readable by R.
...
This allows me to use read.table with minimal changes.
13 years ago
9a3909163f
regressions/ck_ht: Drop secondary sequence table.
...
It is unnecessary, we primarily end up measuring cache miss costs
with it.
13 years ago
0b32fd064e
regressions/ck_ht: Previous commit was in fact shuffled insertion.
...
Add actual shuffled get.
13 years ago
efdae8f77b
regressions/ck_ht: Extend serial benchmark.
...
Add key shuffle get test.
13 years ago
3b5a3c857e
regressions: Update out-of-date comment.
13 years ago
da9e0f58f1
regressions: Use RDTSCP in presence of CK_MD_RDTSCP.
13 years ago
b353ebe8c2
regressions: Remove debug build flags.
13 years ago
314ce3ad81
ck_rwlock: Drop factor argument to trylock API.
...
Remove read-side recursion tests, it is an unsupported
feature to begin with.
13 years ago
512bbc050b
ck_rwlock: Add recursive trylock operations.
...
Unlike ck_rwlock, there is no spin factor associated with these.
13 years ago
380dd410c6
ck_backoff: Saturate geometric back-off.
...
A simple regression test has been added to verify this behavior.
13 years ago
fc2517947b
regressions/ck_ht: Use user-specified hash function for direct benchmark.
13 years ago
18d74f55a1
ck_ht: Add support for user-specified hash functions.
13 years ago
606afe0f23
ck_rwlock: Add ck_rwlock_recursive facility.
...
Allows for write-side recurion. Added upon request by
other users. Regression test has been updated.
13 years ago
402c31ca9a
ck_hp_fifo: Add try{enqueue,dequeue} variants.
...
Regression test has been updated and passes.
13 years ago
a90b402c0c
regressions/ck_fifo: Add test for try{enqueue,dequeue} API.
13 years ago
ff6f4d94f4
ck_stack: Add wait-free trypop and trypush operations.
...
Regression tests have been updated to test these operations as
equivalents to regular looping push/pop operations.
13 years ago
2b2e44203b
Merge pull request #5 from abelmathew/master
...
ck_bag: Merge from abelmathew/master.
13 years ago
6a6f1d53c8
ck_bag: Validate block_size > block_overhead during ck_bag_init.
...
Check ck_bag_init return value in unit test.
13 years ago
bbc6f584fd
ck_bag: Various Bug Fixes and Cleanup.
...
Migrate available block list to CK_LIST.
New blocks are only allocated when the available list is exhausted.
Remove bag->avail_tail.
Print out number of writer iterations for unit test.
Lengthen duration of unit test.
13 years ago
099a6c7b04
ck_bitmap: Add bitmap iterator functions.
...
These can do with some optimization. A majority of this
work is based off an earlier patch submitted by
Shreyas Prasad <sprasad@appne....>.
13 years ago
3cf265cba0
all: Strip trailing whitespaces.
13 years ago
David Joseph
Samy Al Bahra
Samy Al Bahra
Abel Mathew