Thomas Preud'homme
2c228612e5
[commtech] Avoid if in bqv2
...
Uses 2 send() function in order to avoid the end of buffer test and the
mapping test.
2012-01-30 20:13:17 +01:00
Thomas Preud'homme
a20c9a8a21
[commtech] BatchQueue v2
...
Uses 2 mapping to the same structure to avoid prefetching of the
producer semi-buffer by the consumer. The idea is to access everything
through mapping 1 except semi-buffer 2 which is accessed through mapping
2.
2012-01-30 20:09:52 +01:00
Thomas Preud'homme
fba09b60b8
Remove debug informations
2012-01-30 20:08:54 +01:00
Thomas Preud'homme
c6786815cd
Add native algo from OpenMP stream extension
...
Add native algorithm from OpenMP stream extension. This require adding
one function in commtech.h: end_producer(). This function does nothing
for all communication algorithm but gomp_stream (the algorithm added by
this commit).
2012-01-30 20:07:11 +01:00
Thomas Preud'homme
a30a5bfe06
Make all threads are joined
...
in join_threads, nb_thread is the id of the last thread, not the number
of threads to join. Hence the for loop must include this id.
2011-06-01 15:35:08 +02:00
Thomas Preud'homme
f0c75c7570
SINK thread (not INTERM) notify its termination
...
Use !!node_param->type & SINK in likely macro to test wether we are a
SINK node or an INTERM node.
2011-06-01 15:25:08 +02:00
Thomas Preud'homme
bd7379e73a
Propose 2048 and 4096 buffer size for BatchQueue.
2011-05-27 15:42:40 +02:00
Thomas Preud'homme
1ca11d6fc5
[commtech] Ensure real and displayed cmdline match
2011-05-27 15:42:40 +02:00
Thomas Preud'homme
a287e033be
Merge branch 'improve_pipeline'
2011-05-27 15:42:11 +02:00
Thomas Preud'homme
f05cfdcd92
Improve pipeline (cons and prod in //)
2011-05-25 14:33:42 +02:00
Thomas Preud'homme
879769ef32
Fix MCRingBuffer deadlocks
...
* batchSize must be kept in sync with the number of cache line sent.
* Don't ask more data than sent
2011-05-10 17:19:21 +02:00
Thomas Preud'homme
f01db158c2
Use multiples of BUF_SIZE when needed
...
Number of cache line sent and size of reception buffer must be a
multiple of BUF_SIZE.
2011-05-10 11:14:28 +02:00
Thomas Preud'homme
6fcfd60d2d
Fix buffer loop in BatchQueue single data mode
...
The buffer in single data mode in batchQueue was not circular because a
variable was not renamed
2011-05-10 11:02:00 +02:00
Thomas Preud'homme
70f8f95647
Fix option to choose the number of node
...
Option is now in the getopt string and accessible with -l switch.
2011-05-10 11:00:59 +02:00
Thomas Preud'homme
f430cc17a7
Fix bugs coming from refactoring
2011-05-05 19:54:44 +02:00
Thomas Preud'homme
372c36155a
Fix incorrect usage string: --check -> -k
2011-05-05 14:52:41 +02:00
Thomas Preud'homme
756a701466
[commtech] Refactor to chain more than 2 nodes
...
* Refactor the source to be able to chain more than 2 nodes together
* Compile all binaries by default (binList must be set manually in
lancement.sh to run only a subset of the binaries
2011-05-05 14:34:09 +02:00
Thomas Preud'homme
5d71bc53f1
[commtech] Varying size of buffer for BatchQueue
...
Create several variation of BatchQueue, each with a different buffer
size: batch_queue_1024, batch_queue_512, ..., batch_queue_2.
2011-05-05 11:30:00 +02:00
Thomas Preud'homme
9c835d4c46
Add a "sent words == received words" check
2011-05-04 19:35:10 +02:00
Thomas Preud'homme
cea65bd1bc
Support log file without "buf size" message
...
Instead of trying to get the buf size in the first log, grep on all logs
and take the value of the first match.
2011-05-04 19:34:58 +02:00
Thomas Preud'homme
829063f894
Resiliant to empty ${bench}List
2011-05-04 19:34:53 +02:00
Thomas Preud'homme
023552a4de
[commtech] Prevent prefetch in batch queue
...
Prevent prefetch of state variable in BatchQueue by adding padding
between the buffers and state variable.
2011-03-02 13:19:22 +01:00
Thomas Preud'homme
1388b6d556
[commtech] Increase the number of word sent
2011-03-02 13:19:22 +01:00
Thomas Preud'homme
c947c24cf2
[commtech] Handle more than 2 cache hierarchies
...
Rewrite creation of simple gnuplot to handle more than 2 cache
hierarchies (like L2, CPU and mem for sibling cores on same CPU,
non-sibling cores on same CPU and non sibling cores on different CPU).
2011-03-02 13:19:22 +01:00
Thomas Preud'homme
22c97ab418
[commtech] Make BUF_SIZE definition be per tech
...
Don't define BUF_SIZE globally anymore, but per communication technique
2011-03-02 13:19:22 +01:00
Thomas Preud'homme
7c515200e7
[commtech] Remove asm_cache from the comm techs
2011-03-02 13:19:22 +01:00
Thomas Preud'homme
90b7a8007b
[commtech] Rename c_cache to batch_queue
2011-03-02 13:19:22 +01:00
Thomas Preud'homme
f1b3174af3
Fix various bugs in log creation and parsing
...
* Force english locales (esp. for numeric values)
* Handle french and english numeric values
* Handle absence of useless_prod log
* Handle unique cache hierarchy
2011-02-24 21:00:09 +01:00
Thomas Preud'homme
ee71568981
[barrier] Make the bench use perf instead of PAPI
...
- Convert barrier bench from papi+PapiHighLevel to perf framework
- Remove papihighlevel submodule
- Simplify Makefile (include moving some of the code in a separate
script)
2011-01-29 17:51:45 +01:00
Thomas Preud'homme
ff7c38da55
[commtech] Display misses of L1 cache
...
- Remove infos about last level cache
+ Add misses of L1 cache
2011-01-28 23:59:57 +01:00
Thomas Preud'homme
7d7ad0c46a
[commtech] Make WORDS_PER_BUF indep of BUF_SIZE.
...
The number of data sent must be independent of the buffer size chosen
by each algorithm.
2011-01-28 04:56:44 +01:00
Thomas Preud'homme
b1316fb8a4
[commtech] Use param variable for matrice calc.
...
Matrice calculation library should not take parameter value 16 hardcoded
but use the value of the param variable.
2011-01-25 17:27:07 +01:00
Thomas Preud'homme
aef40ff651
[commtech] Fix cache line size
...
Cache line size in current processors (at least the one used for tests:
cerclon) is 64 bytes, not 128. Thus modifying the declaration in
commtech.h
2011-01-25 17:25:38 +01:00
Thomas Preud'homme
c3aad28ad5
[commtech] Add calculation method
...
Add a calculation method which add the value of the first integer of
n consecutive cache lines and write the results in one of the integer of
these cache lines. Next calculation uses the next n consecutives cache
lines and write the result in the next integer.
2011-01-25 17:24:53 +01:00
Thomas Preud'homme
975411a824
Split CSQ in 2 communication techniques.
...
* Divide CSQ in 2 communication techniques: one with 2 slots (as in
BatchQueue aka c_cache) and one with 64 slots (as in the article)
* Rename fake communication technique in none communication technique
and disable any activity (send no longer does anything)
2011-01-25 17:24:53 +01:00
Thomas Preud'homme
5eb7fb50c7
[commtech] CSQ use memcpy in dequeue for fairness
...
Paper about CSQ uses memcpy in enqueue and dequeue. Although it is not
possible to use memcpy in enqueue because of current API, it is possible
to use memcpy in dequeue, hence this commit.
2011-01-19 12:37:14 +01:00
Thomas Preud'homme
7db514a706
[commtech] Fix time metric pattern
...
time metric pattern need to include dots and several elements as a time
is typically of the form 42.42424242
2011-01-16 17:30:00 +01:00
Thomas Preud'homme
1867515152
[commtech] Remove pipe_comm to produce 50000000 l
...
Multiply by 10 the number of cache line send from the producer to the
consumer to have a more accurate mean. This require excluding pipe_comm
as this bench is way too slow to send so much data.
2011-01-16 17:27:21 +01:00
Thomas Preud'homme
35a81bb736
[commtech] Place volatile on the right qualifier.
2011-01-13 14:58:13 +01:00
Thomas Preud'homme
f8ef047808
Remove obsolete TODO.
2011-01-03 11:40:23 +01:00
Thomas Preud'homme
2d879dc3fc
[commtech] Fix idx test in c_cache technique.
...
c_cache watching status value when idx % BUF_SIZE != 0 instead of when
it's equal zero.
2011-01-03 11:35:42 +01:00
Thomas Preud'homme
9f5fb833d7
[commtech] Remove obsolete pattern for total time.
2011-01-03 11:34:47 +01:00
Thomas Preud'homme
5682581335
[commtech bench] Fix the zero test
2010-10-14 00:01:32 +02:00
Thomas Preud'homme
6c2868e20c
[commtech bench] Take the mean over 10 run.
2010-10-13 23:57:58 +02:00
Thomas Preud'homme
006b1b1d94
Simplify and rewrite comm API.
2010-10-01 18:57:46 +02:00
Thomas Preud'homme
c5507590aa
Default to $CWD in all case in parsing.sh
2010-08-12 05:11:46 +02:00
Thomas Preud'homme
5750bea31f
Ignore dacapo bench (for the moment)
2010-05-18 10:35:28 +02:00
Thomas Preud'homme
6899dea515
Allow overriding barriereList and commList
2010-05-18 09:34:25 +02:00
Thomas Preud'homme
e49eaecf18
Allow specifiying the log directory
2010-05-18 09:34:15 +02:00
Thomas Preud'homme
9af26178ba
Add shell options to local function params
...
This requires to switch from bash to sh
2010-05-18 09:34:12 +02:00