bdemsky [Thu, 29 Oct 2009 21:26:18 +0000 (21:26 +0000)]
commit file
hkhang [Thu, 29 Oct 2009 20:23:57 +0000 (20:23 +0000)]
*** empty log message ***
yeom [Thu, 29 Oct 2009 01:26:54 +0000 (01:26 +0000)]
more changes.
yeom [Thu, 29 Oct 2009 01:23:46 +0000 (01:23 +0000)]
changes to handle fixed point analysis properly + bug fix.
jjenista [Wed, 28 Oct 2009 22:26:08 +0000 (22:26 +0000)]
bug fix when applying store results from multiple possible edges to a common target node
jzhou [Tue, 27 Oct 2009 20:30:35 +0000 (20:30 +0000)]
close some debug info
jzhou [Tue, 27 Oct 2009 20:24:04 +0000 (20:24 +0000)]
Restructure the shared memory allocation and fixed multiple bugs in the multicore gc. Now the startegy of shared memory allocation can be configurable as 1) each core can only use its local memory 2) use local memory for lower address space while use global memory for higher address space or 3) all cores allocate globally. The simplest test case can now go through several round of gc correctly. Still need more tests. Also organized the codes a little bit.
yeom [Tue, 27 Oct 2009 17:47:57 +0000 (17:47 +0000)]
added for handling method with child SESE.
It collects pre-effects of callee in the first phase, and then it converts pre-effects into caller's.
bdemsky [Tue, 27 Oct 2009 00:58:20 +0000 (00:58 +0000)]
more stm options
yeom [Mon, 26 Oct 2009 18:41:25 +0000 (18:41 +0000)]
changes towards method context insensitive analysis.
add stall edge mapping.
bdemsky [Sat, 24 Oct 2009 07:10:51 +0000 (07:10 +0000)]
bug fixes
bdemsky [Sat, 24 Oct 2009 06:46:10 +0000 (06:46 +0000)]
changes
bdemsky [Sat, 24 Oct 2009 01:25:38 +0000 (01:25 +0000)]
get Fission w/ array support working
jjenista [Fri, 23 Oct 2009 22:12:22 +0000 (22:12 +0000)]
starting implementation for access paths to improve edge mapping
yeom [Fri, 23 Oct 2009 21:34:50 +0000 (21:34 +0000)]
initial commit for parent/child SESE memory conflicts analysis
bdemsky [Thu, 22 Oct 2009 22:08:43 +0000 (22:08 +0000)]
bug fixes
bdemsky [Thu, 22 Oct 2009 20:58:20 +0000 (20:58 +0000)]
add new features...they don't break the build, but need to check if they work...
jzhou [Wed, 21 Oct 2009 23:19:56 +0000 (23:19 +0000)]
change to use static mapping of core # and block # in multicore garbage modul
jjenista [Wed, 21 Oct 2009 21:36:59 +0000 (21:36 +0000)]
Added feature for nullfying dead variables, which didn't change sharing for benchmarks. It is left in, but commented out, until we decide we need it.
bdemsky [Wed, 21 Oct 2009 04:29:07 +0000 (04:29 +0000)]
try to work on memory usage...have large test case running now
bdemsky [Wed, 21 Oct 2009 01:52:21 +0000 (01:52 +0000)]
working version
adash [Wed, 21 Oct 2009 01:47:22 +0000 (01:47 +0000)]
add changes to release locks early on version mismatch and soft abort
adash [Wed, 21 Oct 2009 01:45:44 +0000 (01:45 +0000)]
fix distributed KMeans bugs and add javasingle version
bdemsky [Wed, 21 Oct 2009 01:01:01 +0000 (01:01 +0000)]
working genome on byte strings
bdemsky [Wed, 21 Oct 2009 00:20:49 +0000 (00:20 +0000)]
changes
bdemsky [Wed, 21 Oct 2009 00:02:57 +0000 (00:02 +0000)]
changes
bdemsky [Tue, 20 Oct 2009 23:46:31 +0000 (23:46 +0000)]
fix bug
bdemsky [Tue, 20 Oct 2009 23:42:29 +0000 (23:42 +0000)]
changes
jjenista [Tue, 20 Oct 2009 21:34:42 +0000 (21:34 +0000)]
revert alias set computation changes, stick with old way
bdemsky [Tue, 20 Oct 2009 01:18:57 +0000 (01:18 +0000)]
seems correct now
bdemsky [Tue, 20 Oct 2009 00:47:56 +0000 (00:47 +0000)]
small differences
bdemsky [Mon, 19 Oct 2009 22:42:03 +0000 (22:42 +0000)]
fix bug for jim
jjenista [Mon, 19 Oct 2009 21:15:42 +0000 (21:15 +0000)]
omputation to determine set of aliased parameter indices was very conservative,
but correct, fortunately. It used to list every parameter pair that could reach
a common node as aliased! Yikes! Actually, we want to see that parameters can
directly reference a common node, but further, they can only be aliased if the
edges to the shared node both have a common reachability state that is also on
the shared node. Otherwise they cannot be referencing a common concrete object.
jjenista [Mon, 19 Oct 2009 17:31:05 +0000 (17:31 +0000)]
little changes to keep, debugging removed
bdemsky [Mon, 19 Oct 2009 09:12:51 +0000 (09:12 +0000)]
readme + input files
more bug fixes
bdemsky [Mon, 19 Oct 2009 06:56:29 +0000 (06:56 +0000)]
more bug fixes
bdemsky [Sun, 18 Oct 2009 10:53:59 +0000 (10:53 +0000)]
bug fixes
bdemsky [Sun, 18 Oct 2009 09:56:13 +0000 (09:56 +0000)]
missing class
bdemsky [Sun, 18 Oct 2009 09:54:59 +0000 (09:54 +0000)]
change
bdemsky [Sun, 18 Oct 2009 09:54:26 +0000 (09:54 +0000)]
allow threads to request abort
bdemsky [Sun, 18 Oct 2009 09:53:53 +0000 (09:53 +0000)]
enough to parse
bdemsky [Sun, 18 Oct 2009 06:02:25 +0000 (06:02 +0000)]
almost done with port
bdemsky [Sat, 17 Oct 2009 22:19:02 +0000 (22:19 +0000)]
changes
bdemsky [Sat, 17 Oct 2009 21:56:43 +0000 (21:56 +0000)]
changes
bdemsky [Sat, 17 Oct 2009 21:15:57 +0000 (21:15 +0000)]
capitalization
bdemsky [Sat, 17 Oct 2009 21:13:41 +0000 (21:13 +0000)]
more bug fixes
bdemsky [Sat, 17 Oct 2009 01:01:43 +0000 (01:01 +0000)]
changes towards finishing yada
jihoonl [Fri, 16 Oct 2009 23:30:49 +0000 (23:30 +0000)]
forgot these files
jihoonl [Fri, 16 Oct 2009 23:25:35 +0000 (23:25 +0000)]
Spider benchmark
bdemsky [Fri, 16 Oct 2009 21:35:34 +0000 (21:35 +0000)]
makefil
bdemsky [Fri, 16 Oct 2009 21:35:22 +0000 (21:35 +0000)]
parses
jjenista [Fri, 16 Oct 2009 17:59:16 +0000 (17:59 +0000)]
working on directto benchmark--removing code like unused constructors that hinder program understanding, and tinkering with the stripped down version for exploring analysis weaknesses
jjenista [Fri, 16 Oct 2009 17:57:13 +0000 (17:57 +0000)]
compiler options for call map debugging
bdemsky [Fri, 16 Oct 2009 10:11:31 +0000 (10:11 +0000)]
more benchmark
bdemsky [Fri, 16 Oct 2009 08:51:15 +0000 (08:51 +0000)]
beginnings of port...won't compile yet
bdemsky [Fri, 16 Oct 2009 03:18:02 +0000 (03:18 +0000)]
remove unused array
bdemsky [Fri, 16 Oct 2009 03:17:25 +0000 (03:17 +0000)]
back out change
bdemsky [Fri, 16 Oct 2009 03:15:08 +0000 (03:15 +0000)]
changes
bdemsky [Fri, 16 Oct 2009 03:13:33 +0000 (03:13 +0000)]
revert commented out change
adash [Fri, 16 Oct 2009 01:28:39 +0000 (01:28 +0000)]
changes to KMeans after bug bix for parsing floats
change manual prefetch placements
rearrange atomic blocks to enable parallel part of the code and not use one big lock
around work function
bdemsky [Fri, 16 Oct 2009 00:42:10 +0000 (00:42 +0000)]
change
bdemsky [Thu, 15 Oct 2009 23:59:00 +0000 (23:59 +0000)]
big problem
bdemsky [Thu, 15 Oct 2009 07:37:23 +0000 (07:37 +0000)]
changes
bdemsky [Thu, 15 Oct 2009 05:32:29 +0000 (05:32 +0000)]
add more lock primitives
bdemsky [Thu, 15 Oct 2009 03:27:59 +0000 (03:27 +0000)]
more bug fixes
bdemsky [Thu, 15 Oct 2009 02:31:12 +0000 (02:31 +0000)]
bug fixes
bdemsky [Thu, 15 Oct 2009 01:11:32 +0000 (01:11 +0000)]
bug fixes
adash [Thu, 15 Oct 2009 00:22:21 +0000 (00:22 +0000)]
change inputs random10000_12
atomic blocks around readFromFile()
bdemsky [Wed, 14 Oct 2009 20:38:04 +0000 (20:38 +0000)]
more work towards supporting arrays
adash [Wed, 14 Oct 2009 18:58:12 +0000 (18:58 +0000)]
added new input file for kmeans random10000_12
jjenista [Wed, 14 Oct 2009 15:09:54 +0000 (15:09 +0000)]
investigating OOPSLA benchmarks further
jjenista [Wed, 14 Oct 2009 15:09:22 +0000 (15:09 +0000)]
updating graph display options
bdemsky [Wed, 14 Oct 2009 09:42:58 +0000 (09:42 +0000)]
more changes towards word level support for arrays
bdemsky [Wed, 14 Oct 2009 08:45:26 +0000 (08:45 +0000)]
back to compiling
bdemsky [Wed, 14 Oct 2009 08:30:24 +0000 (08:30 +0000)]
*** empty log message ***
bdemsky [Wed, 14 Oct 2009 08:25:52 +0000 (08:25 +0000)]
change
bdemsky [Wed, 14 Oct 2009 08:15:51 +0000 (08:15 +0000)]
change
bdemsky [Wed, 14 Oct 2009 08:12:08 +0000 (08:12 +0000)]
change
bdemsky [Wed, 14 Oct 2009 08:05:55 +0000 (08:05 +0000)]
try to organize code more...
bdemsky [Wed, 14 Oct 2009 02:15:36 +0000 (02:15 +0000)]
changes
adash [Wed, 14 Oct 2009 00:14:00 +0000 (00:14 +0000)]
changes to manual prefetching (combine severals sends/recvs)
add logevent to SingleTM microbenchmark
bdemsky [Wed, 14 Oct 2009 00:12:53 +0000 (00:12 +0000)]
changes
bdemsky [Wed, 14 Oct 2009 00:07:36 +0000 (00:07 +0000)]
give nodes unique id's for printing code
bdemsky [Wed, 14 Oct 2009 00:07:12 +0000 (00:07 +0000)]
add comments to made it easier to understand
adash [Wed, 14 Oct 2009 00:03:11 +0000 (00:03 +0000)]
check-in benchmark changes for ipdps paper
bdemsky [Tue, 13 Oct 2009 23:58:58 +0000 (23:58 +0000)]
make build system for benchmarks a little nicer
working toward word-based STM for arrays
bdemsky [Tue, 13 Oct 2009 23:44:03 +0000 (23:44 +0000)]
changes to intruder
bdemsky [Tue, 13 Oct 2009 23:42:13 +0000 (23:42 +0000)]
obvious typo
bdemsky [Tue, 13 Oct 2009 23:39:11 +0000 (23:39 +0000)]
more changes
bdemsky [Tue, 13 Oct 2009 23:37:32 +0000 (23:37 +0000)]
add new option
try to start to organize these options
adash [Tue, 13 Oct 2009 23:26:18 +0000 (23:26 +0000)]
benchmark set for ipdps 2010 paper
adash [Tue, 13 Oct 2009 23:08:36 +0000 (23:08 +0000)]
missed a placement of manual prefetch
adash [Tue, 13 Oct 2009 22:56:15 +0000 (22:56 +0000)]
added manual prefetching kmeans version
bdemsky [Tue, 13 Oct 2009 20:53:48 +0000 (20:53 +0000)]
*** empty log message ***
bdemsky [Tue, 13 Oct 2009 20:33:09 +0000 (20:33 +0000)]
small changes
bdemsky [Tue, 13 Oct 2009 09:45:45 +0000 (09:45 +0000)]
changes
bdemsky [Tue, 13 Oct 2009 08:08:21 +0000 (08:08 +0000)]
changes
bdemsky [Tue, 13 Oct 2009 03:07:18 +0000 (03:07 +0000)]
fixed nasty analysis bug for fission
bdemsky [Tue, 13 Oct 2009 01:16:43 +0000 (01:16 +0000)]
comment out inlineprefetch...wasn't supposed to uncomment
bdemsky [Tue, 13 Oct 2009 01:14:13 +0000 (01:14 +0000)]
changes
projects / IRC.git / log