IRC.git
14 years agomost of the missing context-rewriting code
jjenista [Mon, 15 Mar 2010 19:55:17 +0000 (19:55 +0000)]
most of the missing context-rewriting code

14 years agoa start on reachability, not fully functioning yet
jjenista [Sat, 13 Mar 2010 18:16:13 +0000 (18:16 +0000)]
a start on reachability, not fully functioning yet

14 years agomore changes
adash [Sat, 13 Mar 2010 02:08:34 +0000 (02:08 +0000)]
more changes

14 years agochanges for optimizations
adash [Sat, 13 Mar 2010 01:59:42 +0000 (01:59 +0000)]
changes for optimizations

14 years agochanges
adash [Sat, 13 Mar 2010 01:58:40 +0000 (01:58 +0000)]
changes

14 years agoif an operation on canonical objects wants to use other canonical ops, it must supply...
jjenista [Fri, 12 Mar 2010 20:11:01 +0000 (20:11 +0000)]
if an operation on canonical objects wants to use other canonical ops, it must supply canonical arguments.  that said, if the operation is straightforward and the result isn't hashed, its okay to build the object from scratch and then canonicalize it at the end

14 years agoa test that shows reachabiltiy states are broken
jjenista [Fri, 12 Mar 2010 19:59:57 +0000 (19:59 +0000)]
a test that shows reachabiltiy states are broken

14 years agobunch of bug fixes, graphs appear to be working mechanically over call chains, still...
jjenista [Fri, 12 Mar 2010 19:16:06 +0000 (19:16 +0000)]
bunch of bug fixes, graphs appear to be working mechanically over call chains, still needs attention to reachability as symbols get rewritten over context boundaries

14 years agomaking a big change to call site transform to calculate out-of-context edges correctl...
jjenista [Fri, 12 Mar 2010 01:16:17 +0000 (01:16 +0000)]
making a big change to call site transform to calculate out-of-context edges correctly, but in current form BLEEDS ELEMENTS BETWEEN GRAPHS again, checking in this stable version though so I don't mess up the good stuff while trying to fix it

14 years agodebugging something else, but commiting this bug fix--mapping out-of-context nodes...
jjenista [Thu, 11 Mar 2010 23:38:53 +0000 (23:38 +0000)]
debugging something else, but commiting this bug fix--mapping out-of-context nodes to a unique heap region id requires that the callee method be part of the identifier string otherwise an ooc region can appear in more than one context of a call chain, which is bad

14 years agobug fix yonghun found, return value's region may be a new node for caller context...
jjenista [Thu, 11 Mar 2010 22:18:14 +0000 (22:18 +0000)]
bug fix yonghun found, return value's region may be a new node for caller context, might have to a matching object for caller context

14 years agothere was a simple fix for variable problem, in new model we want to cut out all...
jjenista [Thu, 11 Mar 2010 21:54:35 +0000 (21:54 +0000)]
there was a simple fix for variable problem, in new model we want to cut out all region to region AND var to region edges in caller before we map callee effects back in, things look good for one call but chains any longer get messed up--investigating...

14 years agocaller local variables to in-context regions are modeled as out-of-context edges...
jjenista [Thu, 11 Mar 2010 20:08:50 +0000 (20:08 +0000)]
caller local variables to in-context regions are modeled as out-of-context edges in callee to preserve info, then matched in call site transform.  This approach is running and stable, but the answer doesn't look right with respect to the age of nodes in caller

14 years agoremove some redundant outputs
jzhou [Thu, 11 Mar 2010 19:52:53 +0000 (19:52 +0000)]
remove some redundant outputs

14 years agosome missing check in
jzhou [Thu, 11 Mar 2010 19:23:14 +0000 (19:23 +0000)]
some missing check in

14 years agoNot the most elegant fix, but out-of-context nodes can be uniquely identified by...
jjenista [Thu, 11 Mar 2010 00:13:06 +0000 (00:13 +0000)]
Not the most elegant fix, but out-of-context nodes can be uniquely identified by a string, which is mapped to a particular hrn id to get convergence

14 years agoout-of-context edges prevent convergence...
jjenista [Wed, 10 Mar 2010 23:02:58 +0000 (23:02 +0000)]
out-of-context edges prevent convergence...

14 years agolots of bug fixes, system cannot compute even simple programs without variables as...
jjenista [Wed, 10 Mar 2010 21:59:43 +0000 (21:59 +0000)]
lots of bug fixes, system cannot compute even simple programs without variables as out-of-context edges

14 years agomark before adding zero-overhead linux mode
jzhou [Wed, 10 Mar 2010 18:49:38 +0000 (18:49 +0000)]
mark before adding zero-overhead linux mode

14 years agobig checkin, lots of call site transfer bug fixes, analysis runs for a tiny example...
jjenista [Wed, 10 Mar 2010 00:28:14 +0000 (00:28 +0000)]
big checkin, lots of call site transfer bug fixes, analysis runs for a tiny example correctly

14 years agocallee elements brought into caller get predicates from caller element that satisfied...
jjenista [Tue, 9 Mar 2010 01:05:31 +0000 (01:05 +0000)]
callee elements brought into caller get predicates from caller element that satisfied them, properly, also bug fix for edges from nodes that aren't satisifed from throwing a null pointer exception

14 years agoedge merge
jjenista [Mon, 8 Mar 2010 21:14:09 +0000 (21:14 +0000)]
edge merge

14 years agochanges.
yeom [Mon, 8 Mar 2010 18:19:25 +0000 (18:19 +0000)]
changes.

14 years agochanges to trans.c for performance improvement
adash [Sat, 6 Mar 2010 22:55:04 +0000 (22:55 +0000)]
changes to trans.c for performance improvement

14 years agoadd changes to automate dsm and recovery executions
adash [Sat, 6 Mar 2010 22:51:24 +0000 (22:51 +0000)]
add changes to automate dsm and recovery executions

14 years agochange cache lookup table,
adash [Fri, 5 Mar 2010 10:10:11 +0000 (10:10 +0000)]
change cache lookup table,
remove rcount-> instead use isBackUp field for alignment issues,
add code for signal handler,
some code changes to address cache invalidation-> same as dsm

14 years agochanges. better coding style.
yeom [Fri, 5 Mar 2010 00:27:07 +0000 (00:27 +0000)]
changes. better coding style.

14 years agoinitial commit of new memory queue. still just working with fine-grain cases.
yeom [Thu, 4 Mar 2010 23:15:33 +0000 (23:15 +0000)]
initial commit of new memory queue. still just working with fine-grain cases.

14 years agobug fix, other transfer funcs invoke mutating methods, call site transfer creates...
jjenista [Thu, 4 Mar 2010 22:25:06 +0000 (22:25 +0000)]
bug fix, other transfer funcs invoke mutating methods, call site transfer creates new reachgraph--but new graph wasn't being returned to fixed point computation

14 years agogetting closer, still major bugs in call site transfer function
jjenista [Thu, 4 Mar 2010 08:23:56 +0000 (08:23 +0000)]
getting closer, still major bugs in call site transfer function

14 years agoless transaction
jihoonl [Thu, 4 Mar 2010 08:16:53 +0000 (08:16 +0000)]
less transaction

14 years agosmall bug fix
jihoonl [Thu, 4 Mar 2010 08:15:17 +0000 (08:15 +0000)]
small bug fix

14 years agosmall changes
adash [Thu, 4 Mar 2010 01:54:00 +0000 (01:54 +0000)]
small changes

14 years agoproper placement of atomic blocks
adash [Thu, 4 Mar 2010 01:53:12 +0000 (01:53 +0000)]
proper placement of atomic blocks

14 years agochanges to benchmarks
adash [Thu, 4 Mar 2010 00:17:45 +0000 (00:17 +0000)]
changes to benchmarks
javasingle version of spamfilter

14 years agochanges for prefetch/caching and performance improvements
adash [Wed, 3 Mar 2010 23:53:11 +0000 (23:53 +0000)]
changes for prefetch/caching and performance improvements

14 years agofound a bug where an object of one reach graph could get into another reach graph...
jjenista [Wed, 3 Mar 2010 21:56:55 +0000 (21:56 +0000)]
found a bug where an object of one reach graph could get into another reach graph, which breaks everything

14 years agobetter duplicateLostObject
jihoonl [Wed, 3 Mar 2010 00:56:06 +0000 (00:56 +0000)]
better duplicateLostObject

14 years agoremove hard coded relative path for getting profiling data
jzhou [Tue, 2 Mar 2010 22:52:14 +0000 (22:52 +0000)]
remove hard coded relative path for getting profiling data

14 years agocheck in a debug mode for call site transfer I use a lot, but actually hook it up...
jjenista [Mon, 1 Mar 2010 18:48:38 +0000 (18:48 +0000)]
check in a debug mode for call site transfer I use a lot, but actually hook it up to the build options, whoops

14 years agomodified script for other results
adash [Thu, 25 Feb 2010 22:14:46 +0000 (22:14 +0000)]
modified script for other results
remove unnecessary dstm.conf file
some changes to increase execution time for the Game benchmark

14 years agoenable optimization
adash [Thu, 25 Feb 2010 22:07:44 +0000 (22:07 +0000)]
enable optimization

14 years agojava single version of Matrix Multiply
adash [Thu, 25 Feb 2010 22:06:07 +0000 (22:06 +0000)]
java single version of Matrix Multiply

14 years agofix some bug in the multicore gc
jzhou [Thu, 25 Feb 2010 19:24:26 +0000 (19:24 +0000)]
fix some bug in the multicore gc

14 years agochanges
jihoonl [Wed, 24 Feb 2010 21:44:13 +0000 (21:44 +0000)]
changes

14 years agosystem stable, call site transform wipes out graphs without rebuilding correctly...
jjenista [Tue, 23 Feb 2010 23:46:56 +0000 (23:46 +0000)]
system stable, call site transform wipes out graphs without rebuilding correctly, still working on it

14 years agoa helpful method
jjenista [Tue, 23 Feb 2010 23:46:21 +0000 (23:46 +0000)]
a helpful method

14 years agocommand line parsing change for the script
adash [Tue, 23 Feb 2010 22:52:47 +0000 (22:52 +0000)]
command line parsing change for the script

14 years agochange
jihoonl [Tue, 23 Feb 2010 22:51:58 +0000 (22:51 +0000)]
change

14 years agoadd java version
adash [Tue, 23 Feb 2010 22:11:41 +0000 (22:11 +0000)]
add java version

14 years agochange
jihoonl [Tue, 23 Feb 2010 01:05:19 +0000 (01:05 +0000)]
change

14 years agoadd debugs for now
adash [Tue, 23 Feb 2010 00:12:26 +0000 (00:12 +0000)]
add debugs for now

14 years agochange
jihoonl [Tue, 23 Feb 2010 00:10:10 +0000 (00:10 +0000)]
change

14 years agobug fix, was too aggressive with equals method for canonical objects, have to do...
jjenista [Mon, 22 Feb 2010 22:21:29 +0000 (22:21 +0000)]
bug fix, was too aggressive with equals method for canonical objects, have to do full compare otherwise you can't compare an object to something in the canon in order to retrieve the canonical version

14 years agoadded new barrier and game recovery benchmark
adash [Mon, 22 Feb 2010 19:50:35 +0000 (19:50 +0000)]
added new barrier and game recovery benchmark

14 years agochecking in this test before monkeying around with debugging stuff
jjenista [Fri, 19 Feb 2010 23:00:14 +0000 (23:00 +0000)]
checking in this test before monkeying around with debugging stuff

14 years agosquashing bugs
jjenista [Fri, 19 Feb 2010 19:54:49 +0000 (19:54 +0000)]
squashing bugs

14 years agoanalysis not terminating, looking for things that should equate but don't that preven...
jjenista [Fri, 19 Feb 2010 18:57:57 +0000 (18:57 +0000)]
analysis not terminating, looking for things that should equate but don't that prevent convergence

14 years agobig update--bringing implementation of new analysis into focus
jjenista [Fri, 19 Feb 2010 01:05:18 +0000 (01:05 +0000)]
big update--bringing implementation of new analysis into focus

14 years ago*** empty log message ***
hkhang [Thu, 18 Feb 2010 05:21:30 +0000 (05:21 +0000)]
*** empty log message ***

14 years agowell
jihoonl [Thu, 18 Feb 2010 02:14:55 +0000 (02:14 +0000)]
well

14 years agoto print out byte info
jihoonl [Thu, 18 Feb 2010 02:14:35 +0000 (02:14 +0000)]
to print out byte info

14 years agochange
jihoonl [Thu, 18 Feb 2010 02:05:52 +0000 (02:05 +0000)]
change

14 years agotime stamp
jihoonl [Wed, 17 Feb 2010 06:53:08 +0000 (06:53 +0000)]
time stamp

14 years ago*** empty log message ***
hkhang [Wed, 17 Feb 2010 06:01:26 +0000 (06:01 +0000)]
*** empty log message ***

14 years agoprint out recovery time in micro second
jihoonl [Wed, 17 Feb 2010 01:42:13 +0000 (01:42 +0000)]
print out recovery time in micro second

14 years agoremove debug messages
jihoonl [Wed, 17 Feb 2010 00:39:38 +0000 (00:39 +0000)]
remove debug messages

14 years agono join!
jihoonl [Wed, 17 Feb 2010 00:27:06 +0000 (00:27 +0000)]
no join!

14 years agowell..
jihoonl [Wed, 17 Feb 2010 00:24:44 +0000 (00:24 +0000)]
well..

14 years agodebug
jihoonl [Wed, 17 Feb 2010 00:14:42 +0000 (00:14 +0000)]
debug

14 years agochanges
jihoonl [Wed, 17 Feb 2010 00:14:12 +0000 (00:14 +0000)]
changes

14 years agofix code organization
bdemsky [Mon, 15 Feb 2010 08:07:35 +0000 (08:07 +0000)]
fix code organization

14 years agochanges to improve commit intensive benchmarks and combine invalid messages
adash [Mon, 15 Feb 2010 07:44:04 +0000 (07:44 +0000)]
changes to improve commit intensive benchmarks and combine invalid messages
and commit messages for speedup

14 years agochange
jihoonl [Sat, 13 Feb 2010 00:09:44 +0000 (00:09 +0000)]
change

14 years agomlookup bug fix
jihoonl [Fri, 12 Feb 2010 23:12:52 +0000 (23:12 +0000)]
mlookup bug fix

14 years agomlookup -> altmlookup
jihoonl [Fri, 12 Feb 2010 23:11:23 +0000 (23:11 +0000)]
mlookup -> altmlookup

14 years ago*** empty log message ***
hkhang [Fri, 12 Feb 2010 00:35:17 +0000 (00:35 +0000)]
*** empty log message ***

14 years agoreevaluating abstract garbage collection, for now leave code but don't use it
jjenista [Thu, 11 Feb 2010 23:32:38 +0000 (23:32 +0000)]
reevaluating abstract garbage collection, for now leave code but don't use it

14 years ago*** empty log message ***
hkhang [Thu, 11 Feb 2010 18:33:42 +0000 (18:33 +0000)]
*** empty log message ***

14 years ago*** empty log message ***
hkhang [Thu, 11 Feb 2010 18:10:52 +0000 (18:10 +0000)]
*** empty log message ***

14 years agobug fixing in multicore gc and add profiling code for gc
jzhou [Wed, 10 Feb 2010 17:22:25 +0000 (17:22 +0000)]
bug fixing in multicore gc and add profiling code for gc

14 years agoguard code with sandbox flag to prevent other dsm stuff from breaking
adash [Wed, 10 Feb 2010 02:59:53 +0000 (02:59 +0000)]
guard code with sandbox flag to prevent other dsm stuff from breaking

14 years agoadd sandbox checks for excessive memory allocation and make all exponential
adash [Wed, 10 Feb 2010 02:22:03 +0000 (02:22 +0000)]
add sandbox checks for excessive memory allocation and make all exponential
delay variables as thread locals

14 years agochange analysis so nodes are added on demand, and abstractly garbage collected when...
jjenista [Tue, 9 Feb 2010 22:49:09 +0000 (22:49 +0000)]
change analysis so nodes are added on demand, and abstractly garbage collected when not reachable from roots

14 years agobinary exponential backoff for soft abort
adash [Tue, 9 Feb 2010 20:14:52 +0000 (20:14 +0000)]
binary exponential backoff for soft abort

14 years agomissing checkin
adash [Tue, 9 Feb 2010 01:35:38 +0000 (01:35 +0000)]
missing checkin

14 years ago*** empty log message ***
hkhang [Fri, 5 Feb 2010 23:39:12 +0000 (23:39 +0000)]
*** empty log message ***

14 years ago*** empty log message ***
hkhang [Fri, 5 Feb 2010 23:16:59 +0000 (23:16 +0000)]
*** empty log message ***

14 years agoenable dsm sandboxing
adash [Fri, 5 Feb 2010 02:55:58 +0000 (02:55 +0000)]
enable dsm sandboxing

14 years agocomplete sandbox implementation
adash [Fri, 5 Feb 2010 02:50:47 +0000 (02:50 +0000)]
complete sandbox implementation

14 years agoimplementing
jjenista [Thu, 4 Feb 2010 23:57:11 +0000 (23:57 +0000)]
implementing

14 years agoadd instanceof into multicore version runtime
jzhou [Thu, 4 Feb 2010 18:39:24 +0000 (18:39 +0000)]
add instanceof into multicore version runtime

14 years agoimplementing sandbox
adash [Wed, 3 Feb 2010 02:50:59 +0000 (02:50 +0000)]
implementing sandbox
check-in all the manual prefetch changes

14 years agonew implementation for compiling conflict queues.
yeom [Wed, 3 Feb 2010 02:30:08 +0000 (02:30 +0000)]
new implementation for compiling conflict queues.

14 years ago*** empty log message ***
hkhang [Wed, 3 Feb 2010 01:26:05 +0000 (01:26 +0000)]
*** empty log message ***

14 years agoimplementing
jjenista [Tue, 2 Feb 2010 20:08:01 +0000 (20:08 +0000)]
implementing

14 years agobug fix in multicore gc
jzhou [Tue, 2 Feb 2010 17:38:37 +0000 (17:38 +0000)]
bug fix in multicore gc

14 years agofor test
jihoonl [Tue, 2 Feb 2010 01:22:32 +0000 (01:22 +0000)]
for test

14 years agochange
jihoonl [Tue, 2 Feb 2010 01:11:58 +0000 (01:11 +0000)]
change

14 years agowell..
jihoonl [Tue, 2 Feb 2010 00:35:24 +0000 (00:35 +0000)]
well..