charm.git
8 years agofix immediate for group broadcast
YanhuaSun [Wed, 15 May 2013 22:58:34 +0000 (17:58 -0500)]
fix immediate for group broadcast

8 years agofix the bug that immediate does not work with nodegroup broadcast
YanhuaSun [Wed, 15 May 2013 21:52:28 +0000 (16:52 -0500)]
fix the bug that immediate does not work with nodegroup broadcast

8 years agoAdding additional support for message-logging and multicast/reduction.
Esteban Meneses [Mon, 13 May 2013 21:35:33 +0000 (16:35 -0500)]
Adding additional support for message-logging and multicast/reduction.

8 years agoReplace use of non-portable log2 with bitwise operations
Phil Miller [Fri, 10 May 2013 16:44:23 +0000 (11:44 -0500)]
Replace use of non-portable log2 with bitwise operations

8 years agoadd test to reject CCE 8.1.4
Eric Bohm [Thu, 9 May 2013 20:27:33 +0000 (15:27 -0500)]
add test to reject CCE 8.1.4

8 years agodoc/fem: typo fix
Phil Miller [Thu, 9 May 2013 15:37:18 +0000 (10:37 -0500)]
doc/fem: typo fix

8 years agotests/charm++/xcastredn: Add missing TESTOPTS
Phil Miller [Wed, 8 May 2013 15:50:54 +0000 (10:50 -0500)]
tests/charm++/xcastredn: Add missing TESTOPTS

8 years agoCkLocMgr: add API to request the current host of an element, independent of forwardin...
Phil Miller [Tue, 12 Mar 2013 22:58:24 +0000 (17:58 -0500)]
CkLocMgr: add API to request the current host of an element, independent of forwarding a message

8 years agofix +CmiNoProcForComThread warning by enable it when ppn is larger than physical...
Yanhua Sun [Sat, 4 May 2013 03:46:05 +0000 (22:46 -0500)]
fix +CmiNoProcForComThread warning by enable it when ppn is larger than physical cores

8 years agoIssue #135: PUP: Convert check on negative/excessive sizes to CMK_ERROR_CHECKING
Phil Miller [Fri, 3 May 2013 15:52:30 +0000 (10:52 -0500)]
Issue #135: PUP: Convert check on negative/excessive sizes to CMK_ERROR_CHECKING

8 years agoIssue #135: Don't guard call to setMemoryTypeChare; it's an empty macro when disabled
Phil Miller [Fri, 3 May 2013 15:34:34 +0000 (10:34 -0500)]
Issue #135: Don't guard call to setMemoryTypeChare; it's an empty macro when disabled

8 years agoSolidify distinction between messages for array element initialization and afterwards
Phil Miller [Thu, 2 May 2013 22:51:10 +0000 (17:51 -0500)]
Solidify distinction between messages for array element initialization and afterwards

8 years agoSolidify distinction between messages for (Node)Group initialization and afterwards
Phil Miller [Thu, 2 May 2013 22:47:08 +0000 (17:47 -0500)]
Solidify distinction between messages for (Node)Group initialization and afterwards

8 years agoCleanup: Insert missing indentation
Phil Miller [Thu, 2 May 2013 22:43:17 +0000 (17:43 -0500)]
Cleanup: Insert missing indentation

8 years agoCleanup: Use available member function instead of digging through fields of returned...
Phil Miller [Thu, 2 May 2013 22:42:06 +0000 (17:42 -0500)]
Cleanup: Use available member function instead of digging through fields of returned value

8 years agoUpdate CharmDebug version and corresponding check for new message type
Phil Miller [Thu, 2 May 2013 18:19:50 +0000 (13:19 -0500)]
Update CharmDebug version and corresponding check for new message type

CharmDebug itself has been updated according to the associated directions.

Follow-on to issue #170.

8 years agoDefine bounds of chare arrays at construction time and use to generate IDs
Phil Miller [Wed, 1 May 2013 22:34:05 +0000 (17:34 -0500)]
Define bounds of chare arrays at construction time and use to generate IDs

Arrays constructed with an explicit size ('broadcast construction') or with
bounds specified through CkArrayOptions will now use the bounded size to
generate object IDs from element indices. Any array element being inserted or
referenced will have its index checked against the bound to ensure correctness.

Progress on issue #170, fixes #171.

8 years agoCkLocMgr: Take CkArrayOptions in ctor instead of bits and pieces thereof
Phil Miller [Wed, 1 May 2013 19:19:11 +0000 (14:19 -0500)]
CkLocMgr: Take CkArrayOptions in ctor instead of bits and pieces thereof

8 years agoconst safety for various array-support classes
Phil Miller [Wed, 1 May 2013 19:16:30 +0000 (14:16 -0500)]
const safety for various array-support classes

8 years agoDon't pass global LB group IDs from new arrays to new location managers
Phil Miller [Wed, 1 May 2013 16:54:21 +0000 (11:54 -0500)]
Don't pass global LB group IDs from new arrays to new location managers

8 years agoCkLocMgr: Drop defunct constructor
Phil Miller [Wed, 1 May 2013 16:36:01 +0000 (11:36 -0500)]
CkLocMgr: Drop defunct constructor

8 years agoobjid (issue #170): Initial support for a fixed size object identifier
Ramprasad Venkataraman [Tue, 30 Apr 2013 16:35:11 +0000 (11:35 -0500)]
objid (issue #170): Initial support for a fixed size object identifier

8 years agomodifyScheduler: factor out repeated logic
Ramprasad Venkataraman [Tue, 30 Apr 2013 17:30:51 +0000 (12:30 -0500)]
modifyScheduler: factor out repeated logic

8 years agoenvelope::pup: add missing break statement
Ramprasad Venkataraman [Tue, 30 Apr 2013 17:14:11 +0000 (12:14 -0500)]
envelope::pup: add missing break statement

Culprit was 7a7140bc

8 years agoAdd Mesh Streamer to the Charm++ libraries manual.
Lukasz Wesolowski [Sat, 27 Apr 2013 00:36:38 +0000 (19:36 -0500)]
Add Mesh Streamer to the Charm++ libraries manual.

8 years agobuild #188: Completely clobber all scaffolding that existed to setup CmiBool
Ramprasad Venkataraman [Wed, 24 Apr 2013 21:30:35 +0000 (16:30 -0500)]
build #188: Completely clobber all scaffolding that existed to setup CmiBool

Cmi{Bool,False,True} are simply defined to be {bool,false,true}.
This should stay in place until v6.6 so that we can announce deprecation and finally nuke it.

8 years agodoc #188: Replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 20:11:54 +0000 (15:11 -0500)]
doc #188: Replace Cmi{Bool,False,True} with {bool,false,true}

8 years agocmiobjid #188: Replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 20:08:50 +0000 (15:08 -0500)]
cmiobjid #188: Replace Cmi{Bool,False,True} with {bool,false,true}

8 years agoxlat #188: Replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 20:06:07 +0000 (15:06 -0500)]
xlat #188: Replace Cmi{Bool,False,True} with {bool,false,true}

8 years agolibs #188: Replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 20:05:42 +0000 (15:05 -0500)]
libs #188: Replace Cmi{Bool,False,True} with {bool,false,true}

8 years agoampi #188: Replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 19:55:31 +0000 (14:55 -0500)]
ampi #188: Replace Cmi{Bool,False,True} with {bool,false,true}

8 years agobigsim #188: Replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 19:51:32 +0000 (14:51 -0500)]
bigsim #188: Replace Cmi{Bool,False,True} with {bool,false,true}

8 years agock-ldb #188: replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 05:38:14 +0000 (00:38 -0500)]
ck-ldb #188: replace Cmi{Bool,False,True} with {bool,false,true}

8 years ago#188: fix pointless casts to bool
Ramprasad Venkataraman [Wed, 24 Apr 2013 18:47:51 +0000 (13:47 -0500)]
#188: fix pointless casts to bool

8 years agocomlib #188: more replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 05:15:28 +0000 (00:15 -0500)]
comlib #188: more replace Cmi{Bool,False,True} with {bool,false,true}

8 years agocomlib #188: replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 05:05:29 +0000 (00:05 -0500)]
comlib #188: replace Cmi{Bool,False,True} with {bool,false,true}

8 years agock-perf #188: replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 04:59:21 +0000 (23:59 -0500)]
ck-perf #188: replace Cmi{Bool,False,True} with {bool,false,true}

8 years agopup #188: replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 04:32:56 +0000 (23:32 -0500)]
pup #188: replace Cmi{Bool,False,True} with {bool,false,true}

8 years agoutil #188: replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Wed, 24 Apr 2013 04:32:05 +0000 (23:32 -0500)]
util #188: replace Cmi{Bool,False,True} with {bool,false,true}

8 years agock-core #188: replace Cmi{Bool,False,True} with {bool,false,true}
Ramprasad Venkataraman [Tue, 23 Apr 2013 18:42:17 +0000 (13:42 -0500)]
ck-core #188: replace Cmi{Bool,False,True} with {bool,false,true}

8 years agoCentralLB: don't leak stats count message (fixes #187)
Phil Miller [Tue, 23 Apr 2013 21:27:18 +0000 (16:27 -0500)]
CentralLB: don't leak stats count message (fixes #187)

(cherry picked from commit 41119526be0019304aa8ba1e4803cb6d1770e35a)

8 years agoTypo fix and reformat comment
Phil Miller [Mon, 22 Apr 2013 20:42:50 +0000 (15:42 -0500)]
Typo fix and reformat comment

8 years agoenvelope::pup(): Treat groupinit and group messages separately
Phil Miller [Mon, 22 Apr 2013 20:33:51 +0000 (15:33 -0500)]
envelope::pup(): Treat groupinit and group messages separately

8 years agoDisambiguate sqrt() call for VC++
Phil Miller [Mon, 22 Apr 2013 19:05:17 +0000 (14:05 -0500)]
Disambiguate sqrt() call for VC++

8 years agoInstall new header in include/
Phil Miller [Mon, 22 Apr 2013 19:04:01 +0000 (14:04 -0500)]
Install new header in include/

8 years agowin32: provide a junky replacement for rand48 for various examples
Phil Miller [Mon, 22 Apr 2013 17:27:34 +0000 (12:27 -0500)]
win32: provide a junky replacement for rand48 for various examples

8 years agomsa: convert CMK_OPTIMIZE to CMK_ERROR_CHECKING (issue #135)
Phil Miller [Fri, 12 Apr 2013 20:18:46 +0000 (15:18 -0500)]
msa: convert CMK_OPTIMIZE to CMK_ERROR_CHECKING (issue #135)

8 years agobuild: replace CMK_OPTIMIZE with --with-production (issue #135) and -O with -optimize
Phil Miller [Fri, 12 Apr 2013 20:05:04 +0000 (15:05 -0500)]
build: replace CMK_OPTIMIZE with --with-production (issue #135) and -O with -optimize

8 years agoampi,tcharm: Issue #135 - replace usage of archaic CMK_OPTIMIZE
Phil Miller [Fri, 12 Apr 2013 20:00:54 +0000 (15:00 -0500)]
ampi,tcharm: Issue #135 - replace usage of archaic CMK_OPTIMIZE

8 years agoAdding Jonathan's fix to race condition on restart.
Esteban Meneses [Sun, 21 Apr 2013 00:51:43 +0000 (19:51 -0500)]
Adding Jonathan's fix to race condition on restart.

8 years agoFixing silly printout.
Esteban Meneses [Fri, 19 Apr 2013 21:43:28 +0000 (16:43 -0500)]
Fixing silly printout.

8 years agoFixing a minor bug.
Esteban Meneses [Fri, 19 Apr 2013 21:05:57 +0000 (16:05 -0500)]
Fixing a minor bug.

8 years agoAdding load-balancing independent support for checkpoint.
Esteban Meneses [Fri, 19 Apr 2013 20:02:06 +0000 (15:02 -0500)]
Adding load-balancing independent support for checkpoint.

8 years agockmulticast: nuke internal class IndexPos
Ramprasad Venkataraman [Fri, 19 Apr 2013 04:05:41 +0000 (23:05 -0500)]
ckmulticast: nuke internal class IndexPos

8 years agockmulticast: Some more streamlining of the ::setup() method
Ramprasad Venkataraman [Fri, 19 Apr 2013 04:04:08 +0000 (23:04 -0500)]
ckmulticast: Some more streamlining of the ::setup() method

Avoid unnecessary struct construction / copies etc

8 years agockmulticast: Fix O(numPE) data structure and loop in tree setup
Ramprasad Venkataraman [Fri, 19 Apr 2013 02:29:50 +0000 (21:29 -0500)]
ckmulticast: Fix O(numPE) data structure and loop in tree setup

8 years agoFixing minor bug.
Esteban Meneses [Fri, 19 Apr 2013 02:51:57 +0000 (21:51 -0500)]
Fixing minor bug.

8 years agoCleaning up checkpoint code in message-logging.
Esteban Meneses [Fri, 19 Apr 2013 02:40:37 +0000 (21:40 -0500)]
Cleaning up checkpoint code in message-logging.

8 years agozlib on mpi-bgq
Yanhua Sun [Thu, 18 Apr 2013 20:34:13 +0000 (20:34 +0000)]
zlib on mpi-bgq

8 years agoMerge branch 'charm' of charmgit:charm into charm
Yanhua Sun [Thu, 18 Apr 2013 20:07:46 +0000 (20:07 +0000)]
Merge branch 'charm' of charmgit:charm into charm

8 years agofix the ZLIB path problem on BGQ
Yanhua Sun [Thu, 18 Apr 2013 20:07:40 +0000 (20:07 +0000)]
fix the ZLIB path problem on BGQ

8 years agoarridx: fix operator< to do something sane
Ramprasad Venkataraman [Thu, 18 Apr 2013 17:39:58 +0000 (12:39 -0500)]
arridx: fix operator< to do something sane

8 years agoarridx: add const qualifier to print()
Ramprasad Venkataraman [Thu, 18 Apr 2013 17:39:40 +0000 (12:39 -0500)]
arridx: add const qualifier to print()

8 years agoenvelope: Fix assert to check against sentinel value from enum
Ramprasad Venkataraman [Fri, 12 Apr 2013 17:22:53 +0000 (12:22 -0500)]
envelope: Fix assert to check against sentinel value from enum

8 years agoenvelope: update comments to reflect split between array and arrayinit structs
Ramprasad Venkataraman [Fri, 12 Apr 2013 17:08:20 +0000 (12:08 -0500)]
envelope: update comments to reflect split between array and arrayinit structs

8 years agoenvelope: Replace getsetArrayMgr() with separate get and set methods
Ramprasad Venkataraman [Fri, 12 Apr 2013 17:02:44 +0000 (12:02 -0500)]
envelope: Replace getsetArrayMgr() with separate get and set methods

8 years agoProvide a crude comparison operator< to permit CkArrayIndex in comparison-based conta...
Ramprasad Venkataraman [Tue, 16 Apr 2013 21:05:41 +0000 (16:05 -0500)]
Provide a crude comparison operator< to permit CkArrayIndex in comparison-based containers

8 years agockarrayid: Define a comparison operator (<)
Ramprasad Venkataraman [Mon, 15 Apr 2013 04:26:10 +0000 (23:26 -0500)]
ckarrayid: Define a comparison operator (<)

8 years agomsgqtest: make clean should remove ci.stamp files
Ramprasad Venkataraman [Sat, 13 Apr 2013 02:06:28 +0000 (21:06 -0500)]
msgqtest: make clean should remove ci.stamp files

8 years agomegatest: increase numElements with numPEs in migration test
Ramprasad Venkataraman [Sat, 13 Apr 2013 01:39:59 +0000 (20:39 -0500)]
megatest: increase numElements with numPEs in migration test

8 years agofix the bug in flood test for smp case
YanhuaSun [Sat, 13 Apr 2013 16:37:09 +0000 (11:37 -0500)]
fix the bug in flood test for smp case

8 years agominor fix for trace projection summary
YanhuaSun [Fri, 12 Apr 2013 22:34:16 +0000 (17:34 -0500)]
minor fix for trace projection summary

8 years agosend multiple msgs for flooding
YanhuaSun [Thu, 11 Apr 2013 21:58:43 +0000 (16:58 -0500)]
send multiple msgs for flooding

8 years agoadd flood test to get maximum bandwidth
Yanhua Sun [Thu, 11 Apr 2013 21:16:19 +0000 (21:16 +0000)]
add flood test to get maximum bandwidth

8 years agopose/Makefile: improve safety of make -j
Phil Miller [Wed, 10 Apr 2013 02:45:50 +0000 (21:45 -0500)]
pose/Makefile: improve safety of make -j

8 years agoRefs #123: Refactor headers to reduce exposure of types in charm++.h
Phil Miller [Wed, 10 Apr 2013 02:44:02 +0000 (21:44 -0500)]
Refs #123: Refactor headers to reduce exposure of types in charm++.h

8 years agoDon't export a genericly-named symbol that appears in other libraries (e.g. Cray...
Phil Miller [Tue, 9 Apr 2013 18:17:43 +0000 (13:17 -0500)]
Don't export a genericly-named symbol that appears in other libraries (e.g. Cray's netcdf)

8 years agoMake several headers more standalone and less order dependent
Phil Miller [Mon, 8 Apr 2013 20:56:34 +0000 (15:56 -0500)]
Make several headers more standalone and less order dependent

8 years agoSplit out options appended to the version argument on the command line, e.g,
Jim Phillips [Mon, 8 Apr 2013 20:02:09 +0000 (15:02 -0500)]
Split out options appended to the version argument on the command line, e.g,
treat "net-linux-smp-icc" as "net-linux smp icc".

8 years agoexamples/charm++: Add a bunch of missing TESTOPTS
Phil Miller [Sat, 6 Apr 2013 18:05:05 +0000 (13:05 -0500)]
examples/charm++: Add a bunch of missing TESTOPTS

8 years agoexamples/charm++/leanmd: shorten test run by about 10x
Phil Miller [Sat, 6 Apr 2013 17:48:37 +0000 (12:48 -0500)]
examples/charm++/leanmd: shorten test run by about 10x

8 years agoRename files from .cc to .C to appease MS VC++
Phil Miller [Sat, 6 Apr 2013 17:25:53 +0000 (12:25 -0500)]
Rename files from .cc to .C to appease MS VC++

8 years agoenvelope.h: remove two asserts that are guaranteed to be true
Lukasz Wesolowski [Fri, 5 Apr 2013 19:27:05 +0000 (14:27 -0500)]
envelope.h: remove two asserts that are guaranteed to be true

8 years agoPartial typo fix of 'destory'; references #164
Phil Miller [Fri, 5 Apr 2013 17:36:01 +0000 (12:36 -0500)]
Partial typo fix of 'destory'; references #164

8 years agoFix #160: Document SDAG refnum usage on messages
Phil Miller [Thu, 4 Apr 2013 23:26:07 +0000 (18:26 -0500)]
Fix #160: Document SDAG refnum usage on messages

8 years agoxcastredn: remove a vestigial floor call
Ramprasad Venkataraman [Thu, 4 Apr 2013 15:13:01 +0000 (10:13 -0500)]
xcastredn: remove a vestigial floor call

8 years agoexamples/charm++/leanmd: Fix mismatched new[]/delete
Phil Miller [Thu, 4 Apr 2013 14:40:22 +0000 (09:40 -0500)]
examples/charm++/leanmd: Fix mismatched new[]/delete

8 years agoadd communication statistic summary
YanhuaSun [Thu, 4 Apr 2013 01:07:22 +0000 (20:07 -0500)]
add communication statistic summary

8 years agoxcastredn: Make global variable into Cpv to separate each PE's copy
Ramprasad Venkataraman [Wed, 3 Apr 2013 22:15:02 +0000 (17:15 -0500)]
xcastredn: Make global variable into Cpv to separate each PE's copy

This should fix redmine issue#144. multicore builds exposed the issue
by having the test fail in strange ways.

The test uses a global list of 'local' (onPE) chare array element pointers to
implement delivery for converse broadcasts. Sharing this global list across
multiple PEs caused a list of all chare array elements within a charm node.
Hence a broadcast would have resulted in multiple deliveries to the same chare.
This resulted in multiple contributes to the same reduction and manifested
as reduction msgs appearing for already completed redn epochs.

8 years agockmulticast: make abort msg more informative
Ramprasad Venkataraman [Wed, 3 Apr 2013 21:40:41 +0000 (16:40 -0500)]
ckmulticast: make abort msg more informative

8 years agomake clean should remove .ci.stamp file
Ramprasad Venkataraman [Wed, 3 Apr 2013 19:39:13 +0000 (14:39 -0500)]
make clean should remove .ci.stamp file

8 years agoxcastredn: fix dependency gen, default build flags and args
Ramprasad Venkataraman [Wed, 3 Apr 2013 17:11:35 +0000 (12:11 -0500)]
xcastredn: fix dependency gen, default build flags and args

8 years agoxcastredn: dont sleep filler methods. actually do flops
Ramprasad Venkataraman [Wed, 3 Apr 2013 17:09:29 +0000 (12:09 -0500)]
xcastredn: dont sleep filler methods. actually do flops

8 years agoFixing a bug in a recently added function to resize user size of message.
Lukasz Wesolowski [Tue, 2 Apr 2013 21:22:08 +0000 (16:22 -0500)]
Fixing a bug in a recently added function to resize user size of message.
Adding another helper function that takes as parameter the number of bytes by
which to shrink the message.

NDMeshStreamer: modify the code for shrinking message to use the newly
added function. The previous shrink code was not compatible with recent
envelope reduction changes.

8 years agodocs/charm++: Update completion detection to reflect added doneProducing callback
Phil Miller [Tue, 2 Apr 2013 21:21:52 +0000 (16:21 -0500)]
docs/charm++: Update completion detection to reflect added doneProducing callback

8 years agoReference charmrun by relative path, too
Phil Miller [Tue, 2 Apr 2013 14:10:37 +0000 (09:10 -0500)]
Reference charmrun by relative path, too

8 years agoexamples/charm++: systematically ensure that all binaries are named by relative path
Phil Miller [Tue, 2 Apr 2013 13:58:33 +0000 (08:58 -0500)]
examples/charm++: systematically ensure that all binaries are named by relative path

8 years agoCleanup #74: Updating uses of initcall to initnode
Ronak Buch [Tue, 2 Apr 2013 00:24:35 +0000 (19:24 -0500)]
Cleanup #74: Updating uses of initcall to initnode

8 years agoMerge branch 'charm' of charmgit:charm into charm
YanhuaSun [Mon, 1 Apr 2013 23:19:30 +0000 (18:19 -0500)]
Merge branch 'charm' of charmgit:charm into charm

8 years agoadd +write-analysis-file for trace projections to get some automatic
YanhuaSun [Mon, 1 Apr 2013 23:18:53 +0000 (18:18 -0500)]
add +write-analysis-file for trace projections to get some automatic
summary information