January 2015 - Asg - mailman3.cels.anl.gov

[aesop] #38: bracketed single-line if statements
by aesop 11 May '15

11 May '15

#38: bracketed single-line if statements --------------------------------+------------------------------------------- Reporter: gcdanie@… | Type: defect Status: new | Priority: don't know Component: don't know | Version: Keywords: | --------------------------------+------------------------------------------- When compiling a non-blocking function (e.g., without a __blocking identifier), should that function contain a single line if statement enclosed in brackets (as below), the aesop compiler will error (as below). if (condition) { return something; } aatree.c:79:1: error: stray ‘#’ in program aatree.c:79:7: warning: statement with no effect [-Wunused-value] aatree.c:79:10: error: expected ‘;’ before string constant aatree.c:81:31: error: ‘dir’ undeclared (first use in this function) aatree.c:81:31: note: each undeclared identifier is reported only once for each function it appears in aatree.c: In function ‘__ae_worker_treenode_traverse’: -- Ticket URL: <http://trac.mcs.anl.gov/projects/aesop/ticket/38> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

2 3

[aesop] #46: parallelfor2 test failure
by aesop 20 Mar '15

20 Mar '15

#46: parallelfor2 test failure -------------------------+-------------------------------------------------- Reporter: carns | Owner: Type: defect | Status: new Priority: minor | Component: don't know Version: devel branch | Keywords: -------------------------+-------------------------------------------------- The parser/tests/blocking/parallelfor2 test program fails with the following assertion: {{{ parallelfor2: /home/pcarns/working/aesop-devel/build/include/aesop/ae- ctl.h:229: ae_ctl_done: Assertion `refcount >= 0' failed. Aborted (core dumped) }}} -- Ticket URL: <https://trac.mcs.anl.gov/projects/aesop/ticket/46> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 3

weekly report
by jingxuan 27 Jan '15

27 Jan '15

Hi, everyone, Last week, we were debugging our simulation codes, and preparing presentation for next week. Besides, we were doing more numerical performance analysis using Petri Nets and Markov chains. Jingxuan SUN Holcombe Department of Electrical and Computer Engineering Clemson University Tel:864-643-8210 Email: jingxus(a)g.clemson.edu

1 0

Weekly report
by Guoxin Liu 26 Jan '15

26 Jan '15

Dear All, I am sorry for this late report. In the last two weeks, we were implementing a real distributed testbed on Amazon EC2 in order to measure the load balancing performance in reality, which are affected by both data server's I/O capacity and network bandwidth. Best Regards! Guoxin Liu PHD Candidate ECE, Clemson University

1 0

Weekly Update
by Gim Jongmin 26 Jan '15

26 Jan '15

Dear All, Last week, we uploaded the source code for "Batched Logging" which guarantees the transaction for related operations, i.e. link (update 'from' and 'to' simultaneously), symlink, and rename (delete 'from' and update 'to'). We checked the some use cases and could find some minor issues but it works. The minors will be updated after applying the new APIs. Sincerely, Jongmin -- Jongmin Gim Post-Doctoral Research Associate Dept. of Elecrical and Computer Enginerring Texas A&M University 333B Wisenbaker College Station, TX 77843-3259 Phone: 979-587-8311

1 0

ANL Weekly Status - 20150126
by Harms, Kevin N. 26 Jan '15

26 Jan '15

Triton: * continuing integration with HOSS (formerly SOS) * continuing integration with Clemson security prototype * working on ASG API revisions kevin

1 0

20150126 - Sandia Status
by Danielson, Geoffrey Charles 26 Jan '15

26 Jan '15

All, Last week, Sandia continued working on the HOSS, locking, and networking within Sirocco. Regards, Geoff Danielson gcdanie(a)sandia.gov Geoffrey.Danielson(a)hp.com Cell: 505-263-0751

1 0

branch, master, updated. v0.2-121-gc3101a1
by noreply＠mcs.anl.gov 21 Jan '15

21 Jan '15

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "". The branch, master has been updated via c3101a15861dbf140dded3fcf34f33340ccef7df (commit) from 2d4f309fdfbfb9d9a9f82cdb8dc06e155156ebaa (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit c3101a15861dbf140dded3fcf34f33340ccef7df Author: John Jenkins <jenkins(a)mcs.anl.gov> Date: Wed Jan 21 13:38:21 2015 -0600 put aesop cpp flags after user-provided This helps to prevent stale-file issues when aesop is installed in the same directory as other aesop-using projects, and those other projects are being (re)built. ----------------------------------------------------------------------- Summary of changes: toolchain/bin/aecc-utils.in | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) Diff of changes: diff --git a/toolchain/bin/aecc-utils.in b/toolchain/bin/aecc-utils.in index f95b0af..a86b88c 100644 --- a/toolchain/bin/aecc-utils.in +++ b/toolchain/bin/aecc-utils.in @@ -587,7 +587,7 @@ doCPP() jcppopt="-D__blocking='' -Dpbranch='' -Dpwait='' -Dpbreak=''" fi runCmd $verbose "Preprocessing step failed:" \ - "${ccpath} -DAESOP_PARSER -x c ${jcppopt} -o ${cppout} ${AECC_CFLAGS} ${ccopts} ${cppinput}" + "${ccpath} -DAESOP_PARSER -x c ${jcppopt} -o ${cppout} ${ccopts} ${AECC_CFLAGS} ${cppinput}" return $? } @@ -611,7 +611,7 @@ parseAesopSource() printVerbose "parseAesopSource(): $input" runCmd $verbose "Aesop parsing step failed: " \ - "${AEPARSER} -c ${compiler} ${popts} -r ${namedinput} -o ${output} -i ${input} -- ${AECC_CFLAGS} ${compileropts}" + "${AEPARSER} -c ${compiler} ${popts} -r ${namedinput} -o ${output} -i ${input} -- ${compileropts} ${AECC_CFLAGS}" return $? } @@ -633,7 +633,7 @@ parseAesopHeader() fi runCmd $verbose "Aesop header parsing failed: " \ - "${AEPARSER} -c ${compiler} -j ${popts} -r ${namedinput} -o ${output} -i ${input} -- ${AECC_CFLAGS} ${compileropts}" + "${AEPARSER} -c ${compiler} -j ${popts} -r ${namedinput} -o ${output} -i ${input} -- ${compileropts} ${AECC_CFLAGS}" return $? } @@ -652,7 +652,7 @@ doCompile() printVerbose "doCompile()" runCmd $verbose "Compilation step failed: " \ - "${compiler} -x c -c -o ${outfile} ${AECC_CFLAGS} ${compileropts} ${infile}" + "${compiler} -x c -c -o ${outfile} ${compileropts} ${AECC_CFLAGS} ${infile}" return $? } hooks/post-receive --

1 0

ASG API 1.2.0 available in reference branch
by Phil Carns 19 Jan '15

19 Jan '15

Here is the final 1.2.0 ASG API proposal. The only change from what I sent before is to walk back one of the asg_write() argument order changes (after looking at it closer I realized that particular change wasn't really warranted). The reference implementation and reference test harness have been updated in the "api-1.2.0" branch of the [email protected]:asg/reference repository, including extensive tests for asg_read_sequence() and brief tests for asg_read_one(). We'll be adopting this API in our development branch of Triton shortly as well. Feedback is still welcome; I'm happy to adapt the reference as needed. If there are no objections, though, I would suggest using this API revision for upcoming proposals in a) revising the API to group some related arguments into structs or b) adding transactional extensions. thanks, -Phil On 01/15/2015 04:34 PM, Carns, Philip H. wrote: > Hi all, > > Here is the new header and doxygen document that fixes the > inconsistencies noted earlier (see asg-old.h for the previous version). > > In addition to resolving the things that were listed as "todo" in the > old header, it also has the following changes: > > - corrects documentation of count arguments in > fork/object/container_info_t (per Kevin's comments) > - splits asg_read into two simpler functions (per my email earlier today) > - removes any superfluous arguments (things that can be derived directly > from other arguments) > - the argument to write, punch, and reset that is used to indicate how > many records were successfully written, punched, or reset has been moved > to the last argument position (to be more consistent with how other > output arguments are handled in the API) > > I'll work on updating the asg reference implementation in a branch that > we can merge down to master later. > > thanks, > -Phil

1 0

proposed ASG API 1.2.0
by Phil Carns 15 Jan '15

15 Jan '15

Hi all, Here is the new header and doxygen document that fixes the inconsistencies noted earlier (see asg-old.h for the previous version). In addition to resolving the things that were listed as "todo" in the old header, it also has the following changes: - corrects documentation of count arguments in fork/object/container_info_t (per Kevin's comments) - splits asg_read into two simpler functions (per my email earlier today) - removes any superfluous arguments (things that can be derived directly from other arguments) - the argument to write, punch, and reset that is used to indicate how many records were successfully written, punched, or reset has been moved to the last argument position (to be more consistent with how other output arguments are handled in the API) I'll work on updating the asg reference implementation in a branch that we can merge down to master later. thanks, -Phil

1 0