September 2014 - 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 update
by Gim Jongmin 30 Sep '14

30 Sep '14

Dear All, I apologize for the late update. Last week, I tested the metadata performance of new design, a single byte length record size, even though it was not applied directory cache yet. The performance was 30% improved against previous design if it does not apply directory cache. However, it still has pretty much bugs so that we are trying to make it stable now. 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

[aesop] #34: Initialization + declaration: invalid reorder
by aesop 30 Sep '14

30 Sep '14

#34: Initialization + declaration: invalid reorder ---------------------------+------------------------------------------------ Reporter: dkimpe | Owner: Type: defect | Status: new Priority: major | Component: code generator Version: git repository | Keywords: ---------------------------+------------------------------------------------ In the following (accepted by the compiler): {{{ ... init_a (); int my_var = call_b (); ... }}} my_var gets moved to the context structure (correct); However, it gets initialized at the beginning of the function, effectively moving {{{call_b}}} before {{{init_a}}} in the example. This can be a very hard to track down bug... Possible solution: disallow behavior in source translator or fix. -- Ticket URL: <https://trac.mcs.anl.gov/projects/aesop/ticket/34> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 2

branch, master, updated. v0.2-107-g3bb952c
by noreply＠mcs.anl.gov 30 Sep '14

30 Sep '14

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 3bb952cb6a016208af3c74863f7478f9ba220598 (commit) from 5283f381b66897dbcd37fa2761cf0e60a9132a31 (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 3bb952cb6a016208af3c74863f7478f9ba220598 Author: John Jenkins <jenkins(a)mcs.anl.gov> Date: Tue Sep 30 15:43:41 2014 +0000 test case for trac #34 and #31 (reordering of var decls / defs) Actually, the initializations happen twice as per trac #31 ----------------------------------------------------------------------- Summary of changes: toolchain/tests/parser/Makefile.subdir | 5 +++ .../parser/{pbranch-empty.ae => local-var-def.ae} | 31 +++++++------------ 2 files changed, 17 insertions(+), 19 deletions(-) copy toolchain/tests/parser/{pbranch-empty.ae => local-var-def.ae} (50%) Diff of changes: diff --git a/toolchain/tests/parser/Makefile.subdir b/toolchain/tests/parser/Makefile.subdir index e24c7f0..90046c6 100644 --- a/toolchain/tests/parser/Makefile.subdir +++ b/toolchain/tests/parser/Makefile.subdir @@ -7,6 +7,7 @@ XFAIL_TESTS += \ tests/parser/pbreak-before-blocking.sh check_PROGRAMS += \ + tests/parser/local-var-def tests/parser/local-var-forloop \ tests/parser/basic \ tests/parser/whilebreak \ @@ -75,6 +76,7 @@ check_PROGRAMS += \ tests/parser/cancel-pbranch-loop TESTS += \ + tests/parser/local-var-def tests/parser/local-var-forloop \ tests/parser/basic \ tests/parser/whilebreak \ @@ -143,6 +145,9 @@ TESTS += \ tests/parser/cancel-pbranch-loop AE_SRC += \ + tests/parser/local-var-def.ae + +AE_SRC += \ tests/parser/local-var-forloop.ae \ tests/parser/basic.ae \ tests/parser/whilebreak.ae \ diff --git a/toolchain/tests/parser/pbranch-empty.ae b/toolchain/tests/parser/local-var-def.ae similarity index 50% copy from toolchain/tests/parser/pbranch-empty.ae copy to toolchain/tests/parser/local-var-def.ae index bcad0e2..5840f35 100644 --- a/toolchain/tests/parser/pbranch-empty.ae +++ b/toolchain/tests/parser/local-var-def.ae @@ -1,37 +1,30 @@ /* - * (C) 2009 The University of Chicago + * (C) 2014 The University of Chicago * * See COPYRIGHT in top-level directory. */ -#include <assert.h> +#include <stdio.h> #include <aesop/aesop.h> +#include <errno.h> -static __blocking int pbranch_empty (void) +__blocking int aesop_main(int argc, char **argv) { + int a, b, c; - pwait - { - pbranch - { - } - - pbranch - { - } - } + a = 100; b = 200; c = 300; - return 0; -} + int d = a + b + c; -__blocking int aesop_main(int argc, char **argv) -{ - pbranch_empty (); + if (d != 600){ + fprintf(stderr, + "ERROR: unexpected result for local var decl/def via other locals\n"); + return -1; + } return 0; } aesop_main_set(aesop_main); - /* * Local variables: * c-indent-level: 4 hooks/post-receive --

1 0

Weekly Report
by Luogeng Dai 29 Sep '14

29 Sep '14

Hi All, Last week, I setup a kerberos server and worked on prototype refactoring. I written the automake part, the api and some common code for our prototype. Regards, Luogeng

1 0

Weekly report
by Guoxin Liu 29 Sep '14

29 Sep '14

Dear All, In the last week, we were using queuing theory to derive a simple model for the load balancing problem in a cluster storage system. In this study, we emphasize on minimizing the cost transferring and maximizing energy saving. Best Regards! Guoxin Liu PHD Candidate ECE, Clemson University

1 0

20140929 - Sandia Status
by Danielson, Geoffrey Charles 29 Sep '14

29 Sep '14

All, Last week we worked on integrating a small server and copy client to test performance. Matthew and Lee are both working up their slides for departmental review today, and will release our gathered numbers later. Regards, Geoff Danielson gcdanie(a)sandia.gov Geoffrey.Danielson(a)hp.com Cell: 505-263-0751

1 0

branch, master, updated. v0.2-106-g5283f38
by noreply＠mcs.anl.gov 25 Sep '14

25 Sep '14

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 5283f381b66897dbcd37fa2761cf0e60a9132a31 (commit) from 6aa55147cc76293d58f4a696065cedee9a98b3c0 (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 5283f381b66897dbcd37fa2761cf0e60a9132a31 Author: Phil Carns <carns(a)mcs.anl.gov> Date: Thu Sep 25 17:09:07 2014 -0500 clean up pkgconfig file for aesop ----------------------------------------------------------------------- Summary of changes: toolchain/configure.ac | 8 +------- toolchain/maint/aesop.pc.in | 8 ++++---- 2 files changed, 5 insertions(+), 11 deletions(-) Diff of changes: diff --git a/toolchain/configure.ac b/toolchain/configure.ac index b042139..b251754 100644 --- a/toolchain/configure.ac +++ b/toolchain/configure.ac @@ -88,7 +88,6 @@ LIBS="$C_UTILS_LIBS $LIBS" CFLAGS="$C_UTILS_CFLAGS $CFLAGS" dnl unfortunately pkg-config doesn't separate cppflags/cflags CPPFLAGS="$C_UTILS_CPPFLAGS $CPPFLAGS" -AESOP_PKGCONFIG_LIBS="${AESOP_PKGCONFIG_LIBS} $C_UTILS_LIBS" dnl ==================================================== @@ -149,10 +148,6 @@ LIBS="$OPENPA_LIBS $LIBS" CFLAGS="$OPENPA_CFLAGS $CFLAGS" dnl unfortunately pkg-config doesn't separate cppflags/cflags CPPFLAGS="$OPENPA_CPPFLAGS $CPPFLAGS" -dnl installad aesop lib will depend on opa lib, and installed aesop headers -dnl will depend on opa includes -AESOP_PKGCONFIG_LIBS="${AESOP_PKGCONFIG_LIBS} $OPENPA_LIBS" -AESOP_PKGCONFIG_CFLAGS="${AESOP_PKGCONFIG_CFLAGS} $OPENPA_CFLAGS" dnl ========================================================================== dnl check for unix variants @@ -279,9 +274,8 @@ AC_TRY_LINK( AECC_CFLAGS="" AC_SUBST(AECC_CFLAGS) -AESOP_PKGCONFIG_LIBS="${AESOP_PKGCONFIG_LIBS} ${THREAD_LIB}" +AESOP_PKGCONFIG_LIBS="${AESOP_PKGCONFIG_LIBS} -lpthread" AC_SUBST(AESOP_PKGCONFIG_LIBS) -AC_SUBST(AESOP_PKGCONFIG_CFLAGS) AC_CONFIG_FILES([bin/aecc-utils], []) AC_CONFIG_FILES([bin/aecc], [chmod +x bin/aecc]) diff --git a/toolchain/maint/aesop.pc.in b/toolchain/maint/aesop.pc.in index 99e4f3b..04fc39b 100644 --- a/toolchain/maint/aesop.pc.in +++ b/toolchain/maint/aesop.pc.in @@ -3,7 +3,6 @@ exec_prefix = @exec_prefix@ libdir=@libdir@ includedir=@includedir@ aesop_pkgconfig_libs=@AESOP_PKGCONFIG_LIBS@ -aesop_pkgconfig_cflags=@AESOP_PKGCONFIG_CFLAGS@ db_lib=@DB_LIB@ db_ldflags=@DB_LDFLAGS@ db_cflags=@DB_CFLAGS@ @@ -12,6 +11,7 @@ Name: aesop Description: language for high concurrency system software development Version: @AESOP_VERSION@ URL: http://www.mcs.anl.gov/projects/aesop -Requires: -Libs: ${db_ldflags} -L${libdir} ${db_lib} -laesop ${aesop_pkgconfig_libs} -Cflags: -I${includedir} ${db_cflags} ${aesop_pkgconfig_cflags} +Requires: c-utils openpa +Libs: -L${libdir} -laesop +Libs.private: ${db_ldflags} ${db_lib} ${aesop_pkgconfig_libs} +Cflags: -I${includedir} ${db_cflags} hooks/post-receive --

1 0

[aesop] #78: local var not lifted into __ae_ctl
by aesop 25 Sep '14

25 Sep '14

#78: local var not lifted into __ae_ctl --------------------+------------------------------------------------------- Reporter: jenkins | Owner: Type: defect | Status: new Priority: major | Component: code generator Version: | Keywords: --------------------+------------------------------------------------------- When a variable is declared within the lexical scope of a {{{for}}} loop, it is not put into the {{{__ae_ctl->params}}} structure, instead remaining a local variable. When an aesop function ({{{__blocking}}} or similar) is called and control returns in a different thread, the variable is then potentially in an uninitialized state. The test case for reproduction can be found at tests/parser/local-var- forloop.ae. The important bit is: {{{ for (i = 0; i < 1; i++){ int var = 666; do_something(); // <-- is a __blocking function if (var != 666){ fprintf(stderr, "ERROR: local var has changed (likely uninitialized)"\n); return -1; } } }}} -- Ticket URL: <https://trac.mcs.anl.gov/projects/aesop/ticket/78> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 1