April 2012 - Asg - mailman3.cels.anl.gov

[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

[aesop] #26: OSX warnings during build
by aesop 12 Sep '14

12 Sep '14

#26: OSX warnings during build -------------------------------+-------------------------------------------- Reporter: rross@… | Type: defect Status: new | Priority: don't know Component: don't know | Version: Keywords: | -------------------------------+-------------------------------------------- Lots of warnings when building on my OSX box. -- Rob --- make -C src/c-utils GHC parser/ae-blocking-parser CC opcache.o CC aesop.o DEP lookup3.d DEP triton-string.d CC resource.o DEP libev/event.d In file included from ../resource.c:1: ../src/c-utils/libev/ev.h: In function ‘ev_default_loop_uc_’: ../src/c-utils/libev/ev.h:548: warning: nested extern declaration of ‘ev_default_loop_ptr’ DEP libev/ev.d ../resource.c: In function ‘ae_get_stack’: ../resource.c:738: warning: format not a string literal and no format arguments CC hints.o CC libev/ev.o CC libev/event.o CC triton-string.o CC lookup3.o CC ae-ctl.o CC ae-init.o RANLIB libc-utils.a Linking parser/ae-blocking-parser ... ld: warning: could not create compact unwind for _ffi_call_unix64: does not use RBP or RSP based frame AECC src/aefile/aefile.ae AECC src/socket/aesocket.ae AECC resources/resourcebuilder/resourcebuilder.hae AECC resources/thread/aethread.hae AECC resources/timer/timer.hae CC resources/resourcebuilder/resourcebuilder.o CC resources/thread/aethread.o CC resources/timer/timer.o In file included from ../resources/timer/timer.c:14: ../src/c-utils/libev/ev.h: In function ‘ev_default_loop_uc_’: ../src/c-utils/libev/ev.h:548: warning: nested extern declaration of ‘ev_default_loop_ptr’ /usr/include/assert.h:79: warning: previous declaration of ‘__assert_rtn’ was here /usr/include/assert.h:79: warning: previous declaration of ‘__assert_rtn’ was here ../src/aefile/aefile.hae:23: warning: previous declaration of ‘aefile_pwrite’ was here ../src/aefile/aefile.hae:24: warning: previous declaration of ‘aefile_pread’ was here ../src/aefile/aefile.hae:22: warning: previous declaration of ‘aefile_fsync’ was here ../src/aefile/aefile.hae:23: warning: previous declaration of ‘aefile_fdatasync’ was here ../src/aefile/aefile.ae: In function ‘__ae_worker_aefile_fdatasync’: ../src/aefile/aefile.ae:88: warning: nested extern declaration of ‘fdatasync’ ../src/aefile/aefile.ae: At top level: ../src/aefile/aefile.hae:25: warning: previous declaration of ‘aefile_ftruncate’ was here ../src/aefile/aefile.hae:25: warning: previous declaration of ‘aefile_unlink’ was here /usr/include/assert.h:79: warning: previous declaration of ‘__assert_rtn’ was here /usr/include/assert.h:79: warning: previous declaration of ‘__assert_rtn’ was here /usr/include/string.h:172: warning: previous declaration of ‘swab’ was here /usr/include/stdlib.h:218: warning: previous declaration of ‘mkstemp’ was here /usr/include/stdlib.h:217: warning: previous declaration of ‘mktemp’ was here /usr/include/stdlib.h:235: warning: previous declaration of ‘setkey’ was here /usr/include/stdlib.h:341: warning: previous declaration of ‘valloc’ was here /usr/include/stdlib.h:340: warning: previous declaration of ‘suboptarg’ was here /usr/include/stdlib.h:206: warning: previous declaration of ‘getsubopt’ was here /usr/include/assert.h:79: warning: previous declaration of ‘__assert_rtn’ was here /usr/include/pthread.h:400: warning: previous declaration of ‘pthread_kill’ was here /usr/include/pthread.h:406: warning: previous declaration of ‘pthread_sigmask’ was here /usr/include/unistd.h:710: warning: previous declaration of ‘mknod’ was here ../src/c-utils/libev/ev.h: In function ‘ev_default_loop_uc_’: ../src/c-utils/libev/ev.h:548: warning: nested extern declaration of ‘ev_default_loop_ptr’ ../src/socket/aesocket.ae: At top level: ../src/socket/aesocket.hae:79: warning: previous declaration of ‘aesocket_ready’ was here ../src/socket/aesocket.hae:45: warning: previous declaration of ‘aesocket_accept’ was here ../src/socket/aesocket.hae:56: warning: previous declaration of ‘aesocket_read’ was here ../src/socket/aesocket.hae:67: warning: previous declaration of ‘aesocket_write’ was here RANLIB lib/libaesop.a mcswl174:build rross$ -- Ticket URL: <http://trac.mcs.anl.gov/projects/aesop/ticket/26> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 1

[aesop] #22: overriding CFLAGS in make or configure causes build to fail.
by aesop 12 Sep '14

12 Sep '14

#22: overriding CFLAGS in make or configure causes build to fail. ---------------------------+------------------------------------------------ Reporter: dkimpe | Owner: Type: defect | Status: new Priority: minor | Component: build system Version: git repository | Keywords: ---------------------------+------------------------------------------------ Configuring aesop using CFLAGS="-O2" (i.e. ./configure CFLAGS="-O2") or issuing make CFLAGS="-O2" causes the build to fail. One reason is that it overrides the -I needed to find aesop.h. This most likely is a bug in Makefile.in and/or aecc -- Ticket URL: <http://trac.mcs.anl.gov/projects/aesop/ticket/22> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 2

[aesop] #20: Provide some debugging helper functions for aesop
by aesop 23 Oct '13

23 Oct '13

#20: Provide some debugging helper functions for aesop ---------------------------+------------------------------------------------ Reporter: dkimpe | Owner: Type: enhancement | Status: new Priority: minor | Component: aesop functions Version: git repository | Keywords: ---------------------------+------------------------------------------------ - would be nice if we had a function to poke at aesop and dump a list of what contexts are running, what function they are in, etc. - the problem right now is if you stop an aesop program in gdb you don't see anything helpful; it is usually in the event loop - could trigger with function call, signal, rpc, or whatever -- Ticket URL: <http://trac.mcs.anl.gov/projects/aesop/ticket/20> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 1

[aesop] #19: Add call to feed variable=value keys to the aesop configuration system
by aesop 27 Feb '13

27 Feb '13

#19: Add call to feed variable=value keys to the aesop configuration system ---------------------------+------------------------------------------------ Reporter: dkimpe | Owner: Type: enhancement | Status: new Priority: minor | Component: aesop functions Version: git repository | Keywords: ---------------------------+------------------------------------------------ The goal is to link zeroconf to the more limited aesop equivalent. By adding a {{{processUpdate(key,value)}}} function we can call from zeroconf, we can easily link outside systems to the internal aesop configuration. The other reason for this call is to enable aesop code to change certain system parameters (such as blocking call debugging) easily. -- Ticket URL: <http://trac.mcs.anl.gov/projects/aesop/ticket/19> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 1

[aesop] #18: Enable non-resource code to register configuration items
by aesop 27 Feb '13

27 Feb '13

#18: Enable non-resource code to register configuration items ---------------------------+------------------------------------------------ Reporter: dkimpe | Owner: Type: enhancement | Status: new Priority: minor | Component: aesop functions Version: git repository | Keywords: ---------------------------+------------------------------------------------ With the move to resourcebuilder and branchthreader, we now have non- resource code needing to access the configuration values. Expose the configuration functionality outside of the resources. -- Ticket URL: <http://trac.mcs.anl.gov/projects/aesop/ticket/18> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 2

[aesop] #33: Memory leak
by aesop 26 Feb '13

26 Feb '13

#33: Memory leak ---------------------------+------------------------------------------------ Reporter: dkimpe | Owner: Type: defect | Status: new Priority: minor | Component: code generator Version: git repository | Keywords: ---------------------------+------------------------------------------------ A simple aesop program leaks memory due to the resources not being stopped/finalized. Need to revisit once transition to explicit resource initialization is complete. {{{ ==32292== ==32292== HEAP SUMMARY: ==32292== in use at exit: 1,301 bytes in 8 blocks ==32292== total heap usage: 20 allocs, 12 frees, 84,494 bytes allocated ==32292== ==32292== 6 bytes in 1 blocks are still reachable in loss record 1 of 8 ==32292== at 0x4C28F5D: malloc (vg_replace_malloc.c:263) ==32292== by 0x5552181: strdup (in /lib64/libc-2.13.so) ==32292== by 0x403CE7: ae_resource_init_register (resource.c:122) ==32292== by 0x407EE2: aesop_timer_init_register (timer.c:241) ==32292== by 0x412465: ??? (in /home/lts/git/aesop/parser/tests/blocking/pbranch-empty) ==32292== by 0x401962: ??? (in /home/lts/git/aesop/parser/tests/blocking/pbranch-empty) ==32292== ==32292== 7 bytes in 1 blocks are still reachable in loss record 2 of 8 ==32292== at 0x4C28F5D: malloc (vg_replace_malloc.c:263) ==32292== by 0x5552181: strdup (in /lib64/libc-2.13.so) ==32292== by 0x403CE7: ae_resource_init_register (resource.c:122) ==32292== by 0x408CB8: triton_aethread_init_register (aethread.c:263) ==32292== by 0x412465: ??? (in /home/lts/git/aesop/parser/tests/blocking/pbranch-empty) ==32292== by 0x401962: ??? (in /home/lts/git/aesop/parser/tests/blocking/pbranch-empty) ==32292== ==32292== 8 bytes in 1 blocks are still reachable in loss record 3 of 8 ==32292== at 0x4C28F5D: malloc (vg_replace_malloc.c:263) ==32292== by 0x4C290A7: realloc (vg_replace_malloc.c:632) ==32292== by 0x409C7E: ev_realloc_emul (ev.c:612) ==32292== by 0x409CBC: ev_realloc (ev.c:638) ==32292== by 0x409F30: array_realloc (ev.c:860) ==32292== by 0x40A590: fd_change (ev.c:1043) ==32292== by 0x40EEFE: ev_io_start (ev.c:2674) ==32292== by 0x40AD4E: evpipe_init (ev.c:1317) ==32292== by 0x40FB53: ev_signal_start (ev.c:2920) ==32292== by 0x40DB14: ev_default_loop (ev.c:2095) ==32292== by 0x403E62: ae_resource_register_with_data (resource.c:165) ==32292== by 0x403DBA: ae_resource_register (resource.c:136) ==32292== ==32292== 48 bytes in 1 blocks are still reachable in loss record 4 of 8 ==32292== at 0x4C29042: realloc (vg_replace_malloc.c:632) ==32292== by 0x409C7E: ev_realloc_emul (ev.c:612) ==32292== by 0x409CBC: ev_realloc (ev.c:638) ==32292== by 0x409F30: array_realloc (ev.c:860) ==32292== by 0x411B0D: ev_async_start (ev.c:3689) ==32292== by 0x404052: ae_resource_register_with_data (resource.c:191) ==32292== by 0x403DBA: ae_resource_register (resource.c:136) ==32292== by 0x407F65: aesop_timer_init (timer.c:259) ==32292== by 0x4059F9: ae_resource_init_all (resource.c:978) ==32292== by 0x403625: aesop_init (aesop.c:42) ==32292== by 0x40347B: main (pbranch-empty.ae.s:2528) ==32292== ==32292== 128 bytes in 1 blocks are still reachable in loss record 5 of 8 ==32292== at 0x4C28F5D: malloc (vg_replace_malloc.c:263) ==32292== by 0x4C290A7: realloc (vg_replace_malloc.c:632) ==32292== by 0x409C7E: ev_realloc_emul (ev.c:612) ==32292== by 0x409CBC: ev_realloc (ev.c:638) ==32292== by 0x409F30: array_realloc (ev.c:860) ==32292== by 0x40EE78: ev_io_start (ev.c:2671) ==32292== by 0x40AD4E: evpipe_init (ev.c:1317) ==32292== by 0x40FB53: ev_signal_start (ev.c:2920) ==32292== by 0x40DB14: ev_default_loop (ev.c:2095) ==32292== by 0x403E62: ae_resource_register_with_data (resource.c:165) ==32292== by 0x403DBA: ae_resource_register (resource.c:136) ==32292== by 0x408D27: aethread_init (aethread.c:279) ==32292== ==32292== 168 bytes in 1 blocks are still reachable in loss record 6 of 8 ==32292== at 0x4C28F5D: malloc (vg_replace_malloc.c:263) ==32292== by 0x409804: ae_opcache_init (opcache.c:229) ==32292== by 0x408CFE: aethread_init (aethread.c:272) ==32292== by 0x4059F9: ae_resource_init_all (resource.c:978) ==32292== by 0x403625: aesop_init (aesop.c:42) ==32292== by 0x40347B: main (pbranch-empty.ae.s:2528) ==32292== ==32292== 168 bytes in 1 blocks are still reachable in loss record 7 of 8 ==32292== at 0x4C28F5D: malloc (vg_replace_malloc.c:263) ==32292== by 0x409804: ae_opcache_init (opcache.c:229) ==32292== by 0x407F46: aesop_timer_init (timer.c:253) ==32292== by 0x4059F9: ae_resource_init_all (resource.c:978) ==32292== by 0x403625: aesop_init (aesop.c:42) ==32292== by 0x40347B: main (pbranch-empty.ae.s:2528) ==32292== ==32292== 768 bytes in 1 blocks are still reachable in loss record 8 of 8 ==32292== at 0x4C28F5D: malloc (vg_replace_malloc.c:263) ==32292== by 0x4C290A7: realloc (vg_replace_malloc.c:632) ==32292== by 0x409C7E: ev_realloc_emul (ev.c:612) ==32292== by 0x409CBC: ev_realloc (ev.c:638) ==32292== by 0x40BA5C: epoll_init (ev_epoll.c:242) ==32292== by 0x40CB09: loop_init (ev.c:1752) ==32292== by 0x40DAB0: ev_default_loop (ev.c:2088) ==32292== by 0x403E62: ae_resource_register_with_data (resource.c:165) ==32292== by 0x403DBA: ae_resource_register (resource.c:136) ==32292== by 0x408D27: aethread_init (aethread.c:279) ==32292== by 0x4059F9: ae_resource_init_all (resource.c:978) ==32292== by 0x403625: aesop_init (aesop.c:42) ==32292== ==32292== LEAK SUMMARY: ==32292== definitely lost: 0 bytes in 0 blocks ==32292== indirectly lost: 0 bytes in 0 blocks ==32292== possibly lost: 0 bytes in 0 blocks ==32292== still reachable: 1,301 bytes in 8 blocks ==32292== suppressed: 0 bytes in 0 blocks ==32292== ==32292== For counts of detected and suppressed errors, rerun with: -v ==32292== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 6 from 6) }}} -- Ticket URL: <https://trac.mcs.anl.gov/projects/aesop/ticket/33> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 1

[aesop] #24: add rpc support to aesop repository
by aesop 26 Feb '13

26 Feb '13

#24: add rpc support to aesop repository ---------------------------+------------------------------------------------ Reporter: carns | Owner: Type: task | Status: new Priority: major | Component: parser Version: git repository | Keywords: ---------------------------+------------------------------------------------ This means adding the aercc translator and its network-related dependencies. Make sure that rpc/network support is optional and can be disabled at compile time. -- Ticket URL: <http://trac.mcs.anl.gov/projects/aesop/ticket/24> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 1

[aesop] #17: Enable run-time resource registration
by aesop 26 Feb '13

26 Feb '13

#17: Enable run-time resource registration ---------------------------+------------------------------------------------ Reporter: dkimpe | Owner: Type: enhancement | Status: new Priority: minor | Component: aesop functions Version: git repository | Keywords: ---------------------------+------------------------------------------------ Add support for 'runtime' resource registration. Some notes: - init and finalize might have to be blocking. - Have to take care not to call hard blocking functions (for example pthread_join) from within the init/finalize functions. This is needed to implement the functionality outlined below. ---------------------- - we have some problems with components now: - resources and non-resources are initialized differently in the aesop tree - not clear when to initialize what - have distinction between registering resource and initializing resource - consensus: - simplify things and make the following changes: - every component (even resources) has an explicit, directly callable init function - apps/servers/testprograms will explicitly call the init functions that they need - if a component depends on something else (like rosd -> tosd -> bdb) then the compent should initialize the things it needs - do reference counting in every init/finalize function so that we don't break when init is called twice - have resources "register" with aesop when you initialize them -- Ticket URL: <http://trac.mcs.anl.gov/projects/aesop/ticket/17> aesop <https://trac.mcs.anl.gov/projects/aesop> The Aesop Language

1 1

IASDS workshop
by Phil Carns 04 May '12

04 May '12

Hi guys, This is the CFP for a storage workshop we are holding in conjunction with Cluster. Sorry for the shameless plug, but it is a relevant venue for ASG work :) -Phil --------------------------------------- 4th Workshop on Interfaces and Architectures for Scientific Data Storage (IASDS) http://www.mcs.anl.gov/events/workshops/iasds12 September 24 2012, held in conjunction with IEEE Cluster 2012 in Beijing, China --------------------------------------- Paper Submission Deadline: May 25, 2012 --------------------------------------- High-performance computing simulations and large scientific experiments generate tens of terabytes of data, and these data sizes grow each year. Existing systems for storing, managing, and analyzing data are being pushed to their limits by these applications, and new techniques are necessary to enable efficient data processing for future simulations and experiments. This workshop will provide a forum for engineers and scientists to present and discuss their most recent work related to the storage, management, and analysis of data for scientific workloads. Emphasis will be placed on forward-looking approaches to tackle the challenges of storage at extreme scale or to provide better abstractions for use in scientific workloads. Topics of interest include, but are not limited to: * parallel file systems * scientific databases * active storage * scientific I/O middleware * extreme scale storage Past Workshops: IASDS 2011, Austin: http://www.mcs.anl.gov/events/workshops/iasds11/agenda.php IASDS 2010, Crete: http://www.mcs.anl.gov/events/workshops/iasds10/agenda.php IASDS 2009, New Orleans: http://www.mcs.anl.gov/events/workshops/iasds09/agenda.php Important Dates: Paper Submission Deadline: May 25, 2012 Author Notification: June 22, 2012 Final Manuscript Due: July 7, 2012 Workshop: September 24, 2012 Submission Info: See http://www.mcs.anl.gov/events/workshops/iasds12/submissions/ Workshop Chairs: Philip Carns, Mathematics and Computer Science Division, Argonne National Laboratory (carns(a)mcs.anl.gov) Osamu Tatebe, Center for Computational Sciences, University of Tsukuba (tatebe(a)cs.tsukuba.ac.jp) Program Committee: Francisco Javier Garcï¿½a Blas, Universidad Carlos III de Madrid Philip Carns, Argonne National Laboratory Hideyuki Kawashima, University of Tsukuba Hiroya Matsuba, Hitachi Carlos Maltzahn, UCSC & Ultra-scale Research Center at New Mexico Consortium Ron Oldfield, Sandia National Laboratory Yoshihiro Oyama, University of Electro-Communications Shinji Sumimoto, Fujitsu Osamu Tatebe, University of Tsukuba Andrew Uselton, Lawrence Berkeley National Laboratory Pete Wyckoff, NetApp, Inc.

3 2