DMTCP 2.5.2

@jiajuncao jiajuncao released this Nov 15, 2017 · 296 commits to master since this release

  • All fixes in Release DMTCP-2.4.9 are incorporated in this release.
  • An incompatibility of DMTCP with Open MPI 1.10 when using orterun (mpirun)
    was discovered. This does not affect recent versions, such as Open MPI 2.x.
  • In some rare cases, open files were not properly restored due to
    a use-after-free bug. This is now fixed.
  • In some rare cases, one process had created a SysV shared memory object,
    and a different process was assigned to restore it on restart. This
    was not handled correctly, and is now fixed.
  • Correctly restore CPU affinities of threads
  • Virtualized SysV shared memory keys to avoid race condition on restart
  • Fixed logic for checking if relative path to file was a duplicate
    of another existing path
  • The NSCD area for name service caching daemon was not handled correctly
    in CentOS 6.8 and later correctly. Fixed now.
  • The Linux sched.h include file for scheduling of cores was added to
    satisfy some older Linux distros that needed it for compiling DMTCP.
  • Fixed a regression in which --enable-debug (for verbose debug logs)
    was not being properly written.
  • The DMTCP coordinator was displaying a spurious warning, "Failed to find
    coordinator IP address", because it did not check for a canoncial hostname.
    A related issue prevented DMTCP from working properly on some
    SUSE/openSUSE distros.

DMTCP 2.4.9

@jiajuncao jiajuncao released this Nov 14, 2017 · 400 commits to master since this release

Version 2.4.9 release notes

  • Fixed a regression causing deleted NFS files to be handled incorrectly
  • Fixed handling of glibc for versions greater than glibc-2.24
  • Errors and warnings with gcc-7.x are fixed
  • A rare bug affecting pthread_cancel, etc., created incorrect pid on restart
  • man pages fixed: Description section was always describing dmtcp_command

DMTCP 2.5.1

@jiajuncao jiajuncao released this Sep 5, 2017 · 296 commits to master since this release

Version 2.5.1 release notes

This release mostly provides added robustness. Two notable items of
added functionality are:
i. DMTCP_RESTART_PAUSE and DMTCP_RESTART_PAUSE0 environment variables
for easier debugging upon initial restart
ii. The --debug-logs flag was added to dmtcp_launch/dmtcp_restart.
One can now turn on logging individually for separate plugins,
instead of only turning it on globally.

An incompatibility of DMTCP with Open MPI 1.10 when using orterun (mpirun)
was discovered. This may also affect some other versions of Open MPI 1.10.
This bug will be fixed in a future release.

  • Fixed an issue when starting multiple DMTCP coordinators on same host
    at approximately the same time
  • Fixed issue with PBS scheduler for HPC
  • Fixed issue when restarting on a different host with a larger
    limit on the number of open file descriptors
  • dmtcp_launch/dmtcp_restart now accept '--debug-logs' flag to specify
    which DMTCP plugins should produce logging information
    (It used to be all or nothing.)
  • Improved robustness for IB (InfiniBand) plugin
  • Fixed DMTCP_RESTART_PAUSE and DMTCP_RESTART_PAUSE0 environment variables
    for debugging upon restart
  • The brk() call was failing on restart on Debian due to overly strict assert
  • dmtcp_launch was hanging on some RHEL5 and RHEL6 due to deadlock with
    libc low-level locks. Fixed now.
  • Updated tls_pid_offset in DMTCP to handle newer GLIBc (versions > 2.24)
  • Fixed launch of 32-bit binary when forking/execing from a 64-bit executable
  • Fixed issue that can affect a parent holding a malloc-lock while forking
  • Fixed issue when a user thread calls 'dmtcp_get_coord_ckpt_dir()'

DMTCP 2.4.8

@jiajuncao jiajuncao released this Feb 13, 2017 · 400 commits to master since this release

  • The newest kernels (approximately Linux version 4.0 and later map
    VDSO into memory slightly differently for 32-bit processes running
    in a 64-bit Linux. DMTCP was failing to restart for such 32-bit processes.
    This is now fixed.
  • dmtcp_dlsym() extended to provide more robust wrapper functions.
  • Corner case fixed: signal handlers now restored before plugins restart.
  • Minor bug fixes.

DMTCP 2.5.0

@karya0 karya0 released this Jan 4, 2017 · 96 commits to 2.5 since this release

This release includes a few new plugins and several bug fixes for robustness. Some of the highlights include:

  • Support for InfiniBand UD (in addition to the more common InfiniBand RC).
  • Added support for CMA (Cross-Memory Attach):
    process_vm_readv and process_vm_writev
  • Improved multi-arch (mixed 32-/64- bit) support.
  • Re-added --enable-fast-restart.
  • Added a new commandline option --with-plugin-32 for dmtcp_launch to specify
    32-bit plugins in a 64-bit environment.
  • Added --enable-pthread-mutex-wrappers configure flag to enable
    pthread_mutex_{lock,unlock} wrappers needed for Open MPI.
  • Added ability to specify environment file used in the modify-env plugin.
  • Allow dmtcp_restart to be invoked by root.
  • The following new plugins were added:
    • pathvirt: to virtualize filesystem paths.
    • delayresume: for finer-grained control over resuming of user threads
      during resume/restart.

DMTCP 2.4.7

@karya0 karya0 released this Jan 4, 2017 · 37 commits to 2.4 since this release

If you are using a recent Linux kernel or a recent GLIBC library, and you need to continue to use the 2.4.x branch, then you are recommended to upgrade to this 2.4.7 release.

  • The newest kernels (approximately Linux version 4.0 and later) map vDSO into memory slightly differently for 32-bit processes running in a 64-bit Linux. DMTCP was failing to restart for such 32-bit processes. This is now fixed.
  • In recent versions of GLIBC, libdl/dlerror calls libc/malloc, creating an infinite recursion when we interpose using libdl/dlsym. We now implement our own dlsym and use it in most cases, thus reducing our reliance on libdl.

DMTCP 2.4.6

@karya0 karya0 released this Dec 30, 2016 · 53 commits to 2.4 since this release

  • An --enable-pthread-mutex-wrappers configure flag was added
    to enable pthread_mutex_{lock,unlock} wrappers needed for Open MPI.
  • The signals SIGCANCEL (signal# 32) and SIGSETXID (signal# 33) used by NPTL
    for threads were not being properly restored on restart.
  • Restart was fixed in a case where independent processes had orphaned roots
    of trees.
  • Provides more robust handling of mutexes.
  • Provides more robust handling of InfiniBand.
  • A failure to build on SLES10 was fixed.

DMTCP 2.4.5

@karya0 karya0 released this Jul 19, 2016 · 58 commits to 2.4 since this release

Merge pull request #413 from gc00/2.4.5

DMTCP 2.4.5

dmtcp-2.5.0-rc2

@rohgarg rohgarg released this May 22, 2016 · 171 commits to 2.5 since this release

dmtcp-2.5.0-rc2.tar.gz

Release Notes:


Unsigned downloads (Not recommended)

Please use the links at the top of the page instead.