[PATCH 0/5] addressing races in concurrent process startup

Artemy Kovalyov artemyko at nvidia.com
Tue Dec 12 05:25:11 CET 2023


In the process of initiating multiple processes concurrently, specifically with
automatic detection of the primary process, certain race conditions have been
identified. This patch series introduces a straightforward test that showcases
the issue and subsequently addresses the problems surfaced by the test. These
fixes aim to ensure the robust and secure utilization of DPDK within intricate
solutions that involve starting processes with job orchestrators such as Slurm
or Hadoop YARN.

Artemy Kovalyov (5):
  app/test-pm: add multiprocess test
  eal: fix multiprocess hotplug race
  ipc: fix mp channel closure to prevent message loss
  eal: fix first time primary autodetect
  eal: fix memzone fbarray cleanup

 app/meson.build                     |  1 +
 app/test-mp/main.c                  | 49 +++++++++++++++++++++++++++++++++++++
 app/test-mp/meson.build             |  8 ++++++
 app/test-mp/run.sh                  | 39 +++++++++++++++++++++++++++++
 lib/eal/common/eal_common_memzone.c | 12 +++++++++
 lib/eal/common/eal_common_proc.c    |  4 +--
 lib/eal/common/eal_private.h        |  5 ++++
 lib/eal/common/hotplug_mp.c         |  3 +++
 lib/eal/linux/eal.c                 |  3 ++-
 9 files changed, 121 insertions(+), 3 deletions(-)
 create mode 100644 app/test-mp/main.c
 create mode 100644 app/test-mp/meson.build
 create mode 100755 app/test-mp/run.sh

-- 
1.8.3.1



More information about the dev mailing list