https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1753401#M4770 <P>I noticed our C++ test suite got 50% slower moving from iomp 2023 -&gt; iomp 2026.<BR /><BR />We use std::thread to run code which uses OMP parallel regions.<BR />It looks like that libiomp5 does not clean up the the worker threads properly when the std thread joins. The result is that thousands of threads are being kept around during our test suite.<BR /><BR />Using the iomp library of OneAPI 2025 the expected behavior is restored (the number of observed threads is equal to the OMP_NUM_THREADS specified).</P><P>&nbsp;</P><P>A minimal reproducer and write-up can be found here:&nbsp;<A href="https://github.com/SoShiny/thread_and_omp_issue/tree/main" target="_blank">SoShiny/thread_and_omp_issue: Minimal reproducer for std::tread + OMP issue.</A><BR /><BR />Please advise.</P> Wed, 08 Jul 2026 15:07:55 GMT SoShiny 2026-07-08T15:07:55Z https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1753401#M4770 <P>I noticed our C++ test suite got 50% slower moving from iomp 2023 -&gt; iomp 2026.<BR /><BR />We use std::thread to run code which uses OMP parallel regions.<BR />It looks like that libiomp5 does not clean up the the worker threads properly when the std thread joins. The result is that thousands of threads are being kept around during our test suite.<BR /><BR />Using the iomp library of OneAPI 2025 the expected behavior is restored (the number of observed threads is equal to the OMP_NUM_THREADS specified).</P><P>&nbsp;</P><P>A minimal reproducer and write-up can be found here:&nbsp;<A href="https://github.com/SoShiny/thread_and_omp_issue/tree/main" target="_blank">SoShiny/thread_and_omp_issue: Minimal reproducer for std::tread + OMP issue.</A><BR /><BR />Please advise.</P> Wed, 08 Jul 2026 15:07:55 GMT https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1753401#M4770 SoShiny 2026-07-08T15:07:55Z https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1753433#M4772 <P>Thank you for the reproducer, we will investigate this further.&nbsp;</P> Wed, 08 Jul 2026 20:46:27 GMT https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1753433#M4772 Sravani_K_Intel 2026-07-08T20:46:27Z https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1753549#M4776 <P>I can reproduce the behavior with 2026.1 and have opened a bug report for it.</P><BR /> Fri, 10 Jul 2026 05:42:13 GMT https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1753549#M4776 Xiaoping_D_Intel 2026-07-10T05:42:13Z https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1754084#M4784 <P><a href="https://community.intel.com/t5/user/viewprofilepage/user-id/41891">@Xiaoping_D_Intel</a>&nbsp;Is there a way for me to follow the bug? Or will this thread be updated once the bug is addressed?</P> Thu, 16 Jul 2026 09:25:58 GMT https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/libiomp5-5-0-20260213-worker-thread-leak-with-std-thread/m-p/1754084#M4784 SoShiny 2026-07-16T09:25:58Z