From 247628b112a5e0162e5626ed6de81f28687f31ea Mon Sep 17 00:00:00 2001 From: Weiqun Zhang Date: Sun, 2 Mar 2025 15:54:09 -0800 Subject: [PATCH] Update CHANGES for 25.03 (#4358) --- CHANGES | 48 ++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+) diff --git a/CHANGES b/CHANGES index d087c7686e..77d6157f61 100644 --- a/CHANGES +++ b/CHANGES @@ -1,3 +1,51 @@ +# 25.03 + + -- Make Parser CI less sensitive to roundoff errors (#4321) + + -- FFT: Add raw pointer interfaces (#4345) + + -- Update SUNDIALS GNUMake (#4344) + + -- Change FFTW rapth flag (#4343) + + -- Copy comm flags in make_alike (#4341) + + -- GPU Mapping (#4326) + + -- Fix FillRandomNormal for odd number of length (#4338) + + -- Fix new bug in pure soa communication (#4337) + + -- Expose MLCellLinOP::setInterpBndryHalfWidth for non-hypre cases (#4334) + + -- GMRESMLMG: Support Multi-level composite solve (#4271) + + -- Fix optional component communication for pure SoA particles (#4333) + + -- SYCL: fix '-flink-huge-device-code' compiler flag (#4332) + + -- compile MLCurlCurl for 1D in cmake (#4331) + + -- GitHub Runner: ubuntu-20.04 -> ubuntu-24.04 (#4330) + + -- FFT: Support complex to complex transforms (#4329) + + -- FFT: Add batch support (#4327) + + -- Fix recursion-related nvlink warning in RandomGamma (#4324) + + -- Fix a number of tests for GPU (#4323) + + -- Fix: Names in `PC::make_alike` (#4322) + + -- Enable GPU-aware MPI by default (#4318) + + -- Preparation for deprecation of `__AMDGCN_WAVEFRONT_SIZE` (#4311) + + -- Speed up CUDA Toolkit installation in Windows CI (#4310) + + -- GNU Make: Remove old machines like summit and cori (#4312) + # 25.02 -- SpMatrix: Add constructor for CSR format (#4316)