Commit History

Автор SHA1 Съобщение Дата
  Samuel Thibault 7ba4274ea3 drop spurious include преди 11 години
  Samuel Thibault c7291e5cb2 Use _starpu_callback_func_t instead of explicit function type, which bothers MSVC преди 11 години
  Samuel Thibault 1f1954c43b Add OpenCL kernel submission pipelining, to overlap costs преди 11 години
  Samuel Thibault b7efebe04d Fix task accounting when the scheduler erroneously puts a non-cuda task to a cuda device преди 11 години
  Samuel Thibault 3abd76c814 shutdown StarPU if it was successfully started but the devices we wanted are not there преди 11 години
  Samuel Thibault 86eab72d26 revert r13284 and r13286: probes happen before the TSD is allocated преди 11 години
  Samuel Thibault df782f4617 delete TSD only after fxt is really off преди 11 години
  Samuel Thibault bec10be1cd Use TSD to cache the value of gettid for FXT traces преди 11 години
  Samuel Thibault f313079113 Factorize code to prepare for OpenCL pipelining преди 11 години
  Samuel Thibault 62e0ff89af Fix missing include for windows port of locking преди 11 години
  Samuel Thibault 23636fb4f4 Fix simgrid build преди 11 години
  Samuel Thibault 77e107e4af Warn if the user uses concurrent execution on a device which doesn't support it преди 11 години
  Samuel Thibault 62194c5c86 Fix filling the pipeline преди 11 години
  Samuel Thibault f99ef13c88 Add CUDA kernel submission pipelining, to overlap costs and allow concurrent преди 11 години
  Samuel Thibault 647ccc35a7 Fix estimated allocated memory преди 11 години
  Samuel Thibault 41cfdc75cd free data outside critical section преди 11 години
  Samuel Thibault 3b766361ce free data outside critical section преди 11 години
  Samuel Thibault 206b3adf49 port r13269 from 1.1: free data outside critical section преди 11 години
  Samuel Thibault aa5c3e5c50 Let interfaces declare which transfers they allow with the can_copy methode. преди 11 години
  Samuel Thibault 1ec6838c01 Implement perfmodel locking on windows преди 11 години
  Samuel Thibault e5cf638f86 Pass codelet model when computing footprints for the bound computation преди 11 години
  Samuel Thibault b294c2f942 Fix hardcoded limitation to 64 MPI nodes in traces преди 11 години
  Samuel Thibault f8ff9c7d1c Fix build without memcpy_peer преди 11 години
  Samuel Thibault 89ece6b14f changelog преди 11 години
  Samuel Thibault 2c6e5a75b0 enable gpu-gpu transfers for matrices преди 11 години
  Samuel Thibault d76f89457d Separate bitfields set by starpu from bitfields set by user, since the latter are accessed unsafely by starpu, but the former shouldn't преди 11 години
  Samuel Thibault 2a1bb7bf49 Update 1.1.2 svn revision преди 11 години
  Samuel Thibault 14b8f531a6 Lock performance model files while writing and reading them to avoid issues on parallel launches, MPI runs notably. преди 11 години
  Nathalie Furmento 6d015be440 tests/main/{subgraph_repeat_tag.c,subgraph_repeat_regenerate_tag.c}: stop regenerating tasks when all loops are done, this permits to call starpu_data_unregister(), otherwise it would wait for the tasks преди 11 години
  Nathalie Furmento c232f22ddd examples/spmv/dw_block_spmv.c: fix following #13213 преди 11 години