Hi, > The GLSL compiler can be slowed down significantly by keeping 5x > LLVMContext in memory between compilations in radeonsi. The fact that > radeonsi can indirectly slow down the GLSL compiler (but not LLVM) is > a strong indication that we have a problem.
This will mean there will be two heap regions in virtual adress space. As high-performance malloc implementations are quite greedy releasing memory back to the kernel - wouldn't it be good to have at least some numbers regarding memory footprint before making jemalloc the default? Br, Clemens _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
