Tentative PR #36

Open

FJShen wants to merge 19 commits intoaccel-sim:devfrom

FJShen:tentative_dev

Commits on Mar 21, 2022

end_warp's calculation was not totally accurate
Fangjia Shen
committed
the way how hwtid is assigned looks subobtimal. Is this the way how NVIDIA's GPUs work?
Fangjia Shen
committed

Commits on Mar 23, 2022

added some documentation for core functions relavant to subcore scheduling
Fangjia Shen
committed

Commits on Mar 26, 2022

added debugging support for subcore modelling
Fangjia Shen
committed

Commits on Mar 28, 2022

added debugging support for subcore modelling
Fangjia Shen
committed
Merge branch 'concurrent' of github.com:FJShen/gpgpu-sim_distribution into concurrent
Fangjia Shen
committed

Commits on Mar 29, 2022

implemented RR in find_hwtid
Fangjia Shen
committed

Commits on Mar 30, 2022

added DTRACE to sucore-related code; addressed wrap-around issue
Fangjia Shen
committed
fixed coding mistake in previous commit
Fangjia Shen
committed
use range-based for loop to iterate over hwtids which might have wrap-arounds due to subcore scheduling
Fangjia Shen
committed
fixed some typos; also fixed wrong new[] call
Fangjia Shen
committed
fixed some typo
Fangjia Shen
committed

Commits on Mar 31, 2022

use functional programming to address the id wrap-around issue
Fangjia Shen
committed
include header file to support functional programming
Fangjia Shen
committed

Commits on Apr 1, 2022

added more utility functions to WrappableUnsignedRange
Fangjia Shen
committed
shader_core_ctx can now handle cases when hwtid is coming from a wrapped range of thread ids, thanks to the glamorous WrappableUnsignedRange
Fangjia Shen
committed
fixed a not-so-glamorous bug
Fangjia Shen
committed

Commits on Apr 2, 2022

remove code for a function that was added while discovering ways to make subcore round robin scheduling work, but is no longer relevant
Fangjia Shen
committed

Commits on Apr 3, 2022

disabled SUBCORE debug trace
Fangjia Shen
committed