derek/gem5
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

configs,mem-ruby: Remove old GPU ptls

Browse Source 
These protocols are no longer supported, either
because they are not representative of GPU
protocols, or because the have not been updated
to work with GCN3.

Change-Id: I989eeb6826c69225766aaab209302fe638b22719
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/34197
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Maintainer: Matt Sinclair <mattdsinclair@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This commit is contained in:
Brad Beckmann
2020-09-08 10:51:14 -04:00
committed by Matthew Poremba
parent 173c1c6eb0
commit 80221d7e1d
12 changed files with 9 additions and 8527 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
configs/example/apu_se.py
View File

@@ -236,23 +236,15 @@ shader = Shader(n_wf = options.wfs_per_simd,
voltage_domain = VoltageDomain(
voltage = options.gpu_voltage)))
# GPU_RfO(Read For Ownership) implements SC/TSO memory model.
# Other GPU protocols implement release consistency at GPU side.
# So, all GPU protocols other than GPU_RfO should make their writes
# visible to the global memory and should read from global memory
# during kernal boundary. The pipeline initiates(or do not initiate)
# the acquire/release operation depending on these impl_kern_launch_rel
# and impl_kern_end_rel flags. The flag=true means pipeline initiates
# a acquire/release operation at kernel launch/end.
# VIPER protocols (GPU_VIPER, GPU_VIPER_Region and GPU_VIPER_Baseline)
# are write-through based, and thus only imple_kern_launch_acq needs to
# set.
if buildEnv['PROTOCOL'] == 'GPU_RfO':
shader.impl_kern_launch_acq = False
shader.impl_kern_end_rel = False
elif (buildEnv['PROTOCOL'] != 'GPU_VIPER' or
buildEnv['PROTOCOL'] != 'GPU_VIPER_Region' or
buildEnv['PROTOCOL'] != 'GPU_VIPER_Baseline'):
# VIPER GPU protocol implements release consistency at GPU side. So,
# we make their writes visible to the global memory and should read
# from global memory during kernal boundary. The pipeline initiates
# (or do not initiate) the acquire/release operation depending on
# these impl_kern_launch_rel and impl_kern_end_rel flags. The flag=true
# means pipeline initiates a acquire/release operation at kernel launch/end.
# VIPER protocol is write-through based, and thus only impl_kern_launch_acq
# needs to set.
if (buildEnv['PROTOCOL'] == 'GPU_VIPER'):
shader.impl_kern_launch_acq = True
shader.impl_kern_end_rel = False
else: