configs: Add parameter for GPU scalar cache mandatory queue size
authorKyle Roarty <kyleroarty1716@gmail.com>
Thu, 13 Aug 2020 22:09:45 +0000 (17:09 -0500)
committerKyle Roarty <kyleroarty1716@gmail.com>
Sun, 30 Aug 2020 01:02:16 +0000 (01:02 +0000)
There was a missing option (--buffers-size) used to set the mandatory
queue size for the scalar controllers. This patch renames the option to
be more clear, and adds it to the argument parser.

Default of 128 taken from the implementation on the GCN staging branch

Change-Id: I58b6b57be07498cdf6e39c0bb85982674ec4caa6
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/32676
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Maintainer: Anthony Gutierrez <anthony.gutierrez@amd.com>
Tested-by: kokoro <noreply+kokoro@google.com>
configs/ruby/GPU_VIPER.py

index 50ccd2b259d7cf28a83fcdb3006579deadce55d7..6a6dec53ba987a54a3c32c264be5e0136ddf9de4 100644 (file)
@@ -399,6 +399,9 @@ def define_options(parser):
 
     parser.add_option("--noL1", action = "store_true", default = False,
                       help = "bypassL1")
+    parser.add_option("--scalar-buffer-size", type = 'int', default = 128,
+                      help="Size of the mandatory queue in the GPU scalar "
+                      "cache controller")
 
 def create_system(options, full_system, system, dma_devices, bootmem,
                   ruby_system):
@@ -676,7 +679,7 @@ def create_system(options, full_system, system, dma_devices, bootmem,
         scalar_cntrl.responseToSQC.slave = ruby_system.network.master
 
         scalar_cntrl.mandatoryQueue = \
-            MessageBuffer(buffer_size=options.buffers_size)
+            MessageBuffer(buffer_size=options.scalar_buffer_size)
 
         gpuCluster.add(scalar_cntrl)