mem: Move crossbar default latencies to subclasses

This patch introduces a few subclasses to the CoherentXBar and
NoncoherentXBar to distinguish the different uses in the system. We
use the crossbar in a wide range of places: interfacing cores to the
L2, as a system interconnect, connecting I/O and peripherals,
etc. Needless to say, these crossbars have very different performance,
and the clock frequency alone is not enough to distinguish these
scenarios.

Instead of trying to capture every possible case, this patch
introduces dedicated subclasses for the three primary use-cases:
L2XBar, SystemXBar and IOXbar. More can be added if needed, and the
defaults can be overridden.

1 files changed, 1 insertions, 1 deletions

diff --git a/configs/dram/sweep.py b/configs/dram/sweep.py
index 18a58b2da..f0b20dcc5 100644
--- a/configs/dram/sweep.py
+++ b/configs/dram/sweep.py
@@ -84,7 +84,7 @@ if args:
 # start with the system itself, using a multi-layer 1.5 GHz
 # crossbar, delivering 64 bytes / 5 cycles (one header cycle)
 # which amounts to 19.2 GByte/s per layer and thus per port
-system = System(membus = NoncoherentXBar(width = 16))
+system = System(membus = IOXBar(width = 16))
 system.clk_domain = SrcClockDomain(clock = '1.5GHz',
                                    voltage_domain =
                                    VoltageDomain(voltage = '1V'))