You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is not a bug report, I would like some help reading the implementation.
I would like to study the data cache and memory controller subsystem.
The aim of my research is to investigate how the order (incrementing/decrementing) of stack access sequences affects performance on a cache miss, if the cache requests data from the memory controller as a wrapped burst.
This is not a bug report, I would like some help reading the implementation.
I would like to study the data cache and memory controller subsystem.
The aim of my research is to investigate how the order (incrementing/decrementing) of stack access sequences affects performance on a cache miss, if the cache requests data from the memory controller as a wrapped burst.
For now I was only able to find this code, where the burst is defined as INCR, but the length is a single beat.
This is not really a burst, and I think this is because the request here is done by the load/store unit and not the cache.
https://github.com/chipsalliance/Cores-SweRV/blob/master/design/lsu/lsu_bus_buffer.sv#L879-L880
Could somebody help me find the cache code related to AXI4 burst accesses?
The text was updated successfully, but these errors were encountered: