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
There are several common instruction sequences in RISC-V that can be optimized in the micro-architecture by "fusing" them as a single optimized implementation.
This seems a good opportunity to reduce the number of rows of powdr assembly programs translated from RISC-V. I can think of a few lui uses (also works for auipc) that could be fused:
lui A, <address> addi B, A, <address>
lui A, <address> l{w,h,b} B, <address>(A)
lui A, <address> s{w,h,b} B, <address>(A)
But certainly there are more, as this is RISC-V talked topic. Reference links:
There are several common instruction sequences in RISC-V that can be optimized in the micro-architecture by "fusing" them as a single optimized implementation.
This seems a good opportunity to reduce the number of rows of powdr assembly programs translated from RISC-V. I can think of a few
lui
uses (also works forauipc
) that could be fused:lui A, <address>
addi B, A, <address>
lui A, <address>
l{w,h,b} B, <address>(A)
lui A, <address>
s{w,h,b} B, <address>(A)
But certainly there are more, as this is RISC-V talked topic. Reference links:
The text was updated successfully, but these errors were encountered: