Michael Platzer

Support localparam integer array slicing

2

Hi, when initializing a localparam integer array with a slice of a larger localparam integer array, then Verilator complains that the resulting localparam is not constant. These arrays are initialized...

Internal error when assigning integer array parameter

3

Hi, when attempting to assign the values of an unpacked integer array parameter directly I get an internal error "InitArray type not assigned by AstPattern/Var visitor". Code example: ``` module...

Fix controller FSM's XIF commit tracking signals (#610)

1

Ensure that the controller FSM's XIF commit transaction tracking signals are reset every time a new instruction enters the EX stage (now ensured by simply resetting these signals anytime the...

[XIF] The controller FSM's XIF commit tracking signals sometimes fail to be reset

1

When the XIF interface is enabled, the controller FSM uses the signals `commit_valid_q` and `commit_kill_q` to track whether an XIF instruction already received its commit transaction. These signals must be...

✨ Add symmetric add operation for same source FP format add/sub

3

This PR adds a new operation `ADDS` to the `operation_e` enum, which implements a symmetric addition/subtraction, i.e., with both addends having the same source format. As discussed in #113, currently...

Currently the NONCOMP block sets the result bit when either operand of a "not equal" comparison is a quiet NaN but not for signaling NaNs. In order to treat quiet...

The NONCOMP block sets the result bit when either operand of a "not equal" comparison is a NaN: https://github.com/openhwgroup/cvfpu/blob/4aac6b3e87a30c8567dbe7401eba3274eea18afc/src/fpnew_noncomp.sv#L276-L279 However, it does so only for quiet NaNs, not for signaling...

Proposal: allow the FMA addend to be in either `src_fmt_i` or `dst_fmt_i`

5

Currently, the addend operand of the FMA shares its FP format with the result: https://github.com/openhwgroup/cvfpu/blob/4aac6b3e87a30c8567dbe7401eba3274eea18afc/src/fpnew_fma_multi.sv#L37-L38 This makes sense when considering the various fused multiply-accumulate operations of the RISC-V spec, which...

⚡️ Calculate positive and negative FMA sums in parallel

3

The FMA currently calculates the sum of the product and the addend first and, in case certain conditions are met, it then negates it. https://github.com/openhwgroup/cvfpu/blob/eba1bbebd84121a529a4696c5ce792da3013a8bd/src/fpnew_fma_multi.sv#L474-L475 However, two's complement negation requires...

This PR fixes #123 by correcting the assignment of floating-point status flags on `I2F` and `F2I` conversions, which is currently not compliant with IEEE 754-2008 as detailed in the linked...