Reputation: 3761
I'm trying to reduce a Vec
of Bundle in an UInt
, but got an error:
class DiffPair extends Bundle {
val p = Bool()
val n = Bool()
}
class TMDS extends Bundle {
val clk = new DiffPair()
val data = Vec(3, new DiffPair())
}
val io = IO(new Bundle {
//...
val tmds = Output(new TMDS())
})
val O_tmds_data_p = IO(Output(UInt(3.W)))
If I use this reduce
code :
O_tmds_data_p := gbHdmi.io.tmds.data.reduce((a,b) => Cat(a.p, b.p))
I've got this error :
[info] compiling 1 Scala source to /home/user/GbHdmi/target/scala-2.12/classes ...
[error] /home/user/GbHdmi/src/main/scala/topgbhdmi.scala:72:67: type mismatch;
[error] found : chisel3.UInt
[error] required: gbhdmi.DiffPair
[error] O_tmds_data_p := gbHdmi.io.tmds.data.reduce((a,b) => Cat(a.p, b.p))
[error] ^
[error] one error found
[error] (Compile / compileIncremental) Compilation failed
[error] Total time: 0 s, completed 31 août 2021, 15:04:08
I know that a
and b
are DiffPair
, but a.p
and b.p
are Bool
no ?
Upvotes: 1
Views: 845
Reputation: 3761
The solution I found is to take 'p' value with map
before reduce :
O_tmds_data_p := gbHdmi.io.tmds.data.map(_.p.asUInt).reduce((a,b) => Cat(a,b))
Or, in simplified way :
O_tmds_data_p := gbHdmi.io.tmds.data.map(_.p.asUInt).reduce(_ ## _)
And, that way didn't work, for an obscur reason to me :
O_tmds_data_p := gbHdmi.io.tmds.data.reduce(_.p.asUInt ## _.p.asUInt)
Error :
sbt:GbHdmi> runMain gbhdmi.TopGbHdmiDriver
[info] compiling 1 Scala source to /home/user/GbHdmi/target/scala-2.12/classes ...
[error] /home/user/GbHdmi/src/main/scala/topgbhdmi.scala:71:63: type mismatch;
[error] found : chisel3.UInt
[error] required: gbhdmi.DiffPair
[error] O_tmds_data_p := gbHdmi.io.tmds.data.reduce(_.p.asUInt ## _.p.asUInt)
[error] ^
[error] one error found
[error] (Compile / compileIncremental) Compilation failed
[error] Total time: 0 s, completed Sep 1, 2021, 9:17:15 AM
Upvotes: 1
Reputation: 4051
Have you tried
O_tmds_data_p := gbHdmi.io.tmds.data.asUInt
I think that should work for you.
Upvotes: 0