Incorrect "out of range" values displayed with <enumeratedValue> about svd HOT 10 CLOSED

Correct error message:

❯ svdtools html ./html rk3588.svd 
Processing rk3588.svd
thread 'main' panicked at src/html/html_cli.rs:501:41:
called `Result::unwrap()` on an `Err` value: In file rk3588.svd

Caused by:
    0: In device `RockChip_RK3588`
    1: In peripheral `FSPI`
    2: In register `CTRL0`
    3: Parsing field `IDLE_CYCLE` at 104982:7
    4: `EnumeratedValue error: Value 4 out of range 0..4
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

from svd.

I would propose you to try svdtools instead. At least until you fix all SVD bugs. There a lot of them.

from svd.

Maybe I misunderstood your message, but how is "Value 4 is outside 0..4" in your post correct? The range of values is 0..3, not 0..4.

from svd.

See line 104982 in SVD. There bitWidth = 2 with EnumeratedValue s from 0 to 15.

from svd.

The range of values is 0..3, not 0..4.

Oh. I see. It just we use open range (a..b) and close range (a..=b) in Rust.

from svd.

Error message is here:

from svd.

Ah, thanks for clarifying the open/close range thing. I wasn't aware of that.

I was aware of my bitwidth bug as stated in my first post. I was just pointing out that the open range is a bit misleading, not that the parser didn't find a bug on my side. I fixed the range issue before filing this issue. As for the other bugs (_ as name), I haven't fixed those yet due to a lack of names to use.

The real crime is that ARM Development Studio (= the company that owns the svd format), is happy with the bad bit width, so if I hadn't used the other html tool I wouldn't have been any wiser.

from svd.

You could use svdtools patch after DEF to SVD convert to automatize bug patching.

from svd.

Fixed by #255

from svd.

Thanks!

from svd.