ioregs: field enumerations does not work.

[dpc]

I'm trying to get this to work:

dpc@2b537ce#diff-76fb405a406228f49ec4fac481c36bf4R46

but it produces compiler errors that make no sense.

[bharrisau]

Can you post the errors?

Or can you just make up that PR for the watchdog and we can look at the whole thing.

[bgamari]

@dpc seems like this will need a rebase.

[bgamari]

Meh, never mind; that was trivial enough.

[bgamari]

@dpc oh dear. It seems rustc has a distaste for similarly named variants. The following example demonstrates the issue,

enum Hi1 {Seq1 = 1}
enum Hi2 {Seq1 = 1}

which fails quite similarly,

test.rs:2:11: 2:19 error: duplicate definition of value `Seq1`
test.rs:2 enum Hi2 {Seq1 = 1}
                    ^~~~~~~~
test.rs:1:11: 1:19 note: first definition of value `Seq1` here
test.rs:1 enum Hi1 {Seq1 = 1}
                    ^~~~~~~~

This is quite unfortunate.

[bgamari]

For now I suspect the only solution is to ensure that enum variants have unique names. This could be rather unfortunate moving forward though. Ideally I'd like to leverage namespacing a bit better in ioregs! to ensure things like this don't happen (and clean up the documentation a bit), although this isn't possible at the moment.

[bgamari]

@dpc, I've pushed a compiling variant of your example here. Note that I had to mark std from the zinc crate as pub since app_blink_k20 is outside of this crate.

[dpc]

The problem with colliding names is something else. (but would be nice if these could have namespaces too). I've introduced it before pushing, silly me. After renaming you will get:

rake PLATFORM=k20 build_all
rustc --opt-level 2 -Z no-landing-pads --cfg mcu_k20 --cfg arch_cortex_m4 --target thumbv7em-linux-eabi -Ctarget-cpu=cortex-m4 -C relocation_model=static  -Z lto  --emit obj  -L /home/dpc/lab/rust/zinc/build  -o /home/dpc/lab/rust/zinc/build/intermediate/blink_k20/app_blink_k20.o  /home/dpc/lab/rust/zinc/apps/app_blink_k20.rs 
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:4 error: failed to resolve. Maybe a missing `extern crate std`?
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1 // vim: sw=2
                                                ^~~
note: in expansion of #[deriving]
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:1 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:2:1: 156:1 note: in expansion of ioregs!
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:38:3: 61:5 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:1 error: attempt to implement a nonexistent trait `std::num::FromPrimitive`
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1 // vim: sw=2
                                                ^
note: in expansion of #[deriving]
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:1 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:2:1: 156:1 note: in expansion of ioregs!
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:38:3: 61:5 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:4 error: failed to resolve. Maybe a missing `extern crate std`?
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1 // vim: sw=2
                                                ^~~
note: in expansion of #[deriving]
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:1 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:2:1: 156:1 note: in expansion of ioregs!
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:38:3: 61:5 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:1 error: use of undeclared type name `std::option::Option`
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1 // vim: sw=2
                                                ^
note: in expansion of #[deriving]
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:1 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:2:1: 156:1 note: in expansion of ioregs!
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:38:3: 61:5 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:48:30: 49:2 error: failed to resolve. Maybe a missing `extern crate std`?
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:48         0xa602 => RefreshSeq1,
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:49         0xb480 => RefreshSeq2,
note: in expansion of #[deriving]
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:1:1: 1:1 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:2:1: 156:1 note: in expansion of ioregs!
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:38:3: 61:5 note: expansion site
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:48:30: 48:15 error: unresolved name `std::option::Some`.
/home/dpc/lab/rust/zinc/apps/app_blink_k20.rs:48         0xa602 => RefreshSeq1,
[compier hangs herel; memory consumption increases constantly]

[dpc]

@bgamari: I've reproduced your changes and this works for me now. Thanks. It is weird what is going on here, especially the part of compiler hanging and making system swap insanely after a while.

[bgamari]

@dpc Odd indeed. I saw similar behavior around a month ago when the documentation lint would start spewing lots of allocations. That doesn't seem to be what's happening here though.

[bgamari]

@dpc are you comfortable closing this or do you suspect this issue still lurks?

[dpc]

It seems to me this compiler hang is still there in the erroneous situation
when std::option::Option is not available. I'm not sure if such behavior is
acceptable.

On Thu, Aug 28, 2014 at 3:40 AM, Ben Gamari notifications@github.com
wrote:

@dpc https://github.com/dpc are you comfortable closing this or do you
suspect this issue still lurks?

—
Reply to this email directly or view it on GitHub
#146 (comment).

Dawid Ciężarkiewicz

[bgamari]

@dpc Do you think you could try to come up with a minimal reproduction case? It might be good to bring this upstream.

[dpc]

I will try.

On Thu, Aug 28, 2014 at 12:00 PM, Ben Gamari notifications@github.com
wrote:

@dpc https://github.com/dpc Do you think you could try to come up with
a minimal reproduction case? It might be good to bring this upstream.

—
Reply to this email directly or view it on GitHub
#146 (comment).

Dawid Ciężarkiewicz

[bgamari]

Any progress on this?

[bgamari]

I'm going to close this for now. Let me know if you can produce a test case.