Remove Salt in code generation during build_create
#842
Conversation
|
Updated the test with this use case=) |
|
Please fix the failing UI test so that the CI runs through again: |
…tors with different signature causes compilation error on linux platform.
|
I don't have ideas, why it fails. All builds fine locally. |
…ve compilation error in CI.
|
Seems it is a planned error, but why I don't have it locally?=) |
| #[ink(storage)] | ||
| pub struct Flipper { | ||
| value: bool, | ||
| } | ||
|
|
||
| impl Flipper { | ||
| #[ink(constructor)] | ||
| pub fn new(init_value: bool) -> Self { | ||
| pub fn new2(init_value: bool) -> Self { |
There was a problem hiding this comment.
Instead of doing this I would add a new UI test instead. This way we have the bug isolated
|
The function is marked by the The question is: Why behavior for trait impl block of forwarder logic is not the same as for inherit impl block? Why read-only method can't be forwarded?(here we check that method is I think behavior for read-only functions must be the same as for mutable. But it still will not resolve the compilation issue with constructors=) I agree with idea that the constructor must be not forwarded. |
ink! currently generates code like this in order to provide the user with a compile time warning (upon link time) in case the user specified some incorrect call to a contract. This won't be happening once the new trait codegen has been merged. Until then we have to face this. However, so far it has never been a problem if used properly so I am slightly confused about this error atm. |
Is it a new trait codegen? Yes, I understand why did you add these errors=) My question is related to messages with the Maybe CI doesn't analyze code and doesn't remove the definition of forwarding calls with errors? |
|
I can't reproduce the failure on my local machine either - even with the same @xgreenx Can you try switching the Docker image used by the CI from |
xgreenx
force-pushed
the
bugfix/salt-compilation-error
branch
from
83797db
to
57f1514
Compare
Added `cc --version` in info section
|
So the CI failure for The For the time being I would opt for just merging the PR now and if it still fails on |
Co-authored-by: Michael Müller <mich@elmueller.net>
|
@xgreenx Can you apply my suggestions which revert the debugging? |
|
@cmichi https://gitlab.parity.io/parity/ink/-/jobs/1013394#L997 - the reason why it fails. |
I used the same command to run tests locally, and tests were success=) |
|
@xgreenx the failure is |
* Add `Fixed` entry for #842 * Replace `3.0.0-rc3` with `3.0.0-rc4` * Add bump allocator to release notes
build_create method supports only 2 generic parameters. But code here is passing 3 parametrs.
It causes a compilation error.