Michael
Michael
Currently we are unable to correctly initialize, i.e. `ARRAY[..] OF someStruct` where someStruct has aliased/reference-initialized members. We should adapt the initialization code to initialize each element of an array, ......
When trying to compile the following ``` TYPE Node : STRUCT id : DINT; child : REF_TO Node; parent : REF_TO Node; END_STRUCT END_TYPE ``` the compiler will crash with...
The issue originally reported in #1287 describes the problem for externally declared POUs, which should be fixed with #1290. However, the same issue persists for externally declared structs.
Adds target data-layout and target triple to generated IR. This ensures consistent alignment for both generated code and debug information. Note that this breaks our snapshots for windows builds: ```...
Updates the IR names of methods and actions to adhere to the scheme `__`. Also adds additional `FB_INIT` tests to ensure proper stack-variable initialization. > Can you add a similar...
**Describe the bug** When trying to assign to a dereferenced pointer which is an `RValue` (see example below) we currently do not validate against it and will panic during codegen....
should we allow shadowing parent variables with temp vars ? ``` FUNCTION_BLOCK fb VAR x: INT := 10; END_VAR END_FUNCTION_BLOCK FUNCTION_BLOCK fb2 EXTENDS fb // the following var-block will currently...
**Describe the bug** Currently, assigning any type with an underlying struct-type with a size of exactly 64 bits does not err when assigning the value to a pointer or the...
**Describe the bug** This came up during the implementation of `SUPER` but an equivalent example is also reproducible on master. When trying to dereference the parenthesized result of a `REF`...
**Describe the bug** If I have a pointer with one level of indirection and try to dereference it twice, the compilation ends in a panic in inkwell. **To Reproduce** ```...