[Aldor-l] Fwd: Aldor Compiler

[Laurentiu Dragan]

Hi,

And welcome to the list.

Pippijn van Steenhoven wrote:

> Hello Laurentiu Dragan,
> 
> recently, I have started looking into the Aldor compiler, runtime and library.
> While reading the code I spotted several awkward areas. Specifically, I feel
> that a function that tests for equality should not change the value of its
> arguments. This is the case for "Bool symeEqual (Syme syme1, Syme syme2)" and
> "Bool tformEqual (TForm tf1, TForm tf2)".
> 
> I am interested in Aldor and compiler development but I would like to know
> what this is supposed to mean.

Some of the information is constructed lazily. In order to be able to 
compare two objects sometimes a complete evaluation must be forced.

> Hello Aldor compiler developers,
> 
> I am having issues with the Aldor compiler. When I compile runtime.as
> without optimisations, I get obscure segmentation faults within
> runtime.as (or runtime.c). When I do compile it with optimisation and
> I run the axllib testsuite, there are also several segmentation faults
> in different tests.

Did you compile runtime.as with -Wruntime and optimization?

--
Laurentiu