Re: [Tinycc-devel] 114_bound_signal error when tcc is compiled by tcc on macOS M1, new -arch behavior

[Christian Jullien]

Thank you Grischka.

I understand your point and can perfectly leave with it.
We'll reconsider this option when:
- tcc will be able to generate native arm64 code (probably not too hard, there 
are only issues with reloc)
- when we decide to allow generation of either x86_64/arm64 with two different 
drivers (like -m32/-64 on Windows)

Have a nice day.

-----Original Message-----
From: Tinycc-devel [mailto:tinycc-devel-bounces+eligis=orange.fr@nongnu.org] On 
Behalf Of grischka
Sent: Thursday, August 05, 2021 17:40
To: jullien@eligis.com; tinycc-devel@nongnu.org
Subject: Re: [Tinycc-devel] 114_bound_signal error when tcc is compiled by tcc 
on macOS M1, new -arch behavior

Christian Jullien wrote:
> Hi folks,
>

Hi Christian,

> gcc: error: this compiler does not support arm64                      <= Sad

> Grischka, can you please reconsider this choice as it seems dangerous to
> me to ignore argument? Please note that my error message was the same as
> gcc.

I did make considerations of course and I think the conclusion was
that this is category "fragmentary feature".  The error message
may be the same as with gcc but all the rest required to make it
a feature useful for other people is absent.

There was some aspect of "unnecessarily different" too. tcc -hh
already has a list the ignored options, there was no need to add
an extra line explaining that -arch is ignored now too.

It may be arguable how user-friendly or "dangerous" tcc is with
how it ignores some options, however I do not see the argument
to handle specifically this one differently. Rsp. arguments always
exist of course but do they justify the extra effort, that is
the consideration to be made.

--- grischka

>
>
>
> C.
>
>


_______________________________________________
Tinycc-devel mailing list
Tinycc-devel@nongnu.org
https://lists.nongnu.org/mailman/listinfo/tinycc-devel