Re: Fix configuration: don't use non-functional gcc-ar and gcc-ranlib

[Luca Saiu]

On 2021-02-23 at 01:36 +0100, Bruno Haible wrote:

> Luca Saiu wrote:
>> the only purpose of my
>> definition of AR and RANLIB (quite old, in fact) was to support LTO,
>
> The LTO features of the toolchain are still evolving. Possibly the
> workaround will not be needed any more anyway in a year or two?

Possibly.  However Jitter in fact already uses -fno-lto with compilers
supporting the option, which in my opinion is the right solution --
albeit philosophically contradicting the now deleted workaround.  The
machine-generated vm2.c is and will always be an unusual piece of code
where compiler optimisation needs to be carefully controlled.

-- 
Luca Saiu
* My personal web site:  http://ageinghacker.net
* Jitter:                http://ageinghacker.net/projects/jitter
* GNU epsilon:           http://www.gnu.org/software/epsilon

I support everyone's freedom of mocking any opinion or belief, no
matter how deeply held, with open disrespect and the same unrelented
enthusiasm of a toddler who has just learned the word "poo".

signature.asc
Description: PGP signature