Re: aligned-malloc, rmdir-errno: Sanitize configuration test result

[Paul Eggert]

On 6/15/23 15:40, Bruno Haible wrote:
> I don't see that command substitution in general needs to output "text".

You're right there's nothing explicit about that. However, I doubt 
whether $(X) will work portably if X outputs arbitrary binary data. 
Although the POSIX spec for $(...) says only that X's output cannot 
contain NUL bytes, the spec for sh says that its input must be a text 
file, and arguably that requirement extends to commands like ". X" and 
"$(X)". I suspect that if we asked for clarification here, the POSIX 
folks would say that the behavior is unspecified if X outputs non-text.


> I remember the times, 20-30 years ago, when 'sed' would have problems with
> files that did not end in a newline.

It's a little tricky even today, as 'printf x | grep x' behaves 
differently on GNU platforms than on others (I just checked Solaris 10).


> using "wb" causes problems on z/OS [1].

We could surround the "b" with an appropriate ifdef. Admittedly it's 
klunky. Perhaps its better to leave it alone after all.