Re: [musl] [patch] posix_fallocate.3: add note about error codes for musl.

[Rich Felker <dalias@libc.org>]

On Sun, Sep 13, 2020 at 05:06:25AM -0300, Érico Rolim wrote:
> As can be seen in
> 
> https://git.musl-libc.org/cgit/musl/tree/src/fcntl/posix_fallocate.c?id=73cc775bee53300c7cf759f37580220b18ac13d3
> 
> musl returns the syscall's errors directly, which means it doesn't
> perform the same emulation as glibc, and, more relevant to this change,
> it can return more errors than the ones currently listed in ERRORS.
> ---
>  man3/posix_fallocate.3 | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/man3/posix_fallocate.3 b/man3/posix_fallocate.3
> index 58338d673..2e440b3e1 100644
> --- a/man3/posix_fallocate.3
> +++ b/man3/posix_fallocate.3
> @@ -101,6 +101,10 @@ referred to by
>  .B ESPIPE
>  .I fd
>  refers to a pipe.
> +.TP
> +Libraries that don't perform the emulation shown in NOTES, such as musl libc,
> +may also return the error codes listed in
> +.BR fallocate (2).
>  .SH VERSIONS
>  .BR posix_fallocate ()
>  is available since glibc 2.1.94.
> -- 
> 2.28.0

I think this should be documented the other way around -- adding
EOPNOTSUPP to the list of errors, with a note that applications may
not see the error due to emulation by glibc, but should be prepared to
handle it. The emulation glibc does is highly unsafe (can corrupt the
file -- see https://sourceware.org/bugzilla/show_bug.cgi?id=15661) and
the only reason it wasn't removed was that software had come to rely
on it being there, in part due to lack of documentation otherwise. If
that's fixed, it's possible that glibc could eventually fix this bug
with a new symbol version.

Rich