Eric Blake wrote:
> On 03/30/2011 12:49 AM, Jim Meyering wrote:
>> coreutils' "make check" failed on rawhide due to a kernel change
>> in how readlink and readlinkat treat the empty file name.
>> Before, they'd fail with ENOENT.  Now, it's EINVAL.
>
> That's unfortunate, because it violates POSIX.

Good point.

> Is it worth filing a bug
> against glibc to avoid exposing the kernel's choice to the user?

Actually I don't *know* that this changed in the kernel.
It might have been in glibc.  I didn't look.

In any case, it'd be good to know what motivated the change,
so if no one else gets to it first, I'll track that down tomorrow.

Reply via email to