This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: RFC: x86: Fall back to lazy binding for unrelocated IFUNC symbol [BZ #23240]
On 05/28/2018 05:27 AM, Florian Weimer wrote:
> On 05/27/2018 04:36 PM, H.J. Lu wrote:
>
>> I'd like to avoid these messages.
>
> We could restrict the warning to LD_DEBUG=bindings.
I agree that this would be a better solution.
Disabling BIND_NOW is the wrong thing to do in this case.
However, isn't the *real* problem that we lack correct depth-first
sorting based on relocation requirements?
Could we fix this technically in the future by improving the dynamic
loader? Keep in mind if you say "you can't solve circular dependencies"
the answer is "We can. We are allowed to break them in any way we want."
(which may still leave some applications broken at startup).
Cheers,
Carlos.