Date: Tue, 19 Mar 2019 14:54:24 -0600 From: Dan Allen <danallen46@gmail.com> To: Ed Maste <emaste@freebsd.org> Cc: "freebsd-stable@freebsd.org Mailing FreeBSD-STABLE" <freebsd-stable@freebsd.org> Subject: Re: ifuncs check flawed? Message-ID: <046A9FEB-32C5-40E2-8E9D-1D3503101056@gmail.com> In-Reply-To: <CAPyFy2Dwxz2rgUbsVWhLOYeBZyGQQo6mc88sZ0mLdzw95m_nBQ@mail.gmail.com> References: <35A0AE5B-B25D-4067-86E3-325E65F3C1F4@gmail.com> <CAPyFy2Dwxz2rgUbsVWhLOYeBZyGQQo6mc88sZ0mLdzw95m_nBQ@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> On 19 Mar 2019, at 9:34 AM, Ed Maste <emaste@freebsd.org> wrote: >=20 > There are a few different ways you could address this: > 1. Build either the buildworld or kernel-toolchain targets before make > buildkernel, which will then use the built toolchain including lld. I did this, and it did not work. I always build the system in the = canonical order: buildworld buildkernel installkernel installworld That is why I think the check is flawed! I am using the lld linker. It = is present in the snapshot. The system should be using it, but it = immediately fails due to the aforementioned checks. That is the bug = that I am raising. Dan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?046A9FEB-32C5-40E2-8E9D-1D3503101056>