From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Date: Sat, 2 Sep 2017 18:43:16 +0200 Subject: [Buildroot] [PATCH 2/2] package/libpng: switch to cmake In-Reply-To: <20170902143908.GA3396@scaer> References: <20170902142956.22181-1-bernd.kuhls@t-online.de> <20170902142956.22181-2-bernd.kuhls@t-online.de> <20170902143908.GA3396@scaer> Message-ID: <20170902184316.666a6986@windsurf.lan> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Hello, On Sat, 2 Sep 2017 16:39:08 +0200, Yann E. MORIN wrote: > On 2017-09-02 16:29 +0200, Bernd Kuhls spake thusly: > > We have no arch-specific options to control mips msa and powerpc vsx > > support so we disable these options atm to prevent build errors. > > I'm not even reviewing the patch, because I noticed that a lot of > switch-to-cmake patches actually broke the build one way or the other. > > So I would suggest that we forgo switching the buildsyustem of package > just to switch. > > If upstream has changed, or there is a *compelling* reason to switch, > then so be it. In this case, the commit log *must* explain the rationale > for switching. > > But switching for the sake of switching is a big no-no from me. I agree that there needs to be an explanation for the switch. Is upstream moving to CMake, and getting rid of autotools support in the future ? Other reason ? Also Bernd: for libraries, a test-pkg test is by far not enough to indicate that everything is alright. Look at the switch of cmake in the lzo package that you contributed: it was bogus because it was only building the static library, and no longer the shared library. A test-pkg test doesn't catch this kind of stuff. Thomas -- Thomas Petazzoni, CTO, Free Electrons Embedded Linux and Kernel engineering http://free-electrons.com