From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Date: Tue, 18 Nov 2014 22:55:25 +0100 Subject: [Buildroot] [PATCH 1/2] package/openssl: Enable parallel build In-Reply-To: <1416146446-27060-1-git-send-email-bernd.kuhls@t-online.de> References: <1416146446-27060-1-git-send-email-bernd.kuhls@t-online.de> Message-ID: <20141118225525.6e8b4e02@free-electrons.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Dear Bernd Kuhls, On Sun, 16 Nov 2014 15:00:45 +0100, Bernd Kuhls wrote: > Patch taken from Gentoo, tested 50+ times on buildroot with make -j80. > > Comparison of compile duration: > > make -j1 > real 1m24.464s > user 1m8.620s > sys 0m9.661s > > make -j80 > real 0m27.774s > user 1m34.234s > sys 0m13.441s > > Signed-off-by: Bernd Kuhls > --- > package/openssl/openssl-005-parallel-build.patch | 363 ++++++++++++++++++++++ > package/openssl/openssl.mk | 10 +- > 2 files changed, 368 insertions(+), 5 deletions(-) > create mode 100644 package/openssl/openssl-005-parallel-build.patch Unfortunately, this patch is a fairly large patch, which is more a feature patch than a real fix. Could we instead get it merged upstream? I discussed this with Yann, and I believe Peter will agree that it's a feature patch that we shouldn't take in Buildroot. The solutions are either to take this patch upstream, or in Buildroot build in parallel and then fallback to serial build if the parallel build has failed. Something along the lines of: $(MAKE) -C $(@D) || $(MAKE1) -C $(@D) (Suggested by Yann). So I'll mark this patch as rejected in patchwork. Best regards, Thomas -- Thomas Petazzoni, CTO, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com