Date: Sat, 20 Mar 2021 16:55:47 -0700 From: Mark Millard <marklmi@yahoo.com> To: tech-lists <tech-lists@zyxst.net> Cc: freebsd-arm@freebsd.org Subject: Re: RPI4 clock speeds and serial port Message-ID: <C0FAA127-05CE-43E5-A7D5-B5EB975E7B4F@yahoo.com> In-Reply-To: <YFaGFEXGUCigAJXD@ceres.zyxst.net> References: <20210318170053.GA26688@www.zefox.net> <9FFA0A51-C0B7-4121-95CA-B98669809007@yahoo.com> <E4CF6642-CB70-4495-A865-05469953561C@yahoo.com> <YFaGFEXGUCigAJXD@ceres.zyxst.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2021-Mar-20, at 16:32, tech-lists <tech-lists at zyxst.net> wrote: > On Thu, Mar 18, 2021 at 01:01:06PM -0700, Mark Millard via freebsd-arm = wrote: >>> . . . >>> # >>> # Use of the .clang 's here avoids >>> # interfering with other C<?>FLAGS >>> # usage, such as ?=3D usage. >>> CFLAGS.clang+=3D -mcpu=3Dcortex-a72 >>> CXXFLAGS.clang+=3D -mcpu=3Dcortex-a72 >>> CPPFLAGS.clang+=3D -mcpu=3Dcortex-a72 >>> ACFLAGS.arm64cpuid.S+=3D -mcpu=3Dcortex-a72+crypto >>> ACFLAGS.aesv8-armx.S+=3D -mcpu=3Dcortex-a72+crypto >>> ACFLAGS.ghashv8-armx.S+=3D -mcpu=3Dcortex-a72+crypto >=20 > I used these flags on main/14 no-debug and got this build time for = make > -j6 buildworld && make -j6 buildkernel: >=20 > start: Sat Mar 20 09:12:11 UTC 2021 > finish: Sat Mar 20 15:55:50 UTC 2021 To me the wording is unclear on an important point: Was this a rebuild after installing and booting the kernel and world that had been built with those flags? Or was this running the as yet not updated system and just doing the initial build with the -mcpu assignments? I'd not expect the speed-up until the rebuild stage that was done on a system already running the -mcpu assignment based build. > I think these are quite an improvement over previous build times, so > thanks for suggesting the clang options. >=20 > make cleanworld && make cleandir && make clean is run before the > buildworld/kernel. CCACHE status? > /usr/src and /usr/obj are mounted on usb3 spinning > rust. In /etc/sysctl.conf there is vfs.read_max=3D128 which I = understand > speeds up ufs2 reads. I've never tried vfs.read_max=3D128 but I also use USB3 SSD media instead of spinning rust. > I have the standard unmodified generic-nodebug kernel, and this in > /etc/src.conf >=20 > KERNCONF=3DGENERIC-NODEBUG > WITH_MALLOC_PRODUCTION=3D > WITHOUT_DEBUG_FILES=3D > WITH_CCACHE_BUILD=3D Might CCACHE use be a source of time variations for how long buildworld and buildkernel take? Was this caching effectively cleaned-out/empty for the timing? (I've never used CCACHE so I've no experience with what differences it actually makes.) > WITH_OPENSSL_KTLS=3D >=20 > WITHOUT_APM=3D > WITHOUT_ASSERT_DEBUG=3D > WITHOUT_ATM=3D > #WITHOUT_AUDIT=3D <--- got failure early in make buildworld with this > uncommented >=20 > WITHOUT_BLUETOOTH=3D > WITHOUT_CUSE=3D > WITHOUT_DMAGENT=3D > WITHOUT_FLOPPY=3D > WITHOUT_FREEBSD_UPDATE=3D > WITHOUT_HAST=3D > WITHOUT_IPFILTER=3D > WITHOUT_IPFW=3D > WITHOUT_ISCSI=3D > WITHOUT_LPR=3D > WITHOUT_NIS=3D > WITHOUT_NETGRAPH=3D > WITHOUT_OFED=3D > WITHOUT_PORTSNAP=3D > WITHOUT_PPP=3D > WITHOUT_RADIUS_SUPPORT=3D > WITHOUT_RBOOTD=3D > WITHOUT_ROUTED=3D > WITH_SORT_THREADS=3D > WITH_SVN=3D > WITHOUT_TALK=3D > WITHOUT_TFTP=3D > WITHOUT_UNBOUND=3D > WITHOUT_WIRELESS=3D =3D=3D=3D Mark Millard marklmi at yahoo.com ( dsl-only.net went away in early 2018-Mar)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?C0FAA127-05CE-43E5-A7D5-B5EB975E7B4F>