Date: Thu, 4 Jan 2018 14:40:21 -0800 From: Mark Millard <markmi@dsl-only.net> To: Freebsd-arm <freebsd-arm@freebsd.org>, FreeBSD Ports <freebsd-ports@freebsd.org> Subject: Times to build just devel/llvm50 via poudriere-devel: Pine64+ 2GB, RPi3, RPi2 V1.1 Message-ID: <65D7B16B-E3D7-40F2-BE60-0EE5E5B26B31@dsl-only.net>
next in thread | raw e-mail | index | archive | help
All the below were: NO_ZFS=3Dyes , USE_TMPFS=3Dno , not using cccache, PARALLEL_JOBS=3D1 , ALLOW_MAKE_JOBS=3Dyes . On the RPi2 V1.1 I also set: MAX_EXECUTION_TIME=3D432000 , NOHANG_TIME=3D28800 . Pine64+ 2GB: (so, 2GiBytes of RAM on cortex-a53, eMMC in usdcard slot = via adapter) [05:45:14] [01] [00:00:00] Building devel/llvm50 | llvm50-5.0.0_6 [20:02:56] [01] [14:17:42] Finished devel/llvm50 | llvm50-5.0.0_6: = Success RPi3: (1 GiByte of RAM on cortex-a53, eMMC in usdcard slot via adapter) [03:43:37] [01] [00:00:00] Building devel/llvm50 | llvm50-5.0.0_6 [22:56:56] [01] [19:13:19] Finished devel/llvm50 | llvm50-5.0.0_6: = Success RPi2 V1.1: (1 GiByte of RAM on cortex-a7, USB SSD Stick on powered hub) [04:20:51] [01] [00:00:00] Building devel/llvm50 | llvm50-5.0.0_6 [37:40:02] [01] [33:19:11] Finished devel/llvm50 | llvm50-5.0.0_6: = Success (Somewhat under 2 hr 25 min of that in package.) These were all with default options for devel/llvm50. eMCC performance notes: The rpi3 can get between 10 MiByte/s and 12 MiByte/s, while the Pine64+ 2GB can get between 5 MiBytes/s and 6 MiBytes/s, from what I have observed. Swap partition notes: All 3 had significant swap space configured. The RPi3 and RPi2 needed several hundred MiBytes, I had configured around 1.5 GiBytes. Building devel/cmake used more than building devel/llvm50 , at least on the RPi2: 973 MiBytes was observed in top for devel/cmake on the RPi2. poudriere-devel note: I had adjusted the non-parameterized, hard-coded timeouts in poudriere's scripts for the RPi2 V1.1 so that, for example, package would be allowed to finish. MAX_EXECUTION_TIME and NOHANG_TIME adjustments do not cause some stages to scale the time allowed. top note: I run a modified top that keeps track of and reports the "maximum observed used" figure for the swap usage. So that figure is a low bound on the actual maximum while top was monitoring. For reference: # uname -apKU FreeBSD rpi2 12.0-CURRENT FreeBSD 12.0-CURRENT r327485M arm armv7 = 1200054 1200054 # svnlite info /usr/ports/ | grep "Re[plv]" Relative URL: ^/head Repository Root: svn://svn0.us-west.freebsd.org/ports Repository UUID: 35697150-7ecd-e111-bb59-0022644237b5 Revision: 457579 Last Changed Rev: 457579 =3D=3D=3D Mark Millard markmi at dsl-only.net
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?65D7B16B-E3D7-40F2-BE60-0EE5E5B26B31>