tech-pkg archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: rust and rust-bin confusion



Working on www/firefox on earmv6hf again... The first time I tried to
build it, I built lang/rust first. Of course I had to explicitly set
_MAKE_JOBS_N to 1, but it built.

It seems like huge news if lang/rust is buildable.  Can you describe
your setup?

I used to (try to) build rust on earmv7hf on an NVIDIA Tegra TK1, which is a quad core 32 bit ARM with 2 gigs of memory. It's been a while since I've done earmv7hf, though.

I have an earmv6h system on which I want to be able to run a browser if needed, if just to get past the captive portals on wifi networks that have them, but rust isn't going to build on a system with 512 megs.

I set up an older 1200 MHz Raspberry Pi 3 with a USB SSD, an earmv6hf kernel and userland, and 4 gigs of swap. Aside from "gpu_mem=16" in /boot/config.txt and stopping all the daemons that aren't needed (cron, syslogd, inetd, postfix, dhcpcd, et cetera), I just "unlimit" and "make package" in /usr/pkgsrc/lang/rust.

This last time I saw multiple rust running at the same time in spite of "MAKE_JOBS=1", so I interrupted the build and tried cpuctl offline for three of the cores, but that didn't change anything, so I then explicitly replaced ${_MAKE_JOBS_N} with 1 in Makefile because I didn't know (and still don't know) what actually works to tell rust to not assume the whole system belongs to it.

I considered compiling a uniprocessor kernel, but the build succeeded after changing Makefile. Go figure.

And indeed, rust-bin is now the default on earm where ~everyone believes
that building lang/rust is not possible.

I have great timing :)

Thanks,
John


Home | Main Index | Thread Index | Old Index