Re: strange compiling problems on AMD Ryzen 5 5600G with freebsd-current

On 2022-01-05 14:45, tech-lists wrote:
> On Wed, Jan 05, 2022 at 01:53:40PM -0800, Mark Millard wrote:
>> tech-lists <tech-lists_at_zyxst.net> wrote on
>> Date: Wed, 5 Jan 2022 20:14:54 +0000 :
>> 
>>> On Wed, Jan 05, 2022 at 06:55:38PM +1100, Peter Jeremy wrote:
>>> >1) Can you share more details on "fails to build".  Are you in a position
>>> >   to share a build log.
>>> 
>>> The last part of the build log from llvm13 is here:
>>> 
>>> https://cloud.zyxst.net/~john/FreeBSD/current/amd64/llvm13.log
>>> 
>>> 
>>> It's not the entire build log. i'll try making a better one
>> 
>> Does/did the console output, dmesg -a, or less /var/log/messages show any
>> messages that might be relevant from the time frame? For example:
> 
> bingo. This machine was configured (not by me!) with 2GB swap space.
> It has 16GB RAM.
> 
> this is with make -j6 and hyperthreads disabled.
> 
> Jan  5 21:04:31 r5601g kernel: swap_pager: out of swap space
> Jan  5 21:04:31 r5601g kernel: swp_pager_getswapspace(7): failed
> Jan  5 21:04:40 r5601g kernel: pid 31026 (c++), jid 0, uid 0, was killed: 
> out of swap space
> 
> I'll ask the remote end to reinstall, this time with 16GB swap.
I think the standard math for a system with this much RAM is RAM/2 (8Gb).
IOW 8G should be more than enough. It's been that way for me for ~5yrs on
a server serving ~100 hosts and all supporting services. I've also never 
encountered
swap problems building world/kernel/ports using this formula.

HTH

-- Chris
> 
> (I'm thinking maybe the defaults are set too small. The setup was
> zfs guided install, the "disk" is a WDC PC SN730 SDBQNTY-256G m2 ssd)
> 
> so, the perl problem was down to DTRACE (which might or might not be a chip 
> or a
> perl problem) and the llvm13 issue was down to swap resources.
> 
> thanks everyone for the help