Compiling chrome-browser

[Alain De Vos]

Eating 27GB of ram & using 7GB swap space.

 

[drhowarddrfine]

And hours of your time.
Haven't you already brought this up before?

 

[Alain De Vos]

It's going in the background. So it produces just some heath. I think it will take an hour or 3.
PS1 : vscode is still failing to compile.
PS2: I have 3800 ports compiling fine from source.
PS3: obconf failing to run is annoying as i can't configure openbox.
PS4: I miss some ports like ccl,ldc2,gerbil,haxe,idris2,janet,kawa,odin,shen,v,dotnet

 

[Cath O'Deray]

poudriere bulk -j main -J 3 -Ctv www/chromium

I stopped, after around three five hours, when I realised which two ports were building. Not complaining, just observing.

% pkg info chromium | grep -i PulseAudio
        PULSEAUDIO     : off
%

I do want PulseAudio (for the observed niceness with desktop environments), I don't expect the default to change again in the near future (after on was reverted to off), I can wait for Robert Nagy's enhancement to land in main.

 

[drhowarddrfine]

I think it will take an hour or 3.

The last I read, it took someone 7 hours or so. The last time I tried this--too long ago--it was at least 10 hours.

 

[Alain De Vos]

3 hours was a bit of optimistic.
Using make.conf
OPTIONS_UNSET+=PULSE
OPTIONS_UNSET+=PULSEAUDIO

 

[Cath O'Deray]

I do want PulseAudio

OPTIONS_UNSET+=PULSE
OPTIONS_UNSET+=PULSEAUDIO

Is that not the opposite of enabling?

(Do I misampersand the word UNSET in this context?)

 

[Alain De Vos]

Unset is disabling. But i enable sndio.

 

[Alain De Vos]

It seems like the compilation of Chromium has "stalled/halted". Is something still going on ? Why no log lines ?
[Waiting for Christmas ?]

 

[Cath O'Deray]

Why no log lines ?

Tremendous patience, for the next line, is required. I can't recall how long I waited, during the last successful build, but it was a very long time.

 

[cracauer@]

Eating 27GB of ram & using 7GB swap space.

That does of course depend on how many cores you use for building. Getting too many cores and then not backing them up with RAM is not useful.

 

[Cath O'Deray]

The ZBook that I use daily recently gained an upgrade, an internal SSD alongside the HDD.

In my overexcitement, I made the mistake of creating a single partition and giving it to a ZFS pool, which – of course – I can not shrink.

Nothing irreplaceable – I could destroy, then start afresh – but I'm in no rush. Other priorities.

Spoiler: tree -d -L 4 /internalssd

% tree -d -L 4 /internalssd
/internalssd
└── var
    ├── cache
    │   └── ccache
    │       ├── 0
    │       ├── 1
    │       ├── 2
    │       ├── 3
    │       ├── 4
    │       ├── 5
    │       ├── 6
    │       ├── 7
    │       ├── 8
    │       ├── 9
    │       ├── a
    │       ├── b
    │       ├── c
    │       ├── d
    │       ├── e
    │       ├── f
    │       ├── lock
    │       └── tmp
    └── tmp

23 directories
% ls -hln /internalssd/var/tmp
total 2952
-rw-r--r--  1 1002 0  1.2K  3 May 11:47 CHECKSUM.SHA256-FreeBSD-14.1-PRERELEASE-amd64-20240503-19e335596658-267586
-rw-r--r--  1 1002 0  786M  3 May 11:55 FreeBSD-14.1-PRERELEASE-amd64-20240503-19e335596658-267586-memstick.img.xz
-rw-r--r--  1 1002 0  2.5G  2 May 13:46 GhostBSD-24.01.1.iso
%

 

[Cath O'Deray]

140amd64-default on on beefy12, timeouts after more than 48 hours for chromium and ungoogled-chromium.

rootbert

 

[Alain De Vos]

That does of course depend on how many cores you use for building. Getting too many cores and then not backing them up with RAM is not useful.

PARALLEL_JOBS=6
USE_TMPFS=yes
The Chromium build took 4 hours. Which is still reasonable.

 

[cracauer@]

Why do you use TMPFS when you say that the build drives your computer into swapping?

 

[Alain De Vos]

It speeds up the compiling process. As it's done in memory.

 

[cracauer@]

It speeds up the compiling process. As it's done in memory.

Not if it causes you to swap.