Post ADazT94APvjfxz4m0m by wolf480pl@mstdn.io
(DIR) More posts by wolf480pl@mstdn.io
(DIR) Post #ADaqKoZznqfKWYutDU by wolf480pl@mstdn.io
2021-11-20T13:16:02Z
1 likes, 0 repeats
hmm kernel build log is much less enjoyable to watch than I remembered... it doesn't scroll fast at all...
(DIR) Post #ADaqQClDKI60y4jZ4a by wolf480pl@mstdn.io
2021-11-20T13:17:03Z
0 likes, 0 repeats
I guess it's because there's no compiler warnings...
(DIR) Post #ADarlI80fFxreQ10F6 by wolf480pl@mstdn.io
2021-11-20T13:32:04Z
1 likes, 0 repeats
ok, wtf, this ryzen box is supposed to be 2x faster than my i7-4710MQ, why is clean build of Arch's kernel taking more than 20min?
(DIR) Post #ADarpEhDMp6cAGMYzY by wolf480pl@mstdn.io
2021-11-20T13:32:47Z
0 likes, 0 repeats
(well I guess kernel got bigger, and Arch enabled more features in config since the time I last did this experiment...)
(DIR) Post #ADarxgG2mEitxjgGOG by izaya@social.shadowkat.net
2021-11-20T13:34:16.062990Z
0 likes, 0 repeats
@wolf480pl disk go brr?
(DIR) Post #ADarzzWI8FgiHXNl4K by wolf480pl@mstdn.io
2021-11-20T13:34:43Z
1 likes, 0 repeats
@izaya tmpfs...
(DIR) Post #ADasJV6EwXoHj8V2S8 by wolf480pl@mstdn.io
2021-11-20T13:38:14Z
0 likes, 0 repeats
(also compiler version and flags)
(DIR) Post #ADasLyQOsrqXFy6ncO by izaya@social.shadowkat.net
2021-11-20T13:38:40.493299Z
0 likes, 0 repeats
@wolf480pl I should time it on my 4790 some timeit took more than 20 minutes last I tried so I went and did some reading for a few hours
(DIR) Post #ADasRURzF0KG1vTtZY by wolf480pl@mstdn.io
2021-11-20T13:39:41Z
0 likes, 0 repeats
@izaya I remember it taking 20min on 4710mq in 2015. Should see how much it takes on that lappy nowadays.
(DIR) Post #ADasVZj5ZFKlYEQcr2 by izaya@social.shadowkat.net
2021-11-20T13:40:26.662471Z
0 likes, 0 repeats
@wolf480pl I was building from a dying HDD last time so that probably didn't help, but this was only a few months back
(DIR) Post #ADasXLjwXTNSkjfrSS by wolf480pl@mstdn.io
2021-11-20T13:40:45Z
1 likes, 0 repeats
lol, ran out of /tmpThis puppy has 32G of RAM, but apparently tmpfs is only allowed to eat half of it. How pathetic.
(DIR) Post #ADasf8WVYucY4w6S4e by peter@mastodon.peterbabic.dev
2021-11-20T13:42:00Z
0 likes, 0 repeats
@wolf480pl is the build using all the cores btw?
(DIR) Post #ADasjvFmKIi8N3YmjQ by wolf480pl@mstdn.io
2021-11-20T13:42:57Z
0 likes, 0 repeats
@peter yup, -j12 and they're all 100%, mostly user but a bit of sys on each
(DIR) Post #ADasr1u408NL1PnAbw by wolf480pl@mstdn.io
2021-11-20T13:44:18Z
0 likes, 0 repeats
Unfortunately, this also means I won't be able to make the same test on my 4710MQ in $currentYear because it has only 16G of RAM :/
(DIR) Post #ADat7ibP0zQY5G1Nia by subleq@lonely.town
2021-11-20T13:46:48Z
0 likes, 0 repeats
@wolf480pl linux by default mounts tmpfs with size=50% regardless of ram size
(DIR) Post #ADatD6ngTThya4KuES by novenary@misskey.biribiri.dev
2021-11-20T13:46:18.099Z
0 likes, 0 repeats
@wolf480pl@mstdn.io that's pretty easy to change, mount -o remount,size=24g /tmp should do the trick
(DIR) Post #ADatD7NqJ2dEODFmam by wolf480pl@mstdn.io
2021-11-20T13:48:16Z
0 likes, 0 repeats
@novenary yeah I know, already did that, but it means I have to start building from scratch to measure the time properly.And it was already building modules...
(DIR) Post #ADatNJRELGrgfjhPcG by icedquinn@blob.cat
2021-11-20T13:50:10.304104Z
0 likes, 0 repeats
@subleq @wolf480pl i thought tmp also borrowed out of swap or something
(DIR) Post #ADataO3AVSV2jxg9Cq by wolf480pl@mstdn.io
2021-11-20T13:52:05Z
0 likes, 0 repeats
@icedquinn @subleq tmp is just anonymous pages, and all anonymous pages are backed by swap, unless there is not enough swap
(DIR) Post #ADataOdgJhhsZClJ7Q by wolf480pl@mstdn.io
2021-11-20T13:52:30Z
0 likes, 0 repeats
@icedquinn @subleq (I have zero swap)
(DIR) Post #ADazT94APvjfxz4m0m by wolf480pl@mstdn.io
2021-11-20T14:58:25Z
0 likes, 0 repeats
real 29m39,590son a Ryzen 5 1600 AFsigh.. looks like the kernel got as bigger as much as the CPUs