Re: [regression] oops on heavy compilations ("kernel BUG at mm/zswap.c:1005!" and "Oops: invalid opcode: 0000")

[Piotr Oniszczuk <piotr.oniszczuk@gmail.com>]

> Wiadomość napisana przez Pedro Falcato <pedro.falcato@gmail.com> w dniu 25.08.2024, o godz. 17:05:
> 
> Also, could you try a memtest86 on your machine, to shake out potential hardware problems?


I found less time consuming way to trigger issue: 12c24t cross compile of llvm with „only 16G” of ram - as this triggers many heavy swappings (top swap usage gets 8-9G out of 16G swap part)

With such setup - on 6.9.12 - i’m getting not available system (due cpu soft lockup) just in 1..3h
(usually first or second compile iteration; i wrote simple scrip compiling in loop + counting interations)

Then i switched back to 6.8.2 and…. decided interrupt successful test after 14 iterations (it was 19h of non-stop compiling with heavy swapping)
(sure - i can still keep test going - but builder is also needed for my normal development)

So summarising:
-on 6.9+ just 1..3h seems enough to provoke issue
-on 6.8.2 so far never in past + currently can’t provoke issue (last day in 19h test window nor 5 full distro builds nor last 2 weeks of development)

By above i personally don’t believe issue is in hw (ram)