Monthly News – February 2021

[clem]

https://blog.linuxmint.com/?p=4037

[antikythera]

A big thank you to those responsible for fixing the historic memory leak in cinnamon/nemo. It was the only major bugbear I have been persistently dealing with as I use USB drives a fair amount. My Ctrl+Alt+Esc keys salute you too

However, I wholeheartedly agree with Nick's well worded comment about the concept of limiting memory usage in Cinnamon 5.0
https://blog.linuxmint.com/?p=4037#comment-185763

[Pjotr]

Well, I think limiting Cinnamon's memory usage (automatic restart after reaching the configured max) is a pretty neat workaround.... If it can't be done the way it should, it should be done the way it can. A workaround, but an elegant one.

I'll certainly use it myself, and I'll probably recommend it on my website as well.

[antikythera]

Maybe if it Cinnamon isn't working properly it will be handy for inexperienced users to set and forget the limit but now they've patched some of the problematic processes it is consistently between 67-75 MiB on my NVIDIA machine and no longer skyrockets when after I use a USB drive with Timeshift or just nemo. The reason it's been associated with Timeshift is nemo automatically mounting USB volumes I reckon.

I'm glad it can be configured by end users or disabled (if it's enabled by default on the relevant ISO it will eventually ship on). The lower limit for the RAM setting I trust will be a sensible one.

this is top, chromium and boinc are running but as you can see RAM usage is well within acceptable limits and I have been plugging in and removing USB HDD on and off most of the day for different tasks. I did restart to check if the Plymouth update screwed anything up just under 5 hours ago but have used USB drives again since. I've not once used Ctrl+Alt+Esc since the nemo and cinnamon updates were applied.

Code: Select all

michael@GA-AM1M-S2H:~$ top

top - 00:16:39 up  4:43,  1 user,  load average: 4.40, 3.54, 3.05
Tasks: 208 total,   5 running, 203 sleeping,   0 stopped,   0 zombie
%Cpu(s):  1.4 us, 11.1 sy, 53.9 ni, 33.5 id,  0.0 wa,  0.0 hi,  0.1 si,  0.0 st
MiB Mem :  15952.6 total,  12374.1 free,   1411.3 used,   2167.2 buff/cache
MiB Swap:   2048.0 total,   2048.0 free,      0.0 used.  14168.0 avail Mem 

    PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND  
   1146 boinc     39  19  168804  73904   2572 R  64.2   0.5 135:59.03 wcgrid_+ 
  19199 boinc     39  19  167000 101504   2028 R  64.2   0.6  33:36.92 wcgrid_+ 
   1147 boinc     39  19   75252  74488   2392 R  63.9   0.5 134:42.70 wcgrid_+ 
   1145 boinc     39  19   75172  74308   2392 R  63.6   0.5 133:22.28 wcgrid_+ 
  19821 michael   20   0 5387140 126464  90012 S   3.0   0.8   1:12.31 chromium 
  20017 michael   20   0 5372780 114520  83868 S   1.7   0.7   0:09.20 chromium 
    853 root     -51   0       0      0      0 S   1.0   0.0   2:23.34 irq/44-+ 
  19759 michael   20   0 2593672 216088 146108 S   1.0   1.3   0:25.91 chromium 
    533 boinc     30  10  112252  19000  14292 S   0.7   0.1   0:56.01 boinc    
  20171 michael   20   0   11896   4240   3488 R   0.7   0.0   0:00.12 top      
  17518 root      20   0       0      0      0 I   0.3   0.0   0:11.28 kworker+ 
  19149 root      20   0       0      0      0 I   0.3   0.0   0:01.64 kworker+ 
  19688 root      20   0       0      0      0 I   0.3   0.0   0:03.04 kworker+ 
  20158 michael   20   0  463528  39656  31036 S   0.3   0.2   0:00.52 gnome-t+ 
      1 root      20   0  167508  11848   8660 S   0.0   0.1   0:02.45 systemd  
      2 root      20   0       0      0      0 S   0.0   0.0   0:00.01 kthreadd 
      3 root       0 -20       0      0      0 I   0.0   0.0   0:00.00 rcu_gp   
michael@GA-AM1M-S2H:~$ 

[DAMIEN1307]

https://www.theregister.com/2021/03/02/ ... ygobbling/

[randywatercolor]

Regarding memory leak in Cinnamon, problem exists across Intel and AMD cpus, I have seen no connection to nemo, and both machines have no discrete GPU but only on-board graphics. I have Conky reporting over 92% usage by Cinnamon of memory before crashing Cinnamon (8Gb systems). Ctrl-alt-backspace causes fallback after short delay, then a repeat reloads Cinnamon, often without the leak. Systems run fine as long as they preserve up-time by only suspending thereafter. Problems generally occur within 30min of a restart.
Looking forward to Cinnamon 5.0 with memory usage limits!