[Letux-kernel] Timers (was Re: [PATCH 00/20] A bunch of JZ4730 fixups for letux-kernel)

H. Nikolaus Schaller hns at goldelico.com
Tue Jan 19 15:55:26 CET 2021


> Am 19.01.2021 um 13:35 schrieb H. Nikolaus Schaller <hns at goldelico.com>:
> 
> 
> Only sysbench still gets stuck (or runs forever - or at least more than 50 minutes).

I aborted after 183 minutes...
The top command did say something around 90% processor load.

What I would accept is that it runs something between 10% and 30% of the AM3358.
This did need 359.8562 seconds, i.e. 10 minutes. So 30 - 100 minutes would have been ok.
But not 183.

So the CPU or user-space commands are running pretty slow.

But: what I was now able is to start X11. Also very very slow. But works!

Reaction to mouse events is sluggish, but generally works. I only see this:

(EE) [mi] EQ overflowing.  Additional events will be discarded until existing events are processed.
(EE) 
(EE) Backtrace:
(EE) 0: /usr/bin/X (xorg_backtrace+0x84) [0x557847c4]
(EE) 
(EE) [mi] These backtraces from mieqEnqueue may point to a culprit higher up the stack.
(EE) [mi] mieq is *NOT* the cause.  It is a victim.

/proc/interrupts did not show anything problematic. Even:

root at letux:~# fgrep TCU /proc/interrupts && sleep 10 && fgrep TCU /proc/interrupts
  8:     830167       TCU   0  TCU0
  8:     831204       TCU   0  TCU0
root at letux:~# top

shows that there are just 100 interrupts per second.

Hm. Is the cache not working at all? Can we test that?

BR,
Nikolaus



More information about the Letux-kernel mailing list