[Discuss] Issues with new Linux computer mostly addressed

[Deryk Barker (Honorary Colleague)]

Alan, unless things have changed recently, the reset button simply asserts a reset line/signal on the bus (not that PCIe is really a bus any more).

If the OS is totally locked up and not even sampling the bus lines...

-------
Deryk Barker, Faculty Emeritus
Computer Science Department
Camosun College, Victoria BC

________________________________________
From: Discuss [discuss-bounces at vlug.org] on behalf of Alan W. Irwin [Alan.W.Irwin1234 at gmail.com]
Sent: Saturday, June 09, 2018 10:34 AM
To: discuss at vlug.org
Subject: Re: [Discuss] Issues with new Linux computer mostly addressed

After 3 days of up time, I just experienced another lock up of our new
computer for kernel version 4.16.12.  And this lock up was so harsh
that the reset button gave no response (which I don't understand
because I thought that button was pure hardware and not anything to do
with the Linux kernel or BIOS at all).  As a result, I had to
completely power down the computer, pause for several seconds and
reboot to gain control of it again.  Furthermore, my recent experience
is these lock ups go away if the AMD RX 550 graphics card is not used.
This is consistent with commentary I have just found that was in
response to
<https://www.phoronix.com/scan.php?page=news_item&px=Best-Linux-4.17-Features>.
To summarize that commentary (accessible at the end of the article),
it appears there is an overall issue of immature Linux kernel support
for several recent generations of AMD graphics cards which can cause
lock ups, and these issues will only be resolved by kernel version
4.17.x where x is large enough to work out the bugs in the
just-released 4.17.0.

Given this bad Linux situation with recent generations of AMD graphics
cards, I am frankly surprised that Michael Larabel, the Phoronix guy
who writes many articles lauding Linux support for AMD graphics is not
saying anything about how long (more than a year for the RX 550 and
longer for some older generations) it has been taking to resolve these
issues.  Anyhow, I naively trusted his positive Linux review of the RX
550 that he wrote a year ago just when that card first became
available for sale, and now I am paying the price.

I am confident that at least in the long term these issues will be
resolved since AMD builds their amdgpu-pro driver for Linux right on
top of the open-source amdgpu driver.  So it is in their direct
financial interest to make sure these issues go away.  But the
severity of this recent lock up has convinced me to go back to my
strategy of bypassing this card (using a different computer to display
results in X terminal mode) until Linux kernel support for modern AMD
graphics finally matures.

Alan
__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.sf.net); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

_______________________________________________
Discuss mailing list
Discuss at vlug.org
http://vlug.org/mailman/listinfo/discuss_vlug.org