Santa paid me a visit ...

[jglen490]

That one should work. The one in my service manual appears to be a 9560, which is also an M.2 2230 size.

I haven't been inside mine in a while, but I will be soon to pump up RAM and install a new 1TB M.2 drive.

[GreyGeek]

In case you are curious, here is the inside of the bottom of this HP-17cn1xxx.
The wireless chip is just below and slightly to the left of the fan. It has only a single antenna wire. The second post is unused.
Underneath all the stuff you see is the bottom of the touchpad and the keyboard.
What is not shown is the very flat battery pack, which fits between the storage device connector and the mounting stud on the lower right corner of the mobo.

[GreyGeek]

A short note about the Intel iRIS Xe GPU on this HP-17cn1xxx laptop. During the installation the GPU wasn't really seen and the i915 driver was selected along with the mesa utility for the Xe, but performance wasn't any better than the plain i915. I read on the web that the 5.13 kernel had updated support for the iRIS Xe and to get it all I had to do was install linux-oem-20,04c. After doing that the improvement in DE speed was remarkable. It is now running about twice as fast as my old GT 650M.

Oh, and changing the Linux kernel from 5.11 to f.13 didn't harm the Wi-Fi setup, described in a previous post in this thread, which continues to work like a charm. I swapped the 1 TB spinner for the 2nd Samsung EVO 860 SSD from my Acer and reinstalled KDE Neon and rebuilt my python3 dev tools. When I fired the HP up for the first time after the Neon reinstall, the Wi-Fi could see APs from both the Buffalo 600 N and the cable modem. When I tried the 2.4 GHz AP on the Buffalo it worked, BUT, all the other 2.4 GHz Buffalo 600 N APs showing in the list disappeared, but the 5 GHz APs remained. When I connected with my 5 GHz AP it also worked, but all the other 5 GHz APs from the Buffalo disappeared. When I closed the 5 GHz connection the only APs left were those from the cable modem. It's a mystery I haven't solved yet.

[claydoh]

all I had to do was install linux-oem-20,04c.

Ahh you learn something new every day. So I take it the linux-oem-foo rolls into the next HWE for Ubuntu LTS releases, if I read the docs correctly. In the past I had usually waited for the next HWE kernel to become available early, via the mostly undocumented linux-generic-hwe-20.04-edge route, if I wanted to try out a newer kernel on my laptop. Somehow I never discovered the OEM kernel track. Soooo many kernels......

fwiw, 20.04.d gets you 5.14.....

[GreyGeek]

Claydoh, I am amazed at your constant and high level source of reliable Linux knowledge. I know your quotation says you are a "search bot" but I'll wager much of your information comes from your prodigious memory.

My next problem to tackle is to find a way, if possible, for Neon to communicate to my keyboard backlight and turn it on. The lettering on the keys is so small and faint that in dim light they are impossible to read. That's because they are LEDs, or something similar, and HP is depending on them to make the keys easier to read.
I am a "touch + eyeballs" typist, meaning that I have to occasionally look for a key. Also, my fingers are trained for the Acer keyboard, and the locations of the home, delete, pg up and pg down keys are different, and I have to retrain my fingers.

I've been cruising the web with my new HP and have come to realize this HP lacks a few things, but considering its thinness, the lack is understandable.

One is the lack of a second internal SATA port. In the Acer my second SSD drive (and third SATA via the HDCADDY replacing the CD-ROM) used to store 2nd and 3rd copies of my snapshots via incremental backup. Now, I have to plug in a USB HD CADDY containing a spinner before I run my backup script. The advantage of the HP in this situation is that in an emergency I can take the USB HD CADDY with me quickly, AND, this HP as well because it is a lot easier to handle than the Acer.

The second is the presence of only two USB ports, one on each side. I converted one USB port into four by plugging in an ORICO passive USB 3.0 expander hub. One of the ports is taken up by my HP LaserJet P1606dn printer, leaving three unused and available ports. The fan out of a USB 3.0 port is 4, so if I want to have more USB ports available, I'll have to find my six port active USB 3.0 hub and replace the passive hub, or, plug it into the active hub.

All-in-all, despite the couple of unfixed quirks this HP-17cn1xxx works fantastic with KDE Neon and, I suspect, would work equally well with Kubuntu.

[claydoh]

Claydoh, I am amazed at your constant and high level source of reliable Linux knowledge. I know your quotation says you are a "search bot" but I'll wager much of your information comes from your prodigious memory.

Where does the brane stop and the search engine begin

My next problem to tackle is to find a way, if possible, for Neon to communicate to my keyboard backlight and turn it on. The lettering on the keys is so small and faint that in dim light they are impossible to read. That's because they are LEDs, or something similar, and HP is depending on them to make the keys easier to read.
I am a "touch + eyeballs" typist, meaning that I have to occasionally look for a key. Also, my fingers are trained for the Acer keyboard, and the locations of the home, delete, pg up and pg down keys are different, and I have to retrain my fingers.

You *should* have an Fn-Fkey combo for that. Mine is fn-spacebar to toggle on and the different brightness levels. Yours should have something similar, and also should be OS independent as it is a direct hardware feature.

I am sort of surprised not to spot an NVME slot in there, as it has more or less become de rigueur, even if not being used. Unless it is hiding. But it IS an HP bwahaha

[GreyGeek]

If the NVMe is hiding it is doing a very good job. In similar models it is in the tan area in the lower right corner of that photo, but in mine, as you can see, there is no mounting hardware in that position. That was the first thing I looked for because I was hoping, in the absence of a 2nd SATA port, that there would be an NVMe slot.
Such is life.

I also found that the F5 key turns on the keyboard backlights and controls their brightness. Now, when I touch a key, the lights come on, and I can see what key I'm hitting. After awhile they turn off and stay off until I touch a key. That was the last problem I had to solve and now, with the EKG wine32 software running like a champ this HP has become an awesome laptop!

$ systemd-analyze
Startup finished in 5.046s (firmware) + 3.633s (loader) + 2.514s (kernel) + 924ms (userspace) = 12.119s
graphical.target reached after 914ms in userspace

$ systemd-analyze critical-chain
The time when unit became active or started is printed after the "@" character.
The time the unit took to start is printed after the "+" character.
graphical.target @914ms
└─multi-user.target @914ms
└─ModemManager.service @845ms +69ms
└─polkit.service @750ms +93ms
└─basic.target @731ms
└─sockets.target @730ms
└─uuidd.socket @730ms
└─sysinit.target @726ms
└─systemd-timesyncd.service @663ms +62ms
└─systemd-tmpfiles-setup.service @644ms +16ms
└─local-fs.target @640ms
└─boot-efi.mount @620ms +19ms
└─systemd-fsck@dev-disk-by\x2duuid-C853\x2d95B5.service @513ms +106ms
└─dev-disk-by\x2duuid-C853\x2d95B5.device @507ms

The critical chain has minimal times, and It doesn't look like there is any service to disable that would speed up the boot time.

$ systemd-analyze blame
337ms dev-sda3.device
158ms networkd-dispatcher.service
153ms accounts-daemon.service
152ms systemd-journal-flush.service
147ms systemd-logind.service
142ms udisks2.service
126ms upower.service
117ms dev-disk-by\x2duuid-83cb8dfd\x2daadc\x2d4558\x2d884d\x2deeafac2303bf.s wap
114ms NetworkManager.service
106ms systemd-fsck@dev-disk-by\x2duuid-C853\x2d95B5.service
98ms systemd-resolved.service

So, now it's back to doing Python3/Jupyter Notebook projects and doing other research. Thanks for your help, Claydoh!

[GreyGeek]

It appears that sending snapshots to the Samsung 860 EVO 500GB via the passive USB hub resulted in the hub failure. Just a guess but I suspect that the current drain exceeded the fanout limit and a component failed. It's in the trash can and an Active 6 port 3.0 hub is on its way here.

I'm doing systemd-analyze, +blame, +critical-chain test after every boot to see what services are being loaded and/or take too long to load. This morning networkd-manager was the culprit, so I stopped it, disabled it and masked it. I powered down, waited for the memory to discharge and rebooted. Here is a plot of that boot:


The entire format is new to me. There never was a "zero point" in previous plots. From kernel boot to display of the desktop took 3.4 seconds. Firmware and loader installation took 6 seconds. The delay from hitting the power switch to being presented with the login screen appears to be that 6 seconds. After I enter my password and hit the Enter key, my desktop appears almost instantly, but a couple more seconds pass before any of the task bar icons will respond.
3 seconds or 9 seconds, either one, I'm not complaining. The Acer login to desktop was 5-6 seconds. In line with my observation that this iRIS Xe is about twice as fast as the GT 650M, this makes sense.

Oh, I installed my Steam apps. When Steam starts up it complains that my GPU memory is only 512MB and said performance would suffer. For grins and giggles I ran a simulation in Universe Sandbox^2 that caused the Nvidia to stutter. it ran as smooth as silk. The Xe does not have its own memory but uses the main memory, in my case 12GB. I also ran Minetest 5.4 and Terasology and they both ran like lightening.

Last night I did something that I never did with the Acer because the Acer was too heavy and required a platform and vent fan. I took the HP to bed so my wife and I could watch some xmas love stories on YT. (Roku and YT got in a fight over money and YT cut Roku access off, YT won't allow casting of YT videos from my smartphone to our Roku TV either) The movies on this display were georgeous and seemed like 4K. BTW, we watched the 2019 "Sweet Mountain Christmas" Magen Hytel did a great job.

[GreyGeek]

GOOD NEWS!
Today I went to the HP documentation site and DL's all their manuals relating to the HP-17cn1xxx laptop. In one of them it showed an NVMe 128GB being installed onto the MOBO. I rechecked by photo based on their photo and noticed that there WAS a socket for an NVMe device!!! And, the black spot between the two brass mounting lugs in the lower right corner of the photo is where the screw that holds the NVMe in place belongs. I just ordered a 512GB NVMe device, L85364-005
It will be here in 12 days. It will be my internal storage device. I also purchased a USB to eSATA device that allows the direct connection of a SATA drive to a USB 3.0 port without going through a hub.

[GreyGeek]

Another quick report on Santa's gift.
Today, another kernel upgrade came down the pike:5.13.0-1028-oem, plus a lot of other stuff. I made a snapshot in case of failure and then turned loose the Kraken. When it was done I noticed something I have rarely seen, my wifi symbol showing 100% (all rings showing). I took a look at the details and was stunned to see it running at 434Mb out of 530Mb available on my fiber optic. I have never had a wifi chip that worked this fast. Web pages just snap onto the screen. I'm gonna need a seat belt for my chair, and a helmet with goggles.

[jglen490]

Give your hair that windswept look, too

[GreyGeek]

Give your hair that windswept look, too

What? I have hair?

[jglen490]

Sorry, that was dust on my screen

[GreyGeek]

sorry, that was dust on my screen

lol!!!

[GreyGeek]

The Samsung EVO 860 M.2 NVMe came in today.

This is more like it!

Smart Log for NVME device:nvme0n1p1 namespace-id:ffffffff
critical_warning : 0
temperature : 27 C
available_spare : 100%
available_spare_threshold : 10%
percentage_used : 0%
data_units_read : 89
data_units_written : 10
host_read_commands : 1,588
host_write_commands : 49
controller_busy_time : 0
power_cycles : 2
power_on_hours : 0
unsafe_shutdowns : 1
media_errors : 0
num_err_log_entries : 0
Warning Temperature Time : 0
Critical Composite Temperature Time : 0
Temperature Sensor 1 : 27 C
Temperature Sensor 2 : 27 C
Thermal Management T1 Trans Count : 0
Thermal Management T2 Trans Count : 0
Thermal Management T1 Total Time : 0
Thermal Management T2 Total Time : 0

Here are the front and back photos of the faux HP NVMe. It definitely isn't what was shown on the web page. It has the HP label, L853654-005 on it.