Bionicpup64 8.0 CE

[paulh177]

I found an old BCM4312 adapter and put it in a Bionic64 machine.
I installed the firmware package attached. Remove the fake .gz extension and click-install it. The adapter was detected by SNS and connected to my WiFi router. It worked fine on a brief test.

Ok so this worked, I booted in and the damn wireless worked!

Thanks very much @rcrsn51

BUT (there's a happy ending to this btw) the whole machine was still a slug, with the cpu(s) being hogged by udevd at ~100% ...



so, wtf?

what the screendump of top doesn't show is the intermittent appearance of udevd running a subprocess of hid2hci

A quick google search of 'udevd' 'hid2hci' and 'cpu' threw up this thread on SE

The 43xx chip in my Dell has Bluetooth on (I never use BT), and this is what is causing the massive cpu usage; Someone who knows about kernel stuff + drivers ought to read that thread in detail (? @666philb, ? @rockedge ).

Not only do I never use BT, I didn't even know if it was enabled on my machine, so I rebooted & F2'd into setup, found it was enabled, disabled it & booted into Bionic once more..

Result: instant boot, instant wireless.

Thanks to all who've offered help & suggestions.

Let's hope this is the last problem nailed & I don't have to drop into this thread again except to praise ...

Oh, and this is posted from a very relaxed Bionic64 over wireless.

[rcrsn51]

I checked my own BCM4312 setup and it doesn't show a Bluetooth device.

It may be that your machine has a separate internal BT adapter, possibly on the USB bus.

Run "lsusb". Does it list a BT device? Is it associated with Broadcom or another vendor?

[paulh177]

@rcrsn51
nope, no other BT devices on the machine;
hard to find out much about these old chipsets, but maybe they're different revisions, or maybe it's just a mystery.

[paulh177]

Not sure of *all* the details here, & not sure how many people this might affect, & not sure it is only a Bionic64 issue, but it has been a head-scratcher here for a few days now.

It boils down to this:

In order to compile & run Carla (an audio application for - amongst other things - LMMS), I needed to install PyQt5.

However doing so revealed a conflict between the pre-installed Qt5 library builtins on Bionic64 and PyQt5 Qt5 libraries.

The symptom was being unable to run small Python script that utilised Matplotlib, which triggered the error:

Code: Select all

QFactoryLoader::QFactoryLoader() checking directory path "/usr/local/bin/platforms" ...

Cannot load library /usr/local/lib/python3.8/site-packages/PyQt5/Qt/plugins/platforms/libqxcb.so:
(/usr/lib64/libQt5XcbQpa.so.5: symbol _ZN11QFontEngine14bitmapForGlyphEj6QFixedRK10QTransform version Qt_5_PRIVATE_API not defined in file libQt5Gui.so.5 with link time reference)

QLibraryPrivate::loadPlugin failed on "/usr/local/lib/python3.8/site-packages/PyQt5/Qt/plugins/platforms/libqxcb.so" :
"Cannot load library /usr/local/lib/python3.8/site-packages/PyQt5/Qt/plugins/platforms/libqxcb.so:
(/usr/lib64/libQt5XcbQpa.so.5: symbol _ZN11QFontEngine14bitmapForGlyphEj6QFixedRK10QTransform version Qt_5_PRIVATE_API not defined in file libQt5Gui.so.5 with link time reference)"
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "" even though it was found.

This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.

(googling this will find dozens & dozens of people confused by the error, many with differing solutions because of a variety of different causes; this level of detail in the error is got by setting QT_DEBUG_PLUGINS=1 btw)

If I uninstalled PyQt5, the Matplotlib script would run, but then Carla wouldn't.

I have devx & 32-bit compat sfss loaded, permanently btw. Unloading them did not clear the problem.
My default Python is 3.8.1, compiled and installed by me, but it will be the same for 3.6.x

A search for some of the mentioned offending libs reveals this:

Code: Select all

root# find /usr /lib |grep libqxcb.so
/usr/local/lib/python3.8/site-packages/PyQt5/Qt/plugins/platforms/libqxcb.so
/usr/lib/qt5/plugins/platforms/libqxcb.so
/usr/lib/i386-linux-gnu/qt5/plugins/platforms/libqxcb.so

root# find /usr /lib |grep libqxcb.so|xargs ls -l
-rw-r--r-- 1 root root  9748 Mar 16  2018 /usr/lib/i386-linux-gnu/qt5/plugins/platforms/libqxcb.so
-rw-r--r-- 1 root root 10520 Feb 26  2019 /usr/lib/qt5/plugins/platforms/libqxcb.so
-rwxr-xr-x 1 root root 15648 Jan 15 11:57 /usr/local/lib/python3.8/site-packages/PyQt5/Qt/plugins/platforms/libqxcb.so

root# find /usr /lib |grep libQt5XcbQpa.so.5
/usr/local/lib/python3.8/site-packages/PyQt5/Qt/lib/libQt5XcbQpa.so.5
/usr/lib/i386-linux-gnu/libQt5XcbQpa.so.5
/usr/lib/i386-linux-gnu/libQt5XcbQpa.so.5.9
/usr/lib/i386-linux-gnu/libQt5XcbQpa.so.5.9.4
/usr/lib/libQt5XcbQpa.so.5
/usr/lib/libQt5XcbQpa.so.5.9
/usr/lib/libQt5XcbQpa.so.5.9.5

root# echo $PATH
/bin:/usr/bin:/sbin:/usr/sbin:/usr/local/bin:/usr/X11R7/bin:/root/my-applications/bin:/opt/java/bin

root# echo $LD_LIBRARY_PATH
/lib64:/usr/lib64:/usr/X11R7/lib:/root/my-applications/lib:/usr/local/lib:/opt/jre1.8.0_231/lib/i386:/opt/jre1.8.0_231/lib/i386/client

So as we can see, there are multiple versions of various libs dotted about the place, and my assumption was that the PyQt5 libs probably needed to take precedence.

After a lot of false starts and trailing around on StackOverflow, I came upon this helpful thread, specifically the answer given by "bossbarber"

The clue here is the order in which the lib paths are searched. As you can see on my system, the builtins are located at the top of the order of $LD_LIBRARY_PATH.

If however I reset my $LD_LIBRARY PATH to

Code: Select all

/usr/local/lib/python3.8/site-packages/PyQt5/Qt/lib:/lib64:/usr/lib64:/usr/X11R7/lib:/root/my-applications/lib:/usr/local/lib:/opt/jre1.8.0_231/lib/i386:/opt/jre1.8.0_231/lib/i386/client

Everything (Carla, compiles, Python matplotlib scripts) now worked.

Whether this will lead to a further conflict down the line with a pure Qt5 app remains to be seen.

[rockedge]

I have been having all kinds of problems with python3. I am working with object recognition and deep machine learning and the project
https://github.com/pliablepixels/zmeventnotification

as soon as I attempted to build and use python 3.8 everything broke and just limps along.

[bigpup]

Some bug reports about Bionicpup64 8.0
http://www.murga-linux.com/puppy/viewtopic.php?t=117880
http://www.murga-linux.com/puppy/viewtopic.php?t=117879

[ally]

hi guys

I am struggling to get wifi working on my lenovo w520 4284-4dg using bionicpup4 8.0

works ok in xenialpup64

lspci: 03:00.0 Network controller: Intel Corporation Centrino Ultimate-N 6300 (rev 3e)

dmesg | grep wifi (from xenial 64)
iwlwifi 0000:03:00.0: can't disable ASPM; OS doesn't have ASPM control
iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000-6.ucode failed with error -2
iwlwifi 0000:03:00.0: Falling back to user helper
iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000-5.ucode failed with error -2
iwlwifi 0000:03:00.0: Falling back to user helper
iwlwifi 0000:03:00.0: loaded firmware version 9.221.4.1 build 25532 op_mode iwldvm
iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUG disabled
iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUGFS disabled
iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEVICE_TRACING disabled
iwlwifi 0000:03:00.0: Detected Intel(R) Centrino(R) Ultimate-N 6300 AGN, REV=0x74
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1


I was going to copy over the /lib/firmware from xenial to bionic but got a bit lost

help appreciated!!

[bigpup]

I am struggling to get wifi working on my lenovo w520 4284-4dg using bionicpup4 8.0

What exactly is the problem?
Wifi not recognized?

[Terry H]

hi guys

I am struggling to get wifi working on my lenovo w520 4284-4dg using bionicpup4 8.0

works ok in xenialpup64

lspci: 03:00.0 Network controller: Intel Corporation Centrino Ultimate-N 6300 (rev 3e)

dmesg | grep wifi (from xenial 64)
iwlwifi 0000:03:00.0: can't disable ASPM; OS doesn't have ASPM control
iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000-6.ucode failed with error -2
iwlwifi 0000:03:00.0: Falling back to user helper
iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000-5.ucode failed with error -2
iwlwifi 0000:03:00.0: Falling back to user helper
iwlwifi 0000:03:00.0: loaded firmware version 9.221.4.1 build 25532 op_mode iwldvm
iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUG disabled
iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUGFS disabled
iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEVICE_TRACING disabled
iwlwifi 0000:03:00.0: Detected Intel(R) Centrino(R) Ultimate-N 6300 AGN, REV=0x74
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: L1 Enabled - LTR Disabled
iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1


I was going to copy over the /lib/firmware from xenial to bionic but got a bit lost

help appreciated!!

ally, you posted the dmesg from xenial which works, but not the dmesg form bionic. What does dmesg show in bionic? This will identify what you need.

Copying the firmware from xenial to bionic, if it shows in dmesg as not found is probably OK. If you are using an older kernel in xenial, the version of firmware required for bionic may be later than that included in xenial.

[rockedge]

try switching the Xenial kernel that works into the Bionic you would like to run

[ally]

apologies....

@bigpup, sorry, yes, wifi device not recognised, in hardinfo the iwlwifi module is listed against "03:00.0 Network controller: Intel Corporation Centrino Ultimate-N 6300 (rev 3e)"

dmesg | grep wifi (from bionicpup64 8.0)

iwlwifi 0000:03:00.0: can't disable ASPM; OS doesn't have ASPM control
iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000-6.ucode failed with error -2
iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000-5.ucode failed with error -2
iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000-4.ucode failed with error -2
iwlwifi 0000:03:00.0: no suitable firmware found!
iwlwifi 0000:03:00.0: minimum version required: iwlwifi-6000-4
iwlwifi 0000:03:00.0: maximum version supported: iwlwifi-6000-6
iwlwifi 0000:03:00.0: check git://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git

[Terry H]

ally,
Any of the 3 firmware files listed in the dmesg will work, iwlwifi-6000-6.ucode | iwlwifi-6000-5.ucode | iwlwifi-6000-4.ucode.

Your Xenial install has iwlwifi-6000-4.ucode included, so copy that to /lib/firmware in your bionic install.

[ally]

thanks terry

how do I copy the file(s) over, I'm trying to move from zdrv/lib/firmware but it fails with read-only?

[Terry H]

If you click on the zdrv using ROX. you will be asked whether you want to install SFS or view content. Select to view content, this will mount the zdrv. It will be mounted as Read only, that does not stop files from being copied, So you can then copy the firmware by using drag and drop and selecting copy when the dialog window is displayed when you drag the file to the required location.

[ally]

terry

I must be missing something, opening xenial64 zdrv in rox and bionicpup64 zdrv in adjacent panes after clicking view contents, tried dragging and copy/paste the iwlwifi-6000-4.ucode file

I'm still getting the read only error and no file copied to bionicpup zdrv

[ally]

terry

all sorted - thanks

so I put on my big boy pants, extracted the zdrv sfs, copied on the iwlwifi 6000.4 file then re-squashed the sfs

I'm very proud of myself....

[Terry H]

Just copy it to the /lib/firmware in you file system, so it will be in your save file/folder.

If you want it in the zdrv, you would have to unsquashfs then copy the file to the directory created by when you unsquash for zdrv. Then you would have to use mksquashfs to reproduce the zdrv.

It's much easier to just add it to your save.

Edit: Our posts crossed. Well done.

[ally]

gotcha

I was trying to copy from zdrv to zdrv and not zdrive to file/lib/firmware

the latter is much easier!!

filed away for the next time

appreciate the help

[Mike Walsh]

Morning, all.

Is it my imagination, or do I recall there being an issue with the CPU Frequency Scaling Tool in Bionicpup64? (It may have been one of the other modern Pups; I'm not certain.)

Bionicpup64 is the first Puppy I've got set-up and running the way I like it on this new machine of mine.....but I don't seem to be able to set the 'governor' to 'balanced' (which is what I normally use.) Consequently, this Pentium Gold G5400 is running flat-out @ 3.7 GHz all the time..!! Power's not an issue, since it's a desktop, but I would like to be able to make use of the SpeedStep stuff.

After selecting the 'on-demand' governor, there is nothing showing in the frequency selection boxes. I'm sure I remember there being a 'fix' issued for this (it may have been ozsouth who came up with it?)

In all honesty, I just haven't got time to read all the way through this thread ATM, so any pointers or advice would be very much appreciated, as always.

TIA.


Mike.

[LateAdopter]

Hello Mike

By default your G5400 will be using the intel_pstate driver. It has performance and powersave modes. The default may be sset by the BIOS settings or by the kernel configuration. It is not managed by the Puppy frequency scaling tool.
You can override the default via sysfs.

My intel N3150 defaults to powersave.

Alternatively you can put intel_pstate=disable or intel_pstate=passive as a kernel boot parameter. If you disable then the ondemand governor can take over and work normally but the processor HWP may not be fully configured. passive allows the intel_pstate to configure the processor, but I don't know what happens afterwards because I haven't tried it.

I bought an Athlon 3000G rather than a Pentium Gold G5420 because the AMD processor avoids most of the vulnerabilities