Puppy Linux Discussion Forum

Which fdrv (firmware) do you have installed? Does it include the bcm-4312 firmware?

Do you have a kernel (zdrv & fdrv & vmlinuz) which works with your wifi? Maybe try swapping 5.10 for that?

@charlie6 - the following kernel / driver combo should fix your issue (drivers are kernel-specific):
kernel: https://forum.puppylinux.com/viewtopic. ... ea2#p56329
driver: https://www.mediafire.com/file/4ubv4o63 ... 3.pet/file
(After kernel swapped & reboot, install driver .pet by clicking on it in your Filemanager. You must save on exit to retain it).

Hi peebee,
Hi ozsouth,
Thanks to both of you for your time !
Here running fdrv_LxPupSc64_22.02.sfs so renamed after extracted it from the 263Mb sized kernel-kit-output-5.10.x-x86_64.zip archive, together with vmlinuz, adrv***.sfs, zdrv***.sfs.

to ozsouth: the suggested kernel/combo is related to k5.10.113. Here running on k5.10.123 ... could it work on k5.10.123 although compiled on k5.10.113?

Best regards, Charlie

@charlie6 - unfortunately, drivers must match the kernel exactly.

bcmwl source:
from advice at:
https://forums.linuxmint.com/viewtopic.php?p=1675396

https://pkgs.org/search/?q=bcmwl-kernel-source

@charlie6 you may be better off with the original fdrv from the iso as that has b43 firmware......

kernel-kit 5.10.123 should have the b43 driver at:
/lib/modules/5.10.123/kernel/drivers/net/wireless/broadcom/b43

No reply from Charlie6, who should know better.

LxPupSc64-22.02+2-T works OOTB with the BCM4322 adapter. There is no need to change kernels or switch to the wl driver.

The correct b43 firmware is already present.

Hi rcrsn51, peebee and all,
sorry for my late reply (here still very, very buzy with recent move). Anyway thanks for worrying...

@charlie6 you may be better off with the original fdrv from the iso as that has b43 firmware......

That did the trick! Running it since a few days as now.
In order to summarize the current working configuration:
- to fix the «freezing» upon powerdown: get the adrv_***.sfs, zdrv_***.sfs (NOT the fdrv_***.sfs) and vmlinuz from the 263Mb sized kernel-kit-output-5.10.x-x86_64.zip archive; and rename them replacing the above *** by _LxPupSc64_22.02.

# uname -a
Linux puppypc32254 5.10.123 #1 SMP Mon Jun 20 00:30:18 UTC 2022 x86_64 Intel(R) Core(TM)2 Duo CPU L9600 @ 2.13GHz GenuineIntel GNU/Linux
#

- kept the original fdrv_LxPupSc64_22.02.sfs (wifi connection on BCM4312 wifi interface)

# lspci -nn -d 14e4:
02:00.0 Network controller [0280]: Broadcom Inc. and subsidiaries BCM4312 802.11b/g LP-PHY [14e4:4315] (rev 01)
#

Here then currently running kernel 5.10.123 with b43 driver from kernel 5.16.7-fdrv (I thought all stuff should have been build on the same kernel... )

Again many thanks for your time.
Charlie

Firmware is nothing to do with kernel - its code that the kernel downloads into devices - it's therefore kernel agnostic. Drivers are the code that runs inside the kernel and therefore any out-of-tree drivers added are kernel specific.

In this case the b43 drivers were in-tree in the 5.10 kernel-kit kernel but the b43 firmware was not in the fdrv supplied with the kernel.

Hi peebee, thanks for your explanation !
It will help me to a bit more kernel and drivers understanding.
Best regards from Belgium !
Charlie

Kernel bump from 5.18.14 to 5.18.15, then quickly to 5.19.0. Tested in the current LxPupSc64, VoidPup64, and fossapup64 9.5. All on the Fujitsu S761 hardware, frugal install to SSD.

▶—— BASE SYSTEM ——◀

PC Manufacturer: FUJITSU
Product Name: LIFEBOOK S761

Motherboard Vendor: FUJITSU
Product Name: FJNB225

BIOS Vendor: FUJITSU // Phoenix Technologies Ltd.
Version: Version 1.17
Release Date: 03/14/2012

Intel(R) Core(TM) i5-2520M CPU @ 2.50GHz
Min/Max Speed: 800/3200 MHz
Current Speed of Core 0:3013 MHz, 1:3074 MHz, 2:2990 MHz, 3:3153 MHz
Core Count: 2
Thread Count: 4

Frequency governor   : schedutil
Freq. scaling driver : intel_cpufreq

CPU Vulnerabilities:
 itlb_multihit:KVM: Mitigation: VMX unsupported
 l1tf:Mitigation: PTE Inversion
 mds:Mitigation: Clear CPU buffers; SMT vulnerable
 meltdown:Mitigation: PTI
 mmio_stale_data:Not affected
 retbleed:Not affected
 spec_store_bypass:Mitigation: Speculative Store Bypass disabled via prctl
 spectre_v1:Mitigation: usercopy/swapgs barriers and __user pointer sanitization
 spectre_v2:Mitigation: Retpolines, IBPB: conditional, IBRS_FW, STIBP: conditional, RSB filling
 srbds:Not affected
 tsx_async_abort:Not affected

Voidpup and fossapup run with LXDE ydrives, synaptics drivers for the touchpads on LxPupSc64 and VoidPup64 installed in liu of libinput. In general, all larger programs such as browsers, Office stuff, etc. run as shared portables. Profiles and data storage linked to a data partition so minimal savefiles (256Mb) in each for look'n'feel suffice.

Clean Dmesg in all, very low idle resource use, and excellent glxgears FPS in all. I'm in fossapup64 right now posting using Slimjet Version 35.0.1.0 (based on Chromium 101.0.4951.34) run as a portable.

The only place I have to drop back to an 'older' kernel is on my two Bay Trail desktops. There Ozsouths' 5.15.50 is my kernel of choice in all pups. Currently I have 6 of the Fujitsu laptops and two Bay Trail desktops in service.

Thanks,

c3e305cc205744104b19e9fe989cfa44 LxPupSc64-22.02+4-T.iso

Tested LxPupSc64 version 22.02 +4.

First a kernel note: 5.19.0 is running flawlessly on LxPupSc64 22.02 +2. VoidPup64 22.02 +8, and fossupup64 9.5 on my i5 based Fujitsu S761.

LxPupSc64 22.02 +4 iso generated using the delta and md5sum checked. Install then done, frugal to SSD on the usual Fujitsu S661 laptop. Tested as pristine install first, set up through touchpad config and SNS internet connection using the kernel supported Atheros card. Then as a 'quasi pristine install' using the 5.19.0 kernel, then as an update to my 22.02 +2 install, both with the stock 5.16.7 kernel and with the 5.19.0 kernel.

In all cases, clean boot, clean dmesg, intel video driver used, Atheros card internet connection fine, and normal light use (browser, mail, geany, pcmanfm etc.) is ok.

One thing of note: In 22.02 +4 (both pristine and updates, both kernels, same theme) the idle resource use is much poorer (5% or 6% CPU use vs 1% for 22.02 +2 and memory use roughly 290MB vs 225MB for 22.02 +2). glxgears FPS also is measurably lower in 22.02 +4 (6xxx vs 9xxx in 22.02 +2), also kernel independent.

Looking at lxtask, at idle in 22.02 +2, only lxtask itself shows any CPU %, roughly 1% with a 5 sec update setting. In 22.02 +4 Xorg uses a pretty constant 3% to 4%, then lxtask at 1%. Kernel and install type independent again.

I'm using VoidPup64 pretty much exclusively as my daily now but I'll leave LxPupSc64 at version 22.02 +4 on my personal computer for now in case I want to futz with it. Other supported laptops will stay at 22.02 +2. 22.02 +2 seems happier than 22.02 +4 on this hardware anyway for now.

Cheers,

Update as of October 27,2022: Kernel bumped from 6.02 to 6.05 in LxPupSc64 22.02 +2, +4, +5 and in Voidpup64 22.02 +10c. All frugal installs on the i5 based Fujitsu laptops. Boots are clean in all, no dmesg failures and just the usual expected warnings. Using VoidPup64 +10c with 6.0.5 as my daily with LxPupSc64 22.02 +2 for some other users and some scanning and printing chores. Idle resource is excellent in VoidPup64 +10c and in LxPupSc64 22.02 +2, and glxgears in that LxPupSc64 is stellar.

I still have the higher idle resource use in LxPupSc 22.02 +4 and +5 with notably lower glxgears FPS in those. Two differences noted in them: xorg continues to show up top dog in lxtask using 5% or 6% of CPU, and at boot "making file system usable..." is a 'failure', not 'done' as in +2 or Voidpup. In spite of that; daily operation, saves, reboots, are fine?? I've ruled out any kernel dependence and looked at rc.sysinit changes, xorg, the bootsysinit logs, and have gotten nowhere. The change is definitely between +2 and +4.

Keeping on keeping on with LxPupSc64 +2 and Voidpup64 +10c, both with the 6.05 kernel.

Thanks,

New year, new pup! Thanks, @peebee

Fresh install for this one to the Fujitsu S761 laptop (2nd gen i5, all intel, circa 2012). md5sum correct, frugal pristine install (albeit without the adrv or fdrv). Clean boot, clean dmesg, stellar glxgears FPS, and the same slightly high idle CPU use (5% on this hardware) as in post +2 LxPupSc64 22.02s and in the current S15pup64 running LXDE. Not in Voidpup64 also running LXDE?? Libinput continues to not allow me to turn tap-to-click off on this hardware but my synaptics touchpad driver package installs ok and works perfectly here so that's a non-event. Basically a nice clean pretty complete LXDE platform.

Did this one the honor of a fresh built from scratch savefile. Added the touchpad driver pkg, configurations and lx-panel, .lx-rightclicks and drives_on_desktop, .desktops for my portables, and peasywifi 4.6. No hitches, no glitches. Saved to the fresh savefile and backed it up every few additions. The latest Brave, Slimjet, and un-googled chromium running as portables all are ok. Sylpheed as a portable also as are Master PDF Editor and LibOffice Writer. All present and accounted for. I've been using Audacity 2.1.1 (as a portable) in FP64_CE_radky6 and it wouldn't start here but I haven't chased it at all. Almost certainly a library/slackware/deps thing. Many pups don't always the broth spoil.

FWIW, my synaptics driver package is attached.

Thanks,

Test of 2 most current PUPs (bare metal): S15Pup64 & LxScPUP64

LxPup64

• ISO file Boots rapidly from Ventoy to desktop ... OK

• Boots fine to desktop from SG2D ... OK

• Does NOT have ExFAT support ... Failure (many/most new USB sticks come preformated with ExFAT. Gparted cannot interrogate them while ROX-Thunar cannot read them)

• SAMBA will not start ... Failure

  

  LxPup64 v2301+0T - SAMBA not starting.jpg (60.54 KiB) Viewed 3490 times

S15Pup64

• ISO file will not boot from Ventoy ... Failure

• Boots fine to desktop from SG2D ... OK

• Reads ExFAT without issue ... OK

• SAMBA starts & runs perfect ... OK

• Sessions folder found via PUP SAVESPEC in the ISOs folder allowing the Pristine boot session to save at shutdown

• Saved Sessions founds on reboot

@Clarity

Samba on LxPupSc64-23.01 seems OK for me.........

Thanks @peebee . Could you share the checksum for the version you are using, please? I am running a pristine LxPupSc64 whose md5sum is 97cec761454ce30fab61e5287075b7f3.

Does your version hae ExFAT support too?

Thanks

97cec761454ce30fab61e5287075b7f3 LxPupSc64-23.01+0-T.iso

Haven't investigated exfat yet.....

Clarity wrote: Wed Feb 01, 2023 12:04 pm

LxPup64
[*]Does NOT have ExFAT support ... Failure (many/most new USB sticks come preformated with ExFAT. Gparted cannot interrogate them while ROX-Thunar cannot read them)

Hello @Clarity

I cannot see any differences in the S15Pup and LxPupSc configs to explain why they behave differently wrt ExFAT.

Can you give me some clues as to what you think is missing please?

I guess you could try installing
https://slackware.uk/slackware/slackwar ... 6_64-1.txz

Is the different ExFat support a function of the kernel? LxPupSc64 23.01 has the 6.1.7 kernel while the S15Pup64 +2 has the 5.15.80. Trying a kernel swap would quickly rule this out but I don't have anything ExFat to do it on.

Good thinking @Marv - have added:

# DOS/FAT/EXFAT/NT Filesystems
#
CONFIG_EXFAT_FS=y
CONFIG_EXFAT_DEFAULT_IOCHARSET="utf8"
# end of DOS/FAT/EXFAT/NT Filesystems

to future 6series kernel builds

@peebee :

Your newest 6.1.10 64b kernel swapped into my frugal LxPupSc64 23.01 install on the fujitsu S761. Clean boot and dmesg, idle resource unchanged from the 6.1.7 kernel. Uses the intel video driver, no change there, with Stellar glxgears FPS. Perhaps the highest I've ever seen on this lappie. Looking at the DOTconfig, EXFAT should be supported now but I have no EXFAT drives to test that on. Normal daily use seems fine -basically unchanged from the stock LxPupSc64 kernel-

Cheers,

@Clarity did you manage to test k6.1.10 for EXFAT?

I am worried that one of these configs may be wrong:
CONFIG_EXFAT_FS=y
CONFIG_EXFAT_DONT_MOUNT_VFAT=y
CONFIG_EXFAT_DISCARD=y
CONFIG_EXFAT_DELAYED_SYNC=y
# CONFIG_EXFAT_KERNEL_DEBUG is not set
# CONFIG_EXFAT_DEBUG_MSG is not set
CONFIG_EXFAT_DEFAULT_CODEPAGE=437
CONFIG_EXFAT_DEFAULT_IOCHARSET="utf8"

Hello @peebee, @Marv (or anyone)

Since I have never done this before, I ask guidance:

• Can I just open the ISO file in ISOmaster, replace the kernel with the new one to create an ISO with the changed kernel?
  OR
  Do you prefer some other procedure to create the test distro with the kernel?

Thanks in advance

Clarity wrote: Thu Feb 09, 2023 2:48 am

Hello @peebee, @Marv (or anyone)

Since I have never done this before, I ask guidance:

• Can I just open the ISO file in ISOmaster, replace the kernel with the new one to create an ISO with the changed kernel?
  OR
  Do you prefer some other procedure to create the test distro with the kernel?

Thanks in advance

Yes you can replace kernel file and zdrv inside the iso using ISO master

Thanks for confirmation.

Where can I source the current files that would need testing?

https://sourceforge.net/projects/lxpup/ ... 2/download

Thanks @peebee ,

I am testing this version of LxPupSc64 v23.01+0-T (md5sum checksum 97cec761454ce30fab61e5287075b7f3). I will test bare metal and report as soon as I can free up a test PC. Expect feedback on

1. ExFAT use mounted in running system as well as

2. ExFat reporting on partition usage in GParted.

I need help identifying the 'kernel' location-filename.

Help please

@Clarity ,

Correct, the vmlinuz-xxx-xxx file just gets renamed vmlinuz. The kernel-modules file gets renamed zdrv_LxPupSc64_23.01.sfs to match the LxPupSc64 naming convention. I'd rename them both in whatever location you saved them in after download/extraction, and then use isomaster to replace the two in the stock iso with them. Was that the main question? Basically, in the iso, the kernel files are treated no differently than the other files that make up the pup. Just compressed with the rest in the 'iso' format.

@peebee and @Marv THANKS for your effort and guidance.

Tested ... Results ===> "LxPupSc64-23.02.10Test.iso"

I have added exfatprogs to +1d and GParted now works with EXFAT.....
Kernel > 6.10....
https://sourceforge.net/projects/lxpup/ ... /interims/